IMF 2005 - Dollars Debt and Deficits - Sixty Years After Bretton Woods

Dollars, Debt, and Deficits
Sixty Years after Bretton Woods

Conference Proceedings
Banco de Espaa
International Monetary Fund
International Monetary Fund. Not for Redistribution
2005 International Monetary Fund

Production: IMF Multimedia Services Division
Cover: Jorge Salazar
Dollars, debt and deficits : sixty years after Bretton Woods : conference
proceedings [Washington, D.C.] : International Monetary Fund : Banco de Espaa, [2005]
p.
cm.
Organized by Banco de Espaa and the International Monetary Fund.
ISBN 1-58906-453-4
1. International finance Congresses. 2. Dollar Congresses. 3. Debt
Congresses. 4. Foreign exchange rates Congresses. 5. International Monetary Fund
Congresses.
I. International Monetary Fund. II Banco de Espaa.
HG3881.D67 2005
Price: US$35
Please send orders to:

International Monetary Fund Publication Services
700 19th Street, N.W., Washington, D.C. 20431, U.S.A.
Tel: (202) 623-7430 Telefax: (202) 623-7201
E-mail: publications@imf.org
Internet: http://www.imf.org
recycled paper
Contents
List of Participants
Foreword
Rodrigo deRato and Jaime Caruana
vii
INTRODUCTION
Raghuram Rajan and Jose Vinals
OPENING SPEECHES
Rodrigo de Rato
Pedro Solbes
Jaime Caruana
BLOCK I: GLOBAL IMBALANCES, EXCHANGE RATE

ISSUES, AND DEBT
11
17
21
Session 1; Global Imbalances

A Map to the Revised Bretton Woods End Game: Direct Investment,
Rising Real Wages, and the Absorption of Excess Labor in the Periphery
Michael P. Dooley, David Folkerts-Landau, and Peter Garber
Macroeconomic Dynamics and the Accumulation of Net Foreign
Liabilities in the US: An Empirical Model
Giancarlo Corsetti and Panagiotis Th. Konstantinou
29
47
Session 2; Exchange Rate Issues

Financial Globalization and Exchange Rates
Philip R. Lane and Gian Maria Milesi-Ferretti
What Makes Balance Sheets Effects Detrimental for the
Country Risk Premium?
Juan Carlos Berganza and Alicia Garcia Herrero
93
135
Session 3; Debt in Emerging Economies

Public Debt, Fiscal Solvency, and Macroeconomic Uncertainty in
Latin America: The Cases of Brazil, Colombia, Costa Rica and Mexico
Enrique G. Mendoza and Pedro Marcelo Oviedo
Exchange Rate Regimes and Debt Maturity Structure
Matthieu Bussire, Marcel Fratzscher, and Winfried Koeniger
165
197
Contributions;
Daniel Cohen (discussant)
Stanley Fischer (panelist)
Malcom Knight (panelist)
213
215
219
iii
KEYNOTE SPEECHES
Agustin Carstens
Jean-Claude Trichet
BLOCK II: INTERNATIONAL FINANCIAL ARCHITECTURE
233
237
Session 4; The Role of the IMF

A Model of the IMF as a Coinsurance Arrangement
Ralph Chami, Sunil Sharma, and Ilhyock Shim
The IMF in a World of Private Capital Markets
Barry Eichengreen, Kenneth Kletzer, and Ashoka Mody
249
291
Session 5; Innovations in Private and Multilateral Lending

Dollars, Debt, and the IFIs: Dedollarizing Multilateral Lending
Eduardo Leyy-Yeyati
Optimal Collective Action Clause Thresholds
Andrew G. Haldane, Adrian Penalver, Victoria Saporta,
and Hyun Song Shin
325
363
Contributions;
Lorenzo Bini-Smaghi (discussant)
John Murray (discussant)
Guillermo Ortiz (panelist)
383
3 87
391
IV
List of Participants
Enrique Alberola, Unit Manager, Banco de Espaiia
Juan Carlos Berganza, Economist, Banco de Espaiia
Lorenzo Bini-Smaghi, Director General, Italian Ministry of Economy and Finance
Jorge Blazquez, Advisor, Economic Office of the President of the Spanish Government
Geoffrey H. Board, Board Chief Representative, Reserve Bank of Australia
Jan Brockmeyer, Deputy Executive Director, De Nederlandsche Bank
Ariel Buira, Director of the G-24 Secretariat
Matthieu Bussiere, Senior Economist, European Central Bank
William Calvo, Chief Economist, Banco Central de Costa Rica
Jose Manuel Campa, Professor, IESE
Agustin Carstens, Deputy Managing Director, International Monetary Fund
Jaime Caruana, Governor, Banco de Espaiia
Miguel de las Casas, Economist, Banco de Espaiia
Michael Casey, Head of the Funds Supervision, Central Bank & Financial Services Authorities of
Ireland
Ralph Chami, Division Chief, International Monetary Fund
Roberto Cippa, Director of International Monetary Relations, Swiss National Bank
Kenneth Coates, Director, Centre for Latin American Monetary Studies
Daniel Cohen, Professor, Ecole Normale Superieure
Giancarlo Corsetti, Professor, European University Institute
Daniel Daco, Head of Division, European Commission
Thomas Dawson, Head of the External Relations Department, International Monetary Fund
Guillermo de la Dehesa, Member of the Board, AVIV A
Michael Dooley, Professor, University of California and Deutsche Bank Special Advisor
Sonsoles Eirea, Economist, Banco de Espaiia
Santiago Elorza, Technical Advisor, Spanish Ministry of Economy
Yuchuan Feng, Deputy Governor's Assistant, People's Bank of China
Roger Ferguson, Vice Chairman, Board of Governors of the Federal Reserve System
Vicente Javier Fernandez, Advisor, Spanish Secretariat of State for the Economy
Santiago Fernandez de Lis, Director of the International Economics and International Relations
Department, Banco de Espaiia
Inigo Fernandez de Mesa, Deputy Director General of EMU Affairs, Spanish Ministry of Economy and
Finance
Stanley Fischer, Member of the Board, Citigroup Inc.
Allen Frankel, Head of the CGFS, Bank for International Settlements
Marcel Fratzscher, Senior Economist, European Central Bank
Luis de Fuentes, Deputy Director General, Spanish Ministry of Industry, Trade and Tourism
Peter Garber, Global Strategist, Deutsche Bank
Alicia Garcia Herrero, Head of Division, Banco de Espaiia
Luis Garcia Lombardero, Advisor, Spanish Secretariat of State for the Economy
Jose Garcia Solanes, Professor, University of Murcia
Francisco Gismondi, Director of Macroeconomic Analysis, Banco Central de la Republica Argentina
Giorgio Gomel, Director of International Affairs, Banca d'ltalia
Jane Haltmaier, Division Chief, Board of Governors of the Federal Reserve System
Herve Hannoun, Deputy Governor, Banque de France
Claudio Irigoyen, Director of Monetary and Financial Policy, Banco Central de la Republica Argentina
Zhongxia Jin, Director General of the International Department, People's Bank of China
Alfred Kammer, Deputy Managing Director's Advisor, International Monetary Fund
Kenneth Kletzer, Professor, University of California
Malcolm Knight, General Manager, Bank for International Settlements
Panagiotis Konstantinou, RTN Research Fellow at the University of Rome III
Philip Lane, Director of the HIS, Trinity College Dublin and CEPR
Nuno Leal de Faria, Deputy Director of the International Relations Department, Banco de Portugal
Eduardo Levy Yeyati, Professor, University Torcuato Di Telia

Ruogu Li, Deputy Governor, People's Bank of China
Zhengming Liu, Chief Representative, People's Bank of China Representative Office in Frankfurt
Jose Luis Malo de Molina, Director General of Economics, Statistics and Research, Banco de Espafia
Luis Marti, Executive Director, International Monetary Fund
Manuel Martinez, Economist, Banco de Espafia
Aurelio Martinez Estevez, Professor, University of Valencia
Jose R. Martinez Resano, Economist, Banco de Espafia
Enrique Mendoza, Professor, University of Maryland
Gian M. Milesi-Ferretti, Division Chief, International Monetary Fund
Ashoka Mody, Division Chief, International Monetary Fund
Joaquin Muns, Member of the Board, Banco de Espafia
John Murray, Advisor, Bank of Canada
Michael Mussa, Senior Fellow, Institute for International Economics
Franz Nauschnigg, Head of Division, Oesterreichische National Bank
Emilio Ontiveros, Member of the Board, API
Guillermo Ortiz, Governor, Banco de Mexico
Adrian Penalver, Economist, Bank of England
Jose Perez, President, Intermoney
Kristina Persson, Deputy Governor, Sveriges Risksbank
Jean Pisani-Ferry, Professor, University of Paris-Dauphine
Richard Portes, Professor, London Business School and CEPR
Federico Prades, Advisor, AEB (Spanish Banking Association)
Raghuram Raj an, Economic Counsellor and Director of Research, International Monetary Fund
Rodrigo de Rato, Managing Director, International Monetary Fund
Luis Ravina, Rector, University of Navarra
Carmen Reinhart, Professor, University of Maryland
Rafael Repullo, Director, CEMFI
Jose Juan Ruiz, Head of Department, Santander Central Hispano
Juan Ruiz, Economist, Banco de Espafia
Sinikka Salo, Member of the Executive Board, Suomen Pankki
Javier Santiso, Chief Economist for Latin America and Emerging Markets, BBVA
Jesus Saurina, Head of Division, Banco de Espafia
Miguel Savastano, Division Chief, International Monetary Fund
Miguel Sebastian, Director of the Economic Office of the President of the Spanish Government
Sunil Sharma, Division Chief, International Monetary Fund
Pedro Solbes, Second Vice President of the Spanish Government and Minister for Economy and Finance
Alberto Soler, Advisor, Spanish Ministry of Economy and Finance
Michel Soudan, Assistant of the International Service, National Bank of Belgium
Marc Olivier Strauss-Kahn, Director General, Banque de France
Gyorgy Szapary, Vice President, Central Bank of Hungary
David Taguas, Economist, BBVA
Fernando Tejada, Head of Division, Banco de Espafia
Panayotis Thomopoulos, Deputy Governor, Bank of Greece
Jens Thorn sen, Member of the Board, Danmarks Nationalbank
Jean-Claude Trichet, President, European Central Bank
Juan Jose Toribio, Director, IESE
Jose Uribe, Head of the Management Office, Banco de la Republica de Colombia
Pierre Van der Haegen, Director General of International and European Relations, European Central
Bank
Felix Varela, Head of Research and International Economy, University of Alcala
Manuel Varela, Emeritus Professor, University Complutense Madrid
Fernando Varela, Former Executive Director, International Monetary Fund
David Vegara, Secretary of State for the Economy
Jaume Ventura, Professor, University Pompeu Fabra
Pedro Pablo Villasante, Director General of Banking Supervision, Banco de Espafia
Jose Vinals, Director General of International Affairs, Banco de Espafia
Regine Wolfinger, Economist, European Central Bank
VI
Foreword
Concerns over the major challenges the international monetary and financial system
currently faces and the attendant debate on the policy-related and academic facets of
the international financial architecture have arisen from the lessons imparted by the
financial crises of recent years and from the need to redefine the international financial
community's strategies and strengthen crisis-prevention mechanisms. A new landmark
in this connection has been the international conference Dollars, Debt and Deficits:
Sixty years after Bretton Woods, which took place at the Banco de Espafia headquarters
on 14-15 June 2004. The conference was jointly organized by the Banco de Espafia and
the International Monetary Fund (IMF).
The conference commemorated the 60th anniversary of the Bretton Woods
meetings in July 1944. These meetings were one of the most significant events for the
world economy in the last century and gave rise to the birth of an economic, monetary
and financial system that has played a crucial role in promoting economic and financial
stability, improving global welfare, and alleviating poverty. Representatives from 45
countries met at Bretton Woods against the backdrop of a world economy devastated
by war and marked by protectionist trade policies and unstable exchange rates. The
foundations were laid there for a stable framework of economic cooperation and an
international financial and monetary system, and two institutions which have proved
pivotal for the attainment of these goals were created: the IMF and the International
Bank for Reconstruction and Development, later the World Bank.
Spain's growing ties with the Bretton Woods institutions reached a high in
autumn 1994 when Madrid hosted the IMF/World Bank annual meetings, coinciding
with the 50th anniversary of the agreements. Marking another pinnacle was the IMF's
obliging acceptance of the proposal to commemorate the 60th anniversary with the
organization of an international conference at the Banco de Espafia. Significantly, this
has been one of the few occasions on which the IMF has decided to hold a meeting of
this nature outside its Washington headquarters.
Conference preparations were set in train in December 2003, when the coorganizers made a call for papers to the academic community for contributions on
economic development, policy measures and implications of debt sustainability,
exchange rates and global imbalance in the capital account, debt and exchange rate
crises, crisis-resolution strategies, the relationship between the adopted exchange rate
regime and economic development, the link between fiscal development and exchange
rate regimes, and regional integration and the main international currencies. The
selection of contributions from among those received was by a committee made up of
representatives from the IMF and the Banco de Espafia. The committee was pleasantly
surprised by the number and quality of the contributions submitted, ten of which were
finally chosen. For the session chairs, discussants and expert panels, prestigious
academics and expert staff from central banks, international agencies and economic
research institutes were called upon. The outcome was a high-quality program that
vn
generated notable expectations about the conference, expectations which were more
than fulfilled thanks to the calibre of the participants and the results of the debates.
Among conference speakers and participants, a significant total of 27 countries were
represented.
The conference was therefore an excellent opportunity to reflect on the major
challenges currently facing the international monetary and financial system and on the
future role of the Bretton Woods institutions. It has provided for better understanding
of the development of and changes in the international financial architecture and has
helped offer a perspective as to how a global response should be designed in the future
to avert the perverse effects of crises on the world economy. The conference also
allowed a fruitful exchange of views on exchange rate fluctuations in an increasingly
integrated world economy, the consequences of political uncertainty and other issues
relating to exchange rates, and the term structure of interest rates and the sustainability
of public debt.
The main contributions and debates turned on the dynamics of international
disequilibria, such as the US external deficit, the effects of exchange rate fluctuations,
the impact of uncertainty on public and private external debt, the role of the IMF and of
international financial institutions in reducing emerging economies' foreign currency
financing needs, and the prospects for the international financial architecture in light of
the emergence of new international lending instruments and mechanisms.
In sum, the tight but stimulating conference agenda highlighted once again the
need for national authorities and international financial institutions to coordinate their
work in areas of common interest, harnessing their experience and comparative
advantages, and providing the degree of technical rigour needed to address global
financial problems.
Rod rigo de Rato
IMF Managing Director
Jaime Caruana
Governor of the Bank of Spain
Vlll
Introduction
Raghuram Raj an
Jose Viiials
Banco de Espana
In their opening remarks given to the Conference, Rodrigo Rato, Pedro Solbes and
Jaime Caruana reviewed the history and future challenges of the IMF sixty years after
its creation. They emphasize the role of the IMF in monitoring the financial system, and
in preventing and managing financial crises, stressing the importance of coordination
between international financial institutions (IFIs), national authorities, central banks,
and supervisory agencies. The main discussions gathered in this book focus on issues
related to debt, deficits, exchange rates, and the international financial architecture,
with authors underscoring the importance of these topics for the future configuration of
the international monetary system.
Block I: Global Imbalances, Exchange Rate Issues, and Debt
The dynamics of global imbalances, with special attention to the US and China, is the
first topic of debate in this book. The paper by Michael Dooley, David FolkertsLandau, and Peter Garber outlines the process for a smooth adjustment of global
imbalances, especially in China, in contrast with the alternative scenario of a sudden
and disorderly exchange and interest rate correction.
The key hypothesis at the heart of their prediction of an orderly adjustment is
that periphery governments (in particular, China) will gradually adjust their policies
before the onset of speculative capital flows. Financial policies in these countries are
seen as a component of a more general portfolio management policy with the objective
of creating an efficient domestic capital stock at the end of the process. In this paper's
view, intervention in financial markets is an important part of their development
strategy, and thus it will continue to be large and persistent enough to generate
predictable deviations of exchange rates and nominal wages from normal cyclical
patterns.
In particular, in the case of China, a gradual adjustment of real wages through
inflation and nominal exchange rate targeting will allow China to attain two main
objectives: (i) absorb 200 million rural workers into the industrial sector, which will
ease political pressures and increase those workers' productivity, and (ii) attain an
efficient capital stock at the end of the adjustment period, so that China can be
competitive in international markets without the need of distorted prices.
The challenge of absorbing these vast quantities of labor in China is to set up a
sharing of the benefits with the country importing an increasing flow of Chinese goods
(e.g., the US), as this "beggar-thy-neighbor" policy might disrupt the importing
1
country's labor market. A crucial element, therefore, is the depression of real wages in
China and therefore an increase in the returns of foreign capital directly invested there.
Over time, these returns on investment would converge to normal levels, at the same
time as the Chinese real wage increases and the pool of rural labor is successfully
transferred into the industrial sector. To this aim, financial repression and capital
controls are seen as crucial by the authors, because in order to control the real wage the
government needs to control the rate of inflation and the nominal exchange rate, and
that requires breaking the link between domestic and international interest rates.
On a different front, Giancarlo Corsetti and Panagiotis Konstantinou offer
support to the so-called intertemporal approach to the current account. This approach
has two basic predictions, which the authors confirm using US data. First, a temporary
positive shock to output should, on average, improve the current account. That is,
consumption should not increase as much as the temporary increase in output, therefore
creating a (temporary) current account surplus, which for the US seems to last around
10 years after the shock. Second, a permanent positive shock to productivity would
tend to generate a current account deficit, as consumption would tend to increase more
than with a temporary shock. A permanent shock in productivity would also tend to
raise the return on investment in the US above world levels, thus attracting capital from
abroad. In the US data, a permanent increase in productivity tends to generate an
increasing current account deficit, which starts to close gradually after one year,
reaching balance eight years after the initial shock.
Philip Lane and Gian Maria Milesi-Ferretti explore the interconnections
between financial globalization and exchange rates. In particular, they show how
exchange rate movements affect the dynamics of net foreign asset positions not only
through the traditional effect on trade flows but also through the valuation of inherited
stocks of foreign assets and liabilities. Their paper finds that the co-movement of the
exchange rate and the return on financial assets is crucial to understand valuation
effects. In theory, the impact of exchange rate depreciations depend on: (i) the level of
gross holdings of foreign assets and liabilities, in addition to their net position, as their
rates of return might respond differently to exchange rate movements; (ii) the currency
composition of foreign assets and liabilities (for instance, a depreciation raises the
domestic currency payments on foreign, dollar-denominated debt); and (iii) the
maturity (short- versus long-term) and composition mix (equity, FDI and debt) of a
country's assets and liabilities, as the sensitivity of rates of return to exchange rate
movements might be different across investment categories and maturities.
Lane and Milesi-Ferretti present empirical results pointing to exchange rate
movements as important determinants of rates of return, which gives support to the
valuation channel described before. Furthermore, the composition of the international
balance sheet is important for valuation effects, since the sensitivity of returns to
exchange rates varies across investment categories. For emerging countries, the relation
between domestic currency rates of return and movements in exchange rates is quite
strong, as would be expected since those countries are, in general, unable to borrow or
lend in domestic currency.
The idea that balance sheet effects are relevant for emerging economies is also
the main message of the paper by Juan C. Berganza and Alicia Garcia Herrero. For a
2
sample of 27 emerging economies, they find evidence that real exchange rate
depreciations increase a country's risk premium. In particular, the analysis shows that:
(i) large real exchange depreciations increase a country's risk premium, whereas a real
appreciation does not reduce it significantly; (ii) exchange rate movements affect
country risk mainly through the sudden increase in the stock of external debt and
domestic dollar-denominated debt after a real depreciation; (iii) the increase in export
competitiveness following a depreciation does not outweigh the detrimental impact of
balance sheets outlined before; and (iv) fixed exchange rate regimes appear to amplify
the negative impact of balance sheet effects on the risk premium. These results
highlight that countries with small trade openness, large financial imperfections and
pegged exchange rate regimes should worry most about large real exchange rate
depreciations. Given that these three characteristics can be influenced by economic
authorities, there is a clear role for policy action to mitigate the problem.
Emerging markets seem to have difficulties borrowing in their own currency,
which leads to the dollarization of their liabilities. The last two papers in the first block
analyze different implications of this fact for emerging economies, focusing on why
these countries cannot borrow as much as developed economies and why they
disproportionately borrow at short maturities. Enrique Mendoza and Marcelo Oviedo
explicitly introduce the effect of uncertainty on short- and long-run forward-looking
measures of sustainable public debt. This effect is crucial in the analysis of debt
sustainability for emerging economies, especially after considering three stylized facts:
(i) countries with less variation in their ratios of public revenues to GDP support, on
average, higher ratios of debt to output; (ii) countries with less variation in GDP tend to
support higher average debt to output ratios; and (iii) emerging economies display
significantly lower variation in public revenues and larger fluctuations in economic
activity than industrial countries. In their model, governments' difficulties implied by
low levels of public expenditure lead them to impose "natural debt limit" (NDL) on
themselves that is directly related to the worst-case-scenario for public finances: the
lowest possible level of public revenues and the minimum, indispensable level of
expenditure. If the government borrows above the NDL, it exposes itself to the risk of
lowering expenditures to extremely low and suboptimal levels.
Applying the model to the Mexican economy, the authors show the importance
of uncertainty, financial imperfections and the dollarization of liabilities for the
determination of sustainable debt levels. In particular, Mendoza and Oviedo find that
real exchange rate fluctuations undermine a country's fiscal position. They also find
that the long-run sustainable debt ratio is lower for (i) a higher level of a government's
"basic needs" expenditurebecause of the required slack for additional borrowing in
case of a succession of years of low tax revenues; and (ii) higher variability of adverse
shocks, as it reduces the ability to service debt in the worst scenarios.
Matthieu Bussiere, Marcel Fratzscher and Winfried Koeniger present a
theoretical model that tries to explain the disproportionately high level of liabilities in
the form of short-term debt (debt maturity mismatch) observed in most emerging
economies. In their model, a situation where liabilities are mostly denominated in
dollars and assets are denominated in domestic currency (a balance sheet currency
mismatch) might in turn create and worsen a maturity mismatch. As in the previous
3
paper, the basic mechanism builds on the effect of uncertainty. Solvency problems tend
to tilt financing toward short-term debt: a longer horizon increases uncertainty, and
therefore it lowers the amount of long-term lending that investors are willing to
provide. In turn, a larger share of short-term debt increases the variability of output, as
there is the risk that part of that short-term debt would not be rolled over during bad
times, thus affecting production.
The authors also provide empirical evidence that the effects of the variability of
the real exchange rate on the level of debt and its maturity composition and on the
variability of growth for emerging economies are broadly consistent with the model. In
particular, they find that (i) higher exchange rate variability is associated with lower
total debt and a higher share of short-term debt (maturity mismatch); (ii) larger political
risk is linked to higher short term debt; and (iii) higher maturity mismatch and higher
political and economic risk are linked to higher variability of growth rates.
Comments by Daniel Cohen focus first on the difference found by Lane and
Milesi-Ferretti between two types of countries: those with big capital inflows and
outflows (like China and Taiwan), and those with only big inflows, but no outflows
(like most of Latin America). He interprets the first case as a good signal, reflecting
countries' ongoing process of diversification of risk; whereas he acknowledges that
future research should also focus on why the second group of countries seems unable to
generate those bilateral flows. Regarding the relevance of balance sheet effects, he
agrees with Berganza and Garcia-Herrero on the importance of taking into account the
link between exchange rate depreciations and the increase in sovereign risk, although
he expresses reservations about the relatively big size of the effect found in the paper.
Closing this first block, Stanley Fischer and Malcolm Knight address the
correction of external imbalancesmainly the US external deficitand discuss the
continuation of foreign reserve accumulation in East Asian countries. They both share
an optimistic view of the correction of global imbalances, with a high likelihood of
avoiding a hard landing, although they also recognize that the needed correction will
require a fiscal consolidation in the US and might end up hindering the dynamism of
the global economy. They also agree that it is very unlikely that East Asian countries
will continue the pace of foreign reserve accumulation they have sustained until now.
Fischer, in particular, points out that the main global imbalances correspond to
the US current account deficit, about half of which can be explained by Asia in general
and China in particular. Although there seems to be an implicit strategy from some
Asian countries to build up foreign exchange reserves to avoid speculative attacks like
those at the end of the 1990s, he also acknowledges that it is hard to see Japan and
China continuing the process of reserve accumulation at the current pace. He reckons
that if the process of reserve accumulation in Asia slows down, a US current account
deficit of the order of 2 to 3.5 percent of GDP seems sustainable, and in order to
achieve those levels, the two possible channels are lower growth (through fiscal
adjustment) and exchange rate movements, especially against Asian currencies.
According to his view, this does not necessarily mean a hard landing for the US. He
also points out that a big change in exchange rates need not be disruptive, as the recent
movements of the euro/dollar exchange rate suggest. Finally, he envisions a process
whereby, in the very long run, the world will turn into regional currency blocs
4
(probably starting in Asia) with a tendency towards a global currency, where current
account deficits will not matter much, as they do not matter now for countries inside the
euro area, for example.
Knight, for his part, concentrates on the similarities and differences between the
current US external imbalance and that of the 1980s. He agrees with Fischer that the
current US external deficit now is a reflection of surpluses in developing countries, as
opposed to the 1980s, when those surpluses were mainly in developed regions. As it
happened between 1985 and 1987, the dollar has also depreciated in effective terms in
the last few months. He points out that in the 1980s, that depreciation took some time to
effect an improvement of the US current account. In his view, the experience of the
1980s suggests that the adjustment to reduce the large US current account will take
place without very large disruptions, because the US is at the center of the international
financial system, and therefore can finance its external and fiscal deficits in domestic
currency, which means that the effect of the adjustment will be spread out
internationally, and not be confined to the US. But this also suggests that without a
fiscal consolidation, closing the external imbalance will need a large exchange rate
correction. He also mentions three aspects of the current situation that are different
from the previous exchange rate correction in the 1980s, which make adjustment
harder: (i) the emergence of Asia and the difficulty it imposes on the US's need to shift
resources to an intensely competitive traded goods sector; (ii) the possibility that the
next adjustment could combine both recession and inflation, as prices of commodities
have increased because of high demand and low investment in productive capacity; and
(iii) the lack of growth in other developed areas, as opposed to the 1980s, when Japan
was a clear source of growth.
Block II: International Financial Architecture
A paper co-authored by Ralph Chami, Sunil Sharma and Ilhyock Shim examines the
role of the IMF as a coinsurance arrangement. In particular the paper addresses (i)
whether an IMF-like coinsurance arrangement is beneficial for the global financial
system, and (ii) how should a contract between a country and the IMF be structured in
order to create the right incentives. In particular, should the IMF commit to a contract
before the onset of a crisis, or should the contractual details be decided after the country
is in crisis?
The authors argue that the existence of a coinsurance arrangement between
countries increases their welfare. Furthermore, the best option is the formation of an
IMF-like centralized institution to function as a delegated monitor and to provide
liquidity for temporary balance of payment problems.
The key question is whether an ex-ante agreement on the IMF loan contracts
between the IMF and the member country designed prior to the advent of a crisis is
preferable to an ex-post contract designed once a crisis has already erupted. The paper
demonstrates that in the presence of information asymmetries and given the mandate of
the IMF to safeguard its own resources and to care about the welfare of borrowing
countries, the IMF should pre-commit to a lending contract. Such a pre-commitment
elicits the right policy effort from countries to prevent crises and recover from them.
5
This theoretical exercise also supports the idea that discretion and renegotiations once a
crisis has erupted should be constrained in IMF operations to provide countries the
right incentives for crisis prevention and crisis resolution.
Barry Eichengreen, Kenneth Kletzer and Ashoka Mody contribute to the debate
on the capacity of the IMF to boost private capital flows (acting as a catalyst) in bank
and bond markets. Their paper concentrates on two questions: (i) is it possible to find a
stronger catalytic effect of the IMF on bond issues than on bank loans?, and (ii) does a
country's level of indebtedness affect the capacity of the IMF to catalyze capital flows?
The departing hypothesis is that the IMF is more likely to foster access to bond
markets than to bank lending, since banks already perform some sort of monitoring
function, which is much weaker in bond markets. Thus, the presence of an IMF
program might redress the monitoring and coordination disadvantages of bond markets
vis-a-vis bank lending. Indeed, their empirical exercise finds support for this insight.
First, the presence of an IMF program has a stronger effect on the bond market than on
bank lending. Second, a country's level of solvency appears to have an impact on the
ability of an IMF program to increase access to bond markets. The IMF presence, rather
than its lending, appears to lower bond spreads when countries are not yet insolvent but
are under a risk of liquidity crisis. Beyond a debt ratio of 65 to 70 percent, it is the size
of IMF lending, rather than its mere presence, which seems to contribute to lower
spreads and to improve market access. Finally, they also find that precautionary
programsthose in which, in principle, the borrowing country declares its intention not
to withdraw the funds made available by the IMFare associated with lower
borrowing costs in both loan and bond markets. The results of the analysis suggest that
an IMF-supported program is likely to have a stronger catalytic effect when a sizable
share of the country's external debt takes the form of bonds, and when countries are not
yet insolvent but only face a risk of liquidity crisis. These facts should therefore be
taken into account when designing such a program.
Financial dollarization, and the associated financial fragility it introduces into
emerging economies, has become a crucial issue in the policy debate in those countries
and has shifted the stance towards a more proactive dedollarization. This is the starting
point of Eduardo Levy-Yeyati's paper on the dedollarization of multilateral credit.
In developing countries, as a consequence of the high nominal volatility and
credit risk, a large part of domestic savings moves abroad. This generates a heavy
dependence on foreign credit and specially on multilateral funding from international
financial institutions. This, in turn, becomes one of the most important sources of
dollarization in emerging economies. Levy-Yeyati discusses theoretical and practical
arguments in favor and against IFIs lending in local currency and concluded that IFIs
are natural candidates to launch investment-grade, local-currency credit markets. As
opposed to existing proposals, he argues that an initiative of this kind should and could
rely on the demand from emerging markets residents searching for local currency assets
that minimize the volatility of returns, and which are, at the same time, reluctant to take
on sovereign risk.
According to his view, the intermediation of IFIs as a first step towards the
dedollarization of developing countries' external debt has several advantages: (i) it
would voluntarily dedollarize part of a country's liabilities with no cost either for the
6
local investor or for the IFIs; (ii) it would start an international market in local CPIindexed bonds that can be used in the future by domestic borrowers as a source of
financing that is free from real exchange rate and sovereign risk;; and (iii) it could
eventually attract funds from non-residents willing to hold a speculative position in the
local currency, or in search for currency diversification, without assuming that
country's sovereign risk.
In the context of recent developments in the adoption of collective action clauses
(CACs)provisions introduced in sovereign bond contracts to facilitate the
restructuring of debt in the case of crisis, especially in those jurisdictions where they
had not been traditionally usedAndrew G. Haldane, Adrian Penalver, Victoria
Saporta and Hyun Song Shin present a paper which adds to the debate on the
determination of optimal voting thresholds for collective actions and on the
convenience of standardizing these clauses across issuing countries.
This paper analyzes the factors determining the optimal threshold and the
reasons why different issuers may choose different thresholds. Despite the assurance
that a lower threshold may provide to the debtor, there are disadvantages in following
this strategy. A lower threshold benefits debtors by increasing its ex-post payout in the
event of a crisis. However, it also means higher ex-ante interest charges. Therefore, a
lower threshold might not be preferable since it increases the likelihood of creditors
"running for the door" and, consequently, the probability of liquidity crisis. The choice
of thresholds can depend on the degree of creditor risk aversion and debtor
creditworthiness. In particular, for risk-averse creditors, it may be more convenient to
choose lower thresholds, thereby reinforcing the insurance role of CACs. On the other
hand, the more creditworthy is the debtor the lower is the tolerable CAC threshold. This
occurs because creditors considering a rollover require increasingly more compensation
as creditworthiness declines. Thus, there might be costs to encourage threshold
uniformity.
Finally, the authors point out that their model nests both liquidity runs and debt
restructurings after a solvency crisis. They argue that the tools for dealing with liquidity
and solvency crisiswhich had typically been treated separatelycannot and should
not be considered in isolation. The interaction and spillovers between them need to be
weighted carefully when designing both sets of policy. In particular, how debtors
resolve financial crises affect the likelihood of one.
Lorenzo Bini-Smaghi, commenting on the paper by Chami, Sharma and Shim,
points at time inconsistency as the main problem in IMF programs, especially in those
that involve big emerging countries. This leads to the question of enforcement of IMF
rulesmainly the limits established for access to Fund's resourcesand how to keep
them credible, even if occasionally they have to be violated. In his view, private agents
demand a more strict application of rules if IMF credibility is to be safeguarded. He
also raises the issue of the potential erosion of the preferred creditor status of the IMF.
Regarding the paper by Eichengreen, Kletzer and Mody, Bini-Smaghi highlights the
importance of IMF surveillance and the need to strengthen it both in countries under an
IMF program and in non-program scenarios. As main policy implications, he mentions
the key role of surveillance in highly indebted countries and the possibility of
implementing a non-borrowing facility as a means to provide countries with high
7
frequency monitoring without increasing IMF exposure. This non-borrowing facility

would consist of a voluntary arrangement between the IMF and a country, with similar
characteristics to those of a conventional Fund-supported program (in terms of
conditionally, standards, IMF Board involvement, etc) but without lending.
John Murray comments on the issue of international lending by posing some
important questions. Regarding Levy-Yeyati's proposal, he expresses doubts regarding
(i) the potential of measures put forward in promoting local markets in local currency,
(ii) the possibility of stripping away funds from domestic markets instead of attracting
offshore savings, and (iii) the real attractiveness of the new instruments to be issued for
residents with offshore savings. As a more general comment, he considers whether
dollarization is always such a bad thing. On the paper by Haldane et al., Murray casts
doubts on the real relevance of CAC thresholds for investors, based on the recent
experience of the use of different thresholds. As a future line of research, he also points
out that there are other features of C ACs which also need to be studied in depth.
Guillermo Ortiz, commenting on the future of the international financial
architecture, emphasizes the importance of crisis prevention and resolution. In the area
of prevention he especially highlights the usefulness of transparency and of the
documents jointly prepared by the IMF and the World Bank to strengthen countries'
economies, such as the Reports on the Observance of Standards and Codes (ROSC) and
the Financial Sector Assessment Programs (FS AP). He also regrets the passing of the
contingent credit line (CCL), a former IMF facility aimed at providing insurance to
well-performing countries, and proposes its revival. Regarding crisis resolution, he
supports the IMF's work. Furthermore, he expresses his satisfaction at the
abandonment of the statutory sovereign debt restructuring mechanism proposed by
Anne Krueger of the IMF. He is also confident of the usefulness of CACs, and cast
doubts about the practical consequences of a code of good conduct for sovereign debt
restructuring.
Finally, this book includes the addresses by Agustin Carstens and Jean-Claude
Trichet to participants of the conference. Trichet reviews the changes of the
international financial architecture over the recent years. In the context of a changing
environment he highlights four key areas: (i) the strengthening of the international
institutional set-up that has taken place since the 1990's; (ii) the significant progress
achieved in transparency and in the promotion of best practices; (iii) the improvement
of financial regulation in industrialized countries, in which important developments
have been achieved but still further work is needed; and (iv) the importance of crisis
prevention and resolution initiatives, and the need to draw lessons from the experience
of the last decade. Carstens' remarks focus on the factors that have to be understood in
order to improve the policy response to financial crises. First, he underlines the
complexities involved in the design of these responses in the midst of a crisis,
associated mainly with high levels of uncertainty and volatility. Second, he stresses the
difficult tradeoffs faced by the policymaker, not only in the economic field but also on
political grounds. The persistence of vulnerabilities after overcoming the crisis and the
resulting need for continued reforms constitute the third part of his speech. Finally,
Carstens referred to the role of the Fund in crisis resolution and the ways in which this
institution is working to reduce the frequency and severity of crises.
8
OPENING SPEECHES
This page intentionally left blank
Opening Remarks by Rodrigo de Rato y Figaredo,

Managing Director of the International Monetary Fund
The IMF at 60Evolving Challenges, Evolving Role
Introduction: Welcome and a Few Words on the IMF's Mission

Ladies and Gentlemen, on behalf of the IMF, let me welcome you all to this conference,
which has been organized jointly with the Bank of Spain. Thank you all very much for
your presence here in Madrid, where 10 years ago we celebrated the 50th anniversary of
the IMF and the World Bank.
This is of course a very timely conference from my personal point of view, and I
look forward to hearing your ideas on how the IMF can do its job better. I see the IMF's
main job as promoting financial stability and thereby improving the prospects for sustained
growth. By doing so, the IMF also helps the international community in the global war on
poverty.
An Institution with a History of Responding to Challenges
Safeguarding financial stability has always been at the core of the IMF's mandate. It is
why the institution was set up 60 years ago. But what it takes to promote financial stability
has changed markedly over these 60 years because of global developments. The IMF's
toolssurveillance, lending, and technical assistancehave been developed and
constantly adapted in response to these changes.
The breakdown of the Bretton Woods system, and the adoption of floating
exchange rates in many countries, marked a radical departure from the world of pegged
exchange rates that the IMF was set up to monitor. The IMF's member countries amended
the Articles of Agreement to give the institution a mandate to carry out a regular,
comprehensive analysis of the economic situation and policies of each member country.
This remains the essence of the Fund's surveillance function.
Around the same time, the oil shocks of the 1970s, combined with macroeconomic
instability in much of the industrialized world, created balance of payments problems of
unprecedented severity for a large proportion of the IMF's membership. It was at this time
that developing countries became significant borrowers from the IMF, and that it became
clear that many balance of payments problems were structural rather than cyclical in
nature. This led to the fashioning of new lending policies and instruments geared at helping
developing countries with these problems.
Recent Challenges to Financial Stability: Promise and Perils of Global Capital Flows
Over the last decade or so, the major challenge to promoting financial stability has come
from the growth in the size and sophistication of international capital markets. A large
number of countries have gained access to these markets. In many ways, this financial
11
globalization is a welcome development. It provides opportunities to channel large-scale

private capital flows to finance investment and growth in countries where they can be used
most productively. Capital market integration also, in principle, provides a way for
countries to adjust to external shocks with reduced reliance on official funds.
But this promise of the gains from financial globalization has yet to be fully
realized. In fact, in many emerging market countries, capital flows have themselves been a
source of volatility. This has added a new breed of shockscapital account shocks
which have proved much harder to manage than the current account imbalances with
which the IMF traditionally dealt. Arresting a reversal of capital flows requires measures
that restore investor confidence, supported in some cases by substantial financial help from
the official sector.
Financial globalization has also raised the risks of contagion. It has added new
channelsin addition to the traditional linkages through tradethrough which one
country's vulnerabilities can spread through the global economic system.
The IMF's Response: Improved Surveillance and Better Crisis Prevention and
Resolution
As in the past, the IMF has been adapting its tools to cope with these latest challenges to
global financial stability. The lessons learned from the financial crises of the last decade
are being used to improve the IMF's performance of its key task, namely, surveillance and
crisis prevention.
Surveillance remains at the heart of the IMF's work. Imbalances and vulnerabilities
must be identified and corrected before they can harm not only the country itself, but other
countries and the global system. Surveillance should tell us when economies might be
headed for trouble. It should cover the right issues for each country and the system as a
whole. The findings of this surveillance, as reflected in the conclusions of our Executive
Board, should be expressed clearly: they should signal potential problems with countries'
economic policies both to its policymakers and to markets.
Under my predecessors, Michel Camdessus and Horst Kohler (now President
Kohler), there has already been significant progress toward better crisis prevention.
Transparency has taken a quantum leap in the last decade. The Fund and its members
publish more and better information than ever before. This is promoting greater
accountability and helping markets assess risks more accurately. Intensive health checkups of financial sectors are being carried out through a new program, the Financial Sector
Assessment Program. Active monitoring of international capital markets is now a big part
of our surveillance. And the Fund has sharpened its analytical tools to assess
vulnerabilities and risks faced by countries and regions. Assessments of balance sheet risk
and debt sustainability figure more prominently in our surveillance than in the past.
But however good our surveillance, it is unrealistic to expect that crises will
disappear. Indeed, a dynamic market economy will face occasional crisesand the Fund's
role must then be to help mitigate their impact and shorten their duration through its policy
advice and financial support. This sometimes requires the commitment of substantial
amount of Fund resources. But in most cases, this investment has paid off: it has supported
strong domestic reform programs and helped to limit or avoid contagion. The IMF's loan
to Korea in December 1997$21 billionwas a very large loan by any standards. But it
12
helped restore financial stability by early 1998 and strong growth the following year. And
Korea repaid the IMF ahead of schedule. That was a case where large-scale support was
appropriate and successful. The Fund played a similar role in Mexico in 1995 and Brazil in
1998.
Of course, we do need guidelines for large-scale access to IMF resources, but these
cases also show that exceptional situations can call for exceptional steps to resolve them.
In short, over the past 60 years, the IMF has risen to the challenges faced by its
members. I do not mean to suggest that it has always been right. And the description of the
reforms already undertaken by the Fund is not meant to suggest that we are at the end of
the road. After all, ten years ago here in Madrid, Michel Camdessus spoke of
"strengthening Fund surveillance" and "adapting the Fund's financing facilities to a world
of globalized markets." So this is an ongoing process of adaptation and change, and many
challenges lie ahead.
Promoting Financial Stability: Dealing with the Challenges Ahead
In the area of surveillance, success cannot rely on early warning alone; it must also prompt
early action. There is substantial room for improvement here. IMF surveillance still
consists to some extent of confidential policy advice, coupled with peer pressure from
other countries in the international community. But peer pressure can also mean peer
protection. There is a need to sharpen the incentives for countries to take IMF surveillance
seriously.
Moreover, with increasing financial integration, surveillance must focus not just on
crisis-prone countries but increasingly on the stability of the system as a whole. Even if a
country is not itself at risk, it may be contributing to global imbalances and placing the rest
of the world at risk. The Fund, as the impartial voice of the international community, has a
particularly important role to play here in highlighting major economic challenges that the
world has to tackle. This is why, despite the controversy our advice often provokes, the
IMF calls on the United States, Europe and Japan to contribute to more balanced and
sustained global growth. This means active efforts by the United States to reduce its
deficit, and by the European Union and Japan to promote sustained growth through
structural reforms. Surveillance of the major industrial countries is critical, and multilateral
surveillance, including of global capital markets, needs to be constantly strengthened.
We may also have to adapt our surveillance further in anticipation of some of the
likely developments in the world economy in the coming decades. Current trends imply
that financial globalization will intensify. Emerging markets will represent an even larger
share of the world economy than they do today. The future emerging market giants, India
and China, may pose particular systemic challenges. And the aging of industrial country
populations may also imply higher cross-border capital flows, which will require
monitoring.
In the area of crisis resolution the challenges also are daunting. Even in cases where
we have managed to help resolve crises relatively quickly, the economic and social
disruptions have sometimes been enormous. And the crux of the problem of large-scale
IMF lendingnamely, how to provide large-scale financial support in a way that imparts
incentives for sound economic policiesis still unresolved. Abandoning large-scale
13
financial support entirely is undesirable in view of the volatile character of international

capital flows and the needs of our membership.
Some large IMF-supported programs raise concerns because they appear to suggest
that a country's geopolitical importance or other such factors play a role in IMF loan
decisions. It is important that IMF lending decisions reflect the principle of uniformity of
treatment of member countries in comparable circumstances. But the institution has an
array of financial arrangements to take into account the specific situation facing each
country. In most cases, our normal access policies leave room for the Fund to provide
adequate financial support to ease the adjustment process. And, as in the case of Korea that
I mentioned, in rare cases, exceptionally large access to Fund resources can be necessary to
guard against risks to the global financial system. While such cases draw a lot of attention,
the Fund has also given major support to countries whose situation does not pose systemic
risks or which may not rank high on the geopolitical agenda of our largest shareholders.
For example, our financial support to Uruguay has been quite substantial in relation to the
country's GDP and to its IMF quota. In short, the Fund continues to be a lender that
countries can turn to when they have balance of payments needs, and when other financing
options are drying up.
That said, we also clearly need a Fund that can say "no" selectively, perhaps more
assertively, and, above all, more predictably than has been the case in the past. The
prospect of the Fund declining to provide financial support would help strengthen the
incentives to implement sound policies, thus avoiding the need for Fund support in the first
place. To do this, we may have to think of ways of linking access to Fund resources more
explicitly to a country's policy efforts before the crisis, and perhaps to its responsiveness to
the surveillance process and its adherence to standards and codes. The Fund's proposal for
contingent credit lines took some steps in this direction but it was not found useful by the
membership. But the issues of the design of precautionary arrangements and contingent
access to Fund credit remain on the agenda.
The IMF's Role in the Global War on Poverty: Progress and Challenges
Let me turn now to our role in the global war on poverty, where of course we work closely
with the World Bank. The IMF's mandate here has been shaped partly by the expansion in
our membership over the past 60 years. In the 1950s and 1960s, newly independent
countries in Asia, the Middle East, and Africa became members of the IMF. As a result, by
1970 there had been a four-fold increase in the IMF's membership. Then, in the 1990s,
there was a further expansion to include countries of Eastern Europe and the former Soviet
Union. The concerns of this expanded membership brought the issues of structural
transformation and poverty reduction into the IMF's domain.
The UN conference in Monterrey in 2002 gave some coherence to global efforts to
meet the challenge of reducing world poverty. Under the "Monterrey Consensus,"
developing countries acknowledged that they must help themselves through good
governance and sound policies. Developed countries in turn recognized their responsibility
to provide a helping hand through increased trade and aid. The Monterrey Consensus
provides a common stageand the Millennium Development Goals a common scriptto
enable the many actors involved in the fight against poverty to play their roles better.
14
Within this broad framework, the IMF has taken steps to ensure that while focused
on its core responsibilities of macroeconomic stabilization, it is also actively helping our
members achieve their development goals. The IMF is working to ensure that our financial
support to low-income countries is based on a development and poverty reduction strategy
devised by the country's policymakers after consultation with its stakeholders. The IMF is
examining how its concessional lending window, the Poverty Reduction and Growth
Facility, can help countries get started on the right track, and stay on it, without the need
for prolonged or large-scale financial assistance from the Fund. And, working closely with
the World Bank, the IMF is helping with the global monitoring effort to assess the progress
countries are making toward achieving the MDG.
The many challenges that low-income countries face make rapid progress difficult
to achieve. But where governments have established stable macroeconomic frameworks
and pushed ahead with structural reforms we have begun to see encouraging results.
Tanzania and Uganda, for instance, have seen a sustained improvement in economic
performance. Growth rates have also picked up in other African countries that have made
progress in curbing inflation and establishing better control of the public finances.
But, as with our work on crisis prevention and resolution, we are not at the end of
road in strengthening our work in low-income countries. We are just getting started. Many
of the Fund's initiatives in this area are quite new. They need to be assessed and course
corrections made as required. Our Independent Evaluation Office's forthcoming
assessment of this new approach to helping low-income countries will be very useful in
this regard.
Coming out of these assessments, I would hope to see greater clarity in what the
role of the Fund should be in low-income countries. We need better coordination of our
work with that of other institutionsthe World Bank, the UN and its agencies, regional
development banks, and the WTOso we fulfill our core responsibilities while giving our
member countries the full range of assistance they need. This challenge of defining clearly
what we are trying to achieve when we help a member countryand ensuring that we have
the right tools to achieve itis present in our work with all countries. But it is especially
important for our work in low-income countries, where we are only one partner among
many in helping them achieve their longer-run development goals.
Conclusion
I have only just begun as Managing Director of the IMF. I am proud to lead an institution
that has strong traditions of successful international cooperation, of learning from research
and from experience, and of constantly adapting its tools to meet the needs of a changing
global environment. We cannot foresee the changes that lie ahead in the next 60 years, but
I am certain that the keys to an effective IMF response will be rigor in our analysis, evenhandedness in our treatment, and a cooperative spirit among our member countries. I am
confident these will continue to be the hallmarks of the IMF.
15
Opening Remarks by Pedro Solbes, Second Vice-President of the

Government and Minister of Economy and Finance of Spain
Introduction
Dear Governor, dear Managing Director, ladies and gentlemen, let me first welcome
you all to this seminar on the sixtieth Bretton Woods anniversary, which I have the
pleasure to introduce. I would like in the first place to thank the Spanish Central Bank
and the International Monetary Fund for their efforts in organizing the event.
As you know, the Bretton Woods Meeting in 1944 gave birth to an economic,
monetary and financial system which has, beyond doubt, played a vital role in
promoting both economic and financial stability while contributing to global welfare
and poverty alleviation. As a matter of fact, international economic and financial events
over the past years cannot be analyzed without considering the leading role of the
Bretton Woods institutions.
So first and foremost I would like to express my congratulations to the World
Bank and the International Monetary Fund. Thanks to them we have learnt important
lessons from economic globalization and crisis resolution, let alone the strong
relationship between economic growth, macroeconomic stability and poverty reduction.
I am quite sure that this seminar will provide us all with a good opportunity to
better understand the evolution and changes within the structure of the international
financial architecture. I am also convinced it will allow us to gain insight into the future
global response that should help avert the perverse effects of systemic crises in the
world economy. I will focus my intervention on three aspects that I consider especially
relevant. First I will very briefly share with you my views about the evolution of the
world economy and the Bretton Woods institutions in these last 60 years. Second, I will
briefly address the role my country has played within the Bretton Woods institutions, a
role that we intend to intensify in the future. Third, I will refer to the challenges that, in
my opinion, still lie ahead of us.
Developments of the World Economy and the Bretton Woods Institutions
When the 44 country delegations met some 60 years ago, private financial flows were
much more limited, in both scope and importance, than they are now. Then, the
financial events that occurred from the 1960s onwards and, in particular, the emergence
of the eurodollar and other offshore financial markets and the continuous globalization
of capital movements, hammered out a quite different financial system. And, by the
1990s, international capital flows had already become an essential source of finance for
both industrial and emerging market economies.
Today, financial markets are much more interconnected than they were in the
central decades of the twentieth century, and this clearly improves resource allocation
in terms of the so-called intertemporal trade. At the same time, however, it also
increases concerns about contagion effects and reversal of capital flows.
This new framework brought about various effects on the way the Bretton
Woods institutions do their job. I will mention just a few: first, a stronger relationship
between the Bretton Woods institutions and the private sector in terms of what we
know today as "private sector involvement"; second, the introduction of new financing
policies able to provide the emerging market economies with faster and more powerful
17
responses to external shocks and potential coordination failures linked with systemic
crises; and finally, the consideration of a new approach to the relationship between
economic growth and macroeconomic stability.
Role of Spain in the Bretton Woods Institutions
But let me also briefly refer to the role that my country has played in the financial
institutions that emerged from Bretton Woods.
In 1958, Spain joined the Bretton Woods institutional system and became a
member of both the International Monetary Fund and the International Bank for
Reconstruction and Development. Spain was then an isolated and developing nation,
about to start a process of liberalization of its economy and its trade. A process that,
one could say, ended in 1986 with our accession to what at that time was called the
Common Market of the European Union.
It is easy to understand that the role of Spain in the international financial
architecture has changed significantly since 1958. Such a transformation is related to
two main factors: first, the already mentioned evolution of the international financial
architecture and the role of the multilateral institutions, and second, the overhauling
process within the Spanish economic structure, coupled by a prolonged period of
growth. According to the latter, in the early 1980s Spain became a creditor country and,
since then, it has intensified its participation in the global financial system.
The Spanish quota in the World Bank and the Fund is well below its theoretical
value and we are all aware of the problems this poses. But our contribution to some
initiatives, such as the Highly Indebted Poor Countries (HIPC initiative, is quite a bit
higher. In fact, Spain is, in global terms, the seventh largest donor country under the
HIPC initiative, just behind six G-7 countries.
In the last few years it has assumed a debt relief burden which represents some
3.89 percent of the total debt relief provided by the industrial countries and our per
capita contribution ratios are even higher than those of the main G-7 countries. Besides,
Spain has undertaken bilateral debt relief actions beyond the relief provided under the
Cologne treatment within the Paris Club framework. Along the same line, the Spanish
contribution to the Poverty Reduction and Growth Facility Trust Fund amounts to 4.5
percent, fifth among the donor countries.
Spain has also provided regular funding to other World Bank Group institutions
such as the International Development Association, and to other key initiatives, such as
the Global Environment Facility and the Global Fund to Fight AIDS, Tuberculosis and
Malaria. Other important contributions have also been made to the International
Financial Corporation and the World Bank Institute, while three Consultancy Trust
Funds have been opened within the same group. Moreover, similar actions have been
carried out in other regional development banks.
Additionally, Spain provides some emerging and poor countries with technical
assistance accounts, which will contribute to promote economic growth and strengthen
the poverty reduction strategies in the destination countries. Along this line, it is worth
mentioning the aid planned to some Central American countries under a Special
International Monetary Fund Technical Assistance Account for Costa Rica, El
Salvador, Guatemala, Honduras Nicaragua and Panama.
To close my intervention, let me finally address the issues I think will be pivotal
in the near future. The first would be the need to continue to improve the work of the
international financial institutions, developing a more pre-emptive stance and
18
developing new and better targeted financial instruments. The second relates to low
income countries and how we can improve their financial conditions.
Future Challenges
In the future, the multilateral financial institutions will need to deepen the reform
process as a reaction to the new challenges posed by the size and speed of global
capital flows, especially with respect to those issues related to emerging market
economies and low income countries.
The systemic changes within the international capital markets have underscored
the role of emerging market economies as destinations for capital flows, both in the
short and in the long term. Thus, multilaterals must know to focus not only on crisis
resolution but also on enhanced prevention mechanisms. To this end it is my view that
surveillance must be strengthened, in particular in those areas related to external and
public debt sustainability. This approach should be coupled with the implementation of
stronger transparency criteria and with wider surveillance on the financial sector.
Let me also say that in my view there is a clear need for multilateral institutions
to intensify efforts in terms of both transparency and accountability. By doing so they
will be allowing the general public to better understand the vital role these institutions
play in the international financial system. A role, let me remind you, that is especially
crucial for low-income countries and emerging economies.
On the one hand, and concerning crisis prevention and resolution, I feel that the
international community must define a predictable framework for exceptional access to
IMF resources. The debate about exceptional precautionary access must also be firmly
dealt with.
In this respect I understand that, in order to prevent currency markets from
speculative attacks, an ex-ante or ex-post financial support from the IMF could make a
difference. Of course, in order to avoid potential moral hazard problems it is important
that some conditions are met. Namely, first, the country concerned should not have
debt sustainability problems and therefore should be in a good position to fulfil its
obligations with respect to the IMF. Second, a clear definition of the exceptionality
criteria is also of great importance. And third, it would also be necessary to retain some
degree of tailored level of access so the IMF continues to be able to help the countries
concerned in the design of their macroeconomic policies.
In this very same vein, I think that the debate about future reforms should take
into account the possibility of rewarding the implementation of sound economic
policies in circumstances where the threat of contagion is present. In this sense, the
discussion on a new policy framework aimed at replacing the already expired
contingent credit line and at providing correct incentives in favor of sound policies
might also bear some interesting fruits in the future.
On the other hand, crisis resolution mechanisms must be strengthened via
promotion of both collective action clauses in sovereign emissions, the use of which
should be broadened, and the Code of Conduct. Nonetheless, both tools have their
shortcomings, too, and I believe that further analysis of some other options should not
be put aside.
Undoubtedly, crisis resolution is strongly related to macroeconomic stability,
but surveillance on microeconomic and structural reforms must not be weakened. This
is particularly true given the strong relationship between flexible and competitive
internal markets on the one side, and economic growth together with better response to
shocks on the other.
19
Finally, technical assistance should be included as a key plank of the

multilateral institutions' policy framework, since good governance and institution
building-related assistance have a very positive effect on credibility. This would, in
turn, ease country access to international markets. Moreover, those reforms would help
restore market confidence and trigger new external direct investment inflows, the
benefits of which are known by all of us.
As I mentioned before, the second challenge of the future financial architecture
will be to address the financing of development in low-income countries. In this regard,
I would distinguish three main aspects:
i)
ii)
iii)
First, debt sustainability analysis should be a key issue of the

multilaterals' strategy. In this sense, the access to fresh international
credit, even under concessional conditions, should be graduated in order
to avoid feeding unsustainable progress. From this perspective, the latest
initiative by the International Monetary Fund and the World Bank aimed
at developing new instruments to set reference thresholds on a case-bycase basis is worth mentioning. When it comes to the outstanding stock of
debt, there is little doubt that its size must be adjusted when necessary
through debt relief in order to permanently eliminate the burden on
growth for these countries. On the other hand, it is important that a preemptive stance also be adopted here in order to avoid possible incentives
to excessive indebtedness stemming from implicit bailout clauses. Having
this in mind, I think that the global volume of grants by the multilateral
financial institutions should probably increase, especially for low-income
countries. This increase, of course, should always by accompanied by a
serious study of the debt position of the countries involved in order to
avoid the adverse incentives already mentioned.
Second, we should bear in mind that even if we consider the exit for the
debt burden problem of low-income countries one of our main priorities,
these countries do also, in many cases, lack solid and credible institutional
and regulatory systems. Consequently, the Bretton Woods institutions
should enhance their approach to technical assistance and keep refining
their conditionally in these areas.
Finally, the technical assistance activities of the multilaterals should
continue to converge in order to maximize progress in the area of trade
liberalization. This is a basic step towards a realistic elimination of
external financing gaps in low-income countries. And it should all come
together with an important effort in developed countries to deepen the
removal of trade barriers and subsidies, culminating the Doha Round and
allowing for renewed impulse in the fight for growth and poverty
alleviation. In this sense I would like to finish my intervention by inviting
other developed countries to follow the example of the European
Commission and show enough flexibility so that we can indeed culminate
the Doha Round in a successful manner.
20
Opening Remarks by Jaime Caruana, Governor of Banco de Espana
Mr. Managing Director of the IMF, Mr. Second Vice-President of the Government and
Minister of Economy and Finance, ladies and gentlemen, it is a pleasure for me to
welcome you all to the Banco de Espana and to the inauguration of this conference,
which is jointly organized by the International Monetary Fund and the Banco de Espana
in commemoration of the 60th anniversary of the Bretton Woods agreements. For our
country, which previously hosted the annual meetings of the Fund and the World Bank
on the occasion of the 50th anniversary of these agreements, in autumn 1994, it is a
source of satisfaction that this meeting should be held ten years later in Spain. I am also
particularly pleased to welcome our distinguished audienceboth from Spain and
abroadof representatives from official institutions, from academia and the private
sector, who are going to participate in the busy sessions scheduled over the next two
days.
The Banco de Espana and the International Monetary Fund are tied by longstanding and deep-rooted links. Not only have we worked hand-in-hand for many years
on various matters; we also share a similar mission and vocation, as well as similar
concerns in many respects.
The world has changed in the past 60 years, and the transformations that both
institutions have undergone bear witness to this. We have moved from a system of fixed
but adjustable exchange rate parities to one where floating is the norm. Economic
integration has made enormous headway during these years, both multilaterally and
through regional agreements, the European Union being perhaps the most advanced.
From a world of widespread capital controls and multiple exchange rates, we have
moved to a very different one in which enormous cross-border flows take place daily in
the form of a wide array of financial instruments, among them derivatives, which were
virtually non-existent at the end of the Second World War.
Undoubtedly, free capital movements have contributed to a significant
improvement in the allocation of global saving to more profitable uses, although they
have also entailed a swifter transmission of financial shocks across highly integrated
markets, a phenomenon behind some of the recent crises. Economic policiesfiscal,
monetary and structuralhave been increasingly geared to providing a framework of
stability and appropriate incentives for decision-making by agents. In addition to
contributing to smoothing cyclical fluctuations, these policies help set in place the right
conditions to achieve sustained economic growth based on flexible, competitive and
efficient markets where comparative advantages are tailored to the changing
circumstances of international markets.
The International Monetary Fund has adapted to these changes, seeking to
respond in its capacity as the guardian of international monetary stability to the new
problems that have arisen. It has evolved from focusing essentially on monetary stability
and on countries' external transactions and the management of their exchange rates, to
areas more related to the overall quality of macroeconomic policies and to the design of
structural policies and sound institutions. More recently, the Fund has devoted particular
attention to financial stability.
To a certain extent, central banks have also undergone changes similar to those
at the Fund. Developments in macroeconomic theory and the experience built up over
the years have forged an increasingly broad consensus on the need to apply stability21
oriented macroeconomic policies and, in particular, to conduct an independent monetary

policy with the primary objective of achieving and maintaining price stability, which is
vital for sustained economic growth. Central banks have also focused increasingly on
financial stability issues, both in countries where the central bank is the supervisor of the
banking system and in those where such functions are entrusted to a different
supervisory agency. In the case of the Banco de Espana, these transformations have also
been induced by far-reaching changes in our environment and in the role of the central
bank, including most notably the adoption of the euro as the single currency for an
ample group of European Union countries.
Spain's relationshipand that of the Banco de Espanato the International
Monetary Fund is certainly today very different from what it was in 1958, when our
country joined the Bretton Woods institutions. In the early years of its participation in
the Fund, Spain received financial and, most importantly, technical assistance from this
institution, as part of the process of external openness and liberalization that allowed its
economic take-off in that period. The Fund's excellent technical guidance in this process
supported those economic sectors in favor of liberalization, which nevertheless met
notable resistance in pushing through their program of reforms. I believe that Spain,
since then, has a considerable debt of gratitude with the International Monetary Fund.
Currently, the framework for our relationship to the Fund is very different. As
part of the euro area, monetary policy discussions take place in a broader context
spanning the whole of the euro area. As a member of the Eurosystem however, the
Banco de Espana remains deeply involved in such discussions. In other realms, relations
with the Spanish authorities remain on a bilateral footing, as is notably the case with the
close contact each year on the occasion of the analysis of the Spanish economy via the
Article IV report. That said, the European dimension is becoming increasingly
important. Beyond the single monetary policy, there are various areas of European
coordination in IMF-related matters, although, as is well known, the different European
states are represented on an individual basis in the Bretton Woods institutions.
Spain, which adhered to the New Arrangements to Borrow (NAB) since their
creation in 1998, is today an IMF creditor country. And though we continue to benefit
from the Fund's excellent technical advice, we also provide technical assistance, in some
cases in collaboration with the Fund, to other countries in the process of implementing
systems or procedures in which we have a certain comparative advantage. Moreover,
the Spanish authorities and, in particular, the Banco de Espana, cooperate increasingly
closely with the Fund on matters relating to national and international financial stability.
As I indicated earlier, this issue is becoming more and more important, and the
objectives and the interests of the Fund tend to coincide in this connection with those of
the competent national authorities.
The reasons why national and international authorities pay greater attention to
financial stability can be found in the crises a significant number of countries have
undergone in recent years. These have highlighted the complex linkages between the real
sector and the financial sector and the risk that the latter may not only be a source of
instability, but also amplify the imbalances or shocks generated in other sectors of the
economy. The links between exchange rate crises and banking crisesthe so-called twin
crisesand the feedback mechanisms between the two are, along with the propagation
and contagion channels from one country to another, matters of particular importance
here.
National regulators and supervisors, in addition to their initial aim of
safeguarding the solvency of individual institutions, have given increasing importance to
22
combining the presence of flexible, efficient financial markets and intermediaries with the
absence of excessive volatility on financial markets, and to preventing the so-called
systemic riskswhich affect the whole of the financial systemfrom materializing.
International financial institutions have also become increasingly oriented towards the
need to ensure a proper functioning of international financial markets and to limit the
transmission of shocks from one country to another in a world of ever deeper and
swifter linkages.
The relationship between national and international financial stability is obviously
all the more important for countries, such as Spain, whose banking system has a
significant presence abroad. The expansion of Spanish banks into Latin America explains
why the Banco de Espana has, in recent years, promoted the monitoring and analysis of
financial stability problems in this region in particular, and in the emerging market
economies in general, an area in which the role of the International Monetary Fund has
become increasingly important.
Key lessons may be drawn from the recent crises that have affected emerging
economiesthough not only themwhich should be borne in mind when designing
mechanisms to reduce their frequency and cost. They include, most notably: the need for
a sound and well-designed institutional framework that protects property rights,
prevents excessive public or private sector debt from accumulating and ensures
economic stability; the importance of ensuring consistency between the exchange rate
regime and the domestic economic policies and institutional framework; the risks
entailed by the vulnerability associated with exchange rate instability processes for the
balance sheets of different sectors in the economy, especially when there is a high degree
of polarization; the marked sensitivity of emerging market economies to potentially very
volatile capital flows, the direction of which can change suddenly due, at times, to
factors that may even be exogenous to the country experiencing them; the importance of
an appropriate regulatory and macro-prudential supervisory system to prevent crises
generated in the financial system and to alleviate the ensuing costs if such crises finally
occur, and to cushion the effects of shocks stemming from the real sector of the
economy; and, finally, the importance of distinguishing, to the greatest extent possible,
between liquidity and solvency crises in countries' external debt and in governments'
public debt, as a first step towards applying suitable remedies for resolving each type of
crisis.
As a result of these crises, and in collaboration with other international agencies
and with national authorities, the Fund conducted a far-reaching review of the so-called
"international financial architecture." Numerous initiatives were adopted over recent
years with the aim of lessening the frequency and cost of crises and of improving the
stability of the international monetary system. Accordingly, there has been a notable
effort to increase the transparency of member countries' economic policies and to
improve the information at the disposal of economic agents and, in particular,
international financial markets so that the latter may exert market discipline more
effectively. Specific measures have been adopted to reinforce financial systems. And it
has also been sought to improve the crisis prevention and crisis resolution mechanisms
available to the international financial community, including those particularly difficult
cases in which sovereign debt restructuring proves inevitable. Moreover, instead of
relying on exhaustive international regulations which, as experience has shown, are
difficult to approve and even harder to implement, all these reforms have relied on
voluntarily mechanisms, based frequently on the adoption of codes of best practices,
international standards and systems that offer incentives for their application.
23
Over the next two days the Conference will address these and other matters of
interest to us all, among them the risks associated with the persistent imbalances
between the current account positions of the main economic areas; the consequences of
adopting different exchange rate regimes in the current environment of free capital
flows; the problems that arise from excessive public debt levels both in the emerging and
industrialized countries; and the role of the Fund in the face of the challenges of
economic and financial globalization.
All these issues have interested authorities, the private sector and the academic
community for many years, although the way to deal with them has evolved in step with
changes in the international economy. In what is an exceedingly dynamic environment,
then, it is crucial to identify the root of the problems and possible solutions for them,
which is precisely the chief aim of conferences such as this.
We have an intense and fascinating agenda for discussion. It highlights, once
more, the need for national authorities, international financial institutions and the
academic community to work together in areas of common interest, taking advantage of
their experience and comparative advantages and with the degree of technical excellence
these matters require. I trust that these sessions will shed light on how to improve the
functioning of the global economy and, in particular, the international monetary system
24
BLOCK I
Global Imbalances,
Exchange Rate Issues,
and Debt
Session 1
Global Imbalances
A Map to the Revived Bretton Woods End Game:

Direct Investment, Rising Real Wages, and the
Absorption of Excess Labor in the Periphery
Michael P. Dooley
University of California
David Folkerts-Landau
Deutsche Bank
Peter Garber
Deutsche Bank
The consensus opinion on China's and Asia's currency policy is that it is destabilizing and
must ultimately come to a bad end.1 The "undervaluation" of currencies in order to
generate large trade surpluses, the accumulation of immense foreign exchange reserves, the
distortion of world interest rates, investment patterns, and cross-exchange rates all
evidence the instability.2 The longer it lasts the worse and more abrupt the end will be,
both for Asia and the rest of the world. Among the many ills forecast are:
Currency crises resulting in discontinuous appreciations of periphery exchange

rates vs. USD
Crisis in periphery country banking systems and insolvency of nonfmancial firms
a rapid rise beyond the normal cyclical pattern in US yields with balance sheet
problems across the world
Excessive US government and US foreign debt
Large flow problems as the periphery reallocates resources away from traded goods
and a mirror image reallocation in the center
A stepping-stone along the way to this disaster is an overheating Chinese economy
with rapid inflation. This all adds up to the intensely sour taste of a massive misdirection
of the world's savings. The only question is when the adjustment will occur: a big hit now
or a much bigger one later.
In this paper, we argue that the adjustmentwhich indeed must eventually occur
can proceed smoothly. The catastrophic losses and abrupt price breaks forecast by the
conventional wisdom of international macroeconomics arise from a model of very naive
government behavior. In that model, periphery governments stubbornly maintain a
distorted exchange rate until it is overwhelmed by speculative capital flows.3
In our view a more sensible political economy is the source of current imbalances.
The objectives are the rapid mobilization of underemployed Asian labor and the
accumulation of an efficient capital stock. The mechanism that regulates the mobilization
is a cross-border transfer to countries like the United States that are willing to restructure
their labor markets to accommodate the restructuring of labor markets in Asia. Describing
the policies that generate this transfer and their optimal settings over a finite adjustment
period is equivalent to describing the nature of the current international monetary system,
lr
The discomfort with the current situation was carefully set out several years ago (Mann, 1999; Obstfeld and
Rogoff, 2000). The logic is that although international capital markets were much larger and more resilient
than in the past they could not support a US current account deficit of 5% of GDP for long. Moreover, even
a mild withdrawal of credit from the USfor example a reduction in financing that required a return to
current account balancewould generate a very large and sudden depreciation in the real value of the dollar.
The sensitivity of real exchange rates to changes in current accounts is related to the limited integration of
goods markets across countries. A related concern then and now is that the low level of private and
government savings in the US is generating a perverse flow of world savings to the United States. Summers
(2004) has recently argued, for example, that the single engine for world recovery, US growth and US fiscal
deficits, is a recipe for disaster both for the US and the rest of the world.
2
We add the flight quotes here because we believe that invoking "undervaluation" in the context of profound
underemployment and internal imbalance is more a rhetorical than scientific usage.
3
No one loves these models more than we do; we just do not think the political economy behind speculative
attack models fits the current situation.
31
all its balance-of-payments relationships, and its evolution. We argue below that a simple
but powerful economic analysis of an exhaustible resource problem suggests that current
conditions in international markets are consistent with a stable and properly functioning
system. Our analysis suggests that the system is not headed for a crisis and will serve to
manage the great economic problem of our time, the economic emergence of China.
We Don't See Much Evidence of an Abrupt Break in the System
To some extent the warnings have proven correct. The dollar has depreciated sharply
against the euro and other floating currencies as private investors have reduced their
demand for dollar investments. But, as we have argued elsewhere, the other two legs of
the story are missing to date. US interest rates have not gone up to reflect foreign
reluctance to lend, given the current stage of the cycle, and the flow of world savings to the
US has not diminished. Moreover, economic growth in Asia has not been derailed by
domestic overheating or a breakdown in the international trading system.
The Quick Erosion as the Game Ends
The reason is no mystery; governments in Asia are providing the necessary financing. The
issue now is how long this can continue. The conventional view is that the Asian
governments can fill the gap for only a short interval and, when the wheels fall off, the
adjustment costs for the world economy will be very heavy.4 The mechanism for the
disaster is a familiar refrain. Expectations for the large exchange rate change "needed" to
"correct" current imbalances generate massive private capital flows to the periphery.
Capital controls and financial repression are no match for a determined private sector. If
inflows are not sterilized, the monetary base explodes and the "needed" real exchange rate
adjustment comes through inflation. Faced with this unpleasant reality central banks give
up and revalue nominal exchange rates.
But We Are Not Yet at Half Time
In the alternative interpretation that we will develop here, the mechanism set out above is a
good description of the final days of the original Bretton Woods system. It is relevant for
countries that are ready to graduate to the center. But it ignores the fact that the system
lasted for two decades. The erosion of the effectiveness of capital controls and domestic
financial repression follows the development of domestic financial markets, and this
process typically takes many years.
In a series of papers, we have argued that the current international monetary system
can be understood as a reemergence of the Bretton Woods system.5 In this system, the
center comprises industrial countries with integrated capital markets and floating exchange
rates. An economically important part of the periphery comprises capital account countries
(Latin America) that also allow free capital movements and allow their exchange rates to
float. The periphery also comprises an increasingly ascendant set of trade account
countries (Asia) that fix or manage undervalued exchange rates and are willing as a group
4
See Rogoff (2003). As Rogoff puts it, flying on one engine is easy as compared to landing on one wheel.
Dooley, Folkerts-Landau, and Garber (2003a, 2003b, 2004a, 2004b).
32
to generate official capital flows needed to finance both the consequent current account
surpluses and net private capital inflows. None of the rest of the world's net savings flows
into this periphery.6
The blunt policy instrument used by the trade account region is an "undervalued"
exchange rate that is maintained by capital controls, domestic financial repression and
official intervention. This policy is used during a transition until the periphery country
graduates to the center. The challenge is to explain how these policies generate current
accounts, growth rates and relative prices that resemble those observed today and then to
project these variables into the future. In particular, we are interested in the nature of
optimal policies during the transition and the nature of the endgame.
Exhaustible Resources
The economics underlying the current global imbalance is best viewed through the lens of
an exhaustible resource model. The exhaustible resource is the huge pool of Asian labor
that is underemployed by industrial country standards. Left underemployed, it is
politically dangerous and socially costly. Once employed it produces a stream of product
marginally valued at the global real wage and contributes to social and political stability.
But the faster it is employed, the greater the socially costly dislocation of labor in
other countries. So, to avoid commercial policy retaliations, other countries must be
compensated through one transfer policy or another. Put another way, a larger piece of the
new product stream must be paid to the importing country the faster is the absorption of
the unemployed pool. Since the pool is exhaustible, these transfer policies are temporary
features of the transition.
What Force Drives the Global System?
China has about 200 million unemployed or underemployed workers to bring into the
modern labor force. For political stability, there is a need for 10-12 million net new jobs
per year in the urban centers. A growth rate of around 8+% has served to employ about 10
million new workers each year. About 3 million have been in the export sector.
If the world can absorb politically only the output of an additional 10 million
workers per year (3 million in the export sector), then simple arithmetic indicates that this
surplus is a force for twenty years more in the global system. If it can absorb the surplus
faster, say at a rising absolute rate that will keep the Chinese growth rate constant at 8%
until the surplus is eliminated, then straightforward compounding and linearity
assumptions indicate that this will drive the global system ever more relentlessly for the
next 12 years (see Chart 1).
6
This policy has been criticized as wrongheaded in that FDI should be the source of global finance for a
deficit on current account. The principle behind this argument seems to be that the external accounts should
be properly balanced as a priority over the internal balance. See Goldstein and Lardy (2003). The alternative
argument is that being a net capital exporter seems to work.
7
Exports generate 10% of value added in GDP. The export sector grows twice as fast as the rest of the
economy. So 25% of all growth is from the export sector. Because of a lower capital-labor ratio than in the
rest of the economy, the export sector accounts for about 30% of employment growth.
33
Note that if the growth rate of 8% is maintained, then a growing pressure will be
put on the rest of the world's labor market, peaking out at the end of the process. We
argue below, however, that the optimal adjustment path front loads the absorption of labor.
We do not take a stand on how long this force will drive the global system. But
twelve to twenty years has defined an era for any recent international monetary system.
The Political Economy Tradeoffs
In trying to understand Asia today, it is not sensible to argue that the governments' policy
is to maintain mindlessly undervalued exchange rates indefinitely in a beggar-thy-neighbor
policy. The better approach is to look behind the current exchange rate policy to see what
the government is trying to accomplish.8 Our interpretation is that a sensible desire to
absorb unproductive labor is at the root of the array of policy choices.
We are really referring to China's development strategy. The exchange rates of other Asian countries fly in
formation with the CNY to maintain relative competitiveness.
34
We assume the government has two objectives:
It wants to move workers from an unproductive pool to produce positive

marginal product in the industrial sector. The benefits are both economic
and political. The real wages of workers newly absorbed into the industrial
sector rise, although at the cost of keeping the overall wage of existing
productive workers low. Moreover, the pool of potentially disruptive
excess labor is reduced. The larger the size of the remaining unproductive
labor pool, the greater the political cost to the government. The government
wants to reduce this pool quickly, but faces increasing foreign economic
and political costs that increase with the speed of employment from the
pool.
The government wants the capital stock accumulated as the pool of labor is
absorbed into the industrial sector to be efficient: at the end of the transition
period, the capital stock should be capable, when combined with domestic
labor paid the world real wage, of producing goods going forward that are
competitive with those produced in other countries.
This last objective is a crucial constraint', not just any junk make-work project will
do because the history of development has shown repeatedly that this is the way to endgame crisis. Our guess is that this objective reduces the chances that a sudden change in
relative prices, notably exchange rates and asset prices, at the end of the transition period is
likely to be an outcome.
Optimal Absorption Rates for the Labor Pool: An Exhaustible Resource Problem
The political economy set out above suggests that the challenge is to set up a sharing of
benefits with the receiving country (the destination for this surplus labor product) so that it
will accept the political costs of the rapid restructuring of its own labor market to
accommodate the increased flow of imports.9 We will argue that the structure of this
fundamental problem in international finance is also remarkably analogous to an
exhaustible resource pricing problem.
In the current global system, benefits are shared with importing countries by
initially giving foreign capital access to Asian labor at a low domestic real wage relative to
the world real wage. This gives the capitalist excess profits for some time period and
provides the resources for the capitalist to utilize to keep home country import markets
open. The trick is to set the real wage (real exchange rate) low enough and to adjust it
gradually upward to the expected real wage in the rest of the world until the excess labor
pool is exhausted, all at a minimum cost.
We show in the appendix that the optimal path for the real wage is something like
AB or CD in diagram 1. The important feature of this adjustment path is that domestic
Consumers of cheap imports in the center benefit from rapid import penetration but we observe that
although their willingness to compensate losers is important, it has often failed to outweigh the power of
import competing industries in what could be characterized as a beggar-thy-neighbor policy. Competitive
devaluation or commercial policy has, therefore, often been imposed. A contribution from the periphery to
capital in the center is a possible additional sweetener to avoid this impasse.
35
labor (the real exchange rate) is initially undervalued but converges to world levels at the
end of the adjustment period. This means that the real exchange rate is initially distorted
but converges to a sustainable level. In evaluating the stability of the system, this feature of
the optimal adjustment path is all important.
Diagram 1
Reel wages and Adjustment
A country with a very large stock of labor to employ will want to set a real
exchange rate that appears to be grossly undervalued by conventional measures. But the
undervaluation fades away over time and with it the need for controls over capital flows
and other policies needed to insure internal balance discussed below.
We can summarize this section as follows. The optimal exchange rate and inflation
policy are derived from the exhaustible resource problem. For a fixed exchange rate
regime only one initial real exchange rate is optimal and only one rate of inflation
generates the optimal path for the real wage over time. At the end of the adjustment period
the following conditions hold:
The domestic real wage equals the world real wage in the manufacturing
sector.
The initial pool of surplus labor is employed.
The capital stock has increased to match the world capital/labor ratio in
manufacturing.
The political costs of adjusting displaced labor and capital in the importing
country have been compensated for their costs of adjustment. This co-opts
attempts to use commercial policy to freeze out the exports that are vital to
the development policy.
36
An Indeterminacy: Adjust Nominal Wages or Nominal Exchange Rates?

The optimal adjustment path for the real wage allows the central bank to choose a path for
the nominal wage rate or the nominal exchange rate but not both independently. In fact
Asian central banks use both techniques. For a fixed exchange rate regime the central bank
manages the inflation rate in order to regulate the dollar value of domestic wages and
prices. In this case we would expect wage inflation to be above that in the center so that
domestic real wages rise over time. The alternative would be to set domestic wage and
price inflation at or below that in the center and then allow the nominal exchange rate to
appreciate over time but at a controlled rate.
As long as private market participants understand that policy is driven by the
objectives set out abovethe optimal path for the real wage ratethe same pattern of real
private capital flows and trade account will be generated by either a fixed or managed
float exchange rate arrangement. From the balance-of-payments accounting identity, it
follows that the path of real and nominal official intervention is invariant to whether a
fixed rate or managed float regime is chosen. Those who argue the necessity of switching
to a managed appreciation because of the large accumulation of official reserves are
missing the basic policy problem and its resolution. Moreover, switching from fixed to
managed floating, perhaps in the face of political pressure from the center, would not alter
the real nature of the transition.
The Key Role of Financial Repression
A key to this regime is the ability of the government to repress real wages for an extended
period of time. In our framework, this is equivalent to controlling the rate of inflation and
the nominal exchange rate. Given a foreign rate of inflation and an international interest
rate, this clearly requires that the link between domestic and international interest rates be
broken. In our view China has more than adequate controls on domestic and international
financial transactions to make this possible.
<
Purchases of international bonds are strictly controlled. State owned or

controlled banks provide all the claims available for domestic savers.
The government sets the interest rate on these bank liabilities and rations
bank credit to the private sector growth in the foreign part of the monetary
base is determined by the current account surplus plus targeted net direct
investment inflows.
In this repressed domestic financial system, growth in domestic credit from

the banking system is a residual, that is, the difference between desired
money base growth, (determined by the desired rate of inflation), the
growth in the demand for money and the growth in the foreign part of the
base.
Domestic savings not purchased by the banking system are absorbed by sales of
domestic treasury or central bank securities to households and firms. Note that as long as
37
the real interest rate that clears this market is not above the return on US treasury securities
or other forms of investing the fx reserves the government can absorb domestic savings
and intermediate into foreign bonds at a profit.
The government rations credit to the private sector by forcing the banks to buy
government securities through liquidity and reserve requirements and then rations the
remaining credit to the private sector at fixed lending rates. This of course sets up strong
incentives for private lenders and borrowers to go offshore or to alternative domestic
intermediaries. We assume that the government is an effective counterforce to such
financial innovation for the requisite amount of time.
Internal Balance
The macro management problem for the government in implementing this policy is
daunting but simple enough to set out. In pursing the employment objective, a distorted
real exchange rate will create imbalances in the economy that require an additional policy
instrument. As noted above, the bottom line is that the government must be able to
manage the domestic real interest rate throughout the adjustment period to keep the
domestic economy in balance. The good news is that the problems are large but diminish
overtime.
To make this argument, assume the economy, aside from the 200 million, is in full
employment equilibrium with effective capital controls, no initial net international
investment position, and an exchange rate that balances trade. To set the problem in
motion, now imagine that 200 million unemployed people appear from the provinces. As
discussed above, the path for the real exchange rate that solves the absorption problem
involves a sudden real depreciation that is gradually eliminated. The exchange rate path
that solves the absorption problem therefore subsidizes exports relative to imports and the
trade balance initially moves from balance to surplus.10
The initial current account surplus must equal the amount by which domestic
(government plus private) savings exceeds domestic absorption. It follows that a rise in
the domestic interest rate is needed to reduce absorption relative to savings. But what
happens to the interest rate that insures internal balance over time?
During the adjustment period the trade surplus as a share of GDP will decline and
may move into deficit as the real exchange rate appreciates and domestic income grows
more rapidly than foreign income. A surplus on the service account will appear and grow
as net asset accumulation generates net capital income. But the overall current account as
a percent of domestic GDP will fall for any reasonable set of parameters. It follows that
the domestic interest rate will fall over time as a smaller share of domestic absorption is
crowded out by net transfers abroad. This mitigates the interest differential pressure on
capital controls.
10
An important mitigating factor is that adjustments in commercial policy are likely to encourage imports.
For example, the initial condition for China is a large gap between the effective exchange rate for imports
and exports. In fact, China has not run a large overall trade surplus to date. In part, this probably reflects
large declines in tariffs associated with ascension to the WTO.
38
Sterilization and Inflation

The relevant capital flow "problem" in the face of expected revaluation is large private
capital inflows. Clearly if private capital inflows augment the monetary base and in turn
increase domestic inflation, real wage growth will be too rapid and the transition will be
too short to accomplish the government's objectives. However, if capital inflows are
sterilized, and if domestic financial repression allows the government to finance reserve
creation by issuing low interest domestic securities, the inflationary impact is eliminated.
This is an empirical issue. Capital controls and financial repression do not last
forever but neither does the regime we are describing. We simply observe that to date the
Chinese government has been very successful in hitting an aggressive inflation target.
Some observers have suggested that overheating and an inflationary spiral are already
underway. In our view, that is more of a prediction than an observation. Time will tell,
but we would point out that there are many reasons why inflation may have increased in
recent months. In general, a growth rate of 8+% has not generated inflation in China. In
our view increases in reserve requirements last year, a form of sterilization, have already
reduced the growth in money and credit. Moreover this has been accomplished with no
increase in administered interest rates.
If the capital account is liberalized, expectations of appreciation that are a central

feature of the regime discussed below will generate capital inflows. Moreover, marketdetermined domestic interest rates would make sterilization expensive and so inflation
would be the eventual result. But we do not expect opening of the capital account or
deregulation of domestic interest rates. It follows that the economic linkages between
exchange rate policy and inflation clearly relevant for capital account countries do not now
exist, and we do not expect them to materialize for many years.
39
The Transfer to Foreign Capital

The regime set out so far encourages capital formation in export industries and makes
room for this new investment in the domestic market. But it does not suggest that
nonresident direct investors are the best placed to do the investing. Recall however that the
investor has to expect that the foreign markets for exports remain open and that the
political costs of displaced workers in the importing countries must be compensated.
A transparent but unrealistic example will help make the point. Suppose the right
to supply capital is allocated by the government through licenses on a project-by-project
basis. The gap between the domestic and world real wage would then be captured by
selected capitalists.11 Moreover, the government could lend through domestic balance
sheets to the direct investor and finance this by sales of securities to the domestic market.
The government can reduce the political costs to foreign governments associated with
rapid export growth by allocating some of this capital to foreign investors from countries
that allow the rapid growth of imports. In the present context, with the US absorbing much
of the exports, this allocation would go to US firms. This provides an economic rent until
the convergence of real wages at T, which is not competed away because entry into foreign
direct investment is rationed by the Chinese government.
The foreign investors then become a well-financed and effective lobby to
counteract the resistance to the restructuring of the US labor force away from import
substitutes.12 Each time a worker is matched with foreign capital, the direct investor gets a
benefit equal to the discounted value of the wage differential plus the normal return to
capital. The excess returns are implicitly paid by the Chinese workers accepting the low
but rising real wage. Indeed, from the US balance sheet perspective, there is no real export
of capital from the US to China. All is financed by forced Chinese savings, both the US
current account deficit and the onshore loans to the foreign investor. The US balance sheet
taken as a whole simply intermediates between low yielding Chinese deposits and high
yielding FDI investments.
But perhaps this method of local intermediation is too transparent and difficult
politically. Instead, the government could sell the same domestic security mentioned
above but, rather than make a loan to a direct investor, purchase international reserves in
the direct investor's home credit market. This acquisition of foreign assets favors the
importing country in general rather than just the foreign investor. The foreign investor
then has to borrow in the importing country at his own normal cost of funds, and then buy
yuan to make it investment. Part of the subsidy to the foreigner is then given to the
importing country as a whole, and part to the FDI investor in the form of rents from access
to low real wage labor. Again, no real capital flows from the US to Chinaboth the US
current account deficit and the measured FDI outflow are financed by Chinese savings.
Whether it is booked as FDI or investment managed by foreigners is irrelevant.
More precisely, shared between the investor and the government and perhaps government officials. See
Razin and Sadka (2002) for an interesting discussion of the allocation of rents.
12
We refer to "foreign investors" and not "foreign direct investors" because in this example they are financed
by Chinese saving intermediated through domestic balance sheets.
40
Politically, this is perhaps better because there is an arms length relationship

between the government and the foreign investor. With this more competitive mechanism
we would expect that the surplus generated by access to low wages in China would be
absorbed by adjustment costs. In this case direct investors from countries with open
import markets might enjoy a competitive advantage over other foreign and domestic
investors because they can more effectively mobilize profits to make transfer payments to
their fellow residents.
At this point we do not understand well the mechanism that allocates investment in
the export sector, its profitability or the distribution of those profits. It is also quite
possible that direct investment is restricted and/or the risk that the regime might end
prematurely requires excess profits in order to insure entry. The net profitability of direct
investment is an important ingredient in the evolution of net international investments
positions during the transition. Data on profitability of direct investment in China is
anecdotal at best. We can make a reasonable guess about the gap between the real wage
and marginal product of labor, but we do not have much information about the distribution
of the implied surplus. This is an important topic for further research.
What about the Accumulating Balance Sheet Positions?
Headline numbers for reserve accumulation and the US current account deficits seem to
suggest that the main end game problem is the accumulated net international investment
position of the center and the periphery. But net positions are the difference between two
much larger gross assets and liabilities. Just as in the original Bretton Woods System,
official intervention, that is, large official capital outflows from the periphery are largely
associated with private capital inflows to the periphery. In our view the financial
intermediation and the capital gains and losses generated will substantially mitigate
problems associated with the net international investment positions generated by export led
growth.
At the end of the transition period the government of China will hold a large stock
of US treasury and other securities on which it has earned a relatively low but positive rate
of return. It will also have incurred a large stock of liabilities to domestic claimants. But at
the end of the game, both of these will carry the same international interest rate. The US
will hold a large stock of direct investment which pays the world equity rate going forward
but which has paid a much higher rate during the adjustment interval.
It may be instructive to take another look at the end of the original Bretton Woods
system with these two points in mind. While a careful historical comparison is beyond our
resources at the moment, it is clear that the United States did not run large trade deficits
leading up to the 1971-73 crisis that ended the regime. The "balance of payments deficit"
that observers focused on at the time was the liquidity balance, a concept that put short
term capital inflows below the line. As Depres, Kindleberger and Salant (1966) pointed
out in their celebrated letter to the Economist, this concept of a deficit ignores the
legitimate role of financial intermediation in international financial arrangements. To be
sure, financial intermediation can lead to instability and crises. But the problem is much
more subtle and the "lessons" from countries that have run large and persistent current
account deficits may not be of much use in evaluating the new Bretton Woods.
41
Conclusions
What makes this perpetual motion machine run is, of course, the assumed zero (actually
negative) product of the pool of excess labor that we are implicitly associating with the
outcome of a market-determined real exchange rate and allocation of domestic and
international savings. This provides a free lunch that everyone can share through current
Asian policies.
We have done some simulations with plausible rates of accumulation and returns
and find that the transition to the new steady state need not imply a large continuing net
transfer. So the system ends with a smooth adjustment. The government of China will
have a more productive capital stock and will have managed to employ 200 million people
in world-level wage jobs. The US will own a nice chunk of the Chinese capital stock, and
will have made a fine excess return during its accumulation. There are even mutually
offsetting cross-border claims against each other that can serve as escrow against
confiscation.
During the adjustment period, many dimensions of this development program are
distorted in the periphery. But one thing that is not distorted is the knowledge that at the
end of the transition capital invested in traded-goods industries will have to compete on an
equal basis with capital invested in other countries. We see no practical alternative to
imposing this discipline on an emerging market and at the same time accelerating the
absorption of a large and politically dangerous pool of labor. The feasibility of
maintaining an undervalued exchange rate through monetary policy and controls on
domestic and international capital markets for a long time can, of course, be questioned.
But this is an empirical question. At the moment we do not see a mechanism in the case of
China for significant circumvention of their financial arrangements and regulations.
42
References
Despres, Emile, Charles P. Kindleberger and Walter S. Salant (1966) "The Dollar and
World LiquidityA Minority View", The Economist 5th February; reprinted in
Kindleberger, Charles Poor (1981 ed.) International Money. A Collection of
Essays, London, George Allen and Unwin: pp. 42-52.
Dooley, Michael, David Folkerts-Landau, Peter Garber, "Dollars and Deficits: Where Do
We Go From Here?", (June 18, 2003 a), "An Essay on the Revived Bretton Woods
System", September 2003b, "The Cosmic Risk: An Essay on Global Imbalances
and Treasuries", (February, 2004a), "Asian Reserve Diversification: Does it
Threaten the Pegs?", (February, 2004b), Deutsche Bank, Global Markets Research.
Goldstein, Morris and Nicholas Lardy (2003), "Two-Stage Currency Reform For China"
Asian Wall Street Journal, September 12.
http ://www.iie.com/publications/papers/goldstein0903 .htm
Hotelling, Harold, (1931) "The Economics of Exhaustible Resources," Journal of Political
Economy 39, 137-175.
Mann, Cathrine (1999). Is the U.S. Trade Deficit Sustainable, Washington, DC : Institute
for International Economics,
McKinnon, Ronald and Gunther Schnable, (2003 a), "A Return to Exchange Rate Stability
in East Asia? Mitigating Conflicted Vitue," mimeo, October.
9 (2003b) "The East Asian Dollar Standard, Fear of Floating and Original Sin,"
mimeo, September.
Nurkse, Ragnar (1945) "Conditions of Monetary Equilibrium," Princeton Essays in
International Finance, Spring.
Razin, Assaf and Efraim Sadka, (2002), "Gains from FDI Inflows with Incomplete
Information," NBER Working Paper No. w9008, June.
RogofF, Kenneth (2003), "WorldEconomic Outlook Press Conference, September 18,
2003" in http://www.imf.org/external/np/tr/2003/tr030918.htm
Smarzynsk, Beata and Shang-Jin Wei (2000), "Corruption and Composition of Foreign
Direct Investment: Firm-Level Evidence," NBER WP w7969, October.
Summers, Lawrence (2004), "The United States and the Global Adjustment Process,"
Third Annual Stavros S. Niarchos Lecture , Institute for International Economics,
March, http://www.iie.com/publications/papers/summers0304.htm
43
Appendix: The Logic of the Dynamic Path for Real Wages

The optimal strategy for the government is to set the initial wage and the rate of change in
the wage in order to fully employ the stock of labor at a minimum cost.
Consider first the rate of change for the real wage. An additional unit of labor
employed provides a nonnegative yield to the government b. A unit of unemployed labor
costs government a yield of -r. The yield b can be thought of as tax revenue or political
support for the government. The yield -r might be transfers to the unemployed or political
opposition.
The "price" the government expects to pay the foreign investor to encourage the
global absorption of this labor must fall.
Diagram 1
Real wages and Adjustment
To see this, suppose instead that the government kept the wage constant for two
consecutive time periods. A constant wage generates a constant excess rate of return,
which is sufficient to overcome the resistance of the labor restructured in the importing
country. It therefore generates a constant flow of new employment. If the wage in the first
period was set slightly lower than in the second period, for the same average wage, less
unemployed labor will be carried over into the second period. The carryover is costly so a
constant wage cannot be optimal. The government can get the same increase in
employment at a lower cost by frontloading the adjustment.
The incentive with which the government sweetens the provision of labor to
investors is the difference between the domestic real wage and the world real wage. This
difference must fall, that is, the market wage must be expected to rise.
44
There are two reasons for the flow demand by the foreign direct investor to be
decreasing in the wage rate. First, we make the usual assumption that investment
installation costs rise in the rate of investment over time, the usual bottleneck argument. It
follows that a rapid adjustment requires a greater cost of capital per worker. Second,
investors have to make transfers to offset the political power of displaced workers in the
importing country. Again, it seems likely that the adjustment costs in the country
restructuring its labor market are increasing in the rate of import penetration.
The optimality of the rising price path is the central insight of the exhaustible
resource model. Diagram la is a stylized representation of the problem of absorbing an
exhaustible excess labor pool. Paths AB and CD satisfy this rate of change condition.
Path AB starts from wi, a relatively high initial real wage, path CD begins with W2 and
rises at the same rate. The full solution to the Hotelling (1931) problem requires that the
government set the initial wage so that the initial stock of labor is employed when the
domestic wage rises to the world wage. Clearly a lower initial real wage path CD
generates more total employment over the interval from to to T2 as compared to path AB
from to to TI. It follows that the integral of employment increases as the initial wage
declines and only one initial wage fully employs the initial labor supply. It also follows
that a country with a very large stock of labor to employ will want to set a real exchange
rate that appears to be grossly undervalued by conventional measures.
45
Macroeconomic Dynamics and the Accumulation of

Net Foreign Liabilities in the US: An Empirical
Model*
Giancarlo Corsetti*
European University Institute, University of Rome III, and CEPR
Panagiotis Th. Konstantinou*
University of Rome III
* Paper prepared for the special volume of the conference "Dollars, Debt and Deficits: 60 Years After the
Bretton Woods", co-organized by the IMF and the Banco de Espana. For useful comments we would like
to thank Charles Engel, Soren Johansen, our discussant Jaume Ventura, as well as conference
participants in the aforementioned Conference. This paper is part of the research network on 'The
Analysis of International Capital Markets: Understanding Europe's Role in the Global Economy,9 funded
by the European Commission under the Research Training Network Programme (Contract No. HPRNCT-1999-00067).
t
Correspondence to: European University Institute, Department of Economics, Villa San Paolo, Via
della
Piazzuola 43,1-50133 Florence, Italy. Email: giancarlo.corsetti@iue.it. Tel: +39-055-4685760, Fax: +390554685770.
* Correspondence to: University of Rome HI, Department of Economics, Via Ostiense 139, 1-00154
(RM) Rome, Italy. Email: konstant@uniroma3.it. Tel: +39-06-57374258, Fax: +39-06-57374093.
Introduction
The size of the external deficits run by the US through the second half of the 1990s and
the beginning of the new century has raised controversial issues in the current policy
debate - ranging from sustainability of US current account imbalances, to the main
macroeconomic and international implications of the required adjustment in the next
few years, especially as regards the rate of dollar depreciation in nominal and real terms
compatible with global equilibrium. Macroeconomic and policy exercises are quite
complex, as their results are sensitive to assumptions regarding basic macroeconomic
parameters (e.g., expectations about growth rates and real interest rates), policy
interactions among sovereign domestic policy makers (e.g., explicit or implicit
currency target pursued by central banks, fiscal-monetary policy mix in different
regions of the world), the structure of asset markets (including the possibility of sudden
changes in the desired portfolio composition by international investors, perhaps
reflecting coordination problems in financial markets) and so on. But the current policy
debate has also motivated a thorough reconsideration of economic theories of the
current account and external solvency, as well as of the empirical evidence on the
determinants and dynamic behavior of foreign liabilities.
In this paper, we contribute to the current debate by providing novel empirical
evidence on the dynamic evolution of the stock of US net foreign liabilities. Building
on our previous work (Corsetti and Konstantinou [2004]), we resort to a small set of
assumptions to derive an empirical model that synthesizes the joint dynamics of net
debt and key macroeconomic variables such as consumption, output and investment.
Using the restrictions implied by the external solvency constraint of a country, we
identify empirically transitory and permanent shocks to the variables in our system, and
study the dynamic response to these shocks (on the methodology, see Campbell and
Mankiw [ 1989] and Lettau and Ludvigson [2001, 2004]).l
In the first part of the text, we will briefly summarize the results from the three
variables system developed in our previous work. An important result is that our
empirical model for the US appears to lend support to the basic theoretical propositions
of the theory of the current account. Namely, shocks that raise output and consumption
temporarily also raise the stock of net foreign assets - a pattern implied by the
consumption smoothing hypothesis, but at odds with procyclical views of current
account imbalances. Conversely, shocks that raise output and consumption permanently
- which can be naturally interpreted as a permanent technology shock -are associated
with the build up of foreign liabilities.
Our discussant during the Conference preceding this volume pointed out the
possibility that the behavior of the US macroeconomy in the second half of the 1990s
be affected by an asset market bubble. If this is the case, it may be possible that part of
the US current account deficit has been recently financed with the sale of ultimately
worthless paper to the rest of the world. This observation is relevant for our purposes
insofar as our empirical results will be sensitive to the inclusion of the 'bubble years' in
Relative to the literature on the current account, our analysis of the net foreign position dynamics
requires a minimal set of equilibrium restrictions. In the three-variable model developed in our previous
contribution, the main assumption is that the external solvency constraint holds, complemented by a
stationary distribution for the portfolio share of foreign wealth (see also Kraay and Ventura [2000, 2002]
and Ventura [2003]). In our four-variable model, two additional conditions follow from balanced growth:
the "great ratios" consumption to output and investment to output should be stationary - see King et al.
[1991] among others.
49
the empirical analysis. We therefore run our model on a subsample that exclude the
years from 1998 on, showing that our main conclusions are substantially unaffected.
In the second part of the text, we will develop a four variables system, studying
the joint dynamics of output net of government spending, consumption, investment and
net foreign liabilities. Compared to the three-variable model studied in Corsetti and
Konstantinou [2004], a four-variable system provides a more general empirical model
of the dynamic behavior of foreign liabilities. However, the cost of generality is that it
becomes harder to find an approximate expression for the capital account, which now
cannot be identified separately due to the balanced growth assumptions. While the
stochastic structure of the system changes, we can still identify one permanent shock
(which can be interpreted as a permanent productivity shock), whereas we identify two
permanent shocks in our earlier work.
Our findings in the four-variable model can be summarized as follows. First, we
find strong evidence that three long-run relations exist for the variables in our system,
two of which correspond to stationary 'great ratios' - consumption to output and
investment to output ratios - while a third one is a relation between the logs of net
foreign liabilities and output.
Second, by adopting Generalized Impulse Responses (GIR), we find that
positive shocks that increase consumption and investment also worsen the US net
foreign asset position. On the other hand, a shock that raises output leads to some
improvement in the US net foreign position - although such improvement is not
significant. Interestingly, we find that an exogenous shock to the stock of foreign
liabilities has a marked transitory effect on the level of this variable, which accounts for
a large share of its variance over short- and medium-term horizons.
Third, as in our previous work, we identify a permanent shock as an innovation
with permanent effects on the variables in our system. The dynamic effects of this
shock are similar to the ones studied in our three variable model: a shock with positive
permanent effects on per capita output and per capita consumption also raises net
foreign liabilities. In addition, consistent with the interpretation of the shock as a
permanent improvement in productivity, it is associated with higher per capita
investment. We find that this permanent shock dominates the variation of net foreign
liabilities at horizons of six years and more, whereas transitory shocks dominate its
variation at shorter horizons.
The rest of the paper is organized as follows. Section 2 reconsiders the main
findings of our prior work using a different sample. Section 3 provides a further
theoretical motivation for our work. Section 4 lays out our empirical methodology.
Sections 5 and 6 contain the empirical results from the four variables model. The last
section concludes.
The Response of US Net Foreign Wealth to Temporary and Permanent Shocks: A
Review of Our Main Findings
In this section we briefly review the main features and results of the empirical
methodology developed in Corsetti and Konstantinou [2004] and applied to the US, as
an introduction to the extension of the same analysis presented below. The text will be
developed in a non-technical way, leaving details to Corsetti and Konstantinou [2004]
as well as to the next section, where a more general model is presented.
Many analysts argue that the second half of the 1990s is characterized by an
asset pricing bubble driving the boom of assets markets in the US and elsewhere. A
bubble could affects the intertemporal budget constraint - as the US deficit could at
50
least in part be financed by issuing ultimately worthless assets. This consideration

raises the important issue of whether our results are sensitive to the inclusion of the last
part of our sample in the analysis. To address this issue, in summarizing our main
results from previous work we will refer also to an application of our methodology to a
sample truncated in 1997, i.e., 1963:Q1-1997:Q4. Variable definitions and a discussion
of the data are provided in appendix.
Implications of Budget Constraint
In our analysis, as in Campbell and Mankiw [1989] and Bergin and Sheffrin [2000], we
derive an approximate expression for the budget constraint of a country by taking a
first-order Taylor approximation of the intertemporal budget constraint, imposing the
appropriate transversality conditions and taking expectations. In doing so, we assume
that the portfolio share of foreign wealth in domestic private wealth is stationary - so
that the expected value of this share exist. Domestic private wealth is defined as the
present discounted value of net output Zr, which is GDP net of government spending
and investment. Let Dt denote the stock of net foreign liabilities (the negative of net
foreign wealth), while C, denotes private consumption. Referring to our previous work
for detail, our procedure allows us to obtain an approximate expression for the capital
account in the form:
0)
where lower-case variables denote the log of the corresponding upper-case variables,2
and q>2 is a function of the expected value of the foreign wealth to domestic private
wealth ratio.
The above expression defines the variable to be used in our empirical analysis.
Specifically, Ct is real per-capita expenditure on nondurables and services.3 Net output,
Z,, is gross domestic product net of government expenditure, investment and
expenditure on durables, expressed in real, per-capita terms. The stock of net foreign
liabilities, A, is also expressed in real, per-capita terms. We stress here that A records
more than bonds - as it includes the whole array of assets and liabilities traded
internationally. The variable Dt is derived by cumulating the current account deficits
over the sample period. Data limitations do not allow us to use series of net foreign
liabilities allowing for capital gains and losses on a wide array of assets - as proposed
by Lane and Milesi-Ferretti [2001]. Namely, the series built in this study is at a lower
frequency (annual), and for a smaller sample than the one we adopt in our work. Yet we
should note here that our series and the Lane & Milesi-Ferretti series are quite
correlated.4
Now, it can be shown that, under the weak maintained hypothesis that the real
rate of return rt, the rate of growth of consumption and net output are covariance
stationary, the budget constraint implies that the logs of consumption, net output and
2
Throughout this paper we use lower case letters to denote log variables (e.g., ct = ln(C,), xt = ln(A!i) etc).
Since we are mainly interested in the net foreign liabilities dynamics generated by consumption, we
prefer to exclude expenditure on durables expenditure - as they replace (or add to) capital stock rather
than buying a service flow from the existing capital stock. We include durable expenditure in private
investment.
ê correlation between our measure of net foreign liabilities and that reported in Lane and MilesiFerretti [2001] (based on adjusted cumulative current account) is 0.987. See also the discussion in
Corsetti and Konstantinou [2004].
3
51
net foreign liabilities must be cointegrated. Even if the level of net foreign wealth is
non-stationary in levels (as predicted by standard infinite-horizon intertemporal model),
the transversality condition prevents it to wander away from net output and
consumption. Hence, in line with the literature, KAt is stationary.
As mentioned above, when we log-linearize the intertemporal budget constraint,
we assume that <pz is constant. We note here that this assumption is consistent with
recent work by Kraay and Ventura [2000, 2002] and Ventura [2003], who advocate
models of the current account allowing for international portfolio diversification in
which the portfolio share of foreign wealth is constant. But we also note that our
methodology is valid under a much weaker condition: all we need is a well-defined
expected value of the ratio of net foreign debt to domestic private wealth. For instance,
g>2 is also constant when such ratio varies over time following a stationary distribution.
Indeed, in our econometric study we are unable to reject the hypothesis that q>2
is constant - which may correspond to a portfolio share of foreign assets in wealth that
is either time-invariant, or (more plausibly) follows some stationary distribution. With a
time varying portfolio share, however, it is possible that a country switches its
international net position during the sample period. Since in deriving our log-linear
approximation we have assumed that no variable switches sign, a problem in applying
our methodology to the US is that this country is a net creditor in the first part of our
sample, and becomes a net debtor during the 1980s. Below, we will address this issue
in two alternative ways. In one application of our methodology, we rescale the debt
series so as to make it positive throughout the sample. In another application, we
employ Dt in deviation from its sample mean. Summary statistics for the variables used
are provided in Table 1.
Our empirical approach exploits the cointegrating relation (1) - without
imposing additional structure on, say, preferences of the national representative agents
or technology. As long as budget constraints - which can be derived from the
transversality conditions of the problem of the national representative consumers hold, a country's net output, consumption and net foreign debt should commove in the
long-run and therefore be cointegrated. In fact, as we discuss below, our empirical
findings support this hypothesis.
Specifically, we are unable to reject (marginally) the hypothesis that there is at
most one cointegrating vector among net output, consumption and the stock of net
liabilities (all measured in logs or, in the case of net foreign liabilities, measured in
deviation from average), while we easily reject the hypothesis that there is not
cointegration. The estimates of the cointegrating coefficients have the right sign and
satisfy an important inequality: they imply that the value of net foreign debt is strictly
smaller than the present value of net output.
The log-linearized budget constraint also implies that these coefficients should
sum to minus one. A point stressed by Lettau and Ludvigson [2001, 2004] is that
measurement errors may make it unlikely that these coefficients sum to minus one in
empirical implementations of the model: nondurable consumption flows are not
directly observable and need to be proxied; our measure of net foreign liabilities is a
rough proxy of the true theoretical variable. The restriction on the cointegrating
coefficients is rejected at conventional significance levels when we use our full sample
1963-2002; but it is not rejected when we consider the sample limited to the period
1963-1997. Table 2 shows the cointegrating coefficient estimates in the two
52
applications that we consider, along with the likelihood ratio tests of the
aforementioned parameter restriction.5
Dynamic Effects of Temporary and Permanent Shocks
The empirical approach in Corsetti and Konstantinou [2004] consists of using the
restrictions implied by cointegration to identify the permanent and transitory
components of the three-variable system. Identification is possible because
cointegration places restrictions on the long-run multipliers of the shocks in a model
where innovations are distinguished by their degree of persistence, as shown, for
example, in Johansen [1995], King et al [1991], and Warne [1993]. While this
approach does not identify shocks that are structural in any sense, it yields results that
have some natural structural interpretation.
The steps involved in the procedure are as follows. We first estimate the
VEqCM, and then use the estimated parameters to back out the long-run restrictions.
More specifically, cointegration restricts the matrix of long-run multipliers of shocks in
the system, which identifies the permanent components. The transitory components are
identified in a 'residual' manner. In order to study the dynamic impact of the transitory
innovations, it is assumed that these are orthogonal to the permanent innovations - a
description of our methodology is provided in an appendix.
In our baseline identification, the first permanent shock is the only shock that
has a long-run impact on net output per capita. Hence, it has a natural interpretation as a
permanent technology shock. More generally, this shock can be read as a linear
combination of structural shocks that would have a permanent effect on net output. The
second permanent shock in our baseline model structure has a long-run impact on
consumption and foreign wealth, but no persistent effect on output. The transitory
shock only affects net output in the short run, and can therefore be read as a linear
combination of structural shocks that lead to transitory changes in zt - including
temporary technology shocks.
Referring to our previous work for details on full sample estimation, and to
Table 2 for estimates of the long-run parameters in the 1963-1997 sub-sample, we now
briefly discuss two main results, regarding the response of the system to the first
permanent shock and the temporary shock. The impulse responses for these shocks are
shown in Figure 1 for the case of variables measured in logs, and in Figure 2 for the
case in which we consider Dt in deviation from its sample mean. For the sake of
comparison, Figure 1 also reports the point estimates of the impulse responses for the
full sample (see Corsetti and Konstantinou [2004]).
First, in response to a positive transitory shock raising net output, both
consumption and iiet output rise on impact, but consumption rises by less than net
output. Correspondingly the country runs a current account surplus and accumulates
foreign assets: net foreign liabilities jump down on impact and remain negative for at
least ten years.
The simplest intertemporal models of the current account predicts that national
agents lend abroad to smooth consumption in the face of positive temporary shocks to
5
We estimate the relation

where
It is worth
clarifying that there are three alternative ways of expressing this theoretical restriction: (i) the sum of all
the coefficients of p should be zero; (ii) the sum of the coefficients on z/ and dt should be minus one;
(iii) or solving,
where the sum of the two coefficients equals one. These are
alternative ways of expressing the same parameter restriction.
53
their net output - see e.g. Obstfeld and Rogoff [1996], chapter 2 and especially chapter
3, where in overlapping generation models the effects on foreign assets and
consumption disappear over time. The pattern of our impulse response is strikingly
consistent with this prediction - although our model does not allow us to reach any
conclusion about optimality of the response to the shock.
We stress that our result does not lend support to procyclical current account
deficits. Temporary output expansions are not associated with a widening of the
external imbalance, as implied by traditional models stressing the role of real demand
shocks (e.g. government spending) in generating business cycle fluctuations.
Second, in response to the first permanent shock - that, we argued above, can be
interpreted as a technology shock - consumption increases on impact and keeps
increasing over time, while net output increases more slowly. The permanent shock
increases net foreign liabilities on impact, which keep increasing for a few years after
the shock - although they revert to a lower level in the long run, but strictly above zero.
The standard intertemporal model predicts that permanent productivity shocks that raise per capita net output and consumption in the long run - generate a current
account deficit and raise net foreign liabilities: higher returns to domestic capital attract
foreign investment, higher permanent income tends to reduce domestic saving. Our
empirical results are once again consistent with this prediction. We will reconsider this
specific result in the framework of our four variable model below.
Note that, given that the US is large in the world economy, one may expect a
shock raising US productivity and US demand (i.e. reducing US net saving) to put
upward pressure on the world real interest rate. To the extent that the upsurge in the US
demand translates into a temporary higher real rate of return, the consumption Euler
equation implies that the marginal utility of US consumption should fall gradually
along the transition path. Consistent with this view, as shown by the figure the US
consumption increases gradually in response to the shock.
Variance decomposition is shown in Table 3. Most of the long run variance of
output and consumption is explained by the first permanent shock. Temporary shocks
explain a high share of variance of output and consumption in the short run, but the
influence of this shock remains relevant also over long-horizons. In addition, as in
Corsetti and Konstantinou [2004], the transitory shock accounts for a respectable share
to the variation in the US net foreign position. The second permanent shock explains a
very limited share of the variance of net output and consumption, but its influence on
the variability of net liabilities is sensitive to sample specification.
An important result emerging from our analysis is that our main conclusions
regarding the response of the system to the first permanent shock and the temporary
shock are virtually unaffected by using a shorter sample, and measuring net foreign
liabilities in deviations from its sample mean rather than in logs. They provide the core
of what we interpret as empirical evidence that qualitatively lend support to the
intertemporal approach to the current account.
A Richer Model: Long-Run Relationships Between Consumption, Output,
Investment and Net Foreign Liabilities
We now develop our analysis by specifying a four variable model. Above, we have
defined net output Zt as GDP net of government consumption and total investment. We
now add investment as a separate variable to our system, and defined a new variable for
output as GDP net of government spending - for simplicity, we will refer to it simply
as output. Namely, Yt will now denote GDP net of government spending, while /, will
54
include private investment and consumption expenditure on durables (all in real percapita terms). As mentioned before, a full description of the data is provided in the
appendix.
An important building block of our work is the empirical evidence on comovement, at least in low frequencies, between output, consumption and investment.
This can be clearly seen from Figure 3, where the three series are plotted. Although the
aforementioned variables are well characterized as integrated processes, it is usually
found that the ratios of investment to output and consumption to output are usually
found to be stationary (see e.g. King et al. [1991]). Essentially, there is a long tradition
of empirical support for balanced growth in which output, investment and consumption
all display positive trend growth but the consumption-output and investment-output
"great ratios" do not.
Consistently, DSGE one-sector models restrict preferences and production
possibilities so that balanced growth occurs asymptotically - implying that permanent
shifts in productivity will induce long-run equiproportionate shifts in the paths of
output, investment and consumption. To wit, as in King et al. [1991] suppose that total
factor productivity is well described by a logarithmic random walk with drift:
(2)
Then, yA would be the average growth rate of the economy while t would represent
deviations of actual growth from this trend. In this case realizations of t change the
forecast
of
trend
productivity
equally
at
all
future
dates:
Etlog(At+s) = E M /0g(4+,)+g t . A positive productivity shock raises the expected longrun path and therefore there is a common stochastic trend in the logarithms of
consumption, investment and output. Under these circumstances, the great ratios CJYt
and I/Yt become stationary. Taking logs, we have that:
(3)
(4)
which are two theoretical relations that should hold in the data.
Now, building on the framework of Corsetti and Konstantinou [2004], let us
consider the intertemporal budget constraint:6
(5)
where Bt is the (initial) stock of net foreign assets, rt+j is the world real rate of return at
period t+j, which may vary over time and Rts is the market discount factor for date s
consumption, so that
Then by defining the stock of
net foreign liabilities as Dt = -Bt, the intertemporal budget constraint may be written as:
We have imposed the usual transversality condition
55
(6)
Following Campbell and Mankiw [1989], Bergin and Sheffiin [2000] and
Corsetti and Konstantinou [2004], we derive an approximate expression for the above
intertemporal budget constraint, by taking a first-order Taylor approximation of (6),
imposing three transversality conditions and taking expectations.7 By defining rt ~
ln(l+rf), we obtain:
(7)
where
and,
while
and
Our previous considerations about stationarity of portfolio shares of foreign wealth
(Dt/t) apply also here. By the same token, the existence of the log-linearization
parameters p^9pt and pd also implies stationarity of the present value of consumption
and investment as a share of the present value of net output - corresponding to
stationarity of the great ratios (see also the discussion in the appendix).
Under the realistic assumption that the rate of interest, the rate of growth of
consumption, output and investment are all stationary, the left hand side of the above
expression is also stationary. It should also be stressed that our approximation of the
intertemporal budget constraint implies a restriction on the parameters of the right hand
side of (7). More specifically, the sum of the coefficients on output, investment and net
The three transversality conditions are:
where
the appendix.
. Details of the derivation and discussion are provided in
56
foreign liabilities should equal minus one, i.e.

(see the
appendix for a discussion).
But, as we explain in greater detail in the appendix, estimating a long-run
relation like the left-hand side of (7) is virtually impossible, if the great ratios (3) and
(4) are stationary. Furthermore, the parameters (l//v)>A and/?* cannot be identified
separately, given that (3) and (4) hold. Instead, one may aim at estimating a long-run
relationship of the form:
(8)
which would essentially be a linear combination of (3), (4) and the left-hand side of (7).
Thus, the parameter restriction that -(l//?,p)-p. - pd = -1 cannot be tested, since as we already mentioned - the three parameters are not separately identified. In fact, if
the three long-run stationary relations (3), (4) and (8) are stationary, any linear
combination of the three would also be stationary. I.e., if we estimate (3), (4) and (8),
taking a linear combination of them will yield a stationary relation:
Since the A's are arbitrary, we can chose among an infinity of linear combinations,
including those satisfying the aforementioned restriction.
The empirical approach we describe below simply exploits the above
cointegrating relations. We focus on the dynamics of the US net foreign position and
examine, how the dynamic behavior of the net position is affected by changes in output,
investment and consumption.
Data and Econometric Framework
This section describes the econometric methodology underlying our empirical
approach. The problem is to analyze the dynamic behavior of a vector of variables xt
with n elements. In our application, x, =(it,ct,dt,yty, where // denotes loginvestment, ct log-consumption, dt log of net foreign liabilities and yt log of output. An
important point stressed in the previous section is that our main results are quite similar
whether we use our full sample, or the subsample ending in 1997. Different from the
analysis above, we will now carry out our study using the full sample.
Preliminary Analysis
Table 4 reports the summary statistics of the data. The standard deviation of the
quarterly net foreign liabilities growth is roughly ten times as high as that of
consumption growth, over four times as high as that of output growth and almost twice
as high as the standard deviation of investment. Investment and consumption appear to
have similar average growth rates, whereas output seems to be growing at a higher
average rate. The first order autocorrelation is roughly 0.04 for investment growth, 0.42
for consumption growth, 0.28 for output growth and 0.88 for net foreign liabilities
growth.
The correlations between investment, consumption and net foreign liabilities
and output growth rate are roughly 0.23, 0.07 and 0.80 respectively; the growth rate
57
consumption is positively correlated with the growth rate of net liabilities (0.05) and
output (0.52), while the growth rate of NFL is also positively correlated with output
growth (0.04). Figure 4 plots the series used in the analysis, in level and growth rates.
The Econometric Framework
Here we present the basic elements of our econometric modeling framework and in
order to avoid clutter, we defer the reader to the appendix for a more general
discussion. In our econometric analysis we use a vector autoregression (VAR) with k
lags, which in it Vector Equilibrium Correction (VEqCM) form can be written as:
(9)
The key idea behind cointegration is that the matrix II above, that pre-multiplies the
levels of the variables, may not be of full rank (see Johansen [1995], Hamilton [1994]).
More specifically, the hypothesis that xt is 1(1) is formulated as the reduced rank
hypothesis of the matrix II (see Johansen, [1995]), which can be decomposed into the
product of two matrices ap f , where a, p are eachwxr and have full rank r<n,
(10)
Following this parameterization, there are r linearly-independent stationary relations
given by the cointegrating vectors p; the matrix a gives the speed of adjustment of the
endogenous variables to their 'equilibrium' values (the cointegrating relations), while
there are also n-r linearly-independent non-stationary relations. These last relations
define the common stochastic trends of the system.
For an 1(0) process xr, the effects of shocks in the variables are easily seen in its
Wold moving average (MA) representation,
(11)
Because /* is the forecast error in xtt given past information, the elements of Cs
represent the impulse responses of the components of xt with respect to the ut
innovations. Since these quantities are just the 1-step ahead forecast errors, the
corresponding impulse responses are sometimes referred to as forecast error impulse
responses (Lutkepohl [1991]). In the case where the vector of variables xt is 1(0) (i.e.
stationary), Cs>0 for s*<x>9 hence the effect of an impulse vanishes over time and
hence it is transitory.
Since the components of ut may be instantaneously correlated (i.e., the
underlying shocks may not occur in isolation), orthogonal innovations are often
preferred in impulse response analysis. For example, if a lower triangular Choleski
decomposition is used to obtain B (E u = BBf), the actual innovations will depend on
the ordering of the variables in the vector xt so that different shocks and responses may
result if the vector xt is rearranged.
For nonstationary cointegrated processes the Cs will not converge to zero as
s* in this case. Consequently, some shocks may have permanent effects.
Distinguishing between shocks with permanent and transitory effects can also help in
58
finding identifying restrictions for the innovations and impulse responses of a VEqCM
(see Breitung et al. [2004] for an introduction to structural VEqCMs and Corsetti and
Konstantinou [2004] for an application in a similar context). Specifically, if one finds r
cointegrating relations and thus n-r common stochastic trends, one can identify n-r
permanent and r transitory shocks employing a 'small' number of identifying
restrictions (see King et al. [1991], Warne [1993]). Essentially, taking advantage of the
assumption of orthogonality among structural shocks, one needs (n-r)((n-r)-l)/2
restrictions to identify the permanent shocks and r(M)/2 restrictions to identify the
transitory shocks, relative to the total of n(n-\)!2 usually needed in structural VAR
(SVAR) applications.
As mentioned above, the results of analyses based on orthogonalization
assumptions depend on the ordering of the variables to obtain the orthogonalization.
Recent work by Koop et al. [1996] and Pesaran and Shin [1998] suggested a different
approach to impulse response analysis that does not require ordering of the variables.
Instead of orthogonalized impulse responses from Cholesky decompositions, they
propose generalized impulse responses (GIR henceforth) that are based on a "typical"
shock to the system. The average response of the system to this typical shock is
compared to the average baseline model where the shock is absent. Rather than
examining the effect of a pure shock to, say, investment, GIR analysis considers a
typical historical innovation, which embodies information on the contemporaneous
correlations between the innovations.
Cointegration Analysis and Long-Run Trends
In order to analyze the dynamic behavior of our system of variables, we first need to
determine the cointegrating rank. Based on univariate and multivariate misspecification
statistics (reported in the appendix), we choose an empirical model with three lags. In
Table 5, we report the trace test statistics for cointegration (Johansen, [1995])8 along
with the asymptotic p-values (Doornik [1998]). Notice that the trace test statistics
support a choice of r=3, implying the existence of one common stochastic trend (Stock
and Watson, [1988]). Specifically, we reject the hypotheses that there exist n-r=4
common stochastic trends (no cointegration), n-r=3 common trends, -r=2 common
trends, while we do not reject the hypothesis that there are at most three cointegrating
vectors (see also the associated /7-values reported in Table 5).
Having established the cointegrating rank, we need to impose some restrictions
on the cointegrating vectors in order to achieve identification. The estimates of the
cointegrating parameters p are obtained using maximum likelihood (Johansen, [1995]).
These are summarized in Panel A of Table 6.
The first two cointegrating vectors show that investment and consumption tend
to commove positively and almost at a one-to-one rate with output - consistent with
theoretical results from DSGE models. We also estimate a stationary relation of the
form dr=6.32lyt+stat. error. We can then examine whether the theoretical restrictions
on the cointegrating parameters are valid, namely whether the log differences of
consumption and investment with output- the 'great ratios' - are stationary. Panel B of
Table 6 shows our results, together with the estimated cointegrating vectors. These two
overidentifying restrictions imposed are not rejected at conventional significance levels.
Together with the budget constraint, they imply a stationary relation between net
foreign liabilities and output which now reads dt=lQ.7Q8yt+stat. error. It is useful to
Needles to say that similar results were obtained using the maximum-eigenvalue test statistics.
59
rewrite this relationship asyf*Q.ldt+stat. error: ceteris paribus, an increase in the stock
of foreign liabilities by 1 percent in equilibrium is associated with a 0.1 percent
increase in the flow of output net of government spending.
Given the above estimates of the cointegrating vectors, the VEqCM
representation of xr takes the form:
(12)
where Ax, is the vector of log first differences,
vector of constants,
is a (4x1)
matrix of the adjustment coefficients
and p is the (4x3) matrix of the cointegrating coefficients discussed above (see Panel B
of Table 6).9 In general it is possible to impose some restrictions on the short-run
parameters that are insignificant, and apply standard econometric systems estimation
procedures such as feasible GLS.
The standard f-ratios and F-tests retain their usual asymptotic properties when
applied to the short-run parameters: thus, one can impose individual zero coefficients
based on the f-ratios of the parameter estimators, and eliminate sequentially those
regressors with the smallest absolute values of f-ratios until all f-ratios (in absolute
value) are greater than some threshold value o>. Alternatively, restrictions for individual
parameters or groups of short-run parameters may be based on model selection criteria.
Using our estimated cointegrating relations (see Panel B of Table 6), we have
performed a model reduction exercise starting from a model with two lagged
differences of the variables. The procedure for model reduction is based on a sequential
selection of variables and the AIC. Our results are summarized in Table 7.
First, note that formal statistical testing reveals that our reduction is valid, since
both the Likelihood Ratio and Wald tests do not reject the null of valid model reduction
at conventional significance levels. Second, there is quite some predictability in all the
system's equations, with adjusted R2 ranging from 0.22 (equation for Act) to 0.79
(equation for Adt). This is partly due to the stationary relations 0f XM., but also to the
inclusion of lagged growth rates of all variables in our information set. The equation for
net foreign liabilities displays the highest degree of predictability. Finally, all variables
are shown to 'equilibrium correct' to at least one of the long-run relations. That is, all
variables respond to the previous period's equilibrium error p1 x,_,.
The Dynamic Behavior of Net Foreign Liabilities
In this section, we finally arrive to the core of our empirical contribution, whereas we
derive and examine the dynamics of our system. Although the results presented in
Table 7 already shed some light on it, an exact interpretation of the complicated
dynamic behavior of the four variables is quite cumbersome. A simpler way to present
the adjustment dynamics consists in presenting impulse responses and analyzing the
9
We have examined whether the adjustment coefficients for each of the variables in the analysis, at , are
statistically significant. In Panel C of Table 6, we report two different statistics, a Likelihood Ratio and
Wald statistics, which examine the null that none of the variables 'equilibrium corrects' in order to
restore the equilibrium relations, pf XM , back to their means following a shock to the system. Both types
of tests indicate that all the variables show strong evidence of equilibrium correction, thereby rejecting
the null. Finally, all the adjustment coefficients are reported in Panel D of Table 6.
60
forecast error variance decomposition of net foreign liabilities. In what follows, we will
do so by exploiting different impulse-response methodologies.
We proceed in our analysis by using both the VEqCM - with no restriction
imposed on the short-run parameters - and the subset VEqCM, obtained from our
model reduction exercise (to which we will refer as VEqCM and subset VEqCM
respectively).10 We compute impulse responses from both the full VEqCM and the
subset VEqCM to investigate the dynamic effects of shocks on net foreign liabilities,
providing bootstrap confidence bands for all our experiments.
Generalized Impulse Responses
Since the estimated instantaneous residual correlations are large (see Table 7), it is not
appropriate to consider the forecast-error impulse responses. The first methodology we
adopt is the Generalized Impulse Responses proposed in Pesaran and Shin [1998]. As
mentioned above, GIRs do not require identifying restrictions - they are order free: all
variables are endogenous and affect each other taking into account the historical
distribution of the (reduced form) shocks. So, for example, in analyzing the effect of
investment on net foreign position, GIRs enable us to study the total effect of the
former variable on the latter after taking into account the implications of investment
shocks for output and consumption, as well as all the feedback effects through the
system. Thus our results should not be interpreted as a 'partial effect' of investment on
net foreign liabilities, holding the rest of the variables constant. However, dynamic
responses are by no means structural, in that no shock is identified in a structural sense.
The impulse responses of net foreign liabilities to a one standard deviation
generalized shocks to each of the variables in the system are shown in Figure 5.
Notable results are as follows. A one standard deviation rise in investment ultimately
leads to a permanent increase in net foreign liabilities,11 though with some fluctuations
in the short and medium run. The effect of consumption on net foreign liabilities has
the same sign as investment. Note that the response appears to be significant for a year
following the shock, and it becomes again significant roughly eight years after the
shock. On the other hand, a one standard deviation rise in net foreign liabilities has a
pronounced impact on the level of this variable, but its effects are transitory. One way
to interpret this result is that there might be some inherent short-term dynamics of the
US net foreign position, reflecting the behavior of international financial markets (see
Figure 6). Finally, an increase in output leads to a temporary improvement of the US
net foreign position (decrease in the level of net foreign liabilities).
A similar picture emerges when examining the generalized forecast error
variance decomposition of net foreign liabilities, which is reported in Panel A of Table
8. The variation of net foreign liabilities is dominated by generalized innovations to
itself at all horizons. More specifically, at a horizon of one year, the generalized
innovations in dt account for roughly 95% of its variation. On the other hand, at longer
10
In the first case (VEqCM) we estimate (12) with no restrictions on the short-run parameters and
deterministic terms (a,!"] ,r 2 ,d) as explained above. For the purpose of estimating impulse responses,
results from Briiggemann, Krolzig and Liitkepohl [2002] indicate that strategies using a liberal model
selection criterion (e.g. AIC) are well suited. In this method the significance of adjustment coefficients a
is tested within the Johansen framework, before the VEqCM is estimated. Then further restrictions on
8,r; andF2 are imposed by sequentially deleting variables with /-ratios smaller than a threshold value.
Applying this method to our VEqCM model reduces the number of parameters by 20.
11
We stress that, in light of our description of the GIRs, this result is not a mere reflection of a national
account identity, as it takes into account simultaneously movements in all variables in the system.
61
horizons (about 8 years) the generalized innovations in consumption seem to account

for a non-negligible amount of the variation in dt while investment and output have
only minor contributions to the variation of net foreign liabilities.
The Effect of Permanent (Technology) Shocks
Given that we have found three cointegrating relations in our model, we know that
there is one permanent shock and three transitory shocks in the system. Following the
methodology spelled out in previous work (Corsetti and Konstantinou [2004]), we now
proceed by assuming that these four shocks are orthogonal. This assumption
automatically identifies the permanent shock. As discussed above, since this shock
raises per capita output (as well as consumption and investment) in the long run, it has
a natural interpretation as a permanent technology innovation. In what follows, we
focus exclusively on this shock, for two reasons. First, identifying the three transitory
shocks separately in our four variables system requires further identifying assumptions
- for which theory offers limited or no guidance. Second, the results of the three
variables system show that permanent shocks raise net output, consumption and net
liabilities. We would like to address the important question as of whether investment
dynamics is consistent with our interpretation of this shock as a technology shock, i.e.
whether this shock raises the rate of capital accumulation.
The effects of a one standard deviation permanent shock are shown in Figure 7.
As apparent from the figure, our permanent shock raises consumption (cr), output (yr),
net foreign liabilities (dt) and investment (it) rise. Strikingly, the effects of this shock
are quite similar to those presented in Corsetti and Konstantinou [2004] and reviewed
above - derived in the context of a quite different empirical exercise. Relative to our
previous results, the valuable additional information is that investment indeed
increases, consistent with our interpretation of the shock as a domestic technology
shock, and in accord with standard open economy models.
Consumption responds to the permanent shock by increasing gradually over
time -possibly reflecting equilibrium changes in the real interest rates - as in our three
variable model. Investment, instead, increases rapidly, reaches a peak after
approximately one year, and then converges slowly to a new higher steady state level.
Net liabilities also increase gradually: the change becomes significantly different from
zero after one year.
Panel B of Table 8 reports the fraction of the total variance in the forecast error
of Adt that can be attributed to the permanent shock as opposed to the sum of the three
transitory shocks. For a horizon between one and four quarters, the transitory shocks
account for a portion between 98% and 93% of the variance in net foreign liabilities. At
a horizon of eight to 20 quarters ahead, the transitory shocks continue to contribute a
considerable amount to the forecast error variance of Adt (between 85 and 58 percent).
However, the permanent shock now accounts for roughly 15% - 42% of the variance.
At a horizon of forty quarters, the permanent technology shock accounts for
70% of the variance of net foreign liabilities. Notably, at a horizon of forty quarters, the
transitory shocks still contribute 30% to the variance of Mt, while the permanent shock
accounts for the total of the long-run error variance in dt (and in fact for all variables).12
Again, these findings are similar to those reported in Corsetti and Konstantinou [2004],
12
The property that only permanent shocks affect the variables in the long run, whereas transitory do not,
follows from cointegration and is not specific to the rotation of the shocks we have chosen. See also
Gonzalo and Ng [2001 ] for a discussion.
62
in that permanent shocks account for the majority of the variation in Adt in the
medium to long-run, whereas transitory shocks contribute a non-negligible fraction of
the variation of the US net foreign position.
Conclusions
In this paper we have studied the long- and short-run dynamic behavior of the US net
foreign position. The analysis was performed using quarterly data on real per-capita
output, real per-capita consumption, real per-capita investment and real per-capita net
foreign debt, and applying several recent developments in the econometric analysis of
non-stationary processes and cointegration.
In the first part of the paper we have reviewed results of our earlier contribution
(Corsetti and Konstantinou [2004]) and we have also addressed concerns about a
potential influence on our result of a possible asset price bubble in the asset markets at
the end of the 1990s, by considering a shorter sample. We have found that net output,
consumption and the stock of net liabilities (either in logs or in deviations from the
sample mean) are cointegrated, as implied by the intertemporal budget constraint. We
have then identified transitory and permanent shocks to the system. As suggested by
the intertemporal approach to the current account, a permanent shock that raises percapita net output permanently (which has a natural interpretation as a permanent
technology shock) also raises consumption and net foreign liabilities. Conversely,
transitory improvements in net output lead to a build up of foreign asset - in contrast
with traditional models predicting pro-cyclical current account balances.
In the second part of the paper, we extended our methodology to a four-variable
model, including investment as a separate variable. Using a cointegrated VAR model,
we uncovered three long-run relationships: two correspond to the "great ratios"
discussed in recent DSGE models and are stationary, and the final one results from
combining these two ratios with a log-linear version of the intertemporal budget
constraint, i.e., a long-run relation between the level of US net foreign wealth and US
output.
Second, we have examined the propagation of historical-typical shocks to the
system by means of Generalized Impulse Responses. We found that shocks raising
consumption and investment tend to worsen the US net foreign position; shocks that
increase output tend to improve it mildly, whereas shocks that exogenously increase the
stock of US net foreign liabilities have a pronounced but transitory effect on the level
of this variable. In addition, the later type of shocks accounts for the vast majority of
the fluctuations in the net foreign position at short- and medium-term horizons.
Finally, reconsidering the exercise in the first part of the paper, we have used
the estimated cointegrated VAR to analyze the effect of permanent innovations in
explaining the dynamics of US net foreign position. We found that a permanent
technology shock raising per-capita output and consumption also increases investment
and worsens the US net foreign position. Such result confirms our interpretation of the
permanent shock that in our system raises long run output, as a permanent technology
shock. In the four-variable model, the dynamic behavior of investment is fully
consistent with the dynamics response to a permanent productivity shock that raises the
returns to capital in the US. This shock accounts for the vast majority of the variation of
net foreign liabilities at medium- and long-term, whereas temporary shocks account for
most of its variability at shorter horizons.
63
References
Bergin, P. R. and Sheffrin, S. M. [2000] Interest Rates, Exchange Rates and Present
Value Models of the Current Account, Economic Journal, 110, 535-558
Bruggemann, R., Krolzig, H.-M. and Liitkepohl, H. [2002] Comparison of model
specification methods, Discussion Paper 80, Sonderforschungsbereich 373,
Humboldt-Universitftt zu Berlin
Breitung, J., Briiggemann, R. and Liitkepohl, H. [2004] Structural Vector
Autoregressive modeling and impulse responses, in Liitkepohl, H. and Kratzig,
M. (eds), Applied Time Series Econometrics, Cambridge: Cambridge University
Press, forthcoming.
Campbell, J. Y. [1987] Does Saving Anticipate Declining Future Labor Income? An
Alternate Test of the Permanent Income Hypothesis, Econometrica, 55, 124973
Campbell, J. Y., Lo A. W. and MacKinlay, A. C. [1997] The Econometrics of Financial
Markets, Princeton, NJ: Princeton University Press
Campbell, J. Y. and Mankiw, N. G. [1989] Consumption, Income, and Interest Rates:
Interpreting the Time Series Evidence in Blanchard O.-J. and Fisher, S. (eds.)
NBER Macroeconomics Annual, Cambridge MA: MIT Press, 185-244
Campbell, J. Y. and Shiller, R [1987] Cointegration and Tests of Present-Value
Models, Journal of Political Economy, 95, 1063-1088
Corsetti, G., Dedola, L. and Leduc, S. [2004] International Risk-Sharing and the
Transmission of Productivity Shocks, European Central Bank Working Paper
no. 308
Corsetti, G. and Konstantinou, P. Th. [2004] The Dynamics of US Net Foreign
Liabilities: An Empirical Characterization, CEPR Working Paper (forthcoming)
Doornik, J. A. [1998] Approximations to the Asymptotic Distribution of Cointegration
Tests, Journal of Economic Surveys, 12, 573-593
Doornik, J. A. and Hansen, H. [1994] An Omnibus Test for Univariate and Multivariate
Normality, University of Oxford, Nuffield College Working Paper
Engle, R. F. and Granger, C, W. J. [1987] Cointegration and Error-Correction:
Representation, Estimation, and Testing, Econometrica, 55, 251-276
Glick, R. and Rogoff, K, [1995] Global versus Country-specific Productivity Shocks
and the Current Account, Journal of Monetary Economics, 35, 159-92
Gonzalo, J. and Ng, S. [2001] A Systematic Framework for Analyzing the Dynamic
Effects of Permanent and Transitory Shocks, Journal of Economic Dynamics
and Control, 25, 1527-1546
Hamilton, J. D. [1994] Time Series Analysis, Princeton, NJ: Princeton University Press
Johansen, S. [1995] Likelihood-Based Inference in Cointegrated Vector Autoregressive
Models, Oxford: Oxford University Press
International Monetary Fund (IMF) [2002] Essays on Trade and Finance, Chapter 3 in
World Economic Outlook, 65-107
King, R. G., Plosser, C., Stock, J. H. and Watson, M. W. [1991] Stochastic Trends and
Economic Fluctuations, American Economic Review, 81, 819-840
Koop, G. , M. H. Pesaran and Potter, S. M. [1996] Impulse Response Analysis in
Nonlinear Multivariate Models, Journal of Econometrics, 74,119-147
Kraay, A. and Ventura J. [2000] Current Accounts in Debtor and Creditor Countries,
Quarterly Journal of Economics, November 2000
64
Kraay, A. and Ventura J. [2002] Current Accounts in the Long and Short Run, Working
Paper, Department of Economics, MIT
Lane, P. R and Milesi-Ferretti, G.-M. [2001] The External Wealth of Nations:
Measures of Foreign Assets and Liabilities in Industrial and Developing
Countries, Journal of International Economics, 52, 263-94
Lettau, M. and Ludvigson, S. [2001] Consumption, Aggregate Wealth and Expected
Stock Returns, Journal of Finance, 109, 815-849
Lettau, M. and Ludvigson, S. [2004] Understanding Trend and Cycle in Asset Values,
American Economic Review, 94, 276-299
Liitkepohl, H. [1991] Introduction to Multiple Time Series Analysis, Heidelberg:
SpringerVerlag
Nason, J. M. and Rogers, J. H. [2003] The Present-Value Model of the Current
Account Has Been Rejected: Round Up the Usual Suspects, Federal Reserve
Bankof Atlanta Working Paper 2003-7
Obstfeld, M. and Rogoff, K. [1996] Foundations of International Macroeconomics,
Cambridge MA: MIT Press
Osterwald-Lenum, M. [1992] A Note with Quantiles of the Asymptotic Distributions of
the Maximum Likelihood Cointegration Ranks Test Statistics: Four Cases,
Oxford Bulletin of Economics and Statistics, 54, 461 -472
Pesaran H. M. and Shin, Y. [1998] Generalized Impulse Response Analysis in Linear
Multivariate Models, Economics Letters, 58, 17-29
Quah, D. [1992] The Relative Importance of Permanent and Transitory Components:
Identification and Some Theoretical Bounds, Econometrica, 60, 107-18
Sheffiin, S. M. and Woo, W. T. [1990a] Testing and Optimizing Model of the Current
Account via the Consumption Function, Journal of International Money and
Finance, 9, 220-33
Sheffrin, S. M. and Woo, W. T. [1990b] Present Value Tests of an Intertemporal Model
of the Current Account, Journal of International Economics, 29, 237-53
Stock, J. H. and Watson, M. W. [1988] Testing for Common Trends, Journal of the
American Statistical Association, 83, 1097-1107
Tille, C. [2003] The Impact of Exchange Rate Movements on U.S. Foreign Debt,
Federal Reserve Bank of New York, Current Issues in Economics and Finance,
9, 1-7
Ventura, J. [2003] Towards a Theory of Current Accounts, The World Economy, April
2003.
Wame, A. [1993] A Common Trends Model: Identification, Estimation and Inference,
Institute for International Economic Studies (IIES, Stockholm University)
Working Paper No. 555
65
Appendix
Data Description
This appendix briefly describes the variables employed in the analysis. For all variables
except net foreign liabilities, our source is the FRED II Database of the Federal Reserve
Bank of Saint Louis.
CONSUMPTION C,
Consumption is measured as expenditure on non-durables (PCNDGC96) and services
(PCESVC96). The quarterly series are seasonally adjusted at annual rates, in billions of
chain-weighted 1996 dollars.
OUTPUT NET OF GOVERNMENT SPENDING 7,
This is defined by the identity Yt = GDPt - Gt. GDPt is the real gross domestic product
(GDPC1) and Gt is real government consumption expenditures & gross investment
(GCEC1). All series are seasonally adjusted at annual rates, in billions of chainweighted 1996 dollars.
.
PRIVATE INVESTMENT /,
Investment It is defined as real gross private domestic investment (GPDIC1) + real

change in private inventories (CBIC1) + real personal consumption expenditure on
durable goods (PCDGCC96). All series are seasonally adjusted at annual rates, in
billions of chain-weighted 1996 dollars.
NET OUTPUT Zt
This is Z, = GDPt - Gt - /,. GDPt, Gt and /, are defined above.
NET FOREIGN DEBT A
We build our series of net foreign liabilities by cumulating the negative of the US
Current Account (BOPBCA). We scale the resulting series by 1.000, so that the series
become positive throughout our sample. In the cumulated current account series, the
minimum observation (largest negative in absolute value) is -699.77. So we have
experimented using different additive constants (750, 800, 900, 1000, 1100), verifying
the absence of any qualitative difference in our results.
POPULATION
Our measure of population was obtained by sampling at the end of each quarter the
monthly population series.
PRICE DEFLATOR
To deflate net foreign liabilities, we employ the personal consumption expenditure
chain-type deflator (1996=100), seasonally adjusted (PCECTPI), as a proxy of the
unobserved price deflator corresponding to our measure of consumption.
66
Lag-Length Selection and Misspecification Statistics

Table A.I; Misspeciflcation Statistics of VAR with k=3 lags
Panel A: Univariate Statistics
Equation
a x 100
AR(12)
ARCH(3)
NORM(2]
R2
Ait
Act
2.4337
0.3986
0.875
1.872*
0.515
0.138
3.725
9.608**
0.349
0.275
Adt
2.0146
1.432
17.093**
72.164**
0.823
Ayt
0.9226
0.821
1.591
9.309**
0.342
Panel B: Multivariate Tests

TM'I
LMJ
IM&
LM12
L - B(40)
NORM (6)
17.307
27.986
31.702
11.445
754.558
89.684
[0.366]
[0.032]
[0.011]
[0.781]
[0.000]
[0.000]
Panel C: Lag-Length Selection

k
logi
7i(fc-l/fc)
p- value
0
1
614.8208
1914.306
N/A
2517.752**
N/A
[0.000]
2
3
2057.462
2072.621
270.207**
27.854*
[0.000]
[0.033]
2084.135
20.581
[0.195]
NOTES for Table A.I: The R2 can be interpreted as the fit of the model for each variable relative to
a random walk with drift. AR(12) is an LM test statistic for autocorrelation (F(12,136) distributed),
ARCH (3) is the test for ARCH effects (F(3,142) distributed), and NORM is the Jarque-Bera test for
normality (X2(2) distributed), while * (**) denotes significance at the 5% (1%) level. The LM\ are
tests of i-th order autocorrelation distributed as a %2(9). The NORM (6) is a multivariate Normality test
(Doornik-Hansen, 1994) which is distributed as a X2(6)- The IrB(30) is the multivariate version of the
Ljung-Box test for autocorrelation based on the estimated auto- & cross-correlations of the first [T/4=39]
lags and is distributed as a %2(333). log L denotes the value of the log-likelihood, CfR, is sequential (i.e.
k vs k 1 lags) Likelihood Ratio test statistic corrected by a degrees of freedom adjustment. The lag
order selected is given in boldface.
67
The Log-Linearized Intel-temporal Budget Constraint

The intertemporal budget constraint is given by:
(13)
where Dt is the initial (period t) net foreign debt. We can write (13) as:
(14)
where
Similarly:
Taking logs
(15)
The LHS of (15) can be approximated by taking a first-order Taylor approximation (see
Campbell etal. [1997]):
(16)
Defining
we can rewrite (16) as:
(17)
and the approximate expression for (15) can be written as:
(18)
68
or:
(19)
Notice that:
so
Log-linearizing as above we have:

(20)
where
difference equation in the
Using
the
trivial
identity
and (20), equating the LHS, we obtain a
ratio. Then solving forward:
(21)
where the condition limT^ Pc$(ct+r ~ 0/+T-)-* ^ has been imposed. Observe what the
last condition implies. We have:
<0
so to the extent that (ct+T - 0t+T) is a stationary process, the limit term will go to zero,
at least in expectation (see Campbell et al. [1997]). To see this more clearly, using the
definition of the log-differential:
69
which is a stationary process under the maintained assumption of stationary

consumption growth ( A log Ct) and stationary rates of return (rr).
Notice also, that:
and using similar steps, we obtain:
(22)
while for investment we obtain:
(23)
where pr^ and pI@ defined similarly to the above log-linearization parameters.
Substituting (21), (22) and (23) in (19) we find that:
(24)
or by taking expectations:
(25)
where:
70
Notice that:
Implications for Stationary (Cointegrating) Relations

Under the assumption of balanced growth:
(26)
(27)
are stationary variables. Consider the LHS of (25) which is also supposed to be
stationary:
Reparameterizing we obtain:
(28)
which is also stationary.
It is well known that linear combinations of stationary variables are stationary
variables, hence subtracting from (28) equation (26) and adding p, (/, - yt\ we are left
with:
which will also be stationary.

Notice that it is impossible to test the parameter
restriction
as we are only able to estimate one parameter (p, which is a
71
function of the three parameters in our linearized model. In fact, there is an infinity of
acceptable solutions that would satisfy the above restrictions and can still match the
estimate of (p. To clarify this point, let us write again:
where et is stationary. With a bit of algebraic manipulation, and assuming that

holds, we obtain:
If we knewp,, provided that the long run restriction on parameters holds, then we could
easily determine p* But without knowing piy we would have to solve one equation in
two unknowns (hence we have an infinity of solutions).
Implications for the Coefficients py,pi and pd
The coefficients py, />,, pd defined above all depend on Dt, Yt and 0t. Consider for
instance/ty. Using definitions, we can right it explicitly as:
This expression shows that py depends on the average (steady-state) share of foreign
wealth Dt in domestic wealth
the mean value of the "great ratio" (7/7*)
and the relative growth rates of investment and output. Using similar steps it is also
easy to see that p, and pd will also depend on the share of foreign liabilities in domestic
wealth, as well as the "great ratio" of investment and output and their relative growth
rates.
Finally, using the fact that
72
if the parameter p y exists, with,
the means of CD/IP), (I/Yt) and the growth rates of investment and consumption also
exist-so that
implying that the present value of investment and consumption to output will also be
stationary (with mean value bounded by unity - as far as p^ e [0,l).
Econometric Methodology
In our econometric analysis we employ a vector autoregression (VAR) with k lags,
namely:
(29)
where
is a matrix polynomial in the lag operator, x, is a
H*l vector of variables in the system, and 5 is (n*l) a vector of constants. We assume
that ut is a sequence of independent Gaussian variables with zero mean and covariance
matrix Su and we write the system as an observationally equivalent vector equilibrium
correction (VEqCM henceforth) given by:
(30)
where the coefficient matrices are defined as

The notion of cointegration stems from the fact that the matrix II above, that
pre-multiplies the levels of the variables, may not be of full rank (see Johansen [1995],
Hamilton [1994]). In particular, we first note that the general condition for x, ~ 1(0) (i.e.
stationary) is that II has full rank, so it is non-singular. In this case, |A(1)| = |II| = 0
corresponding to the usual condition that all the eigenvalues of the companion matrix
(or the roots of the characteristic polynomial) should lie within (outside) the unit circle
(see Hamilton [1994], Liitkepohl [1991]). As stationary variables cannot grow
systematically over time (that would violate the constant-mean requirement), if xr ~
1(0), then E[xJ = p*. Taking expectations of (30) yields:
(31)
so when II has full rank, E[X J =-II~'8. Thus, the levels of stationary variables have a
unique equilibrium mean. When x, exhibits 1(1) behavior, II is not full rank and
therefore (31) leaves some of the levels indeterminate. Conversely, when 11=0, we
73
write the VAR in differences,

these are stationary if
has full
rank, in which case xr-I(l).
As we have already mentioned in text, the hypothesis that x, is 1(1) is
formulated as the reduced rank hypothesis of the matrix II, in which case it can be
decomposed into:
(32)
Following this parameterization, there are r linearly independent stationary relations
and n-r linearly independent non-stationary relations, which define the common
stochastic trends of the system. In this case the moving average representation (or
solution) of x, as a function of the disturbances u,, the initial conditions x0, and the
deterministic variables 5 is given by:
(33)
where
and p are
matrices orthogonal to
a and p respectively, C* (L)is a polynomial in the lag operator, and A is a function of
initial conditions, such that p' A = 0. It should further be noted that under the
assumption of cointegration, there are n-r linear combinations of the reduced form
shocks that have permanent effects on the levels variables, while there are also r linear
combinations of the reduced form innovations that do not have long-run effects on any
of the variables in the system. We discuss these issues below.
Under the cointegrating restrictions one can estimate a VEqCM representation
for jLt which takes the form:
(34)
The term p f x,_j gives last period's equilibrium errors; a is the vector of "adjustment"
coefficients (or loadings) that tells us which of the variables react to last periods
equilibrium errors (cointegrating residuals) - that is, which variable, and by how much,
adjusts to restore the equilibrium relations p f x M back to their mean when a deviation
occurs. By virtue of the Granger Representation Theorem (CRT, Engle and Granger
[1987]), if a vector of variables xr is cointegrated, then at least one of the adjustment
parameters in the n*r matrix a must be non-zero in the VEqCM representation (34).
Thus if Xj does at least some of the adjusting needed to restore the long-run equilibrium
subsequent to a shock that distorts this equilibrium, then some of the parameters in the
Ixr vector a/ should be different from zero in the equation for Ax, in the VEqCM
representation (34).
Dynamic Responses
For an 1(0) process xr, the effects of shocks in the variables are easily seen in its Wold
moving average (MA) representation,
(35)
74
The coefficient matrices of this representation may be obtained by recursive formulas

from the coefficient matrices Ay of the levels VAR representation, A(L)X, = 5 + u, (see
Liitkepohl [1991] or Hamilton [1994]). The elements of the Cs's may be interpreted as
the responses to impulses hitting the system. In particular, the (/-th element of Cs
represents the expected marginal response of Xj,t+s to a unit change in xit holding
constant all past values of the process.
Since the components of u, may be instantaneously correlated, orthogonal
innovations are often preferred in impulse response analysis. Using a Cholesky
decomposition of the covariance matrix E(U,U/) = U is one way to obtain
uncorrelated innovations. Let B be a lower-triangular matrix with the property that
E u =BB'.. Then orthogonalized shocks are given by e, =B~X Substituting in (35)
and defining D, = C,B (i = 0,1,2,...) gives:
(36)
Notice that Do=B is lower triangular so that the first shock may have an instantaneous
effect on all the variables, whereas the second shock can only have an instantaneous
effect on X2t to xnt but not on x\t. This way a recursive Wold causal chain is obtained.
The effects of the shocks et are sometimes called orthogonalized impulse responses
because they are instantaneously uncorrelated (orthogonal).
A well-known drawback is that many matrices B exist which satisfy BBf = E u the Choleski decomposition is to some extent arbitrary if there are no good reasons for
a particular recursive structure. Clearly, if a lower triangular Choleski decomposition is
used to obtain B, the actual innovations will depend on the ordering of the variables in
the vector x, so that different shocks and responses may result if the vector x, is
rearranged.
For nonstationary cointegrated processes the Wold representation (for the levels
of the process x/) does not exist. Still the Cs impulse response matrices can be
computed as for stationary processes from the levels version of a VEqCM (see
Liitkepohl [1991]). Generally the Cs will not converge to zero as s > oo in this case and
some shocks may have permanent effects. Distinguishing between shocks with
permanent and transitory effects, as we discussed in text and explain below, can also
help in finding identifying restrictions for the innovations and impulse responses of a
VEqCM.
As we mentioned, the results of analyses based on orthogonalization
assumptions depend on the ordering of the variables to obtain B and hence the
orthogonalized shocks. Recent results of Koop et al. [1996] and Pesaran and Shin
[1998] though have re-examined the concept of orthogonalized impulse responses,
aiming to remove this shortcoming. Instead of orthogonalized impulse responses from a
Choleski decomposition, they suggested generalized impulse responses (GIR
henceforth) that are based on a "typical" shock to the system.
The argument about GIR may be explained as follows. Let the Vector Moving
Average (VMA) representation of the ^-variable cointegrated VAR model be given by:
75
where K O is a vector of constants, KJ are the coefficients of the deterministic trend, and
u, is a vector of unobserved "shocks", where
and let <riy be a
typical element of
Then it holds that:

(37)
where cy is a (wx l) selection vector with element/ equal to unity and zeroes elsewhere.
Then the GIR of the effect of a "unit" shock to \hej-th disturbance term at time t on
Ax,+, is:
(38)
and the GIR of x/+/1 following a shock to they'-f/z variable is:

(39)
where Qh = ]jL0C, and the GIRs are measured h periods after the shock has occurred.
In general, Pesaran and Shin (1998) show that one can interpret generalized
impulse responses for a stationary vector process x, as:
They also explain that in a linear system, the impulse responses will be invariant to
history (the information set on which conditioning is made), and so the GIR will
depend only on the composition of the shocks as defined by
As shown in Pesaran and Shin (1998), the GIR will be numerically equivalent to
the standard impulse response function based on Cholesky decompositions if E u is
diagonal.
Identifying Transitory and Permanent Components: A Short Review of
Methodology
It is useful to review our methodology in some detail, and explain how it is related to
our application. From the GRT it follows that, under the maintained hypothesis that the
growth rates in x* are covariance stationary, there exists a multivariate Wold
representation of the form:
(40)
where C(L) is a nxn matrix polynomial in the lag operator. We want to map these
reduced form innovations into transformed innovations e, that are distinguished by
whether they have permanent or transitory effects. Without loss of generality the
76
shocks cf are ordered so that the first n-r of them have permanent effects; and the last r
of them have transitory effects. Following Gonzalo and Granger [1995], we define a
shock e/p, as permanent if lim^^ dEfo+J/def * 0, and a shock, e,r, as transitory if
7/mâE(O/*r = Applying the methodology of King et al [1991], as extended by Warne [1993]
and discussed in Johansen [1995], the permanent and transitory innovations may be
identified using the estimated parameters of the VEqCM representation of a
cointegrated system. In particular, as explained in Johansen [1995], the matrix C(l) of
the Wold representation (40), admits a closed-form solution in terms of the parameters
of the cointegrated VAR:
(41)
Notice that the structure of this matrix is such that it maps reduced-form disturbances u/
into the space spanned by the columns of a i.e. sp(a). The disturbances o^ L u /
accumulate to the permanent component of x*, whereas transitory disturbances will be
in the null-space of C(l). We can therefore define the permanent disturbances
(permanent shocks) as:
(42)
Then by requiring that the permanent and transitory shocks be orthogonal to each other,
we can define the transitory shocks as:
(43)
Denoting:
(44)
(45)
we have that:
(46)
Notice that in this way we have achieved both the rotation from reduced-form shocks to
permanent and transitory shocks and the orthogonalization. Let
and
;
e, = P~ u,, the transformed Wold representation is:
(47)
77
Thus each element of Ax, has been decomposed into a function of n-r permanent and r
transitory shocks. Essentially, taking advantage of the assumption of orthogonality
among structural shocks, one needs
restrictions to identify the
permanent shocks and r(r-l)/2 restrictions to identify the transitory shocks, relative to
the total of n(n-l )/2 usually needed in structural VAR (SVAR) applications.
Consider the three variable case with =3 and r=l. Notice that we need no
restrictions to identify the transitory shocks: the assumption of orthogonality of the
shocks is sufficient. Instead, we need one extra restriction to identify the permanent
shocks. Next, consider the four variable system with n=4 and r=3. In this case we need
no extra restriction to identify the permanent shock, whereas we need three extra
restrictions in order to identify the transitory shocks.
78
Ikble 1: Summary Statistics for the Sample 1963:Q1 - 1997;Q4

Azt
Act
Adt
A(ag)~
Univariate Summary Statistics

Mean (xlO )
0.5714 0.5535 1.5081
Standard Deviation (xlO 2 ) 1.Q270 0.4714 4.9279
Autocorrelation
0.1306 0.4153 0.8746
Correlation Matrix
2
Ali
Ac*
Adt
A (*-)
1.000
.
.
3.7803
6.3945
0.9447
0.4773 -0.0762
1.000 0.0568
1.000
-0.0344
0.0866
0.7129
1.000
NOTES for Table 1: This table reports summary statistics for quarterly growth of net output AJSJ,
consumption Act j and two measures of net foreign liabilities growth rate Ac?t and A f * ) > where
all variables are expressed in real, per-capita terms. The sample spans the first quarter of 1963 to the
fourth quarter of 1997.
Table 2: Sub-Sample Long-Run Parameter Estimates and Tests

Panel A: Using Logarithms
Panel A.I: Estimated Cointegrating Relationship
Panel A.2: Restricted Cointegrating Coefficients
"*
\ t
\.
Panel B; Using dt = (Dt - D) /D

Panel B.I: Estimated Cointegrating Relationship
Panel B.2: Restricted Cointegrating Coefficients
NOTES for Table 2: Panel A of the table reports our results using dt = log(Dt + ) and Panel B our
results using dt = (Dt D} /D where D is the sample average of Dt* The numbers in square brackets
are the associated t-statistics and the numbers in curly brackets are the associated p-values for the tests.
The sample spans the first quarter of 1963 to the fourth quarter of 1997.
79
Table 3: Forecast Error variance Decomposition for the Sub-Sample

Horizon h
Panel A: Using Logarithms
1~
2
3
4
8
12
16
20
40
0.06
0.07
0.06
0.05
0.07
0.18
0.32
0.46
0.80
0.10
0.12
0.14
0.16
0.20
0.20
0.18
0.15
0.06
0.84
0.81
0.80
0.79
0.73
0.62
0.50
0.39
0.14
0.49
0.54
0.58
0.62
0.73
0.80
0.85
0.88
0.93
0.00
0.00
0.00
0.00
0.00
0.01
0.01
0.02
0.03
0.51
0.44
0.42
0.38
0.26
0.19
0.14
0.10
0.04
0.27
0.36
0.40
0.43
0.46
0.45
0.44
0.42
0.34
0.54
0.50
0.49
0.48
0.49
0.51
0.53
0.56
0.65
0.19
0.14
0.11
0.09
0.05
0.04
0.03
0.02
0.01
0.37
0.34
0.33
0.33
0.33
0.34
0.35
0.36
0.42
0.24
0.16
0.12
0.09
0.05
0.04
0.03
0.03
0.02
Panel B: Using dt = (Dt - D) /D
1
2
3
4
8
12
16
20
40
0.08
0.09
0.09
0.08
0.06
0.12
0.21
0.30
0.63
0.18
0.21
0.24
0.26
0.33
0.36
0.35
0.34
0.21
0.74
0.70
0.67
0.66
0.61
0.52
0.44
0.36
0.16
0.47
0.51
0.55
0.59
0.70
0.78
0.83
0.86
0.94
0.02
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.00
0.51
0.47
0.43
0.39
0.29
0.21
0.16
0.13
0.06
0.39
0.50
0.55
0.58
0.62
0.62
0.62
0.61
0.56
NOTES for Table 3: The table reports the fraction of the variance in the h step-ahead forecast error of
the variable listed at the head of each column that is attributable to innovations in the permanent shocks
(P), and the transitory shock (T). Horizons are in quarters, and the underlying VEqCM is of order 1.
80
Table 4: Summary Statistics

Univariate Summary Statistics
Mean(xl0 2 )
0.5466 0.5472 1.6629
Standard Deviation (xlO 2 ) 2.9138 0.4534 4.6293
Autocorrelation
~it
Ac*
A4
Ayt
0.0447
0.6250
1.0990
0.4163
0.8785
0.2S24
Correlation Matrix
1.000 0.2306
1-000
.
.
0.0719
0.0549
1.000
0.8042
0.5179
0.0429
1.000
NOTES for Table 4: This table reports summary statistics for quarterly growth of private investment
Ait, consumption Act, private output Aj/t> and the net foreign liabilities growth rate Ac?*, where a\i
variables are expressed in real, per-capita terms. The sample spans the first quarter of 1963 to the fourth
quarter of 2002.
Table 5: Trace (Cointegration) Statistics
HQ :r
n - r "
Q(r\n)
Asymptotic rvalue
Q<*s(r\n)
r=0
F"
66.79
[0.000]
47.21
r <1
3
34.72
[0.012]
29.68
r<2
2
15.53
[0.048]
15.41
r <1
1
0.22
[0.639]
3.76
NOTES for Table 5: Q(r\n) denotes the trace statistic as defined in Johansen [1995], i.e. Q(r\n) =
"~^]Clt=r-f 1 In (l ~ -î) The asymptotic p-values reported are calculated using the methods in Doornik
[1998], while the asymptotic critical values are taken from Osterwald-Lenum [1992]. The sample spans
the first quarter of 1962 to the fourth quarter of 2002.
81
Table 6: Cointegrating Parameter Estimates

Panel A: Identified Cointegrating Vectors
Variable
/31
/32 fa
1
(-)
0
(-)
0
<-)
-0.958
it
ct
dt
yt
0
(-)
1
(-)
0
(-)
-0.895
(0.0404)
(0.0157)
0
(-)
0
<-)
1
<-)
-6.321
(0.6728)
Panel B: Over identified Cointegrating Vectors

Variable
fa
fi
/3j
it
1
0
0
<-)
(->
<-)
0
1
0
<->
<-)
<->
dt
0
0
1
<-)
(-)
(-)
yt
-1
-1
-10.708
(-)
(-)
(Q.4682)
Cn: Q(2) = 3.805 {0.149} (bootstrapped {0.311})
ct
Panel Ci No Long-Run Feedback Tests

HQ : QJ= 0
Ait
Act
Mt
Ayt
K,:Q(3)
23.148
10.138
19.084 17.359
b-v<due}
{0-000>
(&017)
{0.000}
{0.000}
WAV:W(3)
29.458
14.264
14.927
25.840
{p-value}
{'QOO>
<aoo3>
{-002>
{-00}
Panel D: Adjustment Coefficients

it - yt
At
-0.163
Act
0.010
Adt
0.118
Ayt
-0.048
ct-yt
[t-atoll
dt-10.708yt
0.142
I1-384)
-0.0003
0.031
I2-963!
-0.0002
0.465
f3-845!
-0.013
0.083
I3'522)
0.002
[*-atoll
[t-tol]
1-3.619]
1-0.133]
[1.574]
[-3.660]
[2.508]
[-4.132]
[-3.301]
[1.205]
NOTES for liable 6: Panel A of the table reports the Full Information Maximum Likelihood (FIML)
estimates of the Cointegrating vectors ft subject to exactly identifying restrictions. The numbers in
parentheses are the associated (conditional) standard errors. Panel B of the table reports the estimates
of the Cointegrating vectors subject to over-identifying restrictions. The likelihood ratio test is distributed
as a x2 (2); the number in curly brackets are the asymptotic and bootstrapped p-vaiue (based on 5000
replications) respectively. Panel C of the Table reports two variants of a test of no long-run feedback
from the levels relations to the growth rates (also referred to as 'weak exogeneity* test). Q (3) is the
likelihood ratio test and W (3) is the Wald test with heteroscedasticity-consistent standard errors, both
distributed as 2 (3). Finally, Panel D of the Table reports the adjustment coefficients oty along with
the associated t-statistics (in square brackets). Statistically significant parameter estimates are given in
boldface. The sample spans the first quarter of 1963 to the fourth quarter of 2002.
82
Table 7: Parsimonious VEqCM Estimates

Panameter Estimates
System Reduction Tests

WAT>: W (20) = 9.979 {0.968}, CU : Q (20) = 17.258 {0.636}
System Statistics
0.285 0.224
0.799
2.462 0.398 2.073
0.271
0.937
2.018 1.924 2.013 1.892

NOTES for Table 7: The Table reports the estimated coefficients from a parsimonious cointegrated vector
autoregressive (VAR) model, where some restrictions have been imposed on the short-run dynamics; tstatistics are given in square brackets. The Table also reports two tests for the reductions (a Wald and
a Likelihood Ratio test) both distributed as a x2 (20) and some specification statistics for each equation
83
Table 8: Forecast Error Variance Decomposition

Panel A: Generalized Innovations
Horizon h
1
2
3
4
8
12
16
20
40
0.01
0.03
0.04
0.05
0.05
0.04
0.04
0.04
0.08
0.00
0.01
0.02
0.03
0.07
0.09
0.09
0.09
0.16
1.00
0.98
0.96
0.95
0.90
0.82
0.69
0.56
0.25
0.01
0.01
0.01
0.01
0.01
0.01
0.03
0.04
0.02
0.01
0.05
0.06
0.07
0.07
0.05
0.04
0.04
0.06
0.00
0.00
0.01
0.01
0.03
0.04
0.04
0.05
0.11
1.00
0.98
0.97
0.96
0.92
0.86
0.76
0.65
0.34
0.01
0.01
0.01
0.01
0.01
0.01
0.03
0.05
0.04
Panel B: Permanent and Transitory Innovations
Horizon h
1
2
3
4
8
12
16
20
40
co
0.01
0.01
0.02
0.02
0.10
0.21
0.33
0.44
0.74
1.00
0.99
0.99
0.98
0.98
0.90
0.79
0.67
0.56
0.26
-
0.02
0.04
0.05
0.07
0.15
0.24
0.33
0.42
0.70
1.00
0.98
0.96
0.95
0.93
0.85
0.76
0.67
0.58
0.30
~
NOTES for Table 8: Panel A of the table reports the fraction of the variance in the h step-ahead forecast
error of the Net Foreign Liabilities dt that is attributable to generalized innovations in w$, tt, u^ and
u%. Notice that by construction these fractions may not sum to unity. Panel B of the table reports the
fraction of the variance in the h step-ahead forecast error Net Foreign Liabilities that is attributable to
permanent (P) and transitory (T) innovations. Horizons are in quarters, and the underlying VEqCM is
of order 2. The sample spans the first quarter of 1963 to the fourth quarter of 2002.
84
Figure 1. Impulse Responses for sub-sample 1968-1997 (o) and the full sample (-)..
Notes for Figure 1: The first permanent shock is assumed to have a long-run impact on all three variables
in the system, the second permanent shock (not reported here) is assumed to have a no long-run effect
on *, but it has a long-run impact on c* and dt and the transitory shock is assumed to have no long-run
effect on any of the variables. The horizon in the figure is measured in years after the shock. The figure
shows the response of each variable to a one-unit shock for the sub-sample 1963-1997 (o) and for the full
sample 1963-2002 (-). The first column shows the responses to the first permanent shock and the second
the responses to the transitory shock, all with the associated bootsrap confidence bands obtained from
estimating the model for the sub-sample.
85
Figure 2. Impulse Responses employing (Dt D) /D as the Net Foreign Liabilities Variable.
Notes for Figure 2: The figure reports impulse responses for a model that includes (Dt JD) /D as
the net foreign liabilities variable, where D is the sample average of Dt. The first permanent shock is
assumed to have a long-run impact on all three variables in the system, the second permanent shock
(not reported here) is assumed to have a no long-run effect on zt, but it has a long-run impact on Ct
and di (Dt .D) /D and the transitory shock is assumed to have no long-run effect on any of the
variables. The horizon in the figure is measured hi years after the shock. It shows the response of each
variable to a one-unit shock for the sub-sample 1963-1997. The first column shows the responses to
the first permanent shock and the second the responses to the transitory shock, all with the associated
bootsrap confidence bands obtained from estimating the model for the sub-sample.
86
Figure 4. The Series employed in the Analysis.
87
ddt/dyt
9dt/dyt
Figure 5. Generalized Impulse Responses of Net Foreign Liabilities,

NOTES for Figure 5: The figure shows the effects of a one-standardrdeviation increase in each of the
variables shown in the denominator. The results are ordering independent. The horizon is in quarters
after the shock. Panel A reports the results for the unrestricted VEqCM(2) and Panel B the results for
the subset VEqCM(2) (see Table 7). The 95% confidence intervals were obtained by bootstrap Monte
Carlo simulation (10000 replications).
88
Figure 6. Generalized Impulse Responses to Shocks in Net Foreign Liabilities.

Notes for Figure 6: The figure shows the effects of a one-standard-deviation increase in net foreign
liabilities on each of the variables in the system. The results are ordering independent. The horizon is
in quarters after the shock. Panel A reports the results for the unrestricted VEqCM(2) and Panel B the
results for the subset VEqCM(2) (see Table 7). The 95% confidence intervals were obtained by bootstrap
Monte Carlo simulation (10000 replications).
89
Figure 7. Impulse Responses to the Permanent (Technology) Shock.

NOTES for Figure 7: The figure shows the effects of the permanent, technology shock on investment z$,
consumption c$, net foreign liabilities dt and output y$. The horizon is in quarters after the shock. Panel
A reports the results for the unrestricted VEqCM(2) and Panel B the results for the subset VEqCM(2)
(see Table 7). The 95% confidence intervals were obtained by bootstrap Monte Carlo simulation (10000
replications).
90
Session 2
Exchange Rate Issues
Financial Globalization and Exchange Rates'

Philip R. Lane
HIS, Trinity College Dublin and CEPR
Gian Maria Milesi-Ferretti
International Monetary Fund and CEPR
*Revised version of a paper prepared for the "Dollars, Debt, and Deficits: Sixty Years After Bretton Woods"
Conference co-organized by the Banco de Espana and the IMF, Madrid, June 14-15 2004. The authors thank
their discussant Daniel Cohen, conference participants in Madrid and Budapest, and participants at the Kiel
Institute for World Economics seminar and the Glasgow-Strathclyde joint seminar for useful comments, and
Abdel Senhadji and Cedric Tille for data on Australia and the United States. Marco Arena, Charles Larkin,
and Vahagn Galstyan provided excellent research assistance. Part of this paper was written while Lane was a
visiting scholar at the International Monetary Fund. Lane also gratefully acknowledges the financial support
of the Irish Research Council on Humanities and Social Sciences (IRCHSS) and the HEA-PRTLI grant to the
HIS. This paper is also part of a research network on "The Analysis of International Capital Markets:
Understanding Europe's Role in the Global Economy', funded by the European Commission under the
Research Training Network Programme (Contract No. HPRN-CT-1999-00067).
Introduction
Financial globalization has been one of the most important trends in the world economy in
recent decades. This process has involved sharply rising foreign asset and liability
positions, whether scaled by GDP or by domestic financial variables (Lane and MilesiFerretti 2003, Obstfeld and Taylor 2004). In addition to larger gross positions, financial
globalization has also allowed a greater dispersion in net foreign asset positions, with a
significant number of countries emerging as either large net creditors or net debtors (Lane
and Milesi-Ferretti 2002a). In general, financial globalization is one of the key trends that
has reshaped the global economy relative to the environment envisaged by the designers of
the Bretton Woods system in 1944 and understanding its macroeconomic implications is
crucial in formulating a view on the appropriate future direction for the international
monetary system.
One consequence of financial globalization is that the international spillovers from
asset price and currency movements have been enhanced. In addition to affecting the
direction and magnitude of net capital flows, asset price dynamics also generate changes in
the valuation of existing investment positions. For instance, the value of the net liability
position of the United States is quite sensitive to the relative movements in the U.S. versus
non-US, equity markets and swings in the value of the dollar. Indeed, such valuation
effects may be as important as current account imbalances in driving the dynamics of net
foreign asset positions (Lane and Milesi-Ferretti 200la, 2002a, Gourinchas and Rey 2004).
Of course, asset price and currency movements cannot be viewed as exogenous
influences on the value of international investment positions, since shifting global demands
for various assets and liabilities are an important driver of financial returns and exchange
rates (e.g. through the determination of country and currency risk premia). Moreover, there
is an obvious interplay between the financial and trade accounts that provides another link
between net foreign asset positions and exchange rates: a long-term debtor may require
real depreciation in order to generate the trade surpluses that are the counterpart of
sustained net investment income outflows (Lane and Milesi-Ferretti 2002b, 2004b).
In this paper we explore the interconnections between financial globalization and
exchange rates. To establish the stylized facts about financial globalization, the first part of
the paper examines trends in gross and net international investment positions and their
components for a large set of advanced and emerging economies. In Lane and MilesiFerretti (2003) we documented for industrial countries an acceleration in the pace of
financial globalization since the mid-1990s; in this paper we update our estimates of
external assets and liabilities for a sample of emerging markets as well. l As noted above, a
central aspect of our analysis is the focus on the factors explaining the changes in external
positions: not only capital flows but also valuation effects, such as those caused by asset
price and exchange rate fluctuations.
In the second part of the paper we first provide an analytical framework that is
useful in understanding the dynamics of net foreign assets, and then explore the
contribution of currency movements to the revaluation component of net foreign asset
dynamics. This relationship depends on a number of factors. For instance, the impact of an
exchange rate depreciation will depend on gross foreign asset and liability holdings (in
addition to the net position); the currency composition of both sides of the international
!
Lane and Milesi-Ferretti (2004c) explain the construction of the data.

95
balance sheet; and the co-movement between exchange rate changes and other financial
returns.2 These factors will vary across countries, according to the level of development,
country size and other characteristics. Along one dimension, a high proportion of the
liabilities of a major industrial country is likely to be denominated in its own currency,
whereas a typical emerging market economy exhibits significant liability dollarization.
Countries also differ as to the mix of short- and long-term debt and the levels of portfolio
equity and FDI holdings in the international balance sheet: the impact of currency
movements on the net external position is undoubtedly sensitive to the external capital
structure.3
Our analysis suggests that theoretical work on open economy macroeconomics
should strive to incorporate elements such as persistent non-zero net foreign asset
positions, large gross asset cross-holdings and mixed portfolios of equity and debt
instruments and illustrate why these features can make a difference to model dynamics and
welfare analysis. Finally, in the last part of the paper we draw out the implications of our
empirical work for policy analysis. In particular, we highlight that the valuation channel is
unlikely to be open to policy manipulation on a sustainable basis.
Trends in International Financial Integration
In Lane and Milesi-Ferretti (2002a, 2003), we documented a number of stylized features of
international capital flows and external positions in industrial countries. Flows to and from
such countries increased substantially in recent years, both in absolute terms and as shares
of GDP and domestic wealth. In this context, the increase in FDI and portfolio equity
investment is particularly noteworthy. The increase in gross external assets and liabilities
means that valuation effects have become more important. We highlight these features
again below, with an updated dataset including both industrial countries and emerging
markets.
Net Flows and Net Positions, Industrial Countries
Figure 1 plots net foreign assets (as a ratio of GDP) against GDP per capita, measured in
current U.S. dollars, for the year 2003. There is a wide dispersion in net external positions
among industrial countries, with Switzerland being by far the largest creditor, and New
Zealand and Iceland the largest debtors in relation to GDP. The positive relation between
net foreign assets and GDP per capita, shown in the Figure to hold in the cross-section,
holds also along the time-series dimensionas a country gets richer, relative to trading
partners, its net foreign asset position tends to improve (Lane and Milesi-Ferretti, 2002a).
Table 1 summarizes net capital outflows from industrial countries over the period
1999-2003, together with changes in their external position. In absolute terms, Japan has
been the largest capital exporter, while Switzerland and Norway had the highest net
outflows relative to their GDP. On the other side, the United States had by far the largest
net inflows in absolute terms, and also as a ratio of GDP.
2
Tille (2003) provides an interesting analysis for the United States.

Lane and Mflesi-Ferretti (200 Ib) analyze some of the determinants of the composition of the international
balance sheet.
3
96
While there is clearly a positive relation between net outflows and change in the net
external position, the Table highlights the importance of valuation effects: for example, the
United Kingdom was a net capital importer during this period, but its net external position
improved by 7.5 percent of GDP; Canada instead was a net capital exporter, but its net
position deteriorated. In absolute terms, the difference between net capital inflows and the
change in the net asset position is particularly large for the United Statesnet inflows
were over $600 billion higher than the net accumulation of liabilities. The reasons for this
discrepancy are further discussed below.
Gross Flows and Gross Positions, Industrial Countries
Figure 2 summarizes the evolution of gross external assets and liabilities in industrial
countries during the past 20 years. The growth in international financial interdependence is
striking: during this period, aggregate assets and liabilities tripled as a share of GDP, FDI
assets and liabilities increased four-fold, portfolio equity assets and liabilities six-fold, and
debt assets and liabilities 2 l/2 times. Focusing on the most recent period, the chart also
shows the effects of the global decline in stock market valuations between end-1999 and
end-2002, which is the main factor behind the reduction in the stock of portfolio equity
assets and liabilities during this period, and the recovery in stock market valuation and
flows in 2003 4
Table 2 summarizes gross capital flows to and from industrial countries during the
most recent period (1999-2003). The size of gross flows is remarkable, particularly to and
from financial centers such as Switzerland and the United Kingdom, but also to and from
the euro area and, relative to GDP, Scandinavian countries. While net flows are also
substantial, the data suggest that portfolio diversification, rather than intertemporal
borrowing and lending, is the dominant motive for international asset transactions among
industrial countries.
The data in Table 2 also confirm the importance of valuation effects, in addition to
gross flows, in explaining the dynamics of external assets and liabilities. For example,
external liabilities (and, to a lesser extent, assets) in the United States increased by
substantially less than the underlying flows. The primary reason was the decline in U.S.
stock market valuations during this period, which reduced the value of both foreign equity
and FDI holdings in the United States. The smaller decline in stock market valuations
(measured in U.S. dollars) in other countries during this period helps explain the smaller
capital losses incurred by U.S. investors on their foreign equity holdings.5
Only a few countries in our sample (including the United States) measure FDI at market value: hence, stock
market fluctuations have a less dramatic impact on FDI stocks compared to portfolio equity holdings. Of
course, the fall in foreign portfolio equity assets and liabilities relative to GDP does not imply a decline
relative to total domestic equity holdings.
5
The difference in stock market performance between the United States and world markets is entirely
accounted for by the year 2003, during which the sharp depreciation of the U.S. dollar raised foreign stock
returns measured in dollars. Between end-1998 and end-2002 the decline in stock market valuations in the
United States and world markets was similar.
97
Net Flows and Net Positions, Emerging Markets

We focus on a sample of 21 emerging markets (listed in the Appendix). Figure 3 plots the
evolution of the average current account balance as a ratio of GDP in our emerging
markets' sample. The key cycles in capital flows to emerging markets stand out clearly
from this picture: the deterioration of current account imbalances in the late 1970s until the
debt crisis, their sharp reversal during the remainder of the 1980s, the increase in
imbalances during the early 1990s, and the new reversal following the 1994-95 Mexican
crisis and especially the Asian crisis. Indeed, both the average and aggregate current
account position of the emerging countries in our sample turned positive in 1998 and
increased further in recent years.
The dynamics of the net external position, expressed as a ratio of GDP, are plotted
in Figure 4. It shows the deterioration caused by the debt crisis and its aftermath, a
subsequent sharp improvement, the stabilization of the net external position from 1990 to
1996, the deterioration caused by the sharp declines in GDP and real exchange
depreciation characterizing the Asian crisis, and the subsequent improvement associated
with current account surpluses and strengthening currencies in Asia. In the data, there is no
evidence of an increased dispersion in current account balances across the countries in our
sample (that is, there is a significant common trend in the net capital flows to this emerging
market group), while the dispersion of the underlying net external positions has increased.
Figure 5 plots the net foreign asset position, scaled by GDP, in relation to GDP per
capita in current US dollars at end-2002. While there is still a positive relation between net
foreign assets and GDP per capita, this relation is much weaker than for industrial
countries. Indeed, creditors include economies with high GDP per capita, such as Taiwan
province of China, but also economies with much lower GDP per capita, such as Russia
and Venezuela.6
Table 3 shows the size of net capital flows among some countries in the sample,
both in absolute terms and as a ratio of GDP, during 1999-2003. The table shows a
number of Latin American and Central European countries as the largest net recipients of
net capital flows, although Turkey and Argentina experienced net outflows if IMF and
"exceptional" financing are netted out. 7As the data on current account dynamics suggest, a
number of emerging marketsparticularly Asian countries, together with Russiahave
been on average net capital exporters to a substantial degree. For example, Thailand's
cumulative net outflows over the 5-year period total over 30 percent of its 2003 GDP.
Gross Flows and Gross Positions, Emerging Markets
Figure 6 provides a longer-term perspective on the size of external assets and liabilities in
these emerging markets. Both assets and liabilities have increased substantially as a ratio
of GDP during the past 20 years. However, there is virtually no increase in average
external liabilities when scaled by exports, rather than GDP, while the trend increase in
external assets is still visible. This stands in contrast with the evidence for the advanced
6
In addition to the standard macroeconomic drivers of net foreign asset positions that were emphasized by
Lane and Milesi-Ferretti (2002a), political risk and natural resource endowments are other variables that may
be important, especially in reference to the countries listed here.
7
See IMF (1993) for a description of balance-of-payments transactions classified as exceptional financing.
98
economies, where the increase, especially since the mid-1990s, is very strong even as a
share of exports.
As noted earlier, an interesting question is whether the composition of external
assets and liabilities has changed over time. Table 4 provides evidence that highlights the
increased relative importance of direct investment and portfolio equity liabilities. The
averages hide substantial heterogeneitycountries such as Chile and the Central European
economies in our sample (Czech Republic, Hungary, and Poland) have external equity
liabilities above 50 percent of GDP, while the levels tend to be lower in Asian economies.
The table also documents the increase in foreign exchange reserves, expressed as a
share of GDP, during the past 20 years. It should be noted, however, that this increase has
gone hand in hand with the increase in other external assets, so that at the end of 2002
reserves in our sample account for the same share of total external assets as in 1982 (over
one third). Direct investment and portfolio equity assets have also increased during the past
two decades, and by 2002 represented around 12 percent of GDP and over 20 percent of
total external assets.
Table 5 characterizes gross capital flows to and from some emerging economies
during the period 1999-2003. The pattern reveals an interesting dichotomy. For a number
of countries, gross flows primarily reflect intertemporal borrowing or lending decisions,
with countries accumulating net assets or net liabilitieseither cumulative inflows or
cumulative outflows are clearly dominant (see also Table 3). Among countries that
accumulated substantial net assets, a number of East Asian countries stand out, particularly
Indonesia, Korea, Malaysia, and Thailand, together with oil-exporting countries such as
Russia. For another group of countries, instead, gross inflows and gross outflows have both
been large and roughly similar in magnitude, reflecting increased financial integration with
the world economy. Examples include Chile and India. China has experienced large
inflows and outflows, but also significant net foreign asset accumulation. Of course,
similar aggregate levels of gross inflows and gross outflows can conceal significant net
imbalances within specific asset categories: for instance, China is a major net recipient of
FDI flows while simultaneously accumulating a significant volume of foreign reserves.
Having provided a broad characterization of the growth in international balance
sheets in recent years for both advanced and emerging economies, we next turn to
providing a simple framework for understanding the underlying drivers.
External Asset Dynamics
In this section, we provide a simple accounting framework that relates the dynamics of net
foreign assets to trade flows, growth, rates of return, and real exchange rates. The goal is to
lay out the various channels by which exchange rates and other macroeconomic
fundamentals can affect the external adjustment process. We then decompose the factors
underlying changes in net foreign assets over the past decade for a set of emerging
markets.
Accounting for External Asset Dynamics
The change in net foreign assets B can be written as the sum of net capital outflows and net
capital gains:
99
(1)
where AFXt is net accumulation of foreign exchange reserves, FA^ is net capital inflows
(excluding reserves), and KG is the net capital gain on the net external position outstanding
(change in stock minus underlying flow). Denote by CA the current account balance, KA
the capital account balance, and EO net errors and omissions.8 Making use of basic
balance of payments identity CA+KA + FA - AFX + EO = 0, we can rewrite (1) as:
(2)
In line with statistical reporting practices, all variables are expressed in US dollars.
Equation (2) can also be expressed as follows:
(3)
where BGST is the balance of trade in goods and services plus net transfers, A and L are
external assets and liabilities, respectively, and i f , if are the nominal yields on these
assets and liabilities. Taking ratios of GDP and indicating such ratios with lower-case
letters, we can express (3) as follows:
(4)
where yt is the growth rate of nominal GDP measured in US dollars. Another way to
re-write the above expression is as follows:
(5)
where g is the economy's real growth rate, n is the rate of inflation (measured with the
GDP deflator), and d is the rate of nominal exchange rate depreciation vis-a-vis the US
dollar. In other words, changes in the net external position can be due to several factors:
1.
2.
3.
4.
The current account;

Net capital gains (measured in US dollars);
The capital account and net errors and omissions;
The effect of exchange rate changes on the past net foreign asset position;9
In balance of payments' statistics (IMF, 1993) the so-called "capital account" measures certain transfers
(such as debt forgiveness); capital flows are recorded in the financial account
9
If external assets and liabilities are all denominated in domestic currency, the capital gain effect will go
exactly in the opposite direction from the exchange rate change effect. Indeed, assume for simplicity that
asset prices in domestic currency do not change. In this case, the capital gain expressed in domestic currency
100
5. the effect of real GDP growth on the past net foreign asset position.
An alternative way to express eauation (5) is in terms of overall rates of return on
external assets and liabilities. Define
as the rate of capital gain on external assets
(liabilities), measured in US dollars, so that
be the real rate of return on foreign assets, measured in US dollars, with an analogous
definition holding for the rate of return on foreign liabilities ftL. In this case we can rewrite (5) as follows:
(6)
Equation (6) shows several factors that can account for the dynamics of net foreign
assets: the adjusted trade balance, the difference between the real rate of return and the
growth rate, adjusted for the bilateral real exchange rate vis-a-vis the US dollar, and
differences in returns between foreign assets and liabilities.
If we express the real rates of return in domestic currency and denote them by
A L
rt ,rt , equation (6) takes the more familiar form:
(7)
This framework delivers several important insights. First, the gap between current
production and current absoiption (i.e. the trade balance) is only one factor in determining
the aggregate evolution of the net foreign asset position: it is vital to also keep track of
valuation and "denominator" effects. Second, as is shown by the third term on the right
hand side (RHS) of equation (6), the difference between the rate of return and the growth
rate, interacted with the inherited net foreign asset position, exerts a potentially powerful
influence on its current dynamics. Third, as captured by the last term on the RHS in
equation (6) , the gross scale of the international balance sheet matters in addition to the
net position: even if the inherited net foreign asset position is zero, the accumulated levels
of gross foreign assets and liabilities will influence the overall dynamics to the extent that
the rates of return differ between the two sides of the international balance sheet.
is zero, but expressed in dollars it becomes :

to the term
similar (with an opposite sign)
in equation (5).
101
The Evolution of Net Foreign Assets in Emerging Markets

In Table 6 we provide a simple decomposition of changes in the ratio of net foreign assets
to GDP between end-1990 and end-2002 for a selection of emerging markets in our
sample. The breakdown follows equation (4), so that changes in net foreign assets are
given by the sum of the current account (itself divided into trade balance and investment
income), capital account and errors and omissions, capital gains (including the effects of
exchange rate changes on the net external position), and the effects of growth on net
external assets.
A number of features are worth highlighting:
Despite a cumulative current account in balance or surplus, countries such as

Indonesia and Thailand experienced a deterioration in the ratio of net foreign assets
to GDP. For both countries, this occurred because of 'capital losses'linked to the
real depreciation of their currencies during the period.
On the other side, the Czech Republic's and Mexico's external position
deteriorated by much less than the large cumulative current account deficits would
suggest, thanks to substantial capital gains on their net external positionlinked to
the real appreciation of their currencies between end-1990 and end-2002.
More generally, the tables highlight the need to focus not only on the current account
(which includes the yield on external assets and liabilities), but also on economic growth
and the overall rates of return on the external portfolio in order to understand the evolution
of net foreign assets. Indeed, growth and especially valuation effects can have an impact on
the evolution of the external position that is of the same order of magnitude as trade
imbalances.
Exchange Rates and the Adjustment Process
The framework summarized in equation (6) highlights the potential contribution of shifts
in exchange rates in determining the dynamics of external asset positions. In this section,
we first briefly review the "traditional" channels by which exchange rates influence the
adjustment process, before focusing on the valuation channel (i.e. the impact of the
exchange rate on the rates of return earned on holdings of foreign assets and liabilities).
Exchange Rates, the Trade Balance and Real Output
The inter-connection between the exchange rate and the trade balance is among the
most-studied questions in international economics, in both academic and policy circles.
From a long-run perspective, the classical transfer problem postulates that persistent
creditor nations should have more appreciated real exchange rates. The mechanism
underlying the transfer problem hypothesis is that the positive international investment
returns earned by long-run creditors have their counterpart in trade deficits and attendant
real appreciation.
Lane and Milesi-Ferretti (2002b, 2004b) find considerable empirical support for the
transfer problem, for both industrial and developing countries. However, they find the
magnitude of the effect differs with country characteristics such as openness, size and the
102
level of development. In relation to financial globalization, important findings are that the
transfer problem is smaller in the absence of current and capital account restrictions and
that equity financing reduces the size of the transfer effect relative to debt financing.
At a shorter horizon, the interplay between the exchange rate and the trade balance
is complex and less well understood. In particular, the cyclical correlation between the
variables will depend on the nature of the shocks hitting the economy, with nominal, fiscal
and real shocks generating different co-movement patterns between the variables.
However, in policy terms, there is a broad consensus that exchange rate depreciation is
typically required if the objective is to engineer an improvement in the trade balance.
Empirical studies of the elasticities of trade volumes to exchange rates and income levels
provide extensive support for this proposition (Hooper et al 2000). Again, the pace of
financial globalization and "real" globalization (in terms of product market integration)
will influence these key elasticities. For instance, the scale of exchange rate adjustment is
eased, as foreign goods become better substitutes for domestic goods. In terms of financial
globalization, wider trade imbalances are more feasible, the more diversified are
international portfolios.
In tracking the dynamics of the ratio of net foreign assets to GDP, real exchange
rates also operate by determining the real value of domestic output in terms of international
price comparisons. For instance, if variables are measured in US dollars, a foreign asset
that is constant in real dollar terms will shrink relative to the constant-dollar value of GDP
if real appreciation vis-a-vis the US dollar occurs. This "denominator" effect is highlighted
in equation (8) in the previous section and is powerful channel by which the real exchange
rate may influence the dynamics of the NFA/GDP ratio.
However, in addition to these well-known channels, exchange rates also potentially
influence the dynamics of international asset holdings through influencing the rates of
return on foreign assets and liabilities.10 We focus on this valuation channel in the rest of
this section.
The Valuation Channel: A Conceptual Framework
As outlined earlier in the paper, the dynamics of net foreign assets depend not only on the
trade balance but also on the rates of return earned on accumulated foreign assets and paid
out on foreign liabilities. In domestic-currency real terms, the net impact is given by
(8)
where the stocks of foreign assets and liabilities A and L are now expressed in domestic
currency. Since these positions are predetermined from a time-/ perspective, the net
valuation impact of a change in the real exchange rate is given by
(9)
10
Clearly, in tracking a ratio, there is some discretion in terms of attributing the impact of an exchange rate
change to the numerator or the denominator via the choice of the reference currency. In the next subsection,
we look at the levels of foreign assets and liabilities in terms of real domestic currency.
103
It is clear from this expression that exchange rate changes can have a non-zero
valuation impact even if the initial net foreign asset position is balanced, so long as the
rates of return on foreign assets and liabilities are differentially affected by a shift in the
exchange rate.11 The magnitude of the valuation channel is directly increasing in the gross
scale of the international balance sheet: the relevance of this channel for aggregate net
foreign asset dynamics is growing in line with the spectacular accumulation of gross
foreign asset and liability holdings in recent years. Relatedly, the valuation channel also
depends on the composition of the international balance sheet, since the sensitivity of
returns to exchange rates will vary across investment categories and will also depend on
the currency composition of foreign assets and liabilities (and on the extent of hedging).
Of course, even if the exchange rate does indeed have a valuation impact, it does
not mean that the net foreign asset position will move one-for-one. First, exchange rate
changes also have a direct impact on the trade balance. Second, a valuation gain represents
a positive wealth effect that will plausibly raise consumption and investment, leading to a
negative comovement between the net returns term and the trade balance (Lane and MilesiFerretti 2002a, 2002b). Third, from another angle, a sufficiently large negative valuation
effect may lead to a sudden stop in capital flows that forces the trade balance to move into
surplus. Finally, it is important to remember that it is the net valuation effect that matters: a
capital gain on foreign assets may be fully offset by a capital gain on foreign liabilities.12
In addition, equation (9) only captures the contemporaneous impact of a change in
the exchange rate. Some returns may respond to the exchange rate only with a lag (for
instance, the future profitability of FDI positions may be affected by current exchange rate
movements). In addition, current exchange rate movements may lead to a revision of
expectations about future exchange rate changes, which in turn feed into the ex-ante
returns required to hold particular foreign asset and liability positions.
As was discussed earlier in the paper, there are polar cases in which the impact of
exchange rate movements on rates of return is straightforward. For instance, the domestic
rate of return on an unhedged foreign asset that offers a fixed foreign-currency return will
fall one-for-one with the rate of real appreciation: a given foreign-currency return will be
diminished by the fall in the real domestic value of foreign currency. Conversely, the
domestic rate of return on a foreign liability that offers a fixed domestic-currency return
will be unaffected by a shift in the real exchange rate. However, the domestic rate of return
on a foreign liability that offers a fixed foreign-currency return (e.g. foreign currency debt
or domestic debt that offers a dollar-linked rate of return) will also fall in proportion to the
rate of real appreciation.
More generally, the net impact of exchange rate movements on the value of
"Strictly speaking, the impact on the returns on foreign assets and liabilities is not the only "valuation" effect
of exchange rate changes. As highlighted by the debate over the Marshal 1-Lerner condition and reemphasized by the current debate about limited exchange rate pass-through, exchange rate movements also
exert a "pure" valuation effect on the trade balance to the extent that import and export volumes are
unresponsive to exchange rate changes. There are also potential valuation effects even on domesticallyowned assets, but we restrict attention to the cross-border positions through which valuation effects have
asymmetric redistribution effects on home and foreign investors.
12
Cross-border hedging could automatically generate such a positive comovement. However, the extent to
which hedging takes place is unclear: much hedging activity occurs between counterparties of the same
nationality, with no net impact on the national risk profile.
104
holdings that carry a variable market return depends on the nature of the co-movements
between exchange rates, asset prices and profitability (in the case of non-market assets
such as FDI positions and some bank claims). In some cases, the inter-connections
between exchange rates and the determinants of market returns can be quite subtle and
complex and may also depend on the underlying source of an exchange rate shock.
For instance, devaluation may be associated with an increase in the rate of return on
foreign liabilities if it is associated with an increase in the profitability of foreign affiliates
operating in the domestic market or, alternatively, if it engenders an increase in the country
risk premium. On the other hand, a devaluation may be generated by a negative domestic
productivity shock that also lowers the return earned by foreign investors. With respect to
foreign assets, domestic real depreciation may be the result of superior overseas economic
performance that raises the overseas rate of return. However, a negative domestic
productivity shock may also reduce the overseas earnings of domestic multinationals, such
that devaluation is accompanied by a decline in the overseas rate of return.13
In view of the range of possible theoretical scenarios, the strength of the valuation
channel is ultimately an empirical issue. We first consider some case study evidence,
before turning to cross-country quantitative exploration in the subsequent subsection.
Case Studies: the United States and Australia
It is possible to gain some insight into the quantitative importance of the valuation impact
of exchange rate movements for those countries that calculate the accounting
decomposition about the relative importance of capital flows, market value capital gains
and exchange rate capital gains in determining the dynamics of foreign asset and liability
positions. This is possible to for two countries in our sample (the United States and
Australia). Of course, an accounting decomposition does not reveal the complete
contribution of the exchange rate valuation channel, since it does not take into account the
potential indirect impact of the exchange rate on market values or on the revaluation of
investment income flows.
Tables 7 and 8 present the decompositions for the United States and Australia
respectively, showing the average annual relative contributions of each component in
proportion to the inherited stocks of foreign assets (liabilities):
13
A further complication is that exchange rate movements may also affect the international tax planning of
multinational corporations that may affect the distribution of reported earnings in different locations. See
Sullivan (2004) on the trend increase in the shifting of reported profits by U.S. multinationals to overseas
affiliates.
105
In addition, the tables display the standard deviations for these components. Figure
7 and Figure 8 present similar data, but scale the size of capital flows and valuation
changes by GDP, so as to provide a ready reckoning of their macroeconomic impact.
Table 7 and Figure 7 show the statistics for the United States over 1990-2003.
While financial flows have traditionally been the dominant source of balance sheet growth,
the 1996-2003 period saw a much greater role for capital gains. Indeed, the contribution of
market-value capital gains exceeded that of financial flows in the growth of the foreign
asset holdings of the United States during the global stock market boom of 1995-1999,
with a similar contribution to the growth in the value of foreign liabilities. Conversely, and
consistently with the evidence in Tables 1 and 2, the global correction in asset prices
during 2000-2003 saw a decline in the market value of both foreign assets and liabilities.
Table 7 highlights the role of the exchange rate valuation channel. The 5.1 percent
average annual dollar appreciation during 1996-2001 was associated with an annual
average 1.7 percent fall in the value of US foreign assets. In contrast, the sharp dollar
depreciation during 2002-03 was associated with an annual average 5.5 percent increase in
its foreign asset position. This gain helped to offset the impact of the growing current
account deficit on the U.S. net external position. (Throughout, in line with expectations,
the exchange rate channel had a near-zero impact on the stock of US foreign liabilities.) In
terms of relative stability, capital flows have been much less volatile than either of the
capital gain components.
Turning now to the Australian evidence, Table 8 and Figure 8 show that the
exchange rate valuation channel has been more important than in the US case. For
instance, the exchange rate valuation term was a bigger contributor than either financial
flows or market-value capital gains in the growth of foreign assets during the 1997.22001.1 period of real depreciation of the Australian dollar. Although, in contrast to the US
case, Australian foreign liabilities are also sensitive to the exchange rate, this side of the
international balance sheet is only about half as sensitive as the foreign asset position. This
is consistent with a larger role for holdings denominated in domestic currency in foreign
liabilities than in foreign assets.
These case studies of the United States and Australia provide suggestive evidence
about the importance of the valuation channel in driving fluctuations in international asset
holdings. We next turn to regression-based analysis for a broader panel of countries.
Regression Analysis, Industrial Countries
In this section, we analyze the sensitivity of rates of return on foreign assets and liabilities
to movements in trade-weighted multilateral exchange rates. 14In addition to the aggregate
positions, we also examine returns for the separate investment categories (FDI; portfolio
equity; portfolio debt; and other (debt)), since the relation between exchange rate
movements and rates of return should depend on the specific characteristics of each
investment class. Our specification is given by:
14
While the most appropriate real exchange rate measure would be a "finance-weighted" index, reflecting the
relative importance of host or source countries in external holdings, the strong correlation between the
geographical pattern of trade and financial flows ensures that a trade-weighted exchange rate is a reasonable
proxy.
106
(10)
where the dependent variable is the real domestic-currency return on foreign assets in
investment category / and the regressor is the log change in the trade-weighted real
effective exchange rate.15 We run an analogous equation for the rate of return on foreign
liabilities.16
Our primary interest is just in establishing the direction and magnitude of the
contemporaneous co-movement between the exchange rate and rates of return.17 We do not
attempt to distinguish between anticipated and unanticipated changes in the real exchange
rate: however, real exchange rates are largely unpredictable at an annual horizon (at least
for our sample of advanced countries), such that this may be a fairly-innocuous
assumption.18
We begin by examining the rates of return on foreign assets in Table 9.19 The
results for total foreign assets are given in column (1). In all cases, the estimated
coefficient is negative: real appreciation is associated with a fall in the domestic-currency
rate of return earned on foreign assets. For a number of countries, the estimated coefficient
is in fact very close to -1: this one-to-one mapping is consistent with a process by which
the foreign-currency real return on foreign assets is orthogonally determined and the
exchange rate just acts to convert the foreign-currency return into domestic terms.20
The smallest estimated coefficient (in absolute value) in the sample is for the U.S.
at -0.37. This admits a number of interpretations. First, some proportion of U.S. foreign
assets is denominated in dollars and hence their value is not directly affected by exchange
rate movements. Second, dollar appreciation could be associated with an increase in
returns on foreign-currency foreign assets. One example would be a positive productivity
shock in the U.S. that both appreciates the dollar and raises returns in U.S. financial
markets. If foreign financial markets positively co-move with the U.S., foreign-currency
asset returns would also rise at the same time. A positive U.S. productivity shock could
15
The rate of return on foreign assets in year t is measured as the sum of investment income and capital
gains earned in that year, divided by foreign assets at the end of year / -1.
16
Clearly, this is a very parsimonious setup. However, in addition to being suited to our short data span,
capturing the simple bivariate relation is an obvious first step, even if it does not rule out the possibility that
any impact of the exchange rate on the rate of return may just be proxying for the role played by some
omitted variable that commonly influences both the rate of return and the exchange rate or may just reflect
endogeneity bias.
17
We could alternatively present the simple correlations between returns and exchange rate changes. See
Lane and Milesi-Ferretti (2003).
18
We return to the issue of the predictability of exchange rates in the discussion of results.
19
In terms of country selection for the regressions, we only include those with at least thirteen years of data
on rates of return. In addition, we rule out observations that may be contaminated by factors such as revisions
in methodology and other corrections.
20
In some cases, we observe a coefficient above unity, which means that real appreciation is associated with a
fall in foreign-currency returns on foreign assets: the domestic investor "loses twice" by suffering both a low
foreign-currency return and an unfavorable conversion rate back into domestic real terms. Such a pattern
could be generated, for instance, if the domestic business cycle is asymmetric with respect to the
international business cycle: the domestic currency appreciates when international partners are doing badly
(as proxied by poor foreign-currency rates of return). This, of course, is a risk-leveraging pattern of comovement between foreign-currency returns and the domestic real exchange rate.
107
also raise the profits earned overseas by U.S. multinationals, such that foreign-currency
return rises in that case as well.
The results for FDI assets are given in column (2).21 For most countries, FDI
positions are still measured at book value, rather than market value, and therefore the
valuation channel is typically understated. However, currency movements should still
matter, since these would affect factors such as the current replacement cost of capital
goods and fixtures, both domestically and overseas. In column (2), the fixed-effects panel
estimate of the impact of real appreciation on the real return on FDI foreign assets is -0.76.
However, there are some cases in which the coefficient is substantially above unity: for
these countries, real appreciation tends to be associated with low foreign-currency real
returns on FDI assets.
We next examine the returns on portfolio equity assets in column (3). The fixedeffects panel estimate is very close to -1, which is consistent with orthogonal contributions
of exchange rate movements and foreign-currency rates of return to the domestic-currency
real rate of return. Two exceptions to this rule are Germany and Switzerland: for these
countries, real appreciations have coincided on average with periods of strongly negative
world stock market returns, and hence disappointing foreign-currency returns on their
equity portfolios.
Column (4) displays the results for foreign assets in the portfolio debt category.
There is strong covariation between the exchange rate and domestic-currency returns and
the estimated coefficients are consistent with the foreign-currency returns on foreign
portfolio debt assets being exogenously determined with respect to the domestic real
exchange rate. However, the United States coefficient is only -0.65, consistent with the fact
that a considerable proportion of its foreign bond holdings are denominated in US dollars.
Finally, we turn to the "other" investment category in column (5). This category
largely comprises bank lending. Since banks do not "mark to market" all assets and
liabilities but rather carry a high proportion at book value, the rates of returns in this
category will be dominated by the yield component, with capital gains and losses
understated. However, on the assets side, the broad picture is quite similar to that for
portfolio debt Again, an important exception is the US, where the coefficient estimate is
insignificant: again, a good candidate explanation is that a high proportion of its foreign
lending is in US dollars.
We turn to the rates of return on foreign liabilities in Table 10. In terms of the
results for total foreign liabilities in column (1), we see quite a mixed pattern in terms of
the estimated exchange rate coefficients across countries. As was shown also in Table 7,
For the United States, the rate of return paid out on foreign liabilities is totally unaffected
by movements in the real exchange rateconsistent with the fact that foreign liabilities are
almost entirely dollar-denominated and offer returns that are not linked to exchange rate
fluctuations (e.g. bank deposits or fixed-interest debt instruments). At the other extreme,
21
In the sample represented in this table, only the Netherlands and Australia record FDI at market value. The
US reports positions measured at both book and market value. For comparability with the countries that only
report book values, the US estimates in these tables refer only to the book value measure of FDI. However,
for the US, if we use the rate of return based on FDI at market value then the exchange rate coefficient in the
FDI asset equation is -0.71 (t-stat 1.38) and it is 0.15 (t-stat 0.37) in the FDI liability equation.
108
the estimated coefficient for Finland is -1.8.22 The fixed-effects panel estimate is -0.68: this
generally indicates that exchange rate appreciation is associated with some deterioration in
domestic real returns on foreign liabilities.23
With respect to FDI liabilities, column (2) of Table 10 suggests that there is little
covariation between exchange rate fluctuations and real domestic returns. In part, this may
be attributed to the fact that FDI positions are mostly measured at book value but the
insignificance of the exchange rate also suggests that the earnings of foreign affiliates in
the domestic market are not (contemporaneously) affected by exchange rate swings. This
pattern is worth exploring further but would require the availability of higher-quality data.
Similar to the case for FDI liabilities, most of the estimated country coefficients for
portfolio equity liabilities are insignificantly different from zero: the domestic-currency
real return offered by portfolio equity liabilities is not systematically affected by the
exchange rate. Again, this is somewhat surprising to the extent that we might expect
domestic stock market booms to be associated with real appreciation.24
Only Canada and Australia show a significant connection between exchange rate
movements and the rate of return paid out on foreign bond liabilities. For the others, the
results support the caricature of bond liabilities that offer a domestic rate of return that is
invariant to exchange rate fluctuations.25 With respect to other liabilities, a number of
countries display significantly negative coefficients, with the estimates far above unity for
Australia and Spain. For this pair, the pattern is akin to that experienced by emerging
22
Albeit significant only at the 10 percent level. Although a pattern of high real returns plus exchange rate
depreciation is also evident in the early 1990s, the most striking period for Finland is the post-EMU 19992002 period: in 1999 the return on its foreign liabilities was large (driven by gains in Nokia's share price
during the equity market boom), while its real exchange rate depreciated (on account of the fall in the
external value of the euro). In contrast, the stock market reversals of 2001-2002 were accompanied by an
appreciating real exchange rate (as the euro's external value recovered). This case is a vivid illustration of the
importance of co-movements between exchange rates and asset prices. It also underlines that exchange rates
need not always move in a "risk-sharing" manner, which applies a fortiori for members of a currency union
that have little influence on the external value of the currency.
23
In no case is the estimated coefficient significantly positive. This is quite surprising, since some of the
mechanisms discussed earlier in order to explain a negative relation between exchange rate appreciation and
the rate of return on foreign assets should symmetrically imply a positive association between exchange rate
appreciation and the rate of return on foreign liabilities. For instance, a positive domestic productivity shock
might raise profitability of foreign affiliates operating in the domestic market and generally boost domestic
asset prices, while at the same time generating real appreciation.
24
Indeed, there is only one significant country coefficient (Germany), but it is negative and large (-2.9)
negative. This means that declines in the German stock market are typically associated with real appreciation.
Since the point coefficient is fairly similar for both assets and liabilities, this implies that German real
appreciation tends to occur during phases of disappointing global stock returns, since the returns on German
overseas assets fall in addition to the returns paid out on domestic stocks owned by foreign investors.
25
Even for Canada and Australia, the pattern of co movement is negative: real appreciation is associated with
low domestic rates of return on their foreign bond liabilities. In part, this may suggest that exchange rate
movements for these countries have a substantial predictable component, since foreign investors would be
prepared to accept a low domestic-currency return if real appreciation were anticipated. The predictability
hypothesis receives some support from the empirical work of Chen and Rogoff (2003), who show that
"commodity" currencies (such as the Australian and Canadian dollars) are more predictable than other
currencies. Of course, another potential contributory factor is the extent to which these countries issue bond
liabilities in foreign currency.
109
markets: real depreciation is associated with an increase in the foreign-currency return paid
to foreign investors.
In summary, the regression analysis in Tables 9 and 10 delivers a number of
interesting lessons. First, especially on the foreign assets side, exchange rate movements
are an important covariate of rates of return, consistent with the operation of a powerful
valuation channel. Second, real appreciation is typically associated not only with lower real
returns on foreign assets, but also lower real returns on foreign liabilities: at least for small
net positions, this implies that the net valuation impact of exchange rate movements on the
net foreign asset position has been limited. Third, the sensitivity of returns to exchange
rates does vary across investment categories: the composition of the international balance
sheet is an important determinant of the aggregate valuation effect. Fourth, the United
States behaves quite differently to other countries in that the rates of return on its liabilities
(in all investment categories) are unaffected by currency movements. Since dollar
depreciation raises the return on its foreign assets, this means that the valuation channel in
the US case may indeed be a powerful adjustment mechanism in correcting its large
external liability position. We return to the feasibility of this option later in this paper.
Exchange Rates and Rates of Return, Emerging Markets
In general, we would expect the relation between domestic-currency rates of return and
changes in the real exchange rate to be even stronger for emerging markets, which in
general have less scope for borrowing or lending in domestic currency. Careful empirical
work has to face the severe difficulties in measuring such rates of return: among these, the
lack of precise historical data on international investment positions; stock-flow
discrepancies; debt reduction and debt forgiveness agreements, and default episodes.
While aware of these limitations, we have constructed rough estimates of rates of
return on external assets and liabilities for our emerging-market sample. The methodology
is based on estimating the stock of external assets and liabilities (Lane and Milesi-Ferretti,
2004c), and using data on interest payments and capital flows to back out rates of return.
A simple panel regression with fixed effects of real domestic-currency rates of
return on external liabilities on changes in the real effective exchange rate gives a
coefficient of -0.86 with a t-statistic of 19.26 As Figure 9 shows, this relation holds not only
along the time-series dimension, but also in the cross-section. The Figure shows a strong
negative relation (plotted for the year 1997, a year of large exchange rate depreciations in
the Asian countries of our sample) between the domestic currency rate of return on
external liabilities and the real effective exchange rate.
The Valuation Channel in International Macroeconomic Models
The preceding empirical analysis has indicated that revaluations are an important
contributor to net foreign asset dynamics. However, it has been standard in both the
traditional Mundell-Fleming approach and contemporary "new open economy
macroeconomics" to consider scenarios in which the initial net foreign asset position is
zero and the gross scale of international balance sheets is ignored. This rules out any
26
A similar regression for assets gives a coefficient of -1, consistent with the fact that overseas assets are
entirely denominated in foreign currency.
110
consideration of the valuation channel in terms of macroeconomic behavior and the

analysis of alternative policies.
However, two important recent exceptions are provided by Benigno (2001) and
Tille (2004). In a two-country model, Benigno (2001) shows that monetary shocks have
much larger real effects if the initial global steady state is characterized by imbalances in
net external positions, since exchange rate movements generate a net valuation effect that
has asymmetric effects on home and foreign countries. One implication is that countries
will disagree about the optimal monetary policy, since the valuation channel acts to
transfer wealth between home and foreign citizens.
Tille (2004) considers the case of initially-balanced net foreign positions but allows
for different levels of scale in terms of gross holdings of foreign assets and liabilities.
Matching the US data, he shows that an increase in gross cross-holdings of domesticcurrency and foreign-currency bonds means that the welfare impact of a surprise monetary
expansion is greatly magnified in the case that the foreign-currency share of foreign assets
is larger than the foreign-currency share of foreign liabilities. Indeed, his calibration
suggests that the welfare impact of the valuation channel is 350 percent more powerful
than the traditional channel, since devaluation confers a sizeable capital gain on the home
country in his setup.
Clearly, much remains to be done to improve the theoretical treatment of the
valuation channel in macroeconomic models. For instance, it would be highly desirable
(albeit extremely challenging) to incorporate a realistic profile of the international balance
sheet (with its mix of FDI, portfolio equity and debt instruments) and jointly determine the
equilibrium response of real variables, asset prices and exchange rates to various shocks
and policies. In this regard, Hau and Rey (2003) and Pavlova and Rigobon (2003) have
made interesting recent attempts to jointly model financial returns and exchange rates.
Finally, another important research question is to assess the contribution of the
valuation channel to persistent shifts in net foreign assets. That is, the valuation effects
induced by currency and asset price movements can generate volatility in the value of
external assets and liabilities but it is an open question as to how important are valuation
effects versus current account imbalances in driving the lower-frequency component of net
foreign asset dynamics. On the agenda for future research is a better characterization of the
relation between exchange rate movements and net foreign asset positions over time: for
instance, the impact effect may primarily operate through the valuation channel, with a
longer-term contribution via gradual adjustment in the current account to a sustained real
exchange rate movement.
Policy Implications
As was emphasized in discussing equation (6) above, financial globalization increases the
empirical relevance of the valuation channel for exchange rate movements. Improving
quantitative understanding of the valuation channel is obviously desirable, in order to keep
better track of the dynamics of net foreign assets and international wealth effects. One
helpful innovation would be for the relevant national statistical agencies to collect more
information on the role played by currency movements in determining rates of return on
foreign assets and liabilities.
From a policy perspective, does the valuation channel offer a reliable method to
address an excessive net external liability position? Gourinchas and Rey (2004) provide
111
some evidence for the US that historical adjustment in its net foreign asset position has
indeed in part relied on the valuation channel, with the exchange rate responding in a
predictable, systematic manner during phases when its external position was
"unsustainable."
However, there is good reason to be skeptical that the valuation channel can be
relied upon to solve adjustment problems. Even for those countries for which a one-time
surprise devaluation may indeed generate a positive valuation effect that improves the net
foreign asset position, such a move would involve a reputational cost: future investors
would require a larger premium in order to compensate for the risk of subsequent
devaluations. Indeed, such manipulation of the exchange rate creates a classic timeconsistency problem, with the standard recommendation that policymakers take steps to
commit to not using the devaluation option as a form of capital levy. While the severity of
this problem is one of the underlying factors behind the prevalence of liability dollarization
and short-maturity debt among the emerging market nations, it may yet have increasing
bite for major debtors among the advanced nations.27
Moreover, as has been highlighted repeatedly in this paper, it is important to
recognize that the U.S. is a special case, in view of its ability to issue a large proportion of
its liabilities in dollars. This capacity is related to the dollar's status as the default reserve
currency and "safe haven." In turn, the dollar status helps explain the systematic positive
difference between the rate of return gained on U.S. external assets and the one paid out on
its liabilities. Nevertheless, further substantial increases in the U.S. net debtor position
would raise the prospect of a substantial U.S. dollar depreciation, with the associated
capital losses inflicted on U.S. creditors. In turn, this may threaten the special status of the
dollar, also in light of the emergence of the euro as an alternative international reserve
currency, and raise the rate of return required by foreign investors on dollar instruments.
It is interesting to speculate on the trend implications of financial globalization for
exchange rate volatility. Along one dimension, if financial globalization improves
international risk sharing, then more similar wealth dynamics could lead to more correlated
aggregate demand patterns and thereby reduce the need for exchange rate shifts. However,
as has been recently emphasized by Kalemli-Ozcan et al. (2003) and Heathcote and Perri
(2004), greater cross-border risk-sharing could also permit increased specialization in
production, with sectoral shocks under that scenario translating into greater real exchange
rate variability.
Along another dimension, the diversification of risks afforded by financial
globalization may also permit greater dispersion in net foreign asset positions, through a
weakening of the association between external imbalances and country risk premia. If that
is the case and the pace of "real" globalization (i.e. the international integration of product
markets) does not proceed sufficiently quickly, then large-scale real exchange rate
movements may increase in frequency, as part of the adjustment process in coping with
enlarged global imbalances.
27
Alberola (2003) discuss the impact of liability dollarization on the path of exchange-rate adjustment in
emerging markets. He argues that the real exchange rate will tend to overshoot its equilibrium level, due to
the need to foster higher current account surpluses in the aftermath of depreciation to make up for to the
increase in liabilities.
112
In light of these opposing forces, it is difficult to make a firm prediction about the
net impact of ongoing financial globalization on exchange rate volatility. In turn, while
financial globalization shifts the terms of the debate about the relative merits of alternative
exchange rate systems, it does not obviously tilt the balance in one direction or the other in
deciding between floating and fixed regimes.
Of course, the acceleration of financial globalization in the 1990s also had a large
impact on exchange rates, by arguably increasing the prevalence and severity of currency
and financial crises. The policy response to the 1990s series of crises has been to
emphasize the importance of adequate domestic financial regulation, the fragility of
pegged exchange rates and robust fiscal control. However, our emphasis on the roles
played by exchange rates and rates of return in driving net foreign asset dynamics also
raises the question of whether national governments should seek to mould the international
balance sheet in some fashion, either directly or by providing incentives to the private
sector to insure against particular financial vulnerabilities. The rapid growth in official
external reserves in many countries in recent years can be interpreted as one response to
the risks associated with financial globalization. In addition, increased direct investment
and portfolio equity flows can in principle improve risk-sharing by tying rates of return on
external liabilities to domestic macroeconomic conditions. Although the literature is
expanding rapidly, more research on this question is clearly needed.
Concluding Remarks
This paper has been concerned with the macroeconomic implications of financial
globalization. Having established recent patterns in terms of gross and net international
asset trade for both advanced and emerging market economies, we have shown that the
dynamics of net foreign asset positions crucially depend on an array of factors beyond the
value of the trade balance: stocks matter, as well as flows. In particular, we have focused
on the importance of the valuation channel of exchange rate adjustment: currency
fluctuations influence the rates of return on the inherited stocks of foreign assets and
liabilities, in addition to operating through the traditional trade balance channel. In turn,
this raises a set of substantive policy questions about the optimal external capital structure
and the exploitability of the valuation channel as an adjustment mechanism.
An open question is how much further the financial globalization process will go: is
the end point the idealized scenario of "perfect market integration", or will barriers such as
trade costs and imperfect information place a limit on the extent of integration? The impact
of financial globalization on exchange rate behavior and the international adjustment
mechanism is likely to remain near the top of the research and policy agendas. We hope
that much clearer answers can be given at the "100 Years After Bretton Woods"
conference in 2044.
113
References
Alberola, Enrique, (2003), Misalignment, Liabilities Dollarization, and Exchange-Rate
Adjustment in Latin America, Banco de Espafla, Documento de Trabajo 0309.
Benigno, Pierpaolo, (2001), "Price Stability with Imperfect Financial Integration," mimeo,
New York University.
Chen, Yu-Chin, and Kenneth Rogoff, (2003), "Commodity Currencies," Journal of
International Economics, Vol. 60, pp. 133-160.
Gourinchas, Pierre-Olivier, and Helene Rey, (2004), "International Financial Adjustment,"
mimeo, UC-Berkeley and Princeton University.
Hau, Harald, and Helene Rey, (2003), "Exchange Rates, Equity Prices and Capital Flows,"
mimeo, INSEAD and Princeton University.
Heathcote, Jonathan, and Fabrizio Perri, (2004), "Financial Globalization and Real
Regionalization," Journal of Economic Theory, forthcoming.
Hooper, Peter, Karen Johnson, and Jaime Marquez, (2000), "Trade Elasticities for G-7
Countries," Princeton Studies in International Economics No. 87.
International Monetary Fund, (1993), Balance of Payments Manual, 5th edition,
Washington, DC.
Kalemli-Ozcan, Sebnem, Bent Sorensen, and Oved Yosha, (2003), "Risk Sharing and
Industrial Specialization: Regional and International Evidence," American
Economic Review, 93, June 2003, pp. 903-918.
Lane, Philip R., and Gian Maria Milesi-Ferretti, (2001 a), "The External Wealth of Nations:
Measures of Foreign Assets and Liabilities for Industrial and Developing
Countries," Journal of International Economics, 55, pp. 263-294.
, (2001b), "External Capital Structure: Theory and Evidence," in H. Siebert (ed.)
The World's New Financial Landscape: Challenges for Economic Policy, BerlinHeidelberg: Springer-Verlag, pp. 247-284.
, (2002a), "Long-Term Capital Movements," NBER Macroeconomics Annual 16,
pp. 73-116.
, (2002b), "External Wealth, the Trade Balance and the Real Exchange Rate,"
European Economic Review, 46, pp. 1049-1071.
, (2003), "International Financial Integration," International Monetary Fund Staff
Papers, 50(S), pp. 82-113.
, (2004a), "International Investment Patterns," IMF Working Paper 04/134, June.
, (2004b), "The Transfer Problem Revisited: Net Foreign Assets and Real
Exchange Rates," Review of Economics and Statistics 86, November.
, (2004c), "The External Wealth of Nations Mark H: Revised and Extended
Estimates of Foreign Assets and Liabilities, 1970-2003," in progress.
Obstfeld, Maurice, and Alan Taylor, (2004), "Global Capital Markets: Integration, Crisis,
and Growth," Cambridge University Press.
Pavlova, Anna, and Roberto Rigobon, (2003), "Asset Prices and Exchange Rates," mimeo,
MIT.
Sullivan, Martin A., (2004), "Shifting of Profits Offshore costs US Treasury $10 Billion Or
More," Tax Notes, 104, pp. 1477-1481.
114
Tille, Cedric, (2003), "The Impact of Exchange Rate Movements on U.S. Foreign Debt,"
Current Issues in Economics and Finance 9(1), Federal Reserve Bank of New
York.
, (2004), "Financial Integration and the Wealth Effect of Exchange Rate

Fluctuations," mimeo, Federal Reserve Bank of New York.
115
Appendix
Industrial countries sample: Australia, Austria, Belgium, Canada, Denmark, Finland,
France, Germany, Greece, Iceland, Ireland, Italy, Japan, Netherlands, New Zealand,
Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom, United States.
Emerging markets sample: Argentina, Brazil, Chile, Colombia, Mexico, Venezuela,
China, India, Indonesia, Korea, Malaysia, Philippines, Taiwan province of China,
Thailand, Czech Republic, Hungary, Poland, Russia, Israel, South Africa, Turkey.
116
Table 1. Net capital outflows and changes in net external position

Industrial countries, 1999-2003
Net outflows
Pet. of 2003 GDP
Billions US$
Japan
Switzerland
Norway
Canada
Euro Area*
Sweden
Denmark
New Zealand
Australia
United Kingdom
United States
558
167
75
74
40
33
24
-7
-88
-139
-2246
13.0
53.8
34.1
8.5
0.5
10.9
11.4
-9.7
-17.3
-7.8
-20.4
Change in net foreign assets

Billions US$ Pet. of 2003 GDP
452
82
86
-43
-205
9
9
-4
-157
134
-1594
10.5
26.6
38.9
-5.0
-2.5
2.8
4.2
-4.9
-30.8
7.5
-14.5
* Change in the net foreign asset position calculated as the sum of the change in the net
positions of Austria, Belgium, Finland, France, Germany, Greece, Italy, Netherlands,
Portugal, and Spain.
Source: IMF, Balance of Payments Statistics, and Lane and Milesi-Ferretti (2004c).
Table 2. Gross capital flows to and from industrial countries (1999-2003)

Capital inflows
billions US$
United States
4167
Euro Area
United Kingdom
Capital outflows
percent of
2003 GDP
trillions US$
percent of
2003 GDP
billions US$
3250
1922
17
3569
38
44
3609
2387
133
2247
44
125
111
26
Sweden
343
223
212
190
165
34
24
74
Denmark
143
68
510
297
124
223
167
Norway
Japan
New Zealand
121
106
23
55
2.5
29
196
664
15
89
15
Switzerland
Canada
Australia
42
63
Change in foreign
liabilities
79
20
percent of
2003 GDP
Change in foreign
assets
biilionsUSS
percent of
2003 GDP
30
1656
15
...
...
...
2794
468
301
346
191
184
148
151
35
68
63
87
551
258
189
199
155
178
30
37
66
167
147
24
75
3
31
193
253
599
20
91
114
14
26
2660
Source: IMF, Balance of Payments Statistics and Lane and Milesi-Ferretti (2004c).
Table 3. Cumulative net capital flows to selected emerging markets, 1999-2003

Total
billions US$
74.8
70.3
37.9
22.1
16.7
14.8
14.1
Brazil
Mexico
Poland
Hungary
Czech Republic
Turkey
Argentina
excluding IMF and

exceptional financing
Percent of
billions
Percent of
2003 GDP
US$
2003 GDP
"Borrowers"
15.2
11.2
18.1
26.7
18.7
6.2
11.1
52.9
78.2
37.9
22.1
16.7
-6.4
-21.2
10.7
12.5
18.1
26.7
18.7
-2.7
-16.7
-131.5
-118.4
-62.5
-40.8
-44.8
-33.5
-27.7
9.3
27.3
21.8
6.7
31.3
16.1
26.7
"Lenders"
China
Russia
Taiwan pr. of China
Korea
Thailand
Indonesia
Malaysia
-131.5
-117.4
-62.5
-56.7
-44.1
-33.3
-27.7
9.3
27.1
21.8
9.4
30.8
16.0
26.7
Source: authors' calculations based on IMF, Balance of Payments Statistics.
119
Table 4. Indicators of International Financial Integration, Emerging Markets

(percent of GDP)
1982
1992
2002
Average net external position
-32.6
-24.5
-26.2
Average external assets

of which:
foreign exchange reserves
FDI + portfolio equity
15.8
24.9
55.5
5.7
1.3
10.9
3.0
19.61
12.1
Average external liabilities

of which:
FDI + portfolio equity
48.4
49.4
81.7
7.4
11.1
32.9
Source: authors5 calculations based on Lane and Milesi-Ferretti (2004c) and IMF, Balance of
Payments Statistics
120
Table 5. Gross capital flows to and from selected emerging markets, 1999-2003
Capital inflows
Total
Excluding IMF and except, fin.
billions
Percentage
billions US$
Percentage of
US$
2003 GDP
of 2003 GDP
Total
Capital outflows
billions US$
Percentage of
2003 GDP
FX reserves
billions Percentage of
US$ 2003 GDP
18.3
259.1
18.3
390.6
27.7
258.6
18.3
24.4
69.9
24.4
132.4
10.9
81.7
13.5
122.5
72.1
91.9
25.2
15.2
63.7
110.4
80.3
86.7
63.5
88.4
88.2
86.5
18.0
3.6
2.1
28.0
4.9
0.4
4.5
26.2
5.7
50.7
54.8
Chile
54.8
32.9
34.1
34.1
14.1
26.2
5.7
50.7
82.1
35.6
10.0
46.3
20.2
111.7
7.2
1.6
16.9
37.9
29.0
8.1
6.5
43.0
2.2
41.0
-1.2
1.1
7.1
-1.8
Indonesia
Thailand
-21.6
-32.2
-10.4
-22.5
-21.8
-32.9
-10.5
-23.0
11.7
11.9
5.6
8.3
10.8
8.6
5.2
6.0
China
Taiwan prov. of China
Korea
Philippines
Brazil
Mexico
Poland
India
259.1
69.9
65.7
22.4
12.8
33.3
Source: authors1 calculations based on IMF, Balance of Payments Statistics and national sources.
W~V~~WWVWI~VM~~VVV~M*~J
Table 6. Decomposition of change in foreign assets for selected emerging markets, 1991-2002
(percent of GDP)
-,,_
.
?ange.ne
roreign assets
0 1 * .
'
Cumulative current account
Cumulative
trartp.
Brazil
Czech Republic
Indonesia
Mexico
Thailand
Turkey
-30.6
-29.4
-6.1
-8.8
-10.0
-21.3
Cumulative
halanre.
2.8
-23.5
59.3
-6.2
31.0
14.2
-30.6
-16.2
-53.3
-36.3
-31.8
-23.1
t
. ,
Cumul.
capital
acct + etrcls&
omissions
Other factors
Growth
K-gainsetc
investm Tnrntnfi
-0.6
-0.2
0.8
-2.9
-6.7
-0.1
effent
10.9
-0.1
26.1
17.7
19.4
11.6
Note: the breakdown of changes in the net foreign asset position reflects equation (5) in the text.
Source: IMF, International Financial Statistics, and Lane and Milesi-Ferretti (2004c).
_
,
Perc. change in
realeff
exchange rate
-13.1
10.5
-39.0
19.0
-21.9
-23.9
-47.9
35.8
-16.1
32.5
-18.2
-2.9
Table 7. US: Relative Contributions of Flows, Market Values and Exchange Rates to
Dynamics of the International Balance Sheet, 1990-2003
Mean
1990-95 1996-2001 2002-03
Standard Deviation
1990-2003
CON FLOW_FA
CON_FLOW_FL
0.056
0.081
0.081
0.108
0.034
0.082
0.030
0.029
CON MV FA
CON_MV_FL
0.031
0.022
0.038
0.034
-0.016
-0.021
0.088
0.066
CON_ER_FA
CON_ER_FL
0.008
0.000
-0.017
-0.003
0.055
0.005
0.030
0.004
Source: US Bureau of Economic Analysis (BEA). We thank Cedric Tille for kindly sharing
in electronic form his history of the BEA data releases.
Table 8. Australia: Relative Contributions of Flows, Market Values and Exchange
Rates to Dynamics of the International Balance Sheet, 1988.3-2004.2
1988.31993.3
Mean
1993.41997.21997.1
2001.1
CON FLOW_FA
CON]_FLOW_FL
0.090
0.017
0.078
0.040
CON MV_FA
CON__MV_FL
0.017
0.014
CON ER_FA
CON] ER_FL
0.020
0.014
Standard deviation
2001.22004.2
1988.3-2004.2
0.081
0.061
0.090
0.016
0.072
0.036
0.040
0.005
0.061
0.024
0.016
0.007
0.128
0.052
-0.023
-0.013
0.083
0.031
-0.022
-0.012
0.11
0.047
Source: Authors' calculations based on data from the Australian Bureau of Statistics.
Table 9. Exchange Rates and Rates of Return on Foreign Assets

I
1
iI
1 USA
(1)
Total
(2)
FDI
-0.37***
-0.57***
~
-1.24
| UK
-1.01***
-0.83***
| Austria
-1.34**
-3.85**
France
(3)
Port Eq
(4)
Port Debt
(5)
Other
-0.65***
-0.11
-0.74
-1.01***
-0.96***
-2.49***
-1.16***
-0.45***
-0.61
-0.37
-0.42
-0.36
Germany
-0.88***
-1.04***
| Italy
-1.08***
-1.17***
| Netherlands
-0.38
-0.33
-0.74*
-0.63**
Sweden
j
! Switzerland
-1.07***
-0.62**
Canada
-0.66***
-0.46*
Finland
-1.09***
-1.07**
Iceland
-0.85
-0.5
Spain
-0.69***
-1.62***
-0.55
-1.62***
-0.61**
I Australia
-0.57***
-0.65**
-0.55*
-0.89***
-0.41*
I Panel
-0.78***
-0.76***
-0.97***
-0.89***
-0.63***
-2.43***
-0.8**
-0.93***
-0.97***
-0.86***
i
Note: Beta coefficients from regression of rate of return on real appreciation. ***^**5* denote
significance at the 1,5 and 10 percent levels respectively. OLS with robust standard errors.
Panel estimation includes country fixed effects (not reported). Full regression results
available from the authors upon request. Data availability varies by country, within
1980-2003 span.
Source: authors' calculations based on IMF, Balance of Payments Statistics, and Lane and
Milesi-Ferretti (2004c).
124
Table 10. Exchange Rates and Rates of Return on Foreign Liabilities
(lj
Total
(2)
FDI
(3)
PortJEq
(4)
PortJDebt
(5)
Other
USA
0.014
0.08
0.36
-0.26
0.05
UK
-0.92***
-0.1
-0.52
-0.46*
-0.78***
Austria
-0.95
-0.35
-2.9**
-0.45
-0.15
-0.48
-0.03
-0.20
-0.57
-0.77
-0.84***
-0.81***
-0.67***
France
0.85
Germany
-0.49**
-1.04
Italy
-0.71***
-0.14
Netherlands
-0.12
-0.33*
Sweden
-0.77***
-0.21
Switzerland
-0.73***
-0.14
Canada
-0.52***
0.06
| Finland
-1.79*
0.91
Iceland
-1.33***
-1.43
Spain
-0.69***
0.15
-1.28
-0.9
-1.52**
Australia
-0.31***
0.02
-0.44
-0.64***
-1.68***
Panel
-0.68***
-0.09
-0.56*
-0.60**
-0.79***
Note: Beta coefficients from regression of rate of return on real appreciation. ***9**5* denote
significance at the 1, 5 and 10 percent levels respectively. OLS with robust standard errors.
Panel estimation includes country fixed effects (not reported). Full regression results
available from the authors upon request. Data availability varies by country, within
1980-2003 span.
Source: authors' calculations based on IMF, Balance of Payments Statistics, and Lane and
125
Figure 1. Net foreign asset position (ratio of GDP) and GDP per capita
Industrial countries, 2003
Sources: IMF, World Economic Outlook (GDP per capita) and Lane and Milesi-Ferretti
(2004c) (net foreign assets).
126
Figure 2. Composition of international portfolio, industrial countries

(sum of assets and liabilities as a ratio of GDP, 1980-2003)
Note: Chart plots the sum of aggregate equity, FDI, and debt assets and liabilities as a share of
aggregate GDP for a sample of industrial countries including: Australia, Austria, Belgium, Canada,
Denmark, Finland, France, Germany, Iceland, Italy, Japan, Netherlands, Norway, Spain, Sweden,
Switzerland, United Kingdom, United States. The sample choice is dictated by data availability.
Source: Lane and Milesi-Ferretti (2004c).
127
Figure 3. Average and aggregate current account to GDP ratio

Emerging markets sample, 1970-2003
Figure 4. Average & aggregate net external position, emerging market sample, 1982-2003*
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
Source: authors calculations based on IMF, Balance of Payments Statistics and Lane and
128
2002
Figure 5. Net foreign assets and GDP per capita

Emerging markets sample, 2002
Sources: IMF, World Economic Outlook (GDP per capita), and Lane and Milesi-Ferretti
(2004c) (net foreign assets).
129
Figure 6. Indicators of international financial integration, emerging markets
1982
1984
1986
1988
1990
1992
1994
1996
1998
Source: Authors' calculations based on Lane and Milesi-Ferretti (2004c).
130
2000
2002
Figure 7. United States: Components of change in external assets and liabilities, 19902003
Source: Authors' calculations based on Tille (2003) and updated data provided by Cedric Tille.
131
Figure 8. Australia: Components of change in external assets and liabilities, 1989-2003
Source: authors1 calculations based on Australian National Statistics
132
Figure 9. Real rate of return on external liabilities and changes in real exchange rate
Emerging market sample, 1997
Note: the real domestic currency rate of return on external liabilities is constructed as the sum of the
yield (interest payments in 1997 divided by the stock of liabilities at end-1996) and the capital gain
rate (change in stock of external liabilities between 1997 and 1996 minus flow, divided by stock of
external liabilities at end-1996). The change in the real exchange rate is the percentage change in the
CPI-based real effective exchange rate between end-1997 and end-1996.
Source: authors9 calculations based on IMF, Balance of Payments Statistics, International Financial
Statistics, and Lane and Milesi-Ferretti (2004c).
133
What Makes Balance Sheet Effects

Detrimental for the Country Risk Premium?*
Juan Carlos Berganza

Alicia Garcia Herrero
Banco de Espafia
*This paper is a draft as of May 2004. Correspondence may be directed to the authors at
jcberganza@bde.es or alicia.garcia-herrero@bde.esY The opinions expressed are those of the authors'
and not necessarily those of Banco de Espana. The authors would like to thank Lucia Cuadro for
excellent research assistance, and Raquel Carrasco, Eduardo Levy-Yeyati, Juan Manuel Ruiz and Jose
Vinals for their useful comments on a previous draft of the paper. Remaining errors in this preliminary
draft are only ours.
Introduction
During the second half of the 1990s emerging countries have experienced very large
swings in the external cost of capital as well as several financial crises, with a large
impact on economic growth. For this reason, academics and practitioners interested in
emerging economies are paying increasing attention to the determinants of a country's
risk premium. An important one is the real exchange rate, since it is particularly volatile
in emerging regions, as compared to industrial ones. Besides, there is a strand of
literature exploring the direct link between real exchange fluctuations and economic
performance, which can serve as a basis to analyze the relation between the real
exchange rate and the risk premium.
Conventional open economy models, and in particular the influential MundellFleming, argue that real depreciations have an expansionary effect by switching global
demand towards domestic production. Already in 1986, Edwards (1986) challenges this
view on several grounds: the possible contractionary effect of a higher price level after
a devaluation as well as a potential negative impact on income distribution. He also
finds some evidence of a small contractionary effect for a sample of 12 developing
countries. More recently, theories based on what has started to be known as the open
economy Bernanke-Gertler-Gilchrist financial accelerator, have challenged the
Mundell-Fleming view. If a country's debt is denominated in foreign currency, a real
depreciation will reduce the country's net worth through a balance sheet effect and, in
the presence of financial imperfections, may increase the cost of capital. This is
particularly relevant for emerging economies given their relatively large share of
foreign currency denominated debt, the frequency of large real depreciations and the
presence of financial imperfections.
In an earlier work, Berganza, Chang and Garcia Herrero (2003) develop a
simple theoretical framework to understand the relation between balance sheets stemming from the increase in the external debt service after a real depreciation - and a
country's risk premium and find evidence a positive relation between the two. This
could have several policy implications, such as the need to reduce foreign currency
indebtness and/or limit, to the extent possible, financial imperfections. It could also
have implications for the choice of the exchange rate regime since avoiding real
exchange rate depreciations becomes crucial for a country's cost of credit.
Given the relevance of the matter, it seems worthwhile investigating the issue
further. In particular, we would like to understand why - and under which
circumstances- balance sheet effects increase a country's cost of borrowing. Among
these questions we shall study: (i) Whether real exchange depreciations are detrimental
for country risk; and to what extent and under which circumstances this is the case, (ii)
Whether real exchange appreciations are beneficial, (iii) Which are the channels of
influence of a real depreciation on country risk; in particular, whether "domestic"
balance sheets, stemming from the increase in domestic foreign currency denominated
debt after a real depreciation are as important as "external" balance sheet effects, (iv)
What is the role of competitiveness, as the most important channel in the traditional
literature of the expansionary effects of real depreciations, (iv) Whether balance sheet
effects are influenced by the existence of financial imperfections, as one would expect
from the financial accelerator literature. And, finally (iv) Whether the exchange rate
regime plays a role in how balance sheets affect country risk, beyond the extent of real
depreciation.
Investigating these issues will help us delimit the extent to which emerging
countries should worry about real depreciations, depending on their own characteristics.
137
In the same vein, it should contribute to identifying which are the most appropriate
policy actions to minimize this problem.
Review of the Literature
Most theoretical models on the impact of balance sheet effects draw from the open
economy version of the financial accelerator, developed by Gertler, Gilchrist and
Natalucci (2003). They generally show that balance sheet effects, related to a sudden
reduction in net wealth, are detrimental either in terms of the cost of capital or output.
However, this result hinges on the existence of financial imperfections. Given these
conditions, the ultimate answer to the question of whether balance sheet effects are
detrimental and when will be an empirical one.
To our knowledge the only work which deals with this issue at the macro level
is that of Berganza, Chang and Garcia-Herrero (2003), who find that balance sheet
effects -stemming from the increase in the external debt service after a real
depreciation - raise a country's risk premium for emerging economies. As for firmlevel data, Forbes (2002) analyzes the impact of 12 major depreciations on a sample of
emerging countries' large firms and finds no significant balance sheet effects on
performance although firms with higher debt ratios tend to show lower net income
growth. It should be noted, though, that Forbes does not take into account the currency
composition of debt. In the same vein, Bleakley and Cowan (2002) show evidence that
the competitiveness effect associated with exchange rate depreciations offsets the
potential contractive balance sheet effect on investment for a panel of Latin American
firms. The authors, therefore, conclude that there is no severe currency mismatch of
output and liabilities in their sample. This optimistic result should, however, be taken
cautiously, since no country fixed effects are considered and Brazilian firms account
for half of the observations. In fact, when each country is analyzed separately, always
with firm-level data, there is evidence of detrimental balance sheet effects on
investment in some countries (namely, Colombia, Mexico and Peru) but not in others
(Brazil, Chile)1. Furthermore, a macroeconomic empirical analysis, such as ours, may
offer a more pessimistic picture of balance sheet effects in as far as it is not only the
tradable sector which is considered but the whole economy. This has fewer possibilities
to hedge its negative wealth in foreign currency than the group of large firms
considered in the firm-level empirical studies.
Objective of the Paper
The objective of this paper is to investigate, at the aggregate level, whether and in
which way real exchange rate depreciations increase a country's risk premium, with
particular attention to balance sheet effects. To this end, a number of questions are
analyzed.
The first is whether an exchange rate depreciation increases a country's
risk premium and under which circumstances this is the case. In principle, this
should happen if balance sheet effects more than counterweigh the expected increase in
competitiveness associated with a real depreciation. The question is why it is so for
some countries and not for others. Identifying these differences is not an easy task but
certainly interesting for policy makers, so as to know to which extent they should worry
about real depreciations.
âlindo, Panizza and Schiantarelli (2003) offer a survey of the results.
138
A second interesting question is whether the impact of real exchange rate

depreciations and appreciations is symmetric. An asymmetry - whereby
appreciations had no significant impact - would make the volatility of the real
exchange rate more of a cause of concern for policy makers since there would be no
instance to benefit from it (i.e., from appreciations). Financial accelerator theories
argue in favor of an asymmetric effect of changes in net wealth since agency problems
may only be binding when the debtor's situation worsens (Bemanke and Gertler, 1989).
Another reason for such an asymmetry could be drawn from the literature on liquidity
constraints, which should only be relevant when a sudden increase in indebtness occurs
and not when there is a net worth gain. A related question is whether the extent of a real
depreciation affects the risk premium more than proportionally; that is, if its impact is
non-linear. If the answer is yes, this may have a bearing on the choice of the exchange
rate regime since there may be no need to worry about small depreciations but only
about large events. Such non linearity could be expected on the basis of the same
arguments as before since large changes in net worth should make financial and
liquidity problems much more binding than relatively smaller ones.
The third question relates to the channels through which real exchange
depreciations affect the risk premium. The most well known channel, the gain in
competitiveness, should reduce the risk premium the more open a country is to trade.
The other crucial channel is that of balance sheet effects, stemming from a sudden
reduction in net financial wealth. In the case of emerging countries, it seems safe to
think of negative net financial wealth because of the generally large stock of debt that
they have accumulated. In the financial accelerator literature, however, balance sheet
effects hinge on the existence of financial imperfections, which we also need to test for.
One interesting issue for policy makers is whether all balance sheets are the same; in
other words, whether an increase in the stock of foreign currency-denominated debt
held by non residents ("external" balance sheets) can have the same detrimental effect
on the risk premium as an increase in the stock of foreign currency-denominated debt
held by residents ("domestic" balance sheet effects). If the former were larger, this
would be an argument in favor of increasing a country's domestic indebtness, even if in
foreign currency, as compared to external indebtness.2 The rationale behind a lower
cost of domestic balance sheet effects may be that having residents as holders of a
country's dollar liabilities, these will benefit from a real depreciation compensating, at
least partially, the loss of wealth of the borrowers. In other words, the real depreciation
will have distributional effects but will not necessarily reduce net financial wealth, as
for external balance sheets. The extent of the wealth effects of domestic balance sheets
may depend on what resident creditors do with their wealth gain. If they are uncertain
about repayment and/or the economic situation deteriorates sharply, they may opt for
capital flight, eliminating the positive impact of the wealth gain on domestic spending
or investment. The extent to which they reinvest their additional wealth may actually
hinge on the existence of financial imperfections.
The fourth question relates to the existence of financial imperfections, a
crucial condition for balance sheet effects to be relevant in the financial accelerator
theories. Given that our sample is composed of emerging countries, one could argue
that they all suffer from financial imperfections. However, the degree to which this is
the case varies from country to country. This is why it seems worth testing whether the
countries with larger imperfections are also those who suffer from larger balance sheet
2
This, however, might not be in the range of options available to policymakers if domestic savings are
very low.
139
effects. In addition, the role of financial imperfections could be different for domestic
and external balance sheets. On the one hand, one could argue that external creditors
are less affected by financial imperfections if the external debt is issued outside the
country, but it is also true that the sovereign debtors have the power to change the rules
of the game even in this case. In addition domestic creditors may be better informed of
their rights, or possible changes in their rights.
The fifth and final issue is the role of the exchange rate regime on how
balance sheets affect country risk, beyond the extent of exchange rate change.
Several authors have developed this idea theoretically but no empirical test exists yet.
Based on the financial accelerator literature, Gertler, Gilchrist and Natalucci (2003)
argue that fixed exchange regimes amplify balance sheet effects because they force the
central bank to adjust interest rates in a manner that enhances financial distress.
Cespedes, Chang and Velasco (2000) show that flexible exchange rates play an
insulating role in the presence of real external shocks so that the output and investment
fall by less than under fixed exchange regimes. The channel is the higher expected real
depreciation under a pegged regime, and thereby the increase in interest rates, since
policy makers will tend to maintain the exchange rate regime during a relatively long
period so as to minimize the size of the change in the relative prices. Another idea for
pegged exchange rate regimes to be detrimental for financial fragility is that agents tend
to feel more protected from exchange rate risk and do not hedge against it (Burnside,
Eichenbaum and Rebelo (2001)). In this line, Ize and Levy-Yeyati (2003) and Broda
and Levy-Yeyati (2003) argue that a pegged exchange regime may induce dollardenominated indebtness, and financial dollarization in general, because it can be taken
as an implicit insurance by the private sector, as well as a demonstration effect from the
part of the government that the exchange rate regime is credible and will be
maintained.3 On the other hand, Elekdag and Tchakarov (2004) show that fixed regimes
can be superior for countries with a high level of indebtness and whose monetary policy
is constrained. This is, therefore, a question worth tackling empirically. We interact
each country's exchange rate regime with external and domestic balance sheets, and
test whether their detrimental effect on the risk premium is larger for fixed regimes.
Both de-jure and de-facto regime classifications are used.
Data Issues and Empirical Strategy
The focus on the country as a whole and, thus, the use of macroeconomic data
substantially limits the number of observations for this study. This is even more the
case given the difficulties of proxying our dependent variable, country risk. The most
widely used proxy in the literature are the returns implicit in the Emerging Markets
Bond Indices (Embi) provided by JPMorgan, after having subtracted total returns of US
treasury bonds4 (from now onwards this variable shall be named Embi). Appendix II
offers details on variable definitions and data sources. The choice of the Embi, together
with the condition we impose that at least four observations of Embi returns exist,
limits our sample to 27 emerging economies and to the period 1993 to 2002 for most
3
Although this idea cannot be fully tested with the data available, Galiani, Levy Yeyati and Schargrodsky
(2003) find indirect evidence that the currency board acted as an implicit insurance for the case of
Argentina.
4
It should be noted that Embi spreads reflect sovereign risk while our objective is broader: country risk in
general since we do not concentrate on public debt only but in all debt denominated in foreign currency,
be it public or private. In any event, the Embi spread continues to be the best available proxy as
sovereign spreads are generally a floor for private sector country risk.
140
countries (for some countries the timeframe is even shorter). This yields an unbalanced
panel with a total of 210 annual observations (Table 1 in Appendix I).
The geographical distribution of the observations among regions can be found
in Table 2 in Appendix I. All major emerging regions are represented although Latin
America is overweighted (with 9 countries and 71 of the observations) and the Middle
East is underweighted.
Apart from the dependent variable (Embi\ the focus of this study is the change
in the real exchange rate. Two different measures are calculated: The first is relevant
for foreign currency indebtness, namely the bilateral nominal exchange rate against the
US dollar adjusted by the domestic inflation (Real Exchange Rate Change). We use the
bilateral exchange rate since we assume that all foreign currency debt is denominated in
US dollar. This is a relatively safe assumption for the countries in our sample. The
second measure is relevant for competitiveness, namely the effective real exchange rate
against the major trading partners (Multilateral Real Exchange Rate Change).
The other crucial concept is that of balance sheet effects, which stem from a
reduction in financial net wealth after a real depreciation. In emerging countries we can
safely assume that financial wealth is negative and corresponds with the increase in the
stock of foreign currency-denominated debt. In other words, although we use a concept
of gross (negative) financial wealth, net financial wealth is bound to be negative,
although probably smaller. The main difference probably lies in the size of
international reserves, which we shall include as a robustness exercise. Our results do
not change. Another interesting issue is whether what matters to measure balance sheet
effects is the change in the stock of debt, because of the depreciation, or the change in
the amount a country needs to pay on that year (the debt service). We shall use the
stock of debt as first option, since it is more in line with the concept of net wealth in the
financial accelerator literature, but robustness test will be conducted with the debt
service. The results do not change.
We differentiate between domestic and external balance sheet effects. External
Balance Sheets are composed by the foreign-currency denominated debt held by non
residents at the end of the previous period (External Debt_l) multiplied by the Real
Exchange Rate Change. In turn, Domestic Balance Sheets are composed of the foreigncurrency denominated debt held by residents at the end of the previous period
(Domestic Debt_l) multiplied by the Real Exchange Rate Change. We take the
previous period to avoid mixing quantity effects, stemming from new indebtness from
t-1 to t, with price effects, from the real exchange rate change. The best available proxy
for Domestic Debt for the sample of countries in this study,5 are the banking system's
dollar denominated deposits. De Nicolo, Honohan and Ize (2003) and Levy-Yeyati
(2004) argue that the banking system's dollar denominated deposits should be very
close to the banking system dollar-denominated credit to the private sector. In fact,
prudential regulations generally oblige banks to maintain very small open positions in
foreign currency. In addition, banks' dollar denominated credit to the private sector
should practically be equal to the total domestic indebtness of the private sector in
foreign currency except for the dollar-denominated debt this sector may issue
domestically. This is bound to be negligible in most emerging countries. As for the case
si External Debt, Domestic Debt is a gross concept of (negative) financial wealth since
the private sector can hold assets in foreign currency and not only liabilities. The
difference between the two, however, is that External Debt includes all sectors of the
3
We would also like to use data on domestic public debt denominated in foreign currency as collected by
Reinhart, Rogoff and Savastano (2003) but it is only available for a small number of the countries we
have included in our analysis.
141
economy and Domestic Debt only the private sector. In any event, it seems reasonable
to think that the public sector will also have negative wealth in foreign currency held by
residents.
Financial imperfections are proxied by a variable measuring the quality of the
institutional setting affecting the risk of investment (Creditor Rights). It is the sum of
three subcomponents: contract viability or expropriation, profits repatriation and
payment delays. Since this definition of creditor rights is more oriented towards
external creditors, we can consider it as a ceiling for the creditor rights of domestic
creditors in as far as emerging countries generally give priority to external debt
payments in case of difficulty.
Competitiveness, the other relevant channel of influence of real exchange rate
depreciations, is measured by the interaction of a country's openness (Openness) and
the change in the effective real exchange rate (Multilateral Real Exchange Rate
Change).
As regards the exchange rate regime, we use both de facto and de jure
classifications. In the former, the underlying exchange rate regime is inferred from the
observed exchange rate movement. The classification by Rogoff and Reinhart (2004) is
the preferred option since it allows us to keep a larger number of observations than
other classifications, such as Levy-Yeyati and Sturzenneger (2003). The de jure
classification is based on the IMF Annual Reports on Exchange Rate Arrangements and
Exchange Restrictions. Given the data limitations, we opt for grouping the
classification into three broad ones: fixed, intermediate and flexible regimes (See
Appendix II for details).
Finally, a number of control variables are included in all specifications. The first
is the lag of the sovereign risk (Embi_l\ to account for its persistence, as will be
shown later. The second is the Embi spread for all emerging countries for which it is
available (Emerging Embi). This should capture a possible similar co-movement
stemming from the market integration of this asset class and potential contagion effects.
At the same time, this control variable allows us to pick up possible time effects in the
regression.
From the statistical tables in Appendix 1 (3, 4 and 5), some stylized facts are
worth mentioning: First, the average of the Real Exchange Rate Change is a small real
appreciation, as opposed to a slight real depreciation in the case of'the Multilateral Real
Exchange Rate Change. Second, the average External Debt is around five times that of
Domestic Debt. Third, the average Real Exchange Rate Change varies only slightly
among different exchange rate regimes, both in the dejure and de facto classifications:
Dejure, flexible exchange rate regimes appreciate slightly on average while the other
two depreciate; de facto, intermediate regimes appreciate slightly while the other two
depreciate. As could be expected, the largest standard deviation is that of de facto
flexible exchange rate regimes. These differences between classifications can be better
understood comparing where each observation stands in the two classifications, as
shown in Table 3 of Appendix 1. From the 203 available observations only 111 find
themselves in the same exchange rate regime in the de facto and dejure classifications.
51 are more flexible de jure than de facto, which we could generally label as "fear of
floating" cases. The remaining 41 are more flexible de facto than announced, which in
16 of cases coincide with "freely falling" experiences of relatively fixed regimes, as
labelled by Rogoff and Reinhart (2004).
Finally, from the matrix of correlations in Table 4, Appendix 1, we can outline
other characteristics of the data. First, the dependent variable (Embi) is very persistent
(with a correlation of 0.71 between t and t-1). Second, the correlation between Embi
142
and either the Real Exchange Rate Change or the External Debt, and therefore External
Balancesheet, is positive, in line with the a priori of the financial accelerator literature.
However, the correlation betwen Embi and Êxternal Debt is negative, which hints at
the idea (confirmed later in our results) that it is not so much the new external
indebtness that matters for country risk, but the sudden increase in the stock of external
debt due to a real depreciation (in other words, the balance sheet effect and not the
quantity effect). Third, while the correlation between Embi and Domestic Debt is
negative but very close to zero, that between Embi and Domestic Balancesheet is
positive and relatively high (higher than for External Balance sheet). Only judging
from these correlations, we should expect a negative net wealth effect also in the case
of domestic balance sheets and not only for external ones. Fourth, the fact that the
correlation between External Debt and Domestic Debt is close to zero seems to indicate
that there is no clear pattern of complementarity or substitution between the two.
Finally, as one would expect, the quality of Creditor Rights, Êxports and Openness
are negatively correlated with the dependent variable but, contrary to the theoretical
literature, the degree of Competitiveness (i.e., the product of Openness and Real
Exchange Rate Change) is positively correlated.
As for the empirical strategy, we opt for a Generalized Method of Moments
(GMM), following Arellano and Bover (1995). We prefer this option to using OLS so
as to (i) remove unobserved time-invariant country-specific effects; (ii) account for the
potential endogeneity arising from the inclusion of the lagged dependent variable in
addition to other possibly endogenous right-hand side variables (particularly the real
exchange rate); and (in) deal with the possibility that the dependent variable is not
stationary. The second reason is particularly important since there might be instances of
reverse causality (from country risk to the real exchange). The GMM empirically
strategy allows us to take our results on safer grounds.
The Arellano-Bover estimator, or GMM system estimator combines the
regression expressed in first differences (lagged values of the variables in levels are
used as instruments) with the original equation expressed in levels (this equation is
instrumented with lagged differences of the variables)6. The disadvantage with this
empirical strategy, though, is the relatively small number of observations while the
conditions to use GMM should be complied with asymptotically. As a robustness test,
we run all regressions in OLS, with robust standard errors. The results remain
unchanged.
Results
(i)
The net impact of real exchange depreciations and appreciations
As a first step, it seems important to confirm whether real exchange rate depreciations
raise a country's risk premium. Controlling exclusively for the persistency of the risk
premium (Embi_J) and the evolution of the asset class (Emerging Embi), a statistically
significant positive relation is found between the change in the real exchange rate and
the risk premium (Table 1, column I)7. Although this first approximation is very
general and does not specify the channels through which the real exchange rate
^n all the estimations we present results for a Sargan test of over-identifying restrictions that checks the
overall validity of the different moment conditions and in all the cases we fail to reject the null
hypothesis.
7
The result holds when the change in the real exchange rate is defined in effective terms and not in
bilateral ones, exclusively against the US dollar.
143
influences country risk, the result could be understood as a net effect. Such negative
relation, more in line with the recent open-macro financial accelerator models than with
the more traditional literature, offers a warning signal to emerging countries, which
often suffer from real exchange rate depreciations.
It seems interesting to test whether the effects of real exchange rate changes on
a country's risk premium are symmetric, in other words, whether real appreciations
lower the country risk premium in the same way as real appreciations raise it. As Table
1, column II indicates, real exchange rate appreciations (Appr*Real Exchange Rate
Change) do not seem to contribute to reducing country risk since we cannot reject the
hypothesis that their coefficient is equal to zero8. This result is in line with the models
of financial imperfections, which expect detrimental effects of balance sheets only for
negative shocks to productivity, based on the argument that agency problems may only
be binding on the down side (Bernanke and Gertler, 1989). Another plausible
explanation is liquidity constraints. The asymmetric impact of real depreciations and
appreciations may have an important policy implication: other things given, it should
make emerging countries more reluctant to allow for fluctuations in the real exchange
rate, not being able to profit from the "good times" (real appreciations) while suffering
from the bad ones (real depreciations, particularly if sharp) . In particular, a real
exchange depreciation of one percentage point has an immediate impact on the risk
premium of 25 basis points.
The question is whether the impact of a real exchange rate depreciation is
linearly proportional to the size of the latter. In other words, whether it is the same in
terms of the country's risk premium to experience small depreciations over time or a
sudden large real one. Our results offer a negative answer. Table 1, column III shows
evidence of a non-linear effect of real exchange rate depreciations, accounted for as the
square of this variable, and the country risk premium.
It should be noted that the asymmetry is a short-run effect, which may disappear in the long run. If the
current coefficients could be interpreted as long-run ones (dividing them by one minus the estimated
coefficient formfc/__7), the asymmetry could disappear.
144
Table 1: Impact of real exchange rate changes on the country risk premium17
Specifications
Number of obs
Dependent variable: Embi
Embi 1
Emerging Embi
Real Exchange Rate Change
Appr * Real Exchange Rate Change (fr)
1
183
II
183
III
183
0.78 ***
(0.10)
0.64***
(0.18)
1533.62**
(606.38)
0.64 ***
(0.11)
0.33*
(0.18)
0.68 ***
(0.11)
0.39**
(0.17)
Diff Effect Dep * Real Exchange Rate Change (2)
-97.28
(604.54)
120.95
(649.11)
2474.57 ***
(634.17)
-892.38
(623.40)
[Real Exchange Rate Change] 2

Constant
4170.78***
(545.02)
-260.77**
(119.04)
Appr Constant
-62.98
(114.57)
-99.01
(85.62)
Diff Effect Dep Constant

Sargan test
25.56(1.00)
22.49(1.00)
H0:p1+p2=0
(p-value) 0.00
HQ can be rejected
-122.95
(95.72)
1 29.59 *
(75.57)
19.69(1.00)
The dynamic panel estimation uses one step GMM system estimators with heteroskedasticlty-consistent standard errors.
Lags dated t-2, t-3 and t-4 for Real Exchange Rate Change, Appr * Real Exchange Rate Change, Diff Effect Dep * Real Exchange Rate Change
and [Real Exchange Rate Change]2 were included as instruments.
Standard errors in parenthesis (p-values for the Sargan tests).
Significance of coefficients: * at 10% ; - at 5%; *** at 1%
I/ Results are maintained (i) using OLS with robust standard errors instead of GMM, (ii) including the debt service instead of the
stock of debt, and/or (ii) subtracting a country's international reserves to the stock of debt
Although real depreciations tend to be detrimental for a country risk premium,

we find a few observations where the opposite is true. The question is what makes
these cases different. As a tentative answer, since the small number of observations
does not allow us to explore the issue more rigorously, we look at the commonalities in
the observations in which exchange rate depreciations lead to a reduction in a country's
risk premium (23 out of a total of 75 depreciations). We refer to this group as the
optimistic case. Taking the general case as a benchmark (namely the 52 observations in
which real exchange rate depreciations lead to an increase in the risk premium) and
making them equal to 100, the optimistic case is characterized by a lower external debt
(about 20% lower than in the general case), higher tradability (15% higher), and better
creditor rights, all as expected (Figure 1). However, they also have a much higher
domestic dollar-denominated debt (50% more on average than in the general case). It is
important to notice that exchange rate depreciations are much smaller in the optimistic
case, which mitigates the relevance of the previously mentioned differences. We shall
analyze this issue in more detail later.
145
Figure 1: Characteristics of the optimistic case17 against the general one27
I/ Real depreciations reduce the cost of borrowing

21 Real depreciations increase the cost of borrowing.
In the case of real appreciations, there are a few observations where we find the
expected positive impact (i.e., a reduction in country risk). We call this the "optimistic
case", since it is not generally confirmed in our empirical results, and compare it with
the "pessimistic one" (where real exchange rate appreciations increase the risk
premium). As one would expect, the former has more debt (both external and domestic
dollar denominated) so that it can profit more from its reduction in value after the
appreciation. It is also less open to trade, so that it is less damaged by the appreciation.
Creditor rights are lower but this is probably a less relevant variable than for
depreciations since we are not in a binding situation, when net wealth falls.
146
Figure 2: Characteristics of the pessimistic case v against the optimist one 2/
I/ Real appreciations increase the cost of borrowing

2/ Real appreciations decrease the cost of borrowing.
(ii)
Channels for a real exchange depreciation to influence the risk premium
We, now, specify the channels through which the real exchange rate may influence a
country's cost of borrowing, based on the existing literature. The most important ones
might be balance sheet effects, from external and domestic dollar denominated debt,
and competitiveness. Also financial imperfections could be a potential channel in as far
as financial accelerator theories make balance sheet effects dependent on the existence
of such imperfections. Focusing exclusively on the external debt, we find that the
External Balance sheets after a depreciation clearly increase the risk premium while
they are not significant after an appreciation. (Table 2, column I). The same result is
found for Domestic Balance sheets (Table 2, columns II and III).9 The latter seems to
indicate that domestic private sector indebtness in foreign currency has negative wealth
effects and not only redistributive ones. In turn, competitiveness affects country risk
symmetrically and in the expected direction (reducing it with a real depreciation and
increasing it with an appreciation)10. Better creditor rights tend to lower country risk.
Finally, we try to separate quantity effects from price ones by including in the
regression the increase in external and domestic dollar denominated debt and export
growth, all in US dollar. None of the quantity effects are found significant. In the case
9
The significance of domestic balance sheets, after a depreciation, is weakened (from a level of
significance of 1% to 10%) when both external and domestic balance sheets are included in the
regression (Column III). This is probably due to the collinearity between the two variables (Table 5 in
Appendix I show a correlation of 0.52)
10
The correlation between External Balancesheet and Competitiveness is very high, pointing to
collinearity problems. An analysis of the correlation between parameters confirms this problem. This is
why we shall exclude Competitiveness in the following regressions, substituting it for Increase Exports
which accounts mainly for the quantity effect, as Exports are measured in dollars.
147
of external and domestic debt, this result can be understood as if the country risk
premium were not affected by new indebtness but rather by the sudden reduction in net
wealth, due to real depreciation. This is in line with financial accelerator theories.
Table 2: Channels of influence of changes in the real exchange rate l/
Specifications
Numberofobs
1
179

Embi 1
Emerging Embi
Appr* External Balancesheet
Diff Effect Dep * External Balancesheet
Increase External Debt
II
152
0.65 ***
(0.11)
0.25*
(0.14)
-39945
(776.30)
491 7.25 ***
(1.567.73)
Appr * Domestic Balancesheet

Diff Effect Dep * Domestic Balancesheet
Increase Domestic Debt
Appr* Competitiveness
Diff Effect Dep* Competitiveness
Increase Exports
Creditor Rights
Appr* Constant
Diff Effect Dep* Constant
Sargantest
2205.09 **
(977.31)
-5048.16***
(1.792.62)
-37.04*
(21.30)
283.62
(237.24)
-55.37
(49.99)
0.65 ***
(0.07)
0.20
(0.14)
0.61 ***
(0.08)
0.15
(0.14)
-427.05
(834.89)
3324.59 **
(1.673.84)
-466.38
(388.54)
14455.44 ***
(2.111.46)
-262.96
(497.92)
7826.1 1 *
(4.819.81)
1846.57 **
(784.68)
-898.91
(1.045.50)
2096.28 **
(968.15)
-4062.94***
(1.500.05)
-48.43**
(23.01)
393.01*
(227.28)
21.35
(55.48)
19.42(1.00)
III
152
15.39(1.00)
-47.41**
(21.44)
428.82*
(223.46)
-18.03
(45.17)
6.96(1.00)
IV
122
0.61 ***
(0.06)
0.25***
(0.16)
-357.98
(1.036.92)
3496.77 ***
(1.737.62)
-0.46
(1.89)
-583.40
(808.42)
-1758.74
(4.752.85)
0.02
(0.03)
2277.10 *
(1.246.20)
-4441.31***
(1.672.54)
-495.61
(311.58)
-56.76*
(30.01)
-7.90
(46.84)
480.10
(291.02)
11.98(1.00)
The dynamic panel estimation uses one step GMM system estimators with heteroskedasticity-consistent standard errors.
Lags dated t-2. t-3 and t-4 for Appr * External Balancesheet, Off Effect Dep External Balancesheet Increase External Debt,
Appr * Domestic Balancesheet, Diff Effect Dep" Domestic Balancesheet, Increase Domestic Debt,
Appr Competitiveness and Diff Effect Dep Competitiveness were included as instruments.
Significance of coefficients: * at 10%; " at 5%; at;1%
(Hi)
Haw do financial imperfections influence balance sheet effects
In the previous set of regressions we have found direct evidence of the detrimental
effect of financial imperfections on the risk premium. However, financial accelerator
theories consider financial imperfections more as a condition under which balance
sheet effects can increase the cost of borrowing than as a separate channel. To test this
hypothesis, we interact each country's financial imperfections -proxied with the quality
of creditor rights - with balance sheet effects, both external and domestic. We separate
countries in three groups, those with the best creditor rights, those with intermediate
ones and those with the poorest. Balance sheet effects are clearly larger in the last
group, followed by the intermediate one (Table 3, columns I and II, respectively). In
particular, for Domestic Debt only do countries with the poorest creditor rights see their
risk premium increase because of domestic balance sheet effects. In the case of
intermediate creditor rights we cannot reject the hypothesis that domestic balance
sheets have no effect on the risk premium or is even negative for good creditor rights
148
(Table 3, bottom of Column III). This could be explained by the fact that domestic
creditors in the countries with the poorest creditor rights do not trust the system enough
to use - or keep - their additional net worth at home.
Table 3: Financial imperfections and the influence of external
and domestic balance sheet effects on the risk premium ll
Specifications
Number of obs
I
174

EmbM
0.78 ***
(0.11)
0.51 ***
(0.18)
2514.32***
(667.56)
-659.22
(662.47)
-1483.08 *
(653.56)
Emerging Embi
Low Creditor Rights * External Balancesheet (YI)
Diff Effect Medium Creditor Rights * Exteranal Balancesheet (Y2)
Diff Effect High Creditor Rights * External Balancesheet (v3)
Low Creditor Rights * Domestic Balancesheet (6,)
Diff Effect Medium Creditor Rights Domestic Balancesheet (63)
II
151
0.63 ***
(0.04)
0.49 ***
(0.19)
15868.14***
(1.706.31)
-1 51 48.1 9 ***
(1.901.49)
-18861.95 ***
(2.705.76)
-440.31
(343.59)
86.13
(166.94)
-237.62 **
(119.19)
-279.09 **
(123.16)
Diff Effect High Creditor Rights * Domestic Balancesheet (63)

Increase Exports
-493.69**
(242.69)
-37.79
(169.61)
-1 56.65
(99.92)
-1 74.69 *
(109.67)
Low Creditor Rights Constant

Diff Effect Medium Creditor Rights Constant
Diff Effect High Creditor Rights Constant
Sargan test
20.25(1.00)
H0:Yi+Y3=0
(p-value) 0.01
HO can be rejected
17.33(1.00)
H0: 6^62=0
(p-value) 0.46
HO cannot be rejected
H0: 6^63=0
(p-value)
0.04
HO can be rejected
Lags dated t-2, t-3 and t-4 for Low Creditor Rihgts * External Balancesheet, Diff Effect Medium Creditor Rights * External
Balancesheet, Diff Effect High Creditor Rights * External Balancesheet, Low Creditor Rights * Domestic Balancesheet, Diff Effect
Medium Creditor Rights * Domestic Balancesheet, and Diff Effect High Creditor Rights * Domestic Balancesheet were included as
instruments.
Significance of coefficients: * at 10% ; - at 5%; *' at 1 %
149
(iv)
How does the exchange rate regime influence balance sheet effects
After identifying when balance sheet effects are particularly a problem, we now
analyze to what extent they are influenced by the exchange rate regime in place. This is
particularly interesting if we consider that the exchange rate regime is an important
policy variable for the economic authorities.
As previously mentioned, several theoretical models argue that a fixed
exchange rate regime amplifies balance sheet effects on the risk premium. This is
confirmed in our results, when interacting the exchange rate regime and domestic and
external balance sheet effects. The exchange rate regime is lagged one period to avoid
that what was originated by a certain regime is assigned to another one. We use both de
jure and de facto classifications and compare the results.
Starting with external balance sheet effects, fixed exchange rate regimes, de
jure, amplify their detrimental impact on the cost of borrowing (Table 4, column III).
This is so when compared with the average balance sheet effect, i.e., when the
exchange rate regime is not considered (Table 4, column I). The flexible regime is
clearly superior since we cannot reject the hypothesis that external balance sheets under
this regime leave the risk premium unchanged (Table 4, bottom of column III). When
taking the de facto classification, fixed regimes are also the most detrimental (Table 4,
column II), with a larger coefficient than the average case (Table 4, column I). This
time the differential effect of the flexible exchange rate is not significant but the
intermediate one is clearly better than the pegged, although not to the extent of
eliminating the detrimental effect of external balance sheets on the risk premium. In
sum, although the results are relatively similar in the two classifications for the fixed
exchange rate regime, this is not the case for the intermediate and flexible ones. One
possible explanation is that the de facto classification has twice as many observations
under the intermediate regime than the de jure classification. The opposite is true for
flexible exchange rate regimes. This difference is probably explained by the wellknown phenomenon of "fear of floating", as countries tend to announce that the
exchange rate will move more flexibly than they actually allow for.
150
Table 4: The exchange rate regime and external balance sheets "
Specifications
Number of obs
1
178

EmbM
II
170
DE FACTO
0.78 ***
(0.13)
0.55 ***
(0.16)
2489.84 ***
(769.48)
Emerging Embi
External Balancesheet
Fixed. 1 External Balanacesheet (0,)
Diff Effect Intermediate..! * External Balancesheet (02)
Diff Effect Flexible.! ' External Balancesheet (03)
Creditor Rights
-50.21*
(29.69)
-470.18 ** ^
(249.37)
'
194.83
(324.66)
Increase Exports
Constant
Constant Fixed. 1
Diff Effect Intermediate^ Constant
Diff Effect Flexible.! Constant
Sargan test
22.84 (1 .000)
III
177
DEIURE
0.71 ***
(0.15)
0.59 ***
(0.14)
0.73 ***
(0.12)
0.57
(0.17)
3333.64**'
(1.158.73)
-2741.70"
(1.283.98)
-1844.16
(1.591.65)
-29.35**
(12.18)
-366.91 (178.14)
3145.73***
(1.226.52)
-439.06
(1.633.45)
-2488.72 **
(1.244.84)
-44.53**
(22.60)
-544.85 "
(276.72)
147.72
(171.55)
-140.66**
(66.87)
-1 70.01 ***
(64.07)
23.20 (1 .000)
238.28
(276.96)
-38.73
(60.47)
-1 28.84 **
(58.78)
18.1 1 (1 .000)
^10^02=0
(p-value) 0.06
H0 can be rejected
H.:G,+Q3-0
(p-value)
0.16
H0 cannot be rejected
Lags dated t-2, t-3 and t-4 for External Balancesheet, FixecM * External Balancesheet, Diff Effect Intermediate.1! External Balancesheet, Oiff Effect
Ftexibte.1 External Balancesheet were included as instruments.
Significance of coefficients: at 10% ; - at 5%; at 1%
In the case of domestic balance sheets, pegged regimes are clearly worse on the
basis of the de lure classification with double the coefficient than for the average case
(Table 5, columns III and I, respectively). Intermediate and flexible regimes are clearly
superior since we cannot reject the hypothesis that balance sheet effects under any of
these two regimes leave the risk premium unchanged (Table 5, bottom of column III).
The differences among de facto regimes are not significant (Table 5, column II).
151
Table 5: The exchange rate regime and domestic balance sheets17

Specifications
Number of obs
1
151

EmbM
II
143
DE FACTO
0.61 ***
(0.07;
0.43 **
(0.18)
8100.45 *
(4614.37)
Emerging Embi
Domestic Balancesheet
FixedJ* Domestic Balancesheet (p,)
0.54 ***
(0.12)
0.52 ***
(0.15)
Diff Effect FlexibleJ * Domestic Balancesheet (p3)

-411.67
(374.42)
-67.37**
(46.55)
500.59
(477.63;
Creditor Rights
Constant
Constant FixecM
Diff Effect Intermediate.! Constant
Diff Effect Flexible_1 Constant
Sargan test
20.32(1.000)
0.63 ***
(0.06)
0.51 ***
(0.15)
6710.74
(5.110.24)
-2323.34
(5.545.48)
-674.53
(9.472.57)
-236.83
(175.40)
-35.85**
(10.29)
Diff Effect Intermediate.. 1 * Domestic Balancesheet (p2)
Increase Exports
III
177
DEIURE
309.68
(156.25)
-149.44
(104.02)
-132.10
(105.82)
17.51(1.000)
16319.13***
(717.04)
-15153.64*'*
(1.605.39)
-13334.29***
(1.987.78)
-495.17
(332.18)
-48.43*
(28.94)
267.86
(278.76)
80.70
(62.24)
-37.61
(38.23)
15.10(1.000)
He:pi+Pz"0
(p-value) 0.41
Ho cannot be rejected
H0: p^pa'O
(p-value) 0.22
HQ cannot be rejected
Lags dated t-2, t~3 andt-4 for Domestic Balancesheet, Fixed_1 * Domestic Balancesheet, Diff Effect lntermediate_1 Domestic Balancesheet, and Diff Effect
Ftexible_1 * Domestic Balancesheet were included as instruments.
Significance of coefficients: at 10%; ** at 5%; at 1%
Given its policy implications, it seems worth exploring why it is the case that
pegged regimes behave worse than others. As a tentative answer (since the small
number of observations does not allow us to explore the issue more rigorously) we look
at the commonalities in the observations under a fixed regime and compare them with
those for intermediate and flexible regimes11.
Fixed exchange rate regimes tend to accumulate more external debt and
domestic dollar-denominated debt, as argued by Ize and Levy-Yeyati (2003) and Broda
and Levy-Yeyati (2003)12 This is more the case in de facto than de jure classification
(Figure 2 and 3)13, which might be explained by the fact that some of the announced
pegged regimes are not expected to be maintained (in fact there are much fewer
observations for de facto pegs than de jure). The same might be true for some of the
intermediate regimes which are announced (particularly crawling pegs). No clear trend
appears for Domestic debt.
11
The number of observations for each group can be found in Table 3, Appendix 1.
This is the case not only in levels, as shown in Figures 3 and 4, but much more so when we look at the
rates of change of external debt form t-1 to t. This is not included in the graph because of the differences
in scale.
13
In the specific intermediate regimes, dejure, domestic dollar denominated debt is actually lower.
12
152
Another plausible explanation, other than the accumulation of foreign currency

debt, could be that real exchange rate depreciations are larger under fixed exchange rate
regimes. As Table 3 in Appendix 1 shows, this is not the case either in the dejure orde
facto classifications, since the observations under the pegged regime do not have the
largest average real depreciation. It could, nevertheless, happen that pegs suffer more
frequently from events of very large depreciations, which we have shown to be more
detrimental. Looking at the 5% extreme values of the right tail of our distribution (i.e.,
the largest real depreciations), this does not seem to be the case. In fact, most of the
extreme observations fall under intermediate regimes both in the de jure or de facto
classifications.
Figure 3: Characteristics of managed and flexible

exchange rate regimes against fixed ones: De facto classification
153
Figure 4: Characteristics of managed and flexible

exchange rate regimes against fixed ones: Dejure classification
In sum, from this cursory exploration of the data, the most plausible explanation
for the more detrimental balance sheet effects under pegged regimes is the relatively
larger accumulation of dollar-denominated debt, coupled with the existence of poorer
creditor rights, and not so much the accumulation of a larger depreciation or extreme
depreciation events under pegged regimes. This is in line with the idea that fixed
exchange rates tend to be perceived as an implicit insurance by the private sector and
that public authorities may increase their dollar-denominated indebtness as a
demonstration effect that the regime will be maintained.
Conclusions and Policy Implications
This paper builds upon the empirical literature on the impact of real exchange rate
depreciations for the economy as a whole. In particular, it confirms Berganza, Chang
and Garcia Herrero (2003)'s finding of a positive relation between changes in the real
exchange rate and a country's risk premium for a sample of 27 emerging economies
and explores additional questions to determine what makes balance sheet effects so
detrimental for the risk premium.
We show evidence that the effect of a real depreciation is neither symmetric nor
linear. On the former, real appreciations are not found significant in reducing a
country's risk premium, while real depreciations clearly increase it The immediate
effect of a real depreciation of one percentage point is an increase in the country risk
premium by 25 basis points. On the latter, sharp real depreciations have much larger
negative effects than smaller ones. This should make policy makers wary of real
exchange rate volatility, particularly if large, since there is no period when they clearly
benefit from it. There are, however, a few cases in our sample, where exchange rate
depreciations reduce the risk premium. A cursory look at the characteristics of these
observations points to the importance of having a relatively low level of external debt,
154
higher trade openness and better creditor rights, for real exchange rate depreciations to
be beneficial.
We also show that the main channels for the exchange rate to affect country risk
are external and domestic balance sheets, stemming from the sudden increase in the
stock of external debt and domestic dollar-denominated debt after a real depreciation.
In the case of domestic balance sheets, this can be interpreted as evidence of the
presence of wealth effects and not only redistribution ones. In addition, the same
asymmetric impact is found for balance sheets as for the real exchange rate; that is, the
reduction in the stock of foreign-currency debt after a real appreciation does not reduce
country risk. On the contrary, the degree of competitiveness appears to have a
symmetric effect - and with the expected sign - on country risk. In any event, The
evidence of a positive and highly significant relation between the exchange rate change
and country risk, which can be considered a net effect, indicates that competitiveness is
not an important enough factor to outweigh the detrimental impact of balance sheets.
New external and domestic dollar denominated indebtness is not found significant,
suggesting that what matters is not so much the amount of new borrowing but rather the
sudden reduction in net financial wealth because of a price change.
When financial imperfections are considered (proxied by the quality of creditor
rights) our results confirm the a priori of the financial accelerator literature: the poorer
creditor rights are, the more external and domestic balance sheet effects increase the
risk premium. Finally, fixed exchange rate regimes appear to amplify the negative
impact of balance sheet effects on the risk premium. This seems to be related to the fact
that pegged regimes have a bigger (and faster growing) stock of external debt, on
average, and not so much to the extent of the real depreciation. The latter is not larger,
on average, under this regime, not even the number of events of large depreciations,
which have been found to be particularly detrimental through the result of nonlinearity. A plausible explanation for the potential causal relation between a pegged
regime and a larger external debt is that this regime is perceived as an implicit
insurance by the private sector. In the same vein, public authorities may increase their
dollar-denominated indebtness as a demonstration effect that the peg will be
maintained.
In sum, a number of policy conclusions can be drawn from these results. The
volatility of the real exchange rate, especially if large, is something to worry about in
emerging countries. This is because it tends to increase country risk, a key variable for
economic growth, in an asymmetric and non-linear way. The main channels through
which the real exchange rate affect the risk premium are external and domestic balance
sheet effects and, to a lesser extent, competitiveness, in the opposite direction.
Therefore, the countries that should worry most are those with small trade openness,
large financial imperfections and pegged exchange rate regimes, which are associated
with bigger and faster growing external indebtness. The combination of these three
characteristics can make real exchange rate depreciations particularly detrimental for a
country's risk premium, an extremely important variable for emerging countries in need
of external financing because of its strong impact on economic growth. Given that these
three characteristics can be influenced by economic authorities, there is clear a role for
policy action to mitigate the problem.
155
References
Arellano and Dover (1995). "Another Look at the Instrumental-Variable Estimation of
Error-Components Models", Journal of Econometrics, Vol. 68, pp. 29-52
Berganza, Chang and Garcia Herrero, (2003) "Balance Sheet Effects and the Country
Risk Premium: An Empirical Investigation", Working Paper No. 0316, Banco
deEspana
Bemanke and Gertler, (1989) "Agency costs, Net Worth, and Business Fluctuations",
The American Economic Review, Vol. 79, pp. 14-31.
Bleakley and Cowan, (2002) "Corporate dollar debt and depreciations: much ado about
nothing?". Working Papers No. 02-5, Federal Reserve Bank of Boston.
Broda and Levy-Yeyati (2003), "Endogenous Deposit Dollarization", Staff Report 160,
Federal Reserve Bank of New York.
Burnside, Eichenbaum and Rebelo (2001), "Hedging and Financial Fragility in Fixed
Exchange Rate Regimes" European Economic Review, Vol. 45 (7), pp. 11511193.
Cespedes, Chang and Velasco, (2000) "Balance Sheet Effects and Exchange Rate
Policy". NBER Working Paper No. 7840.
De Nicolo, Honohan and Ize (2003), "Dollarization of the Banking System: Good or
Bad?", IMF WP/03/146.
Edwards (1986), "Are Devaluations Contractionary?", The Review of Economics and
Statistics, Vol. 68, Issue 3, pp. 501-508.
Elekdag and Tchakarov (2004), "Balance sheets, Exchange Rate Policy and Welfare",
IMFWP/04/63.
Forbes, (2002) "How do large depreciations affect firm performance?", NBER
Working Paper No. 9095.
Galiani, Levy Yeyati and Schargrodsky (2003), "Financial Dollarization and Debt
Deflation under Currency Board", Emerging Markets Review, Vol. 4, pp. 340367.
Galindo, Panizza and Schiantarelli, (2003) "Debt composition and balance sheet effects
of currency depreciation: a summary of the micro evidence", Emerging Markets
Review, Vol. 4, pp. 330-339.
Gertler, Gilchrist and Natalucci, (2003) "External Constraints on Monetary Policy and
the Financial Accelerator", mimeo.
International Monetary Fund, "Annual Report on Exchange Rate Arrangements and
Exchange Restrictions-, various issues.
Ize and Levy-Yeyati (2003), "Financial Dollarization", Journal of International
Economics. 59, pp. 323-347.
Levy-Yeyati (2004) "Financial Dollarization: Evaluating the Consequences", mimeo,
Universidad Torcuato di Telia.
Levy-Yeyati and Sturzenneger (2003), "A de facto Classification of Exchange Rate
Regimes: A Methodological Note", American Economic Review, vol. 93, pp.
1173-1193.
Reinhart, Rogoff and Savastano (2003). "Addicted to Dollars", NBER Working Paper
No. 10015.
Rogoff and Reinhart, (2004) "The Modem History of Exchange Rate Arrangements: a
Reinterpretation", Quarterly Journal of Economics, Vol.119, pp. 1-48.
156
APPENDIX I:
STATISTICAL ISSUES
Table 1
Countries and years included
Country name
Algeria
Argentina
Brazil
Bulgaria
Chile
China
Colombia
Cote D'lvoire
Croatia
Ecuador
Malaysia
Mexico
Morocco
Nigeria
Panama
Peru
Philippines
Poland
Republic of Lebanon
Russian Federation
Slovakia
South Africa
South Korea
Thailand
Turkey
Venezuela
Zimbabwe
Years
Number of years
1999-2002
1993-2002
1993-2002
1994-2002
1999-2002
1994-2002
1997-2002
1998-2002
1996-2002
1995-2002
1996-2002
1993-2002
1993-2002
1993-2002
1996-2002
1997-2002
1993-2002
1994-2002
1998-2002
1997-2002
1993-2002
1994-2002
1993-2002
1997-2002
1996-2002
1993-2002
1997-2002
4
10
10
9
4
9
6
5
7
8
7
10
10
10
7
6
10
9
5
6
10
9
10
6
7
10
6
vJo. of observations
210
157
Table 2
Geographical distribution of the sample
Region
Number of
countries
Number of
observations
as a % of total
sample
5
9
6
6
1
27
42
71
48
44
5
210
20.0
33.8
22.9
21.0
2.4
100
Asia
Latin America
Eastern Europe
Africa
Middle East
TOTAL
Table 3:
Descriptive Statistics of the regression variables
Variable
Embi
Emerging Ernbi
Fixed real exchange rate change de facto
Intermediate real exchange rate de facto
Flexible real exchange rate de facto
Fixed real exchange rate change de iure
Intermediate real exchange rate de iure
Flexible real exchange rate de iure
Effective real exchange rate change
External Debt
Domestic Debt
r\M*VLa.eti/
Dal'K/' **
uomesoc baiancesneei
Dpenness
uompeimveness
Increase Exports
Creditor rights
No.Obs.
Mean
Sfd. Deviation
Minimun
Maximun
210
210
208
55
109
38
73
68
67
210
209
208
207
155
143
172
208
207
203
208
560.40
617.47
-0.019
0.009
0.011
-0.036
-0.009
-0.012
0.0159
0.0044
0.5683
3.75
0.0018
0.1132
69.68
-0.0024
0.3642
-0.0017
0.0714
7.21
515.95
143.61
0.1561
0.152
0.115
0.231
0.164
0.167
0.135
0.1477
0.2589
10.43
0.0928
0.2721
478.83
0.0248
0.2107
0.0388
0.1485
2.11
60.233
352.72
-0.8126
-0.319
-0.257
-0.813
-0.448
-0.813
-0.266
-0.3746
0.1473
-17.43
-0.3071
3925.75
1007.55
0.895
0.895
0.415
0.616
0.895
0.529
0.415
1,137
1,561
41.66
0.6432
2,109
5091.47
0.163
1,195
0.2254
0.7998
-100
-0.1485
0.05903
-0.1348
-0.3651
12
Table 4
Relation between the classification of de jure and de facto exchange rate regimes
158
Table 5:
Matrix of correlation
Real
Multilateral
Exchange
Real
Increase
Emerging
Rate
Exchange External External
External
Domestic
Increase
Domestic
Increase Creditor
Embi EmbiJ Embi
Change Rate Change Debt
Debt
Balancesheet
Debt Domestic Debt Balancesheet Openness Competitiveness Exports
Rights
Embi
1.00
EmbiJ
0.71 1.00
Emerging Embi
0.1B
0.07
1.00
0.31 0.00
0.00
1.00
Multilateral Real Exchange Rate ( 0.35 0.06
0.16
0.63
1.00
External Debt
0.36
0.35
0.01
-0.02
-0.02 1.00
-0.29 -0.33
-0.02
-0.22
-0.13 -0.20
1.00
0.32 0.02
0.07
0.93
0.87 0.00
-0.17
1.00
Domestic Debt
-0.07 -0.11
0.01
-0.09
-0.08 0.04
0.09
-0.09
1.00
0.06 0.02
0.02
-0.10
-0.15 -0.02
-0.03
-0.10
-Q.05
1.00
Domestic Balancesheet
0.35 0.14
-0.04
0.51
0.45 -0.02
-0.13
0.52
-0.81
0.01
1.00
Openness
-0.17 -0.09
-0.03
-0.04
-0.06 0.36
-0.02
-0.04
-0.22
-O.D5
0.10
1.00
Competitiveness
0.20 -0.10
0.04
0.71
0.84 -0.13
-0.17
0.75
-0.02
-0.11
0.26
-0.12
1.00
Increase Exports
-0.10
0.11
0.15
-0.17
-0.10 -0.14
0.16
-0.15
0.05
-0.05
-0.16
0.08
-0.10
1.00
Creditor Rights
-0.40 -0.32
-0.16
-0.07
-0.14 -0.01
0.01
-0.11
0.09
-0.20
-0.08
0.21
-0.10
-0.10
1.00
APPENDIX II:
DATA SOURCES AND VARIABLE DEFINITIONS
Below we list the variables and sources used for this study, as well as the
transformations made to the data. The data are annual and cover the periods and
countries shown in Table 1.
Dependent variable;
* Embi Country risk premium or spread in the external cost of borrowing: equals
returns for U.S. dollar-denominated Brady bonds, loans, Eurobonds, and U.S. dollardenominated local markets instruments for emerging markets minus total returns for
U.S. Treasury bonds with similar maturity (the stripped yields of the Emerging Markets
Bond Index, Embi, for each country). The spreads are measured in basis points.
Source: JP Morgan.
Objective variables;
* External Debt: equals the total debt in convertible currencies owed to nonresidents, as
the end of the reporting year in U.S. dollars divided by the nominal GDP in 1995 in
U.S. dollars, so as to take into account the relative size of the country.
Source: The Institute of International Finance (IIF).
* Domestic Debt: proxied by the domestic deposits in U.S. dollars divided by the
nominal GDP in 1995 U.S. dollars to take into account the relative size of the country.
Source: International Financial Statistics (IFS) of the International Monetary Fund
(IMF) and Levy-Yeyati (2004).
* "Real" Exchange Rate: equals the average number of units of local currency per U.S.
dollar during the year adjusted by the inflation price index (with 1995=1) divided by the
nominal exchange rate in 1995. Thus, in 1995, Real Exchange Rate is equal to 1 and an
increase (decrease) in Real Exchange Rate is a depreciation (appreciation).
Source: IIP.
* Multilateral Real Exchange Rate: is an annual average index of the nominal effective
exchange rate of the local currency with respect to six leading trading partners, deflated
by the relative consumer prices. An increase (decrease) in Multilateral Real Exchange is
a depreciation (appreciation)
Source: IIF.
* "Real" Exchange Rate Change: equals the changes in "Real" Exchange Rale
between year t and year t-1. (A/ "real" exchange rate).
* Multilateral Real Exchange Rate Change: equals the changes in Multilateral Real
Exchange between year t and year t-1. (bdn effective real exchange rate).
* Exports', equals the total value of export of goods and services to nonresidents, valued
at market prices in millions of U. S. dollars.
Source: IIF.
* Openness: is defined as the ratio of Exports to the nominal GDP in 1995 U.S. dollars.
Source: The IIF.
160
* External Balancesheet: equals the product of External Debt in year t-1 and "Real"
Exchange Rate Change between the years t-1 and t.
* Domestic Balancesheet: equals the product of Domestic Debt in year t-1 and "Real"
* Competitiveness: equals the product of Openness in year t-1 and Multilateral Real
* Creditor Rights: measure the quality of the institutional setting affecting the risk of
investment. The rating assigned is the sum of three subcomponents, each with a
maximum score of 4 and a minimum score of 0. A score of 4 indicates a very good
environment for creditors and 0 a very poor. The subcomponents are: contract
viability/expropriation, profits repatriation and payment delays. Countries are divided
into three groups: tow, medium and high creditor rights.
Source: International Country Risk Guide.
Control variables;
* Emerging Embi. equals the average of the stripped yields of the Emerging Markets
Bond Index, Embi.
Source: JP Morgan.
* Appreciation (Appr): is a dummy variable that takes on a value of one if Real
Exchange Rate Change is negative and zero otherwise. Real Exchange Rate Change is
never zero throughout our sample.
* De facto classification of exchange rate regimes: From the 15 groups considered in
Rogoff and Reinhart (2004), we group them in three groups: (infixed, which includes
codes such as "no separate legal tender", "pre announced peg or currency board
arrangement", "pre announced horizontal band" and "de facto peg"; (ii) intermediate,
composed of "pre announced crawling peg", "pre announced crawling band", "de facto
crawling peg", "de facto crawling band", "moving band" and "managed floating"; and
(inflexible, including "freely floating" and "freely falling". The group "dual market in
which parallel market data is missing" (7 observations in our sample) is left out of the
classification. A dummy variable is defined for each group.
Source: Rogoff and Reinhart (2004).
* De jure classification of exchange rate regimes: Every IMF member country is
required to report and publish each year the stated intentions of the central bank yielding
a de jure classification. From the 8 groups considered, we group them in three groups:
(i) fixed, which includes "exchange arrangement with no separate legal tender",
"currency board arrangement", "conventional pegged arrangement" and "pegged
exchange rate within horizontal bands"; (ii) intermediate^ composed of "crawling peg",
"crawling band" and "managed floating with no pre-announced path for the exchange
rate"; and, (iii) flexible, including "independently floating". A dummy variable is
defined for each group.
Source: Annual Reports of Exchange Rate Arrangements and Exchange Rate
Restrictions (IMF).
161
Session 3
Debt in Emerging Economies
Public Debt, Fiscal Solvency, and Macroeconomic

Uncertainty in Latin America: The Cases of Brazil,
Colombia, Costa Rica and Mexico*
Enrique G. Mendoza
University of Maryland and the National Bureau of Economic Research
Pedro Marcelo Oviedo
Iowa State University
* This paper is based partly on material we developed for the World Bank's project on Assessing Fiscal
Sustainability in Latin America and in work on a method to evaluate fiscal sustainability that we did while
visiting the Research Department of the Inter-American Development Bank. We are grateful to Manuel
Amador, Andres Arias, Guillermo Calvo, Umberto Delia Mea, Arturo Galindo, Santiago Herrera, Claudio
Irigoyen, Alejandro Izquierdo, Guillermo Perry, Tom Sargent, Alejandro Werner and participants at the XIX
Meeting of the Latin American Network of Central Banks and Finance Ministries for helpful comments and
suggestions. The views expressed here are the authors5 only and do not reflect those of the World Bank or
the Inter-American Development Bank. Addresses for the authors are: Enrique G. Mendoza, Department of
Economics, University of Maryland, College Park, MD 20742; Pedro Marcelo Oviedo, Department of
Economics, Iowa State University, Ames, IA 50011.
Introduction
The ratios of public debt to GDP have been rising steadily in the economies of Latin
America. Comparing averages for the period 1996-2002 with those for the period 19901995, the average debt ratios of Brazil, Colombia, and Mexico increased by about 10
percentage points in each country. In Costa Rica, the public debt ratio did not change
much, but it was already at a relatively high level of around 50 percent at the beginning of
the 1990s. Given that growing public debt has traditionally been an indicator of financial
weakness and vulnerability to economic crisis in the region, there is concern for assessing
whether these high levels of debt are in line with the solvency of the public sector or
should be taken as a warning signal that requires policy intervention.
The goal of this paper is to assess the consistency of the public debt ratios of Brazil,
Colombia, Costa Rica and Mexico with the conditions required to maintain fiscal solvency.
This assessment is based on a variant of the framework proposed by Mendoza and Oviedo
(2004). In particular, we apply their one-sector model with exogenous government
expenditure rules. This methodology produces estimates of the short- and long-run
dynamics of public debt ratios in a setup in which public revenues are subject to random
shocks and the government aims to maintain its outlays relatively smooth in the face of this
uncertainty. The government is handicapped in its efforts to play this insurer's role
because markets of contingent claims are incomplete. The government can only issue oneperiod, non-state-contingent debt.
Mendoza and Oviedo (2004) show that, in this environment with incomplete
contingent-claims markets, a government averse to a collapse in its public outlays and
facing revenue uncertainty will impose on itself a "natural debt limit" (NDL) determined
by the growth-adjusted annuity value of the primary balance in a state of "fiscal crisis."
They define a state of fiscal crisis as the one at which a country arrives after experiencing a
sufficiently long sequence of adverse shocks to public revenues and the government
adjusts its outlays to minimum admissible levels.
An important implication of the NDL is that it allows the government to offer its
creditors a credible commitment to remain able to repay "almost surely" at all times
(including during fiscal crises). It is important to note that this commitment is not an adhoc assumption but an implication of the assumptions that (a) government is averse to
suffering a collapse of its outlays, (b) public revenues are stochastic, and (c) markets of
contingent claims are incomplete. However, the commitment is in terms of an "ability to
pay criterion," and as such it does not rule out default scenarios that may result from
"willingness to pay" or strategic reasons.
The NDL sets the upper bound for public debt but is not, in general, the same as the
"sustainable" or equilibrium level of debt. The model does not require public debt to
remain constant at the level of the NDL. Instead, the path of "sustainable" or equilibrium
public debt depends on the initial conditions of debt and revenue, the probabilistic process
driving revenues, and the policy rules governing public outlays. In the short-run, the
dynamics of the distribution of public debt are driven by the government budget constraint.
In the long run, depending on alternative sequences of realizations of public revenues, the
government can end up retiring all the debt or hitting the debt limit. This extreme behavior
of the long-run debt dynamics is particular to this basic version of the Mendoza-Oviedo
167
model. The two-sector general equilibrium variant of their framework with endogenous
government outlays features a unique, invariant long-run distribution of public debt (see
Mendoza and Oviedo (2004) for details).
The results suggest that current debt ratios in Brazil and Colombia are near the
natural debt limits that would be consistent with fiscal solvency only if one assumes
perceived commitments to large reductions in non-interest outlays (in excess of 6
percentage points of GDP) in a fiscal crisis. Mexico was also near its debt limit in 1998.
However, in Mexico's case the reduction in outlays that would have been needed to keep
the government solvent if revenues had continued to suffer adverse shocks was smaller (at
1.5 percentage points of GDP). Still, in all three cases the implied level of adjustment in
outlays to keep the government solvent in a state of fiscal crisis is more than two standard
deviations below recent averages, indicating that the likelihood that the countries could
produce the required adjustment in outlays if they were called to act on their perceived
commitment is low. In contrast, the model indicates that Costa Rica's largest public debt
ratio of the last 12 years could be financed in a state of fiscal crisis with a modest cut in
outlays equivalent to 1.2 times the standard deviation of outlays.
The above results are very sensitive to underlying assumptions regarding the longrun real interest rate and growth rate. Current debt ratios in all four countries are found to
be unsustainable for plausible reductions in growth rates to the averages of the last 20
years, instead of the average of the last 40 years used in the baseline scenario. Similarly,
the current debt ratios are found to be unsustainable if the long-run real interest rate is set
at 8 percent instead of the 5 percent value of the baseline estimates.
The Mendoza-Oviedo framework was designed as a forward-looking policy tool
that intentionally sets aside the risk of default. This was done because the framework is
intended to produce the sustainable debt ratios that a government that does not consider the
option of defaulting on its obligations could support. This is in line with the assumptions
of the traditional approaches to assess debt sustainability based on deterministic steadystate estimates or empirical applications of the intertemporal government budget
constraint. However, while this approach is the ideal benchmark in a forward-looking
policy analysis, it does have the drawback that it does not take into account how default
risk considerations could affect sustainable debt dynamics. To address this issue, we study
in this paper how estimates of natural debt limits and simulated debt dynamics vary when
the basic Mendoza-Oviedo model is modified to incorporate exogenous default risk.
Introducing default risk results in marked reductions in the levels of natural debt limits.
We also compare the results of the Mendoza-Oviedo model with those produced by
the conventional methodology based on calculations of steady-state debt ratios (or
"Blanchard ratios"). In the countries where the average of revenues exceeds that of outlays
by a large enough margin (Costa Rica and Mexico), the Blanchard ratio yields higher debt
ratios than the natural debt limits of the Mendoza-Oviedo model. This finding shows that
assessments of sustainable debt based on steady-state calculations that use averages of
revenues and outlays and fail to take into account aggregate, non-insurable fiscal shocks
can lead countries to borrow more than what is consistent with fiscal solvency. In the
countries where the averages of revenues and outlays are very similar (Brazil and
Colombia), the Blanchard ratios yield negligible debt ratios. The Mendoza-Oviedo model
can explain high levels of debt in these cases if the government can commit (credibly) to
large enough cuts in outlays in a state of fiscal crisis. To be credible, however, these cuts
168
must not represent unusually large deviations relative to the historical mean (i.e., they
should not be larger than two standard deviations in percent of the mean).
The paper is organized as follows. Section 2 is a short survey of the existing
methods for calculating public debt ratios consistent with fiscal solvency. Section 3
summarizes the one-good variant of the Mendoza-Oviedo model. Section 4 applies the
model to the cases of Brazil, Colombia, Costa Rica and Mexico and discusses the results.
This Section includes sensitivity analysis and an extension to incorporate exogenous
default risk. Section 5 reflects on important caveats of the analysis and provides general
conclusions.
Computing Public Debt Ratios Consistent with Fiscal Solvency: A Survey
A central question in fiscal policy discussions is how to determine whether the stock of
public debt is "sustainable," in the sense of being consistent with the fiscal solvency
conditions implied by current and future patterns of government revenue and outlays.
Developing effective tools for computing these sustainable public debt ratios has proven a
difficult task.
The first problem that studies in this area face is how to give operational content to
the notion of fiscal sustainability. There is a tendency to associate the notion of
unsustainable public debt with failure to satisfy the government budget constraint, or with
the government holding a negative net-worth position. However, from an analytical
standpoint these can be misleading because the "true" government budget constraint (as an
accounting identity relating the overall public sector borrowing requirement to all sources
and uses of government revenue) must always hold. Thus, an analysis that shows that a
given stock of public debt fails a "particular" definition of the budget constraint is
ultimately reflecting the fact that the analysis failed to incorporate important features of the
actual fiscal situation of the country under study. How this failure translates into a
judgment about the sustainability of public debt depends on assessments (typically implicit
in the analysis) about the macroeconomic outcomes associated with the different
mechanisms that can be used to maintain fiscal balance. Arguments about sustainability
are therefore implicitly arguments about the pros and cons of these alternative
mechanisms, not about whether the intertemporal budget constraint of the government
holds.
Consider a basic example. In the canonical long-run analysis of public debt
sustainability, we consider long-run, average levels of public revenue and expenditures,
and view as the sustainable debt-output ratio the annuity value of a long-run target of the
primary balance-output ratio. If a government has a large stock of contingent liabilities
because of the high risk of a banking crisis, the stock of public debt may be judged to be
unsustainable because, once these contingent liabilities are added, the debt-output ratio
exceeds the long-run indicator of sustainability.
However, the government budget
constraint must hold, and thus if a banking crisis does occur the government will ultimately
have to adjust the primary balance or rely on other "sources" of financing such as the
inflation tax or a debt default. Adjustment via the primary balance is generally judged as
consistent with this canonical view of sustainability, while adjustment via inflation or
default would not because these would be viewed as alternatives inferior to adjustment of
the primary balance in terms of social welfare.
169
Beyond the problem of defining an operational concept of public debt

sustainability, there are also problems in the design of methods for calculating sustainable
debt levels. These difficulties reflect the gap between the aspects of fiscal policy
emphasized in the different methods and those aspects that seem empirically relevant for
explaining the actual fiscal position of governments. The literature on methods for
assessing public debt sustainability reflects the evolution of ideas on these issues. In the
lines below we review the main features of the different methods. The intent is not to
conduct a comprehensive survey of the literature (as there are already a number of
excellent surveys, see, for example, Chalk and Hemming 2000 or IMF 2002, and 2003 a)
but to highlight the central differences among the existing methods.
The starting point of most of the existing methods for calculating sustainable public
debt-output ratios is the period budget constraint of the government. This constraint is
merely an accounting identity that relates all the flows of government receipts and
payments to the change in public debt:
0)
where Bt+i is the stock of public debt issued by the end of period /, Bt is maturing public
debt on which the government pays principal and the real interest rate rr, Tt is total real
government revenue and Gt are current real government outlays (i.e., Tt- Gt is the primary
fiscal balance).
Long-Run Methods
As the example of the long-run approach to debt sustainability given earlier illustrated, the
long-run approach is based on steady-state, perfect-foresight considerations that transform
the government's accounting identity (1) into an equation that maps the long-run primary
fiscal balance as a share of output into a "sustainable" debt-to-output ratio that remains
constant over time (see Buiter 1985, Blanchard 1990, and Blanchard, Chouraqui,
Hagemann and Sartor 1990). In particular, if we define the long-run rate of output growth
as y and after some basic algebraic manipulation, the accounting identity in (1) yields:
(2)
where b is the long-run debt-to-GDP ratio, t and g are the long-run GDP shares of current
revenue and outlays, and r is the steady-state real interest rate. Condition (2) can be read
as an indicator of fiscal policy action (i.e., of the "permanent" primary balance-output ratio
that needs to be achieved by means of revenue or expenditure policies so as to stabilize a
given debt-output ratio), or as an indicator of a "sustainable" debt ratio (i.e., the target
debt-output ratio implied by a given projection of the long-run primary balance-output
ratio)
170
Intertemporal Methods
An important shortcoming of the long-run approach is that it fails to recognize that the
"long run" is essentially a theoretical construct. In the short run governments face a budget
constraint that does not reduce to the simplistic (albeit significantly more tractable)
formula of the long-run analysis. There can be temporarily high debt ratios, or temporarily
large primary deficits, that are consistent with government solvency, and indeed incurring
in such temporarily high debt or deficits could be optimal from a tax-smoothing
perspective. To force a country into the straight jacket of keeping its public debt-output
ratio no larger than the level that corresponds to the long-run stationary state can therefore
be a serious mistake. The realization of these flaws in the long-run calculations led to the
development of intertemporal-budget-constraint methods that shifted the focus from
analyzing directly the debt-output ratio to studying the time-series properties of the fiscal
balance, so as to test whether these properties are consistent with the conditions required to
satisfy the government's intertemporal budget constraint. This intertemporal constraint
serves as a means to link the short-run dynamics of debt and the primary balance with the
long-run solvency constraint of the government.
In their original form (see Hamilton and Flavin 1986) the intertemporal methods
aimed to test whether the data can reject the hypothesis that the condition ruling out Ponzi
games on public debt holds. This condition states that at any date / the discounted value of
the stock of public debt /+/' periods into the future should vanish as j goes to infinity:
In other words, in the long run the stock of debt cannot grow
faster than the gross interest rate. If this no-Ponzi-game (NPG) condition holds, the
forward solution of (1) implies that the intertemporal government budget constraint holds:
The present value of the primary fiscal balance is equal to the value of the existing stock of
debt, and hence the exiting public debt or public debt-output ratio is deemed "sustainable."
A number of articles tried different variations of this test by testing for stationarity
and co-integration in the time series of the primary balance and public debt, and produced
different results using U.S. data (Chalk and Hemming, 2000, review this literature). These
intertemporal-budget-constraint methods also began to introduce elements of uncertainty
into public debt sustainability analysis, but mostly in an indirect manner as sources of
statistical error in hypothesis testing, or by testing the NPG condition in expected value or
as an orthogonality condition that considers the fact that at equilibrium the sequence of real
171
interest rates used to discount the "terminal" debt stock must match the intertemporal
marginal rate of substitution in private consumption.l
The article by Bohn (1998) provided an alternative interpretation of intertemporal
methods that reduces them to testing if the primary balance responds positively to
increases in public debt. In particular, // the primary balance-output ratio and the debtoutput ratio are stationary, the following regression can be used to test for sustainability:
(3)
where st is the ratio of the primary fiscal balance over GDP, st is a well-behaved error term
and Zt is a vector of determinants of the primary balance other than the initial stock of
public debt. In Bohn's case, he included in Zt the cyclical variations in U.S. GDP and a
measure of "abnormal" government expenditures associated with war episodes. He found
strong evidence in favor of p > 0. Thus, controlling for war-time spending and the
business cycle, the debt-output ratio is mean-reverting in U.S. data. Moreover, the positive
coefficient p, indicating that the primary balance displays a linear response that is both
positive and systematic to increases in debt, is sufficient (albeit not necessary) to ensure
that the intertemporal government budget constraint holds. This is because, as the more
recent methods for evaluating fiscal sustainability under uncertainty emphasize, ensuring
fiscal sustainability requires that above a certain critical maximum level of debt 6*, the
primary balance responds positively (in either linear or non-linear fashion) to changes in
public debt. The intuition is that if this is the case any large increase in debt due to " large
negative shocks" is eventually reversed through primary surpluses.
Chapter HI of IMF (2003b) applied Bohn's method to a panel of industrial and
developing country data. The results identified a group of developing countries where the
sustainability condition p > 0 holds, and others where it does not. Moreover, the study
also shows evidence of non-linearities in the relationship between debt and primary
balances, indicating that countries that were able to sustain larger debt ratios in the data
also displayed a stronger response of the primary balance to debt increases.
Recent Development: Probabilistic Methods and Methods with Financial Frictions
Recent developments in public debt sustainability analysis follow two strands. The first
strand emphasizes the fact that governments, particularly in emerging markets, face
significant sources of uncertainty as they try to assess the patterns of government revenue
and expenditures, and hence the level of debt that they can afford to maintain. From the
perspective of these probabilistic methods, measures of sustainability derived from the
long-run approach or the intertemporal analysis are seen as inaccurate for governments that
In expected value the NPG condition is
, in the "equilibrium" tests the
condition becomes
172
hold large stocks of debt and face large shocks to their revenues and expenditures. The
key question here is not whether public debt is sustainable at some abstract steady state, or
whether in a sample of a country's recent or historical past the NPG condition holds. The
key question is whether the current debt-output ratio is sustainable given the current
domestic and international economic environment and its future prospects.
The second strand aims to incorporate elements of the financial frictions literature
applied to the recent emerging-markets crises. For example, public debt in many emerging
markets displays a characteristic referred to as "liability dollarization" (i.e., debt is often
denominated in foreign currency or indexed to the price level.). As a result, abrupt
changes in domestic relative prices that are common in the aftermath of a large
devaluation, or a 'sudden stop' to net capital inflows, can alter dramatically standard longrun calculations of sustainable debt ratios and render levels of debt that looked sustainable
in one situation unsustainable in another. Calvo, Izquierdo and Talvi (2003) evaluate these
effects for the Argentine case and find that large changes in the relative price of
nontradables alter significantly the assessments obtained with standard steady-state
sustainability analysis. The general version of the Mendoza-Oviedo model also
incorporates these effects (see Mendoza and Oviedo 2004).
The probabilistic methods for assessing fiscal sustainability propose alternative
strategies for dealing with macroeconomic uncertainty. A method proposed at the IMF by
Barnhill and Kopits (2003) incorporates uncertainty by adapting the value-at-risk (VaR)
principles of the finance industry to debt instruments issued by governments. The aim of
this approach is to model the probability of a negative net worth position for the
government. The method requires estimates of the present values of the main elements of
the balance sheet of the total consolidated public sector (financial assets and liabilities,
expected revenues from sales of commodities or other goods and services, as well as any
contingent assets and liabilities), and an estimate of the variance-covariance matrix of the
variables that are viewed as determinants of those present values in reduced form. This
information is then used to compute measures of dispersion relative to the present values
of the different assets and liabilities that determine the value at risk (or exposure to
negative net worth) of the government.
A second probabilistic method recently considered for country surveillance at the
IMF (2003c) modifies the long-run method to incorporate variations to the determinants of
sustainable public debt in the right-hand-side of equation (2), and also examines short-term
debt dynamics that result from different assumptions about the short-run path of the
variables that enter the government budget constraint in deterministic form. For example,
deterministic debt dynamics up to 10 periods into the future are computed for variations of
the growth rate of output of two standard deviations relative to its mean.
The same IMF publication proposes a stochastic simulation approach that computes
the probability density function of possible debt-output ratios. This stochastic simulation
model, like the VaR approach, is based on a non-structural time-series analysis of the
macroeconomic variables that drive the dynamics of public debt (particularly output
growth, interest rates, and the primary balance). The difference is that the stochastic
simulation model produces simulated probability distributions based on forward
simulations of a vector-autoregression model that combines the determinants of debt
dynamics as endogenous variables with a vector of exogenous variables. The distributions
173
are then used to make assessments of sustainable debt in terms of the probability that the
simulated debt ratios are greater or equal than a critical value.
Xu and Ghezzi (2002) developed a third probabilistic method to evaluate
sustainable public debt. Their method computes "fair spreads" on public debt that reflect
the default probabilities implied by a continuous-time stochastic model of the dynamics of
treasury reserves in which exchange rates, interest rates, and the primary fiscal balance
follow Brownian motion processes (so that they capture drift and volatility observed in the
data). The analysis is similar to that of the first-generation models of balance-of-payments
crises. Default occurs when treasury reserves are depleted, and thus debt is deemed
unsustainable when the properties of the underlying Brownian motions are such that the
expected value of treasury reserves declines to zero (which occurs at an exponential rate).
The Basic Version of the Mendoza-Oviedo Model
The probabilistic methods summarized in the last section make significant progress in
incorporating macroeconomic uncertainty into debt sustainability analysis but they are
largely based on non-structural econometric methods. In contrast, the Mendoza -Oviedo
(MO) method aims to provide an explicit dynamic equilibrium model of the mechanism by
which macroeconomic shocks affect government finances. The MO method also differs
from the other probabilistic methods in that it models explicitly the nature of the
government's forward-looking commitment to remain solvent, instead of focusing on
computing estimates of exposure to negative net worth or depletion of treasury reserves.
As explained below, the MO method determines sustainable debt ratios that respect a
natural debt limit consistent with a credible commitment to repay similar in principle to the
one implicit in the long-run and intertemporal methods.
The structural emphasis of the MO approach comes at the cost of the reduced
flexibility and increased complexity of the numerical solution methods required to solve
non-linear, dynamic stochastic equilibrium models with incomplete asset markets. At the
same time, by proceeding in this manner the MO framework seeks to produce estimates of
sustainable public debt that are robust to the Lucas critique. The non-structural or reducedform tools used in the other probabilistic methods to model the dynamics of public debt are
vulnerable to the policy instability problems resulting from the Lucas critique. This is not a
serious limitation when these methods are used for an ex-post evaluation of how well/7orf
debt dynamics matched fiscal solvency conditions, but it can be a shortcoming for a
forward-looking analysis that requires a framework for describing how equilibrium prices
and allocations, and hence the ability of the government to raise revenue and service debt,
adjust to alternative tax and expenditure policies or other changes in the environment
(including, for example, a Sudden Stop limiting the access to international capital markets
or a collapse in the real exchange rate triggered by a devaluation of the currency).
The basic principles of the MO method are as follows. Assume that output follows
a deterministic trend (so that it grows at a constant, exogenous rate y) and the real interest
rate is constant. Public revenues follow an exogenous stochastic process. The government
is extremely averse to suffering a collapse in its outlays. Hence, it aims to keep its outlays
smooth unless the loss of access to debt markets forces it to adjust them to minimum
tolerable levels. The government can only issue non-state-contingent debt. The
government budget constraint in (1) can then be re-written as:
174
0+r)**, =*,(!+')-(',-*,)
(4)
where lowercase letters refer to ratios relative to GDP.

Since the government wants to rule out a collapse of its outlays below their
tolerable minimum levels, it would not want to borrow more than the amount of debt it
could service if the primary balance were to remain forever (or "almost surely" in the
language of probability theory) at its lowest possible value, or "fiscal crisis" state. A state
of fiscal crisis is defined as a situation reached after a "sufficiently" long sequence of the
worst realization of public revenues and after public outlays have been adjusted to their
tolerable minimum. This upper bound on debt is labeled the "Natural Debt Limit" (NDL),
which is the term used in the precautionary-savings literature for an analogous debt limit
that private agents impose on themselves when they can only using non-state-contingent
assets to try to smooth consumption (see Aiyagari 1994). The NDL is given by the
growth-adjusted annuity value of the primary balance in the state of fiscal crisis.
The "history of events" leading to a fiscal crisis has non-zero probability (although
it could be a very low probability) as long as that crisis state is an event within the support
of the probability distribution of the primary balance, and as long as there are non-zero
conditional probabilities of moving into this crisis state from other realizations of the
primary balance. Under these conditions, the government knows that there is some
probability that it could suffer a fiscal crisis in the future, and to ensure it can pay for its
minimum level of outlays it must not hold more debt than it could service then.
Since the NDL is a time-invariant debt level that satisfies the government budget
constraint with revenues and outlays set at their minimum, it follows that the NDL implies
that the government remains able to service its debt even in a state of fiscal crisis. Thus,
the NDL that a government imposes on itself in order to self-insure so as to rule out a
collapse of public outlays below its tolerable minimum also allows that government to
offer lenders a credibly commitment to remain able to repay its debt in all states of nature.
To turn the above notions of the NDL and its implied credible commitment to
repay into operational concepts, we need to be specific about the factors that determine the
probabilistic dynamics of the components of the primary balance. On the revenue side, the
probabilistic processes driving tax revenues reflect the uncertainty affecting tax rates and
tax bases. These processes have one component that is the result of domestic policy
variability and the endogenous response of the economy to this variability, and another
component that is largely exogenous to the domestic economy (which typically results
from the nontrivial effects of factors like fluctuations in commodity prices and commodity
exports on government revenues). The one-sector version of the MO model used in this
paper incorporates explicitly the second component.2
On the expenditure side, government expenditures adjust largely in response to
policy decisions, but the manner in which they respond varies widely across countries.
Industrial countries tend to display countercyclical fiscal policies while fiscal policy in
2
Note that the exogenous determinants of public revenue dynamics can be important even in economies that
have successfully diversified their exports away from primary commodities. In Mexico, for example, oil
exports are less than 15 percent of total exports but oil-related revenues still represent more than 1/3 of public
sector revenue.
175
emerging markets is generally procyclical (as the recent literature on procyclical fiscal
policy has shown). In addition, the "adjustment" or minimum level of public outlays to
which the government can commit to adjust in a fiscal crisis is particularly important for
determining the NDL and the sustainable debt ratios in the MO model.
Labeling the fiscal-crisis level (or lowest realization) of the government revenueGDP ratio as f mm and the minimum level of the ratio of outlays to GDP that the government
can commit to deliver as g"in (for g^m < tmn\ it follows from the government budget
constraint in (4) that the NDL is the value of b* given by:
(5)
This NDL is lower for governments that have (a) higher variability in public
revenues (for example, if fiscal revenue follows a symmetric Markov chain that obeys the
rule of simple persistence, the value off" can be written as a multiple of the standard
deviation of public revenues and hence lower values of tmm reduce the debt limit), (b) less
flexibility to adjust public outlays, and (c) lower growth rates or higher real interest rates.
The NDL represents a credible commitment to repay in the sense that it ensures
that the government remains able to repay even in a state of fiscal crisis for a given known
stochastic process driving revenues and a given policy setting the minimum level of
outlays. However, this should not be interpreted as suggesting that the need to respect the
NDL rules out the possibility of sovereign default. Default triggered by "inability to pay"
remains possible if there are large, unexpected shocks that drive revenues below what was
perceived to be the value of tmm or if the government turns out to be unable to reduce
outlays to g^m when a fiscal crisis hits. In addition, default triggered by "unwillingness to
pay" remains possible since the NDL is only an ability to pay criterion that cannot rule out
default for strategic reasons. Later in Section 4 we explore an extension of this framework
that incorporates default risk into the basic MO setup.
Consider a government with exogenous, random fiscal revenues (say, for example,
oil export revenues) and an ad-hoc smoothing policy rule for government expenditures,
such that gt = g (for g^g*71"7) as long as &,+; > &*, otherwise gt adjusts to satisfy condition
(4). By (4) and (5), if at a particular date the current debt ratio is below 6* and the
realization of the revenue-output ratio is tmm, the government finances g by increasing bt+j.
In contrast, if at some date the current debt ratio is at b* and the realization of revenues is
f", (4) and (5) imply that g^g*'".
In a simple example with zero initial debt, it is straightforward to show that if the
government keeps drawing the minimum realization of public revenue, it will take the T
periods that satisfy the following condition for the government to hit the NDL:
(6)
This result indicates that, in the worst-case scenario in which revenues remain
"almost surely" at their minimum, the government can access the debt market to keep the
ratio of public outlays at the level g for a longer period of time the larger is the excess of
176
"normal11 government outlays over minimum government outlays relative to the excess of
normal outlays over the minimum level of revenues. Thus, the government uses debt to
keep its outlays as smooth as possible given its capacity to service debt as determined by
the volatility of its public revenues, reflected in the value of tmm , and its ability to reduce
public outlays in a fiscal crisis, reflected in the value of g7""2.
It is critical to note that the key element of the expenditure policy is not the level of
g"m per se but the credibility of the announcement that outlays would be so reduced during
a fiscal crunch. The ability to sustain debt and the credibility of this announcement depend
on each other because a government with a credible ex ante commitment to major
expenditure cuts during a fiscal crisis can borrow more and access the debt market for
longer time, and hence, everything else the same, this government faces a lower probability
to be called to act on its commitment. In a more general case in which public revenue is
not an exogenous probabilistic process but it is in part the result of tax policies and their
interaction with endogenous tax bases, the credibility argument extends to tax policy.
Governments that can credibly commit to generate higher and less volatile tax revenueoutput ratios will be able to sustain higher levels of debt, and to the extent that this helps
the economy produce stable tax bases it helps support the credibility of the government's
ability to raise revenue.
The condition defining the NDL in (5) has a similar form as the formula for
calculating sustainable debt ratios under the long-run method:
However, the
implications for assessing fiscal sustainability under the two methods are sharply different.
In general, the long-run deterministic rule will always identify as sustainable a debt-output
ratio that is unsustainable once uncertainty of the determinants of the fiscal balance and the
NDL are taken into account. This is because the long-run method ignores the role of the
volatility of the elements of the fiscal balance, while in the MO model one finds that,
everything else the same, governments with less variability in tax revenues can sustain
higher debt ratios.
Consider the case of two governments with identical long-run averages of tax
revenue-output ratios at 20 percent. The tax revenue-output ratio of government A has a
standard deviation of 1 percent relative to the mean, while that of Government B has a
standard deviation of 5 percent relative to the mean. Assuming for simplicity that the
distributions of tax revenue-output ratios are Markov processes with tmm set at two standard
deviations below the mean, the probabilistic model would compute the natural debt limit
for A using a value of tmm of 18 percent, while for B it would use 10 percent. The
deterministic long-run method yields the same debt ratio for both governments and uses
the common 20 percent average tax revenue-output ratio to compute it. In contrast, the
MO method would find that debt ratio unsustainable for both governments and would
produce a debt limit for B that is lower than that for A.
Another important difference between the two methods is in the way that the MO
method views the debt limit vis-a-vis the way in which the long-run method views the
steady-state debt ratio. In the long-run analysis, the debt ratio is viewed as either a target
ratio to which a government should be forced to move to, or as the anchor for a target
primary balance-GDP ratio that should be achieved by means of a policy correction. In
contrast, in the MO method condition (5) defines only the maximum level of debt. Unless
the NDL binds, this maximum level of debt is not the equilibrium or sustainable level of
177
debt that should be issued by the government (although it clearly plays a central role in
determining both). Depending on the probabilistic and policy assumptions driving
revenues and expenditures (and, generalizing condition (5), depending also on the
stochastic processes of the real interest rate and output growth), a country can have levels
of debt lower (even much lower) than this critical level, and may take a very long time on
average to enter a state of fiscal crisis or even never arrive at it.
The MO methodology models uncertainty in the form of discrete Markov
processes. Given information on the current stock of public debt, the current tax revenueGDP ratio and the assumed behavioral rules for government outlays and statistical
moments of the public revenue process, the model produces conditional one-period-ahead
and unconditional long-run distributions of the debt-output ratio, as well as estimates of the
average number of periods in which b* is expected to be reached from any initial bt.
Again, depending on the nature of the stochastic processes and policy rules of revenues
and outlays, it may take a few quarters to hit the debt ceiling on average or it may take an
infinite number of quarters to do it
Results of the MO Method for Brazil, Colombia, Costa Rica, and Mexico
This section applies the MO method described in the last section to the cases of Brazil,
Colombia, Costa Rica and Mexico. The section begins with a brief review of the recent
growth performance and evolution of public debt, public revenues and public expenditure
ratios in these countries. This review serves to identify key parameters needed to simulate
the debt dynamics and solve for natural debt limits.
Review of Growth Performance and Fiscal Dynamics
The growth performance of the four countries examined in this study over the last two
decades was weak. As shown in Table 1, average growth in GDP per capita for the period
1981-2000 ranged from a minimum of 0.5 percent in Brazil to a maximum of 1.25 percent
in Costa Rica. These countries grew at faster rates in the past. Taking averages starting in
1961, the smallest average per-capita GDP growth rate was 1.8 percent for Costa Rica and
the largest was 2.6 percent for Brazil. Given the apparent structural breaks in the trend of
GDP per capita of these countries since the early 1980s, we conduct our public debt
analysis under a baseline growth scenario that uses 1961-2000 average growth rates, and
compare the results with a scenario that views the growth slowdown of the last two
decades as permanent by using 1981-2000 average growth rates.
Reliable cross-country estimates of public debt stocks and interest rates on public
debt at the general government level are hard to obtain. We use the figures documented in
IMF (2003b) for ratios of public debt to GDP for the period 1990-2002. As shown in
Table 1, the mean debt-GDP ratios for the full sample range from 33.7 percent in
Colombia to 49.5 percent in Costa Rica. However, these full-sample averages hide the fact
shown in Figure 1 that debt ratios have been growing rapidly (except in the case of Costa
Rica, where the ratio has been fairly stable but also fairly high). If we split the sample to
create averages for 1990-1995 and 1996-2002, we find that the mean debt ratios of Brazil,
Colombia and Mexico increased by about 10 percentage points between the first period
and the second periods. By the end of the sample period, all four countries displayed debt
178
ratios of about 50 percent (see Figure 1). The key question to answer is therefore whether
a debt ratio of 50 percent is consistent with fiscal solvency given the pattern of growth and
interest rates and the volatility of fiscal revenues that these countries face.
Real interest rates on public debt are hard to measure because of differences in
maturity, currency denomination, indexation factors, and residence of creditors for the
various debt instruments that each country issues, and also because of the existence of
time-varying default and exchange-rate risk premia. One useful proxy is the measure of
sovereign risk proposed by Neumeyer and Perri (2003), which is the spread of the EMBI+
index relative to the U.S. T-Bill rate deflated by an estimate of expected inflation in the
U.S. GDP deflator. The sample period of this measure is relatively short (starting in 1994)
and biased because it includes mainly observations for a turbulent period in world capital
markets. Still, this measure of real interest rates on public debt shows substantial premia
over the world's risk free rate (the average real interest rates including the risk premia for a
quarterly sample from 1994:1 to 2002:2, are 12.9 percent for Brazil and 10.3 percent for
Mexico). The real interest rate on debt instruments will deviate from these figures to the
extent that (a) debt instruments are domestic rather than external, (b) debt is denominated
in domestic currency rather than in foreign currency or in terms of an indexation factor,
and (c) the debt instruments have liquidity or transactions value for the holders. The lower
bound of the interest rate on public debt is set by the real interest rate on U.S. public debt.
The 1981:1-2000:4 average of the U.S. 90 day T-bill rate deflated by observed U.S. CPI
inflation is about 2.8 percent.
Given the above considerations, we consider two interest rate scenarios in our
calculations. The baseline scenario considers a real interest rate of 5 percent, which
incorporates a small long-run default risk premium of about 2.2 percentage points above
the U.S. T-bill rate. The alternative is a high-real-interest-rate scenario in which the real
interest rate is set at 8 percent (i.e., the long-run default risk premium is about 5.3
percentage points above the T-bill rate). In both scenarios we remain relatively optimistic
about growth prospects, using average growth rates for the period 1961-2000.
The measure of public revenues that we need to isolate for conducting the debt
analysis is the total of all tax and non-tax government revenues excluding grants. For
government expenditures we require the total non-interest government outlays (including
all expenditures and transfer payments and excluding all forms of debt service). Once
again, limitations of the existing international databases make it difficult to retrieve
consistent measures of these variables that apply at the level of the entire non-financial
public sector and, in the case of the outlays, that include the annuity values of all
contingent liabilities resulting from obligations like banking- or pension-system bailouts.
We put together estimates of the revenue and outlay ratios for the four countries under
study by combining data from IMF (2003 b) and World Development Indicators (see the
footnotes to Table I for details). The former shows data for the non-financial public sector
but it does not separate interest and non-interest expenditures for some countries. The latter
separates interest and non-interest expenditures but reports data only for the central
government. Thus, the revenue and expenditures data are not exactly comparable across
countries.
The average ratios of total public revenue to GDP during the period 1990-2002
ranged from 12.6 percent in Colombia to 23 percent in Mexico (Table 1). Turning to
examine the volatility of public revenues, Costa Rica displays the lowest coefficient of
179
variation in the revenue-output ratio, at 5.4 percent of the mean. The coefficient of
variation of revenues is similar in Colombia and Mexico (8.9 and 8 percent respectively).
Brazil shows the highest coefficient of variation of public revenues at 14.1 percent.
Interestingly, the country with the lowest coefficient of variation in public revenues (Costa
Rica) is also the one with the highest average public debt-GDP ratio.
The average of total non-interest outlays as a share of GDP is very similar in
Brazil, Costa Rica and Mexico, at about 19 percent. The average outlays ratio for
Colombia is much lower, at 12.8 percent, but this is one of the measures that applies only
to the central government, so the actual ratio for the non-financial-public sector must be
higher. With regard to the variability of non-interest outlays, Mexico shows the smallest
coefficient of variation (about 4 percent), followed by Costa Rica (10 percent). Brazil and
Colombia show the highest coefficients of variation in the outlays ratio, at about 14 percent
for both.
Natural Debt Limits: Baseline Scenario and Two Alternatives
Table 1 reports three sets of calculations of natural debt limits for Brazil, Colombia, Costa
Rica and Mexico. The baseline scenario considers the 1961-2000 average growth rates of
GDP per capita and a 5 percent real interest rate. The growth- slowdown (GS) scenario
uses the 1981-2000 average growth rates (with the real interest rate still at 5 percent). The
high-real-interest-rate (HRIR) scenario uses a real interest rate of 8 percent (with the
growth rates set at the 1961-2000 averages).
The baseline scenario differs from the other two because it is designed to produce a
coefficient of fiscal adjustment that yields an NDL equal to the largest debt ratio observed
for each country in the 1990-2002 period. Table 1 reports this coefficient of "implied
fiscal adjustment" in terms of how many standard deviations relative to the mean of noninterest outlays are needed to yield a debt limit equal to maximum observed debt (given the
data for means and coefficients of variation of revenues and outlays, the average growth
rates, the real interest rate, and assumed floors of public revenue equal to two standard
deviations below the corresponding means). The Table also shows the implied minimum
ratio of outlays to GDP resulting from the coefficient of fiscal adjustment. The GS and
HRIR scenario keep the same fiscal adjustment coefficient and just alter either the growth
rate or the real interest rate.
The public debt-GDP ratios of the four countries under study peaked at similar
levels during the 1990-2002 period (ranging from 50 percent in Colombia to 56 percent in
Brazil). The coefficients of implied fiscal adjustment reported in Table 1 show that, in
order to produce NDLs that can support these high levels of debt, three of the four
countries (Brazil, Colombia and Mexico) need credible commitments to undertake large
cuts in outlays if they were to hit a fiscal crisis. The adjustment measures 2 standard
deviations below the mean of non-interest government outlays in Mexico, 2.3 in Colombia
and 2.6 in Brazil (which are equivalent to cuts relative to the average outlays-GDP ratios of
1.5, 4 and 6.7 percentage points respectively). The fact that these adjustments are outside
the two-standard-deviation threshold of the data for non-interest outlays of the previous
decade suggests that the probability that the governments could produce these large
expenditure cuts if they needed to act on their commitment is low.
180
The situation in Costa Rica seems more favorable. Costa Rica can support a
credible commitment to repay a debt-output ratio as large as 53 percent of GDP with a
fiscal adjustment of 1.2 standard deviations below the mean of non-interest outlays (which
is equivalent to about 2 percentage points of GDP). Thus, the model is consistent with the
data in predicting that Costa Rica, the country with less volatility in public revenues,
should be the one that has a better chance of sustaining a high debt ratio. This is the case
even though Costa Rica's growth rate is the same as Colombia's and is lower than the
growth rates in Mexico and Brazil by 0.4 and 0.7 percentage points respectively.
The potential dangers of using the Blanchard ratios for conducting debt
sustainability analysis are illustrated in the baseline results. The case of Mexico is
particularly striking. The Blanchard ratio, which would compute a sustainable debt ratio
using the difference between the average public revenue and the average government
outlays, yields a debt ratio of about 132 percent, nearly 2.5 times larger than the NDL
produced by the MO model. In the case of Costa Rica, the Blanchard ratio also exceeds
the NDL but by a very small margin. In contrast, in the cases of the two countries where
complying with the NDLs would require the largest expenditure cuts (i.e., where the
consistency of the high debt ratio with fiscal solvency is more questionable), the Blanchard
ratio yields public debt ratios so low that are of little practical relevance. The Blanchard
ratio for Brazil is 3.5 percent and for Colombia the ratio is negative, at -5.3 percent
(because in the data the average outlays exceed the average revenues by 0.16 percentage
points of GDP). In these cases the Blanchard ratio would be switched to its alternative
interpretation, where a target debt ratio would be chosen and the Blanchard ratio would
then be used to set an average primary fiscal balance that the governments should aim to
obtain by adjusting fiscal policy.
Consider next the natural debt limits in the GS and HRIR scenarios. If the growth
slowdown of the last two decades persists, and even assuming that the coefficients of fiscal
adjustment were to remain as high as estimated in the baseline scenario, the current debt
ratios would exceed the natural debt limits of all four countries by large margins. The
situation of Brazil would be particularly compromised, because the current debt ratio of 56
percent would exceed the maximum debt ratio consistent with a fully credible commitment
to repay in the GS scenario by nearly 26 percentage points of GDP.
The HRIR scenario, in which for example a retrenchment of world capital markets,
increased long-run default risk, or the pressure of large fiscal deficits in industrial countries
push the real interest rate on emerging markets public debt to 8 percent, has even more
damaging effects. In this case, even if the growth rates recover to the 1961-2000 averages
and even with the large fiscal adjustment coefficients set in the baseline scenario, the
natural debt limits of all four countries fall to a range between 25.2 percent for Brazil to
27.4 percent for Costa Rica. Notice, however, that the prediction of the model is not that
an increase of the interest rate to 8 percent would trigger immediate fiscal crises in all four
countries. For a fiscal crisis to occur immediately, the increase in the interest rate would
have to be once-and-for-all and permanent. A transitory hike in the real interest rate could
be absorbed in an analogous manner as a transitory downturn in public revenues, and
hence a fiscal crisis would only be triggered after a sufficiently long sequence of adverse
shocks.
181
This last observation highlights again the fact that the NDL is not (in general) the
same as the sustainable or equilibrium level of debt, which is determined by the dynamics
driven by the government budget constraint. We turn now to study these dynamics.
Debt Dynamics
The simulations of debt dynamics consider a grid of public debt-GDP ratios that spans the
interval from 0 to the NDL (the model assumes that if the government budget constraint
yields negative debt at any time, the corresponding fiscal surplus is instead rebated to the
private sector as a lump-sum transfer). The dynamics of sustainable debt can be traced
from any initial public debt ratio in this interval. However, one needs to be careful in
studying the long-run dynamics of debt ratios because this basic version of the MO model
features at least two long-run distributions of public debt, one converging to 0 and the
other to the NDL. Which of these two distributions is attained in the long run depends on
initial conditions.
The prediction that the long-run debt ratio is not determined within the model (i.e.,
that the long-run debt ratio depends on initial conditions) is not a peculiarity unique to the
MO model. An outcome like this is the key prediction for debt dynamics of the classic
tax-smoothing framework of Barro (1979), and it is also in line with the findings on
Ramsey optimal taxation problems in which smooth taxes are optimal taxes (see Chapter
12 of Ljungqvist and Sargent 2000).
The stochastic processes of public revenues used in the simulations are
characterized by time-invariant Markov chains. The Markov chains are defined by three
objects: an /7-element vector of realizations of the revenues, /, an nxn transition probability
matrix, P, and a probability distribution for the initial value of the realization of revenues,
Tio. The typical element of P, Py , indicates the probability of observing revenues f=f ; in
the next period given that revenues are tt\ in the current period.
For each country, the vector of realizations of revenues has 5 elements (w=5). The
lowest value oft is set two standard deviations below average revenues in each of the four
countries under analysis. We use Tauchen's (1986) univariate quadrature method in order
to construct the rest of the elements of / and the matrix P so as to approximate the firstorder autocorrelation and standard deviation of public revenues observed in the data.
The stochastic simulations require also that we generate a 7-period time series of
realizations of revenues, i.e. /;, t2, ... tT, drawn from the Markov vector t. This time series
is constructed using realizations of a uniform random variable u in [0,1] that indicates how
to pick one realization of tax revenue in period t+1 given the realization at time t. The tax
revenue at t+1, tj say, is determined according to
or
The first application of the stochastic simulations is illustrated in Figure 2, which

shows the number of periods that it takes to hit a fiscal crisis (i.e., to hit the NDL) for the
country-specific calibrated parameters and seven initial debt ratios that span the range
between 0.25 and the maximum debt ratio observed in each country in the 1990-2002
sample. From each initial condition of the debt ratio, there are different stochastic paths
that public debt, revenues and outlays can follow in the future, and each of these paths
182
features a different number of periods to hit a fiscal crisis. The tall bars in the figure report
the average time to a fiscal crisis from 1000 simulations of the public-debt dynamics
induced by the government budget constraint in equation (4). The simulated tax revenues
correspond to the realizations of revenues drawn from the Markov chains calibrated to the
revenue processes of each country (see Table 1 for the corresponding averages and
coefficients of variation). Depending on initial conditions, there are simulation in which a
fiscal crisis never occurs (particularly for low initial debt ratios). In these cases, the
average measure of time to fiscal crisis would go to infinity, and hence is omitted from the
charts.
The short bars in the plots show the minimum possible number of periods before
hitting the fiscal crisis. This extreme adverse scenario corresponds to simulations of debt
dynamics under the assumption that, in every period, the government draws the lowest
realization of the public revenue contained in the vector of realizations of revenues.
The top two graphs in Figure 2 show that in Brazil and Colombia the mean time to
a fiscal crisis is high (15 years or more) for an initial debt ratio equal to 0.25. A fiscal
crisis can occur much sooner on average (in less than 2.5 years) when then initial debt ratio
is around 0.5. The difference between the mean and extreme time to hit a fiscal crisis is
larger in Brazil than in Colombia for all initial debt ratios. Furthermore, the extreme
scenario indicates a shorter time to a fiscal crisis in Brazil than in Colombia regardless of
the initial debt ratio. This is because the lowest realization of revenues are calibrated to be
two standard deviations below the mean, so the higher volatility of revenues in Brazil
implies that the crisis-level of revenues for Brazil is more distant from the mean than for
Colombia.
The bottom two graphs in Figure 2 (for the cases of Costa Rica and Mexico) serve
to illustrate how uncertainty affects the dynamics of public debt and the fact that the NDL
is not in general the same as the equilibrium or sustainable debt. The fact that these plots
only show bars for the mean time to a fiscal crisis for high debt ratios (above 50 percent)
indicates that in the sampling of 1000 stochastic simulations starting from debt ratios
below 0.5, there were many cases in which the debt ratio converged to zero instead of
reaching the NDL. Consider for example the initial debt ratio of about 0.35 in the plot for
Costa Rica. The extreme measure of time to a fiscal crisis shows that a fiscal crisis will
occur after 10 years of experiencing realizations of revenues two-standard-deviations
below the mean. However, in the 1000 stochastic simulations used to calculate the mean
time to a fiscal crisis, only about half of the simulations reached the NDL and the rest
converged to a zero public debt position in the long run. Similarly, for an initial debt ratio
of 0.4, the debt ratio never reached the NDL in 127 out of 1000 simulations. Thus, the
Costa Rican and Mexican cases illustrate examples in which the sustainable debt dynamics
deviate sharply from the NDL.
Figure 3 shows a sample of simulated time series of the public debt ratio and
illustrates further how much the NDL and the sustainable debt ratios differ. The figure
shows 50 simulations of debt ratios all starting from a common initial ratio of 0.35, using
the parameter values calibrated to Costa Rica. At each date /, a random draw of public
revenues, along with the date-/ debt and the fiscal rules for public outlays, are used to
determine the value of the debt at t+L Notice that, whereas for some paths the debt ratio
increases rapidly to reach the NDL, for other paths it takes a long time to reach it and for
other paths the debt goes to zero. As explained above, for a large range of initial values of
183
public debt the model predicts that debt ratio will reach the debt limit while for other initial
values the debt ratio goes to zero. This implies that for starting values of the debt ratio
above 0.35, the fraction of paths driving the debt to its maximum level increases and for
starting values below 0.35 that fraction decreases.
Default Risk
Up to this point we followed the methodology proposed by Mendoza and Oviedo (2004) in
which sovereign default was set aside to focus on modeling the optimal debt policy
consistent with fiscal solvency and an ad-hoc rule smoothing public outlays. Default risk
was only taken into account in setting the value of the long-run real interest rate used to
solve for the NDLs and the debt dynamics (5 percent in the benchmark case, 8 percent in
the high-interest-rate scenario). However, time-varying default risk premia are an
important feature of public debt in emerging markets. It may make sense for a government
to conduct a forward-looking debt sustainability exercise in which it assumes that default
risk is time invariant as a benchmark scenario, but it is important to study how the results
change when time-varying default risk is introduced.
One important limitation of the analysis of default risk is that existing theoretical
models of default on optimal sovereign debt contracts face serious challenges in explaining
observed debt ratios. The canonical model of Eaton and Gersovitz (1981) considers a riskneutral lender and a risk-averse borrower that has the option of defaulting at the cost of
facing permanent exclusion of the debt market. The lender is willing to take on the risk of
default by charging a rate of interest that incorporates a premium consistent with the
probability of repayment (as implied by a standard arbitrage condition between the
expected repayment on defaultable debt and the return on risk-free bonds). There are wellknown theoretical problems with this setup, related to the classic Bulow-Rogoff critique
showing that the threat of exclusion may not be credible because of the option to enter in
deposit contracts with lenders, but even if the model were not affected by these problems,
recent quantitative studies show that models in the Eaton-Gersovitz tradition support very
small debt ratios of less than 10 percent of GDP (see Arellano 2004). This is because the
models yield probabilities of default that increase too rapidly at very low levels of debt.
Given the above difficulties with the theory of endogenous default risk, we follow a
pragmatic approach that takes into account the same risk-neutral lender of the EatonGersovitz model but incorporates an exogenous probability of repayment calibrated to
match observed default risk premia in emerging markets. The arbitrage condition of the
risk-neutral lender implies:
(?)
In this expression, R is the gross world risk-free real interest rate and Afbj is the
probability of repayment (i.e., l-h(bd is the probability of default). The repayment
probability is modeled as an exponential probability distribution expf-a6^, where the
curvature parameter a determines the speed at which the repayment probability falls as
debt increases.
184
The exponential formulation of default risk has the advantage that it is consistent
with two key properties of the endogenous default probability of the Eaton-Gersovitz line
of models: (a) the probability of default is increasing and convex on the level of debt and
(b) the probability of default is zero if the stock of debt is zero. The formulation fails to
reproduce the property of the Eaton-Gersovitz setup that the probability of default
approaches 1 for a finite "rationing" level of debt at which debtors always find it preferable
to default than to repay (the exponential probability of default approaches 1 asymptotically
as debt goes to infinity). However, the formulation still allows for values of a that yield
very large risk premia for high levels of debt comparable to those observed in the data.
We calibrate the value of a so that the arbitrage condition in (7) holds taking as
given the EMBI+ country risk premium and the public debt ratio observed in Mexico in
1998 (the year of Mexico's maximum debt ratio in the 1990-2002 sample). Mexico's debt
ratio in 1998 was bt = 0.549 and the real interest rate that the country faced on this debt,
measured as the U.S. 90-day T-bill rate plus the EMBI+ spread, was R(bt) = 10.48 percent.
The risk-free rate (i.e., the real U.S. 90-day T-bill rate) was 7?w =3.2 percent. Plugging
these figures into (7) and solving for a yields a = 0.124. Interestingly, the observed
maximum debt and the real interest rate with default premium for Brazil in the year 2002
are very similar to Mexico's 1998 figures (Brazil's debt ratio was 0.56 and the real interest
rate with default risk was 10.8 percent). This suggests that using a common value of a for
the four countries under study is not a bad first approximation.
As we show below, default risk has two important implications for the analysis of
sustainable debt based on the MO model. First, it lowers the levels of NDLs, since the real
interest rates considered in Table 1 are lower than those resulting in the worst state of
nature with default risk. Second, it alters the dynamics of public debt, since the rate of
interest now increases with the level of debt. These two effects result in lower NDLs,
reduced levels of sustainable debt and faster convergence to states of fiscal crisis.
Table 2 shows the effects of introducing time-varying default risk in the
calculations of NDLs.3 In all the estimates shown in this Table, the risk free rate is set at
the 1990-2002 average of the real 90-day T-bill rate, which is 2.36 percent, and the
curvature parameter of the probability of repayments is kept at a = 0.124.
The first panel of Table 2 shows how the benchmark estimates of NDLs change
when default risk is introduced. These benchmark estimates take the same growth rates
and minimum levels of public revenues and outlays as in the benchmark scenario of Table
1. The resulting NDLs are significantly smaller (by 17 to 26 percentage points of GDP)
than those in the benchmark case without default risk. Note that this sharp decline of the
NDLs occurs despite the fact that the risk-free rate at 2.36 percent is about half the longrun real interest rate used in the benchmark scenario of Table 1. The repayment
probabilities at 96 percent and the default risk premia around 4.35 percent are similar
across countries. The NDLs in this case ensure that governments would be able to repay
even during a fiscal crisis, but they still may choose to default on debt ratios about 0.33
with 4 percent probability.
Note that with default risk, the constant rate of interest in the denominator of the formula for the NDL is
replaced with the interest rate including default risk defined in equation (7). Since this interest rate depends
on the level of debt, the NDL is now the solution to a non-linear equation.
185
The second panel of Table 2 shows how NDLs change in the growth slowdown
scenario (which is perhaps more relevant since the data used to calibrate the risk free rate
and the repayment probability function are for the 19901-2002 period). Again, relative to
the growth slowdown scenario of Table 1, we are lowering the risk-free rate from 5 to 2.36
percent and introducing time-varying default risk. To isolate the contribution of the latter,
the third panel of Table 2 shows the NDLs obtained using the growth rates of the growth
slowdown scenario but assuming that there is no default risk so that countries can borrow
at the 2.36 percent risk free rate. Since this rate is about half the one used in Table 1, we
obtain very high NDLs (ranging from 73 to 152 percent of GDP).
Two comparisons are interesting to make using the second and third panels of
Table 2. First, the fact that the NDLs of the growth slowdown scenario in Table 1 (ranging
from 30 to 45 percent of GDP) are much smaller than those of the no-default-risk case in
panel 3 of Table 2, shows that our strategy of setting a long-run real interest rate of 5
percent as a proxy for default risk in the estimates of Table 1 was not a bad approximation.
Second, the calculations of NDLs of the second and third panel differ only because the
second incorporates the time-varying default risk premium (i.e., both have the same riskfree rate of 2.36 percent). Since the NDLs without time-varying default risk are 3 to 5
times larger than those with default risk, this comparison shows that default risk has major
implications for estimates of NDLs.
The last panel of Table 2 re-computes the required adjustments in outlays (i.e. the
values of g71"") needed to support the observed maximum debt ratios of each country in the
1990-2002 sample as NDLs taking into account time-varying default risk. The
adjustments in outlays are significantly larger than those reported in Table 1. Measured in
standard deviations, the required adjustment in outlays exceeds the two-standard deviation
threshold for all countries and it is larger for Mexico than for the other countries.
Measured in terms of percentage points of GDP, the adjustments rank from 9.4 and 6
percentage points for Brazil and Colombia respectively to about 7 percentage points for
Costa Rica and Mexico. This ranking suggests again that the debt positions of Brazil and
Colombia are more difficult to reconcile with fiscal solvency considerations than those of
Costa Rica and Mexico.
Figure 4 illustrates the implications of default risk for the dynamics of public debt
reflected in the measures of time to hit a fiscal crisis. The average time to hit a fiscal crisis
is generally lower once the fact that the default risk premium rises with the level of debt is
considered. Moreover, default risk produces debt dynamics that hit NDLs more often.
This can be seen in the plots for Costa Rica and Mexico. In Costa Rica's plot, for
example, an initial debt ratio of 40 percent produced at least some stochastic simulations in
the sampling of 1000 runs in which the debt vanished when we use a fixed interest rate.
With default risk, however, all 1000 stochastic simulations eventually hit the NDL and on
average it takes about 9 years for the economy to hit a fiscal crisis.
Conclusions, Caveats and Extensions
This application of the basic version of the MO model to the cases of Brazil, Colombia,
Costa Rica and Mexico shows that, with the exception of Costa Rica, public debt ratios are
already close to the natural debt limits that the governments of these countries should
respect if they wish to preserve smooth access to debt markets, and protect the credibility
186
of their perceived commitment to be able to repay their debts. This is the result assuming a
relatively optimistic scenario in which the growth slowdown of the last two decades is
reversed to recover the average per-capita growth rates observed between 1961 and 2000,
and the real interest rate remains at a relatively low level of 5 percent. This baseline
scenario is also optimistic in that it requires credible commitments to large cuts in
government outlays which recent experience indicates are a low-probability event (in terms
of a coefficient of adjustment defined in units of standard deviations relative to the mean of
the ratio of outlays to GDP). Considering less optimistic scenarios in which growth
continues at the low trend of the last two decades or the real interest rate increases to 8
percent, current debt ratios would exceed NDLs even with the same tough stance to cut
outlays in a state of fiscal crisis assumed in the benchmark scenario.
The model predicts that the long-run dynamics of the public debt ratio are
undetermined (or more precisely that there is no unique, invariant limiting distribution for
the debt output ratio). This result needs to be considered carefully. On one hand, the
result does not require very strong assumptions: stochastic revenue, relatively inflexible
outlays and some limit to debt market access (whether an NDL or some an ad-hoc debt
limit). Also, the same outcome would result if we take outlays as given and consider
instead arguments for tax smoothing as in Barro (1979). If these are the maintained
assumptions of fiscal solvency analysis, the estimates of time to a fiscal crisis and the
stochastic simulations of debt ratios shown in Figures 2-4, together with the natural debt
limits, summarize all relevant information for assessing whether observed public debt
dynamics are sustainable. In addition, we would need to worry about how this nonstationarity property affects other statistical methods of debt sustainability analysis that
rely on the existence of a well-defined long-run distribution of debt.
On the other hand, we may question the validity of the assumptions that led to the
result of an indeterminate long-run distribution of debt. In Mendoza and Oviedo (2004)
we propose a setup in which government chooses its outlays optimally instead of setting
ad-hoc rules. In particular, we assume that government outlays yield utility to the private
sector, with the added feature of a Stone-Geary-like utility function that sets minimum
levels of government expenditures. We preserve the spirit of the "tormented insurer's"
problem by keeping the assumption that the government aims to maximize its contribution
to private utility given the randomness of fiscal revenue and the fact that it can only issue
non-state-contingent debt. In this setup, the government has a precautionary-savings
motive that yields a unique, invariant limiting distribution of public debt (resulting from
the trade off between the need to self-insure against the risk of persistent revenue shocks
and the desire to smooth government outlays). The role of the NDL is made clearer
because the desire to respect it emerges from the fact that otherwise the government is
exposed to the risk of experiencing states in which its outlays can be very low, and the
government is very averse to these states because of the constant-relative-risk-aversion
nature of the utility function of public expenditures.
Most of the analysis was conducted giving a limited role to default risk (by simply
setting a long-run, time-invariant real interest rate with a premium above the risk-free rate).
This was done following the approach of the MO model to provide a forward-looking tool
to design fiscal programs with the explicit intention of preserving the government's ability
to fulfill its financial obligations. However, default risk is an important feature of
187
emerging markets of sovereign debt, and hence it is worth adding it to the analysis of debt
sustainability.
We introduced default risk by adopting an exogenous, exponential probability of
repayment that varies with the level of debt. We assumed that lenders are risk neutral and
hence are willing to take default risk by lending at a rate that incorporates the premium that
equates the expected return of risky lending with the risk-free interest rate. The risk
premium function was calibrated to match observed debt ratios and EMBI+ spreads on
sovereign debt. Introducing this change into the basic version of the M-O model produces
smaller debt limits and speeds up the dynamics that lead to states of fiscal crisis in which
NDLs are reached. NDLs that completely ignore default risk support dynamic paths of
sustainable debt with much higher debt ratios than those obtained when default risk is
introduced. However, since the basic M-O model approximated the long-run component
of default risk by adding a constant premium above the risk-free interest rate, it yields
results for debt dynamics that are a much closer approximation to those produced by the
model with time-varying default risk than those of a model that ignores default risk
completely.
The application of the MO model undertaken in this paper did not consider two
other important elements of the dynamics of public debt in emerging economies: the
endogeneity of the tax bases and fiscal policy choices and the role of financial frictions like
liability dollarization. The endogeneity of the tax bases can be incorporated into the
structure of the MO model. This requires introducing the decisions of the private sector
with regard to the variables that determine the allocations and prices that conform tax bases
(such as labor supply, consumption, the current account and capital accumulation).
Similarly, liability dollarization can be introduced by modifying the model to incorporate
tradable and nontradable goods. The dynamic stochastic general equilibrium model
studied in Mendoza and Oviedo (2004) tries to make progress in these two directions.
188
References
Aiyagari, S. Rao (1994), "Uninsured idiosyncratic risk and aggregate saving," Quarterly
Journal of Economics, 109(3):659-684.
Arellano, Cristina (2004), "Default Risk, the Real Exchange Rate and Income Fluctuations
in Emerging Economies", mimeo, Duke University.
Barnhill Jr., Theodore M. and George Kopits (2003), "Assessing Fiscal Sustainability
Under Uncertainty," IMF Working Paper No. WP/03/79.
Barro, Robert J, (1979), "On the Determination of Public Debt," Journal of Political
Economy.
Blanchard, Olivier J., Jean-Claude Chouraqui, Robert P. Hagemann and Nicola Sartor
(1990), "The Sustainability of Fiscal Policy: New Answers to an Old Question,"
OECD Economic Studies, No. 15, Autumn.
Blanchard, Olivier J., (1990), "Suggestions for a New Set of Fiscal Indicators," OECD
Working Paper No. 79, April.
Bohn, Henning, (1998), "The Behavior of U.S. Public Debt and Deficits," Quarterly
Journal of Economics, v. 113, August, 949-963.
Buiter, Willem H. (1985), "Guide to Public Sector Debt and Deficits," Economic Policy:
An European Forum, Vol. 1., November, 13-79.
Calvo, Guillermo A., Alejandro Izquierdo and Ernesto Talvi (2003), "Sudden Stops, the
Real Exchange Rate and Fiscal Sustainability: Argentina's Lessons," mimeo,
Research Department, Inter-American Development Bank.
Chalk, Nigel and Richard Hemming (2000), "Assessing Fiscal Sustainability in Theory and
Practice," IMF Working Paper No. WP/00/81.
Eaton, Jonathan, and Mark Gersovitz (1981), "Debt with Potential Repudiation:
Theoretical and Empirical Analysis," Review of Economic Studies, v. XLVn, 289309.
Hamilton, James D. and Marjorie A. Flavin (1986), "On the Limitations of Government
Borrowing: A Framework for Empirical Testing," American Economic Review, v.
76, September, 809-819.
International Monetary Fund (2002), "Assessing Sustainability," SM/02/166.
, (2003), "Sustainability-Review of Application and Methodological
Refinements," mimeo, Policy Development and Review Department.
Ljungqvist, Lars and Thomas J. Sargent (2000), Recursive Macroeconomic Theory, MIT
Press.
Mendoza, Enrique G. Assaf Razin and Linda L. Tesar (1994), "Effective Tax Rates in
Macroeconomics: Cross-Country Estimates of Tax Rates on Factor Incomes and
Consumption," Journal of Monetary Economics, December, v. 34:3, 297-323.
Mendoza, Enrique G. and Pedro Marcelo Oviedo (2004), "Fiscal Solvency and
Macroeconomic Uncertainty in Emerging Markets: The Tale of the Tormented
Insurer," mimeo, Iowa State University.
Reinhart, Carmen, Kenneth S. Rogoff and Miguel A. Savastano (2003), "Debt
Intolerance," mimeo, IMF Research Department.
189
Table 1. Fiscal Sector Statistics and Natural Debt Limits

(in percent of GDP)
Brazil 1/
Colombia 21
Costa Rica 21
Mexico 3/
Public debt
average
maximum
year of maximum
1990-2002
40.68
56.00
2002
1990-2002
33.71
50.20
2002
1990-2002
49.46
53.08
1996
7990-2002
45.92
54.90
1998
Public revenue
average
coeff. of variation
two-standard dev. Floor
1990-2002
19.28
14.13
13.83
7990-7999
12.64
8.86
10.40
7990-2000
20.28
5.41
18.09
1990-2002
22.96
8.04
19.27
Non-interest outlays
average
coeff. of variation
1991-1998
19.19
13.76
7990-7999
12.80
13.55
7990-2000
18.54
9.98
1990-2002
19.27
3.96
2.55
12.46
2.30
8.81
1,16
16.40
2.02
17.73
1.86
50.49
-5.29
1.83
53.31
54.80
2.20
54.92
131.54
Fiscal sector statistics
Implied fiscal adjustment 4/

Implied minimum non-interest outlays
Benchmark Natural Debt Limits

(1961-2000 growth rates, 5 percent real interest rate)
2.55
Growth rate
56.09
Natural debt limit
3.52
Blanchard ratio
Growth Slowdown Scenario

(1981-2000 growth rates, 5 percent real interest rate, benchmark fiscal adjustment)
1.25
0.48
1.05
Growth rate
45.10
40.10
30.34
Natural debt limit
46.36
4.20
1.90
Blanchard ratio
0.83
36.96
88.52
High Real Interest Rate Scenario

(1961-2000 growth rates, 8 percent real interest rate, benchmark fiscal adjustment)
2.55
1.86
1.83
Growth rate
25.81
27.39
25.19
Natural debt limit
28.16
1.58
-2.70
Blanchard ratio
2.20
26.53
63.55
Note: The source of public debt data is IMF, World Economic Outlook, October 2003. The sources
of revenue and non-interest outlays are as noted in country footnotes.
1/ Revenue data from IMF, World Economic Outlook, Oct. 2003., non-interest outlays data derived
from data for the Central Government in World Development Indicators.
21 Revenue and non-interest outlays data for Central Government from World Dev. Indicators.
3/ Revenue and non-interest outlays data from IMF, World Economic Outlook, October, 2003.
4/ Implied fiscal adjustment is the number of standard deviations relative to the mean needed to
obtain a benchmark natural debt limit equal to the largest public debt ratio observed in the data.
Table 2. Natural Debt Limits with Default Risk

Brazil
Colombia
Costa Rica
Mexico
33.28
95.96
4.31
33.18
95.97
4.30
34.14
95.86
4.42
33.88
95.89
30.38
96.31
3.93
32.12
96.10
4.16
29.12
96.46
3.76
NDLs in the growth slowdown scenario without default risk and risk free rate of 2.36 percent
121.60
Natural debt limit
72.95
152.30
100.80
Required fiscal adjustment to support observed maximum debt ratios as NDLS 2/

50.20
56.00
Natural debt limit
93.97
Probability of repayment
93.30
6.57
7.35
Default risk premium
6.85
9.82
Implied minimum non-interest outlays
-5.95
-9.37
relative to average outlays
3.43
3.55
in number of St. devs.
54.90
93.43
7.20
15.23
-4.04
5.30
Benchmark NDLs with default risk 1/

Natural debt limit
NDLs in the growth slowdown scenario with default risk

Natural debt limit
26.12
96.82
3.37
53.08
93.64
6.95
14.12
-4.42
2.39
4.39
Notes: Calculations done as described in the text, using a risk free rate of 2.36 percent, which is the 1990-2002
average of the inflation-adjusted 90-day U.S. T-bill rate, and a curvature parameter for the risk function of
a =0.124, which was calibrated to match the EMBI+ spread and the debt ratios observed in Mexico in 1998.
17 Based on the benchmark values of growth rates and minimum revenue and outlays shown in Table 1
21 Values of minimum outlays required to support maximum debt ratios shown in Table 1 as NDLs in the setting
with default risk, using growth rates from the benchmark scenario.
Figure 1. Total Public Debt as a Share of GDP
-* Brazil
"-afir- Costa Rica
-*- Colombia
**%&*, Mexico
Figure 2. Number of Periods before Hitting a Fiscal Crisis
Figure 3. Simulations of Debt-to-GDP in Costa Rica. Starting Value bg= 0.35
Time periods
Figure 4. Time to Hit a Fiscal Crisis with and without Default Risk
Exchange Rate Regimes and

Debt Maturity Structure*
Matthieu Bussiere
European Central Bank
Marcel Fratzscher
European Central Bank
Winfried Koeniger
IZA, Bonn; Finance and Consumption in the EU, European University Institute
*We would like to thank for their very helpful comments Carmen Reinhart and seminar participants at
the IMF-Banco de Espaiia conference on "Dollars, Debt, and Deficits - Sixty Years after Bretton Woods"
in Madrid 2004. The views expressed in this paper are those of the authors and do not necessarily reflect
those of the European Central Bank. This draft is as of October 2004.
Introduction
Do exchange rate regimes matter for macroeconomic performance? This is an old
question of major importance for policy-makers which explains the persistent scientific
debate on the topic. The recent empirical findings of Husain et al. (2004) suggest that
exchange-rate regimes imply quite heterogeneous outcomes in countries with different
economic development. In order to understand better this heterogeneity, this paper
focuses on a specific interaction between exchange-rate regimes and economic
performance which has been relatively unexplored so far. We investigate whether and
how the exchange-rate regime affects the maturity mix of private debt in emerging
markets that only have access to credit denominated in foreign currency.
The importance of credit in foreign currency in emerging markets can hardly be
overemphasized. For example, Levy-Yeyati (2003) documents that more than 70% of
bonded debt in emerging markets is denominated in foreign currency. At the same time
many emerging markets have substantial maturity mismatches which are frequently
identified as one major culprit of the Asian financial crisis of the 1990s (see, for
instance, Chang and Velasco, 2000, Corsetti, Pesenti and Roubini, 1999, or Rodrik and
Velasco, 1999).1 Yet, while many papers have explained how imbalances in the asset
and liability structure of emerging markets can cause currency and financial crises, the
factors that trigger such imbalances in the first place have received relatively little
attention so far. In particular, few papers have considered the possibility that the
exchange-rate regime influences the debt structure. Instead of the more prevalent view
that maturity mismatches lead to volatile exchange-rates, we stress that highly flexible
and volatile exchange rates may shift the debt profile towards short-term maturities,
thereby increasing the vulnerability of emerging markets to liquidity crunches.
First, we provide a model which links exchange-rate regimes and maturity
mismatch in open emerging market economies. We show theoretically how currency
mismatch may lead to and exacerbate maturity mismatch in economies with flexible
exchange rates resulting in higher output volatility. Compared with much of the
literature on the subject, our model abstracts from asymmetric information and moral
hazard (see, e.g., Diamond, 1991, Jeanne, 2000, and Tirole, 2003), and focuses instead
on the role of market incompleteness. Although we recognize that asymmetric
information and moral hazard may be important, we show that such model ingredients
are not necessary to explain the joint phenomena of currency depreciation and asset
liquidation accompanied by high short-term debt ratios. Thus, the removal of market
failures may not suffice to tilt the debt profile towards safer, long-term debt. Instead our
model assigns a crucial role to the development of financial markets or instruments that
allow agents to insure better against risk.
Second, we provide empirical results that support the predictions of the model
for a set of 28 open emerging market economies. We add to the literature on exchangerate regimes and macroeconomic performance (see, for example, Husain et al., 2004,
and their references) by analyzing the influence of exchange-rate regimes on the debt
structure in some detail.2 In our analysis we take into account that the choice of the
exchange-rate regime or the degree of intervention in the currency market is influenced
by macroeconomic factors (the fear of floating discussed in Calvo and Reinhart, 2002).
We use annual data from the World Economic Outlook (WEO) for macro-economic
!
Arteta (2002) establishes the additional result that floating regimes seem to increase currency
mismatches in domestic financial intermediation.
2
For a more on the related literature see Bussiere, Fratzscher and Koeniger (2004).
199
variables and the Bank of International Settlements (BIS) for debt variables. Our main
finding is that countries with more flexible exchange rates hold a higher share of debt
with short maturity. This maturity mismatch of foreign debt is associated with more
volatile output.
The remainder of the paper is structured as follows. Section 2 presentw a simple
model to show how exchange-rate regimes can change the debt-maturity structure and
output volatility. We empirically test the main predictions of the model more formally
in Section 3. Finally, we discuss policy implications and conclusions in Section 4.
A Simple Model
In this section we illustrate how flexible exchange rates might shift the debt structure
towards short-term maturities.3 More specifically, we model how the exchange-rate
regime influences solvency and the choice of debt maturity (the model draws on work
in Bussiere, Fratzscher and Koeniger, 2004). Forward-looking and impatient riskneutral agents choose whether to consume or to invest, financing their investment with
short- or long-term foreign debt. We assume that debt (i) can only be obtained in the
international capital market, (ii) is denominated in foreign currency and (iii) is
constrained by solvency, which requires that agents can always repay.
Agents face a simple trade-off in their choice of debt maturity. Since flexible
and volatile exchange-rates tighten solvency constraints relatively more for long-term
debt, borrowers have an incentive to raise the share of short-term debt. However, shortterm debt is risky and the investment project can be liquidated before the investment
return materializes so that agents have smaller collateral if they borrow short term As a
consequence, a larger share of short-term debt raises the share of investment projects at
risk. In our model, liquidation of the collateral, and a larger fraction of short-term debt
are the result of optimal choices of individual agents.
We now present the structure of the model in more detail. Consider an economy
with 3 periods. Impatient risk-neutral agents with endowment K either can invest K into
a project or immediately consume it. If agents invest they can decide whether to finance
consumption with short or long-term debt. Agents with short-term debt need to rollover their debt in the second period. The decision to invest yields an immediate flow
jK, 1 > / > 0 . Instead, the project income Y only becomes available in period 3.
All debt is denominated in foreign currency. This is realistic for emerging
market economies in which most of the debt is in foreign currency (see Levy-Yeyati,
2003, or Corsetti, Pesenti and Roubini, 1999). The different currency denomination of
investment returns and debt liabilities implies that exchange-rate flexibility influences
the borrowing choices of investing agents. If financial markets are incomplete,
borrowers will have to bear at least some of the exchange-rate risk. We assume as
benchmark case that all risk is in fact borne by borrowers: markets are incomplete in
that borrowers only have access to a risk-free asset with interest rate r. This implies that
investing agents face constraints that ensure repayment of their debt in all states of the
world. These constraints differ with respect to maturity and crucially depend on the
pledgeable income and the maximum depreciation of the exchange rate. We now
characterize these constraints for flexible and fixed exchange-rate regimes.
3
Of course, the choice whether to adopt flexible exchange rates might depend on the stock of short-term
debt in the economy: countries with a large amount of short-term debt might fear to float the exchange
rate. Since this endogeneity is extensively discussed in the literature (see Calvo and Reinhart, 2002), we
focus entirely on the direct effect of flexible exchange rates on debt maturity in the theoretical part.
Instead, the empirical estimation will try to account for the endogeneity of the exchange-rate regime.
200
Solvency Constraints
Flexible Exchange Rates
Given the endowment K and nroiect income 7, agents that invest and borrow short term
cannot borrow more than
in the first period where R is the interest factor,
Xi denotes the maximum possible depreciation of the exchange rate until period 2 and
the exchange-rate in the first period is normalized to 1.
If the agent invests and borrows long-term instead he cannot borrow more than
in the first period where Xs denotes the maximum possible
depreciation of the exchange rate until period 3. Note that access to credit becomes
relatively tighter for long-maturity debt, the larger R and the larger X3 relative to X2
. Since long-term debt needs to be repaid in period 3 the pledgeable income K+Y is
worth less in present-value terms. This is especially so, if the maximum possible
depreciation of the exchange rate increases over time. Instead, long-term debt increases
the pledgeable income from AT to K+Y, since the agent earns the project returns with
certainty. Thus, more credit is available short-term if the difference in pledgeable
income Y is relatively small compared with the additional possible depreciation of the
exchange rate X3 / Xi . Note that if exchange rates are characterized by a stochastic
process with increasing variance for longer time horizons,
Fixed Exchange Rates
As a stylized conparison let us consider exchange rates that are fixed with certainty. In
this case the pledgeable income is independent of debt maturity and agents cannot
borrow more than (K + Y)/R2 in the first period where the exchange-rate in the first
period is normalized to 1. Note that agents can always borrow more than KIR in the
first period since arbitrage implies that the return to the project 7+y, where y^Y/K ,
needs to be larger than R2 . That is, for agents to find it optimal to invest, projects have
to yield a return at least as high as the risk-free asset over two periods. More
importantly, more credit is available if exchange rates are fixed rather than flexible.
Optimal Maturity
Assuming that the debt roll-over in the second period implies an infinitesimally small
transaction cost e, all debt has long maturity if exchange rates are fixed with certainty.
Instead, flexible exchange rates imply a non-degenerate choice of debt maturity. We
analyze this choice focusing in particular on the role of different degrees of exchangerate volatility captured by the paramete]
This corresponds to investigating the effect
of different exchange-rate regimes on the debt-maturity structure in the empirical
section.
We assume that agents are impatient: the discount factor
. This
together with risk neutrality implies that agents prefer to bring forward consumption as
much as possible to the present. Thus, the solvency constraints determine the
consumption profile of investing agents and also the optimal maturity choice.
201
Table 1: Consumption profile
No debt
K
Period 1
Period 2
no depreciation
depreciation
Periods
no depr. in
period 1
LTdebt
ST debt
0
*
no depr. in period 2
depr. in period 2
depr. in period 1 no depr. in period 2
depr. i n period 2
^
^
^
0
0
0
We assume that in each period the exchange rate depreciates by factor X >\
with probability p and remains constant with probability 1-p. That is the worst possible
exchange rate from the perspective of the borrower is X in period 2 and X in
period 3. For simplicity we assume that the agent always needs to liquidate the project
if the exchange rate depreciates and the agent borrows short-term. This implies the
following restriction on the parameter space: 1 + y < RX .5
Table 1 summarizes the consumption profiles for the different choices of the
agent. Using these consumption profiles we can easily spell out utility for the different
choices as a function of the model's parameters. Normalizing by K and denoting utility
with Uj , j = nj,s , with the subscript n for no investment, / for investment financed
with long-term debt and s for investment financed with short-term debt, we get
Note that
and
, i=/,$ , as long as agents are
impatient. Figure 1 plots utility as a function of X. Table 2 contains the parameter

values that were chosen for illustration purposes.
In Bussiere, Fratzscher and Koeniger (2004) we endogenize the probability of liquidation. This makes
the algebra more cumbersome but does not change the basic insights.
In the second period the agent cannot roll-over the debt if ^f RX-j^
<0
RX
RX
which implies the
condition in the text. Note that in the second period the pledgeable income is always K+Y since the
project returns realize in period 3.
202
Table 2: Parameter values
Figure 1: Utilities as a function of X
In Figure 1, both us and ul decrease as the maximum possible depreciation X

increases: more volatility of future exchange rates (a larger X) tightens credit
constraints and shifts consumption of impatient agents into the future. Of course, un,
the utility if the agent does not invest and borrow, does not depend on X and is flat at
the normalization value 1. If X is not too high, however, it is optimal for the agent to
invest: Uj > un and us > un . Interestingly, long-term debt is preferred for small X,
ul>us , but higher X can shift debt towards short maturity: HS>UI . Moreover, as
mentioned above the scope for borrowing short-term becomes smaller for high project
returns: in Figure 1, the region of the X fs in which the agent finds it optimal to borrow
short-term becomes smaller as y increases from 0.1 to 0.15 .
203
Linear utility implies that for given X and y the agent either borrows only
long-term or only short-term if he invests. In order to generate predictions about the
debt structure in the aggregate economy in a simple way, we add heterogeneity in the
project return y. Consider an economy populated by a continuum of agents with
heterogeneous project returns yf in the interval [0;^]. One can show that for each X
there exists a critical value of .y at which the agent finds it optimal to borrow longterm6 As the intuition obtained from the discussion of the solvency constraints and
illustration in Figure 1 suggest, this critical value is higher for larger X so that the
share of projects financed with short-term debt increases for larger JSf(see Bussiere,
Fratzscher and Koeniger, 2004, for an elaboration of this point). Of course, also the
total number of projects decreases because investment becomes less attractive.
However, relative to a benchmark with a fixed exchange rate, a flexible exchange-rate
regime results not only in a smaller level of total debt but also in larger share of debt
with short maturity. This increases volatility (as long as some projects are financed
short-term): projects financed on a short-term basis are liquidated with probability p in
which case the project return Y is lost.
This is the basic insight we want to convey: if a country adopts a flexible
exchange-rate regime, additional macroeconomic volatility can result from an
endogenous shift of the country's debt structure towards short-term maturities. We now
try to find some empirical support for this hypothesis.
Empirical Evidence
To test empirically whether the choice of the exchange rate regime has an impact on
debt structure in emerging markets, we use financial and macroeconomic data for 28
emerging market economies: 9 in Asia, 8 in Latin America, 8 Central and Eastern
European Countries (CEECs), as well as South Africa, Russia and Turkey. Time series
on debt were taken from the BIS and start in the 1980s. The country sample was
selected to include mostly open emerging markets (i.e., countries that opened up their
financial account during or before the period under consideration), and based on data
availability criteria. For CEECs, the first part of the 1990s had to be dropped due to
data unavailability and because the first years of the transition to a market economy
were characterized by very high volatility. Such high volatility can be considered as a
one-off event, not representative of the mechanisms we aim at analyzing.
Of course, if this critical y is larger than y, no debt is borrowed long-term at all.

204
Table 3: Exchange rate regime, distribution by period (%).

Total ! 1960-69 1970-79 1980-89 1990-99 2000-2004
0.4
A. No separate legal tender
3.2
B. Pre announced peg or currency
board arrangement
249
50.5
33.2
7.3
14.1
22.6
C. De facto peg
5.8
2.7
2.3
5.0
9.5
9.7
0.6|
D. Pre announced crawling peg
1.4
0.9
0.7
E. Pre announced crawling band
1
narrower than or equal to +/-2%
1.9|
5.6
3.2
F. De facto crawling peg
0.5
16.4
10.2
4.0
6.7I
G. De facto crawling band narrower
I
than or equal to +/-2%
16.41
8.2
22.7
25.9
13.8
8.9
1
H. Pre announced crawling band that is
_
_
wider than or equal to +/-2%
0.9l
0.5
1.6
3.2
1
1. De facto crawling band narrower than
or equal to +/-5%
10.9
14.1
16.4
4.6
11.2
6.5
J. Managed floating
11.7
15.9
6.4
14.1
4.9
25.8
K. Freely floating
1.8
2.6
9.7
L Freely falling
15.1J
4.1
13.6
23.6
22.7
3.2
M. Dual market in which parallel
I
market data is missing
2.9,
4.6
3.6
1.8
3.0
Number of observations
1088)
220
220
220
304
124
ii
Source: Reinhart and Rogoff, 2004. The sum of each column is by construction equal to 100%.
Data on exchange-rate regimes were provided by Reinhart and Rogoff (2004).

This classification allows distinguishing 13 types of exchange-rate regimes, listed in
Table 3 (the original classification included 15 categories but two of them never
appeared in the sample)7. The Reinhart-Rogoff classification is a de facto classification
in that the authors analyzed to what extent announcements of de jure regimes truly hold
de facto. They find that many of the countries1 de jure regimes deviate significantly
from the actual exchange-rate behavior.
Table 3 shows that for the countries in our sample, pre-announced pegs or
currency board arrangements are the most common regimes over the entire period,
representing nearly one fourth of the observations. However, the proportion of
countries with such regimes fluctuates considerably over time: whereas it was above
50% in the 1960s, it fell to a third in the 1970s with the end of the Bretton Woods
system, and less than 10% in the 1980s. Since then, the number of countries with a preannounced peg has increased again, to reach 22.6% of the country sample in the years
2000-2004. Conversely, the most extreme form of floating exchange rate ("freely
falling") has increased regularly between the 1960s and the 1980s, to become the most
common category in the 1990s with 22.7% of the cases in the sample. After 2000
however, this category has fallen considerably to slightly above 3%, whereas managed
floating exchange-rate regimes have become in the past four years the most common
form of arrangement, with more than a fourth of the observations.
These categories were (a) pre announced horizontal band that is narrower than or equal to +/-2% and (b)
moving band that is narrower than or equal to +/-2% (i.e., allows for both appreciation and depreciation
over time).
205
Figure 2: Share of pegged exchange rate regimes (categories A,B,C in Table 3) and
de facto exchange-rate volatility in the sample.
Note: The scale is on the left and right vertical axis, respectively.
In addition to this classification, we also consider an alternative classification
based on actual exchange-rate volatility.8 Such measure is in fact strongly correlated
with the above mentioned de jure classification, as presented for instance on Fig. 2 9.
To make the Reinhart and Rogoff measure amenable to econometric testing, we
attribute values to the categories presented in Table 3 (1 for category A, 2 for B, etc,
assuming that each increment is equal).10
In the Reinhart and Rogoff (2004) paper, two measures are presented: a socalled "fine measure" with the categories of Table 3, and a "coarse measure" with only
six categories, each one aggregating two or three levels of the "fine measure". By and
large, these two measures yielded similar results in the specifications we have
estimated (see below), suggesting that results do not depend on the mapping of the
categories into numbers. A simple regression of the de facto volatility measure on the
Reinhart-Rogoff measure using fixed effects - not reported here - shows a positive
coefficient, significant at the 1% level, suggesting that more flexible exchange-rate
regimes are linked to higher actual exchange-rate volatility. Clearly, the link between
the Reinhart-Rogoff measure and the de facto volatility measures of exchange-rate
8
We computed, for each year, the standard deviation of the first-differenced real effective exchange rate,
using monthly data.
9
In Figure 2, pegged exchange rate correspond to the first three categories A, B and C of Table 3.
10
We tested this assumption using a set of 0/1 dummy variables, one for each of the categories
represented in Table 3 (we took the first category as benchmark). Results suggested that the increment
may actually be declining in the case of short-term debt: moving from regime A to regime B has a
stronger effect on short-term debt than moving from B to C, etc. However, this effect is difficult to
measure with precision given the degrees of freedom, which is why we resorted to the more
straightforward linear specification.
206
regimes are more complex than this simple regression suggests. For instance, a fixed
peg regime is likely to exhibit low exchange-rate volatility but this does not mean that
exchange-rate uncertainty is equally low: economic agents may anticipate a speculative
attack against the peg which would have the effect to transform the regime into a
floating exchange rate, accompanied by higher exchange-rate volatility. To account for
this phenomenon, we computed four different measures of de facto exchange-rate
volatility, measuring the standard deviation of the exchange rate over (i) the past three
years, (ii) the past year, (iii) the year ahead and (iv) the three years ahead.
Table 4: summary statistics, total and regional breakdown
Lat. Am.
Asia
New EU M. S.
198019901980199019941989
2003
1989
2003 < 1994 2003
Pegged ER (% total)
11.3
21.4
16.4
27.3
NA
51.1
4.8
3.3
Exchange rate volatility
1.8
2.1
NA
1.9
53.2
48.9
38.5
38.3
NA
42.8
Debt/GDP (%)
17.6
14.8
Short-term/total debt (%)
19.0
15.9
NA
24.5
7
Growth volatility
0.7
0.6
1.3
1.1
NA
0.6
all EMFs
1980-19901989 2003
12.3
2.9
33.9
19.4
1.0
Note: unweighted average over sample period.

1/Normalised to 1 for all EMEst 1980-1989.
Turning to the dependent variables, Table 4 presents key stylized facts, broken
down by region and by time period. All debt variables presented in Table 4 are debt to
banks, taken from the BIS database.11 This particular definition of debt is not without
caveat as it excludes in particular other debt instruments which also play an important
role as well; however, it has the advantage to be available for many countries over a
long period of time. In addition, this source offers a convenient breakdown by maturity:
short-term debt is defined as debt whose maturity is below one year. Interestingly, for
Latin America and Asia a larger ratio of exchange-rate pegs over time is negatively
correlated with the amount of short-term debt. For all emerging market economies this
decrease is less pronounced but instead the higher ratio of pegged exchange-rate
regimes is positively associated with total debt. This is suggestive for the mechanisms
highlighted by the simple theoretical perspective. However, Table 4 also shows
pronounced heterogeneity across regions. Overall, Latin American countries hold more
debt, as a percentage of GDP, than Asian countries and new European Union Member
States. Such differences do not seem to be directly linked to the above described
measures of exchange-rate regimes. As we will see below in the econometric results,
other factors such as fixed country effects and other (time varying) control variables
have an impact on the total debt level of a country and need to be controlled for in the
econometric analysis.
To estimate the impact of exchange-rate volatility on the debt structure, we
followed the GMM methodology developed by Arellano and Bond (1991) using lagged
differences as instruments. The results reported below are robust to using a standard
fixed-effect estimator (see Bussiere, Fratzscher and Koeniger, 2004, for further
discussion of the econometric methodology in our application). In order to control for
other factors that may influence debt, we included on the right-hand side a set of
control variables: real PPP adjusted GDP per head (as a proxy for catching up effects),
11We focus on private rather than government debt since the theoretical model outlined in the previous
section applies to the former.
207
27.1
26
41.7
18.7
0.8
the government budget balance (to account for the possible presence of non-Ricardian
agents), investment (see Bleakley and Cowan, 2002), a dummy variable for capital
account liberalization and a dummy variable for currency crises. The presence of the
latter in the specification stems from the specific problem arising from currency crises,
relatively frequent in our sample of emerging markets: as the debt ratios to GDP are
computed using the exchange rate to convert foreign into national currency, such ratios
mechanically jump up during crises times. Arguably, this variable is then not strictly
exogenous. Likewise, the assumption that exchange-rate volatility is exogenously given
is debatable and the endogeneity issue is only partly dealt with in the GMM
methodology. Ideally, one would use instrumental variables in this context, that is,
variables correlated with exchange-rate volatility but not with the debt ratios. However,
we are not aware of a good instrument for our application. One should note however
that if endogeneity plays a role, the link between exchange-rate volatility and debt that
we measure in our regressions is actually a lower bound. If policy makers intervene in
the foreign exchange markets to reduce to reduce exchange-rate volatility when the
debt maturity structure is tilted towards short-term debt (fear of floating), the positive
correlation between volatility and the share of short-term debt should be downward
biased and thus less positive.
Table 5: Exchange rate regime and debt
S. T. debt
Dependent variable: (% total debt)
Reinhart-Rogoff:
Fine measure
S. T. debt
(% GDP)
0.23**
Coarse measure
0.32***
0.09
0.10
0.63 **
0.88 ***
0.28
0.25
-0.02
-0.04
0.08
0.04
One year lag
0.02
0.01
Three years lead
0.14
One year lead
0.17*
De facto definition
Three years lag
0.08
0.18
0.11
Total debt
(% GDP)
0.82***
0.17
2.28 ***
0.49
-0.46 ***
0.16
-0.29 ***
0.05
0.12
-0.24 *
-0.46 *
0. 16
0.27
-0.05
-0.21
0.10
0. 16
Standard errors in italics

*, **, *** denote significance at the 10%, 5%, 1% level, resp.
Table 5 reports the results. Each of the three columns corresponds to one of the
three following dependent variables: short-term debt as a share of total debt, short-term
debt as a percentage of GDP, and total debt as a percentage of GDP. According to the
simple theoretical results presented in the previous section, more exchange-rate
volatility (measured as either de jure or de facto) should have a positive impact on
short-term debt (first two columns), and a negative one on total debt. The table is
synthetic in that it reports only the coefficient in front of the exchange-rate regime
variables and not the control variables (estimated in different regressions, one at a
208
time). The full results are available upon request. Regressions including the de facto
exchange-rate volatility measures are estimated over 304 to 336 observations (taking
three years to compute lags implies losing some of the observations). Those including
the Reinhart and Rogoff measure are estimated with around 350 observations. Overall,
the results provide substantial evidence in favor of the model's predictions. The
coefficient of the Reinhart and Rogoff measures of exchange-rate volatility are positive
and significant in the first two regressions, when the dependent variable is short-term
debt. The de facto volatility measures are not as clear-cut, as many of the variables we
used do not show up significantly, particularly in the second specification. However,
interestingly, the de facto variable using the forward-looking one-year ahead volatility
measure, which accordingly is the most relevant measure for debt with maturity under
one year, enters the specification significantly and with a positive sign. Turning to the
last column of Table 5, the impact of the de facto volatility measure on total debt is
clearly negative and significant for three of the four measures, whereas it is negative
but insignificant for the last one. This result is very much in line with the results of the
theoretical section: more exchange-rate uncertainty is associated with lower total debt.
More surprisingly, the impact of the Reinhart and Rogoff measure appears to be
positive instead of negative. This result is counter-intuitive and seems to be driven by
the currency crisis episodes: in the immediate aftermath of crises, countries often move
to a more flexible exchange-rate regime but at the same time the debt ratio jumps up
since the exchange rate depreciates and GDP falls. In fact, when currency crisis
episodes are dropped from the sample, the impact is no longer significant. Although our
currency crisis variable should pick up this effect, it does not completely solve the issue
because the increase of the debt ratio often lasts a couple of years, whereas the crisis
variable was set equal to one only at the time of the crisis.
Finally, Table 6 reports results using output volatility as the dependent variable
(again with one of the reported regressors at a time). The results suggest that more
exchange-rate volatility is indeed correlated with larger output volatility, although it
seems mostly through the impact on the debt maturity structure.
Table 6: Exchange rate regime and output volatility
coef.
S.E.
Reinhart-Rogoff:
Fine measure
0.18
0.24
Coarse measure
0.22
0.69
De facto definition
Three years lag
One year lag
Three years lead
One year lead
S.-T. debt (% tot. debt)
S.-T. debt (% GDP)
Total debt (% GDP)
-0.15
0.30
0.31
0.32
2.85
0.01
0.004
0.23
0.17*
0.62
0.20
0.98**
0.01
0.01
Dependent variable is volatility of real output

*, **, *** denote significance at the 10%, 5%, 1% level, respectively.
209
Conclusions
This paper has investigated the role of exchange-rate regimes for macroeconomic
stability, focusing in particular on the influence of the exchange-rate regime on the
maturity structure of debt. A simple theoretical model suggests that a more flexible
exchange-rate regime may shift the debt structure towards short-term debt, which in
turn can increase output volatility. Taking the model to the data, we find broad support
for the model. In particular, more exchange-rate uncertainty seems to be associated
with a larger share of short-term debt.
A possible extension for future research would be to allow for the possibility of
default in the model and to introduce moral hazard and asymmetric information, since
these phenomena are important for capital markets in reality. However, better data are
necessary to test predictions of such a model. Data on interest rates across different
maturities for private debt would allow to analyze the interaction of credit prices and
credit demand, instead of focusing only on credit supply and credit volumes.
Our model emphasizes the importance of market incompleteness and does not
rely on asymmetric information or moral hazard to explain the debt structure and the
inclination of emerging markets to be subject to financial crises and substantial real
volatility in the economy. If market incompleteness is important, it is crucial to develop
financial markets or instruments that allow agents to insure better against risk so that
financial crises in emerging markets can be avoided. Concrete policy proposals to
address this issue have started to emerge. Some of the proposals call for the
development of domestic financial markets for local-currency substitutes to dollarized
debt (Levy-Yeyati, 2003) or for the issuance of bond contracts denominated in units of
a basket of emerging-market currencies (Eichengreen and Hausmann, 2003).
210
References
Arteta, Carlos (2002), "Exchange Rate Regimes and Financial Dollarization: Does
Flexibility Reduce Bank Currency Mismatches?", International Finance
Discussion Papers, Number 738, September 2002, Board of Governors of the
Federal Reserve System.
Bleakley, Hoyt and Kevin Cowan (2002): Corporate Dollar Debt and Depreciations:
much ado about nothing?, UC San Diego, mimeo.
Bussiere, Matthieu, Marcel Fratzscher and Winfried Koeniger (2004): Currency
Mismatch, Uncertainty and Debt Maturity Structure, forthcoming ECB
Working Paper.
Calvo, Guillermo A. and Carmen M. Reinhart (2002): Fear of Floating, Quarterly
Journal of Economics, vol. 67, 379-408.
Chang, Roberto and Andres Velasco (2000): Banks, Debt Maturity and Financial
Crisis, Journal of International Economics, vol. 51, 169-94.
Corsetti, Giancarlo, Paolo Pesenti and Nouriel Roubini (1999): What Caused the Asian
Currency and Financial Crisis?, Japan and the World Economy, vol. 11, 305-73.
Diamond, Douglas W. (1991): Debt Maturity Structure and Liquidity Risk, Quarterly
Journal of Economics, vol. 106, 709-37.
Eichengreen, Barry and Ricardo Hausmann (2003): Original Sin: the road to
redemption, in: Eichengreen, Barry and Ricardo Hausmann, eds., Other
People's Money: Debt Denomination and Financial Instability in Emerging
Market Economies, University of Chicago Press, Chicago.
Husain, Aasim, Aska Mody and Kenneth S. Rogoff (2004): Exchange Rate Regime
Durability and Performance in Developing Versus Advanced Economies,
NBER Working Paper No. 10673.
Jeanne, Olivier (2000): Debt Maturity and the Global Financial Architecture, CEPR
Discussion Paper No. 2520.
Levy-Yeyati, Eduardo (2003): Financial Dollarization: where do we stand?, Conference
on Financial Dedollarization: Policy Options, IADB, Washington D.C., mimeo.
Reinhart, Carmen and Kenneth S. Rogoff (2004): The Modern History of Exchange
Rate Arrangements: A Reinterpretation, Quarterly Journal of Economics, vol.
69, 1-48.
Rodrik, Dani and Andres Velasco (1999): Short-Term Capital Flows, Annual World
Bank Conference on Development Economics, 1999.
Tirole, Jean (2003): Inefficient Foreign Borrowing: a dual- and common-agency
perspective, American Economic Review, vol. 93, 1678-702.
211
COMMENT
Financial Globalization and Exchange Rates
(Philip R. Lane and Gian Maria Milesi-Ferretti)
Daniel Cohen
Ecole normale superieure
This is a wonderful paper, crisp and full of insights. The core idea is the following:
countries are rarely in the comer position of being either strictly debtor or creditor.
They are most often simultaneously both. They hold claims on the rest of the world and
they are in debt with respect to it.
The most obvious example that comes to mind is the US. It is not only the
largest debtor on earth but the largest creditor as well. Over the period 1999-2002
alone, the US has borrowed the equivalent of 31% of its GDP. But over the same
period, it also accumulated claims over the rest of the world worth about 15% of GDP.
This pattern is frequent for industrial countries, but is also visible for emerging
countries as well. In the case of Korea, for instance, the inflows represent 10% of GDP,
the outflows stand at 12% of GDP. In Taiwan, the numbers are 25% of GDP both ways
(again from 1999 to 2004).
The benefit of being in and out simultaneously is obvious within any theoretical
framework where countries want to diversify their portfolio. For countries which are
high in debt and do not cash in the benefits of being on both sides of the fence, this
adds to the ills of debt crises.
For countries which do cash in the benefits of being on both sides, the
macroeconomics of exchange rate fluctuations become quite interesting. What happens
when the exchange rate depreciates? For most industrial countries, the paper shows that
the return on foreign assets appears to be almost orthogonal to the exchange rate of the
country. What happens on the liability side? The surprising result of the paper is that
liabilities appear to follow pretty much the same pattern which implies that "the net
valuation impact of exchange rate movements on the net foreign asset position has been
limited"
The exception to this rule is the US. Contrary to other industrialized countries,
liabilities are unaffected by currency movements, a fact which quite obviously results
from the US borrowing in dollar. But the surprising result is that assets follow almost
the same pattern. It all appears to be as if the US was living in a dollar zone, with little
implications of exchange rate fluctuations on assets and liabilities. This is not
necessarily good news if one counts on the dollar depreciation to solve, through the
sheer force of wealth effects, the debt problem of the US.
213
The only exceptions to the rule that flows are important in and out come from
countries which are deep in debt. This is the case of Brazil, for instance, that brought in
17% of GDP over 1999-2004 and brought out -1.4% of GDP (i.e., decumulated over its
gross position abroad) over the same period. I would like to suggest here some
connection with the work that we have done with Richard Fortes. We explore the
dynamics of debt for middle income debtors and try to differentiate the role of "risk",
growth, and policies on the pattern of debt accumulation on emerging countries.
We characterize risk through real interest rate, which involves both interest cum
spread and deviation of exchange rate to PPP. The bulk of this latter term is quite often
enormous. We have decomposed the debt dynamics into the following identity:
Increase of the Debt-to-GDP ratio =
real interest rate * Debt-to-GDP ratio
- Growth rate of the economy * Debt-to-GDP ratio
- Primary Surplus/GDP
The real interest rate is the nominal rate (risk free rate + spread) adjusted for the
deviation of the exchange rate from PPP. The dynamics are computed up to the year of
the debt crisis itself. We present this decomposition below by dividing each of the three
terms of the right-hand side by the sum of their absolute values (the sum of absolute
value then adds to one). I take here four cases of importance.
Interest+Change
Growth
Deficit
Brazil
0.47
-0.51
0.02
India
0.35
-0.49
0.16
Russia
+0.50
-0.50
0
Turkey
0.52
-0.10
-0.39
Each item expressed as a fraction of the sum of absolute value
The first term is roughly interpreted as a confidence premium, the second term
as a measure of the underlying fundamentals and the third term as a measure of the
policy choices. We see that in all four cases the countries are heavily burdened by the
interests-exchange rates term, which almost entirely cancels the (beneficial) growth
factor. The fact that these countries appear to be in the Lane-Milesi-Ferretti comer
situation has certainly much to do with it. Countries which are deep in debt and have no
precautionary assets are likely to be more vulnerable to confidence shocks. More work
is needed to thorough this connection, but at this stage much already has been learnt.
214
COMMENT
Stanley Fischer
Member of the Board, Citigroup, Inc.
It is a great pleasure to participate in this panel, surrounded to left and right by former IMF
colleagues, each of whom has moved on to better things - whether inside the Fund or out of
it. I will discuss six points, very briefly. First, what are the imbalances about which we
worry? Second, do they need to be fixed? Third, how? Fourth, what about the role of
China? Fifth, is the US dollar going to have a hard landing and do we need to worry about
that? And finally, I would like to talk a bit about currency blocks.
1. The imbalances. Typically people worry about one imbalance, the US current
account deficit, which exceeds five percent of GDP. Of course there are corresponding
current account surpluses in other countries, though there seems to be less concern about
that side of the equation.
Consider date for 2003. The US current account deficit was about 530 billion
dollars in 2003. The corresponding surpluses were in Japan ($140 billion) and the rest of
East Asia (approximately another $140 billion). The mysterious statistical discrepancy
accounted for 120 billion dollars. The remaining counterpart of the US deficit consisted of
the surpluses of the EU, Russia, other oil producers, and other countries. Fundamentally,
then, over half the US current account deficit was with Japan and East Asia.
2. Do the deficits need to be fixed? There is a certain inconsistency in the way we
all talk about current account imbalances. Almost every country prefers to have a current
account surplus. Many countries that are quite happy with their surpluses complain that the
US is running a large deficit and acting irresponsibly in the global economy.
In the aftermath of the Asian crisis, in which countries with more reserves on the
whole surmounted the crisis better, many countries wanted to build up their reserves.
These countries were content to run current account surpluses. In the midst of its lengthy
recession, Japan has not wanted the yen to appreciate. So it too has been content to run
large current account surpluses and build up reserves.
Dooley, Folkerts-Landau and Garber have described this as a new Bretton Woods
system, one that will enable the world to absorb the massive supplies of labor in China that
are now entering the global economy. The only problem with that argument is that the
current level of the US current account deficit is unsustainable. Under reasonable
assumptions, the US external debt to GDP ratio will continue rising without limit. That
means the process cannot continue.
That is not to say that the US current account deficit has to be reduced to zero.
Rather it needs to be reduced to a sustainable level, which would be about 2.5-3 percent
of GDP. That would still leave room for countries that want to do so to accumulate
reserves, and for foreign capital to flow to the United States. Similarly, with China
receiving foreign direct investment inflows, it could run a current account deficit.
215
The question is sometimes raised of whether it is appropriate for the world's richest
economy to run a current account deficit. Shouldn't capital flow from the rich to the poor
countries? If risk-adjusted rates of return in poor countries were higher than those in the
United States, capital would flow towards the poor countries until rates of return equalized.
However we need also to recognize that individuals want to invest where markets are
reasonably liquid, and where they have the assurance that their property rights will be
protected. In those respects, individuals from around the world will continue to want to
invest in the United States. So it is likely that capital will continue to flow to the United
States.
3. How will the adjustment occur? The adjustment can take place either through US
exports growing more rapidly than imports as a result of more rapid growth abroad, or
through a depreciation of the dollar. It would be bad for the US and for the world economy
if the adjustment takes place as a result of slower growth in the US - though the fiscal
correction that will be needed in the United States will tend to reduce US growth. If
growth rates remain close to potential, i.e. if there is no major recession in the US, the
dollar will have to depreciate. And it will have to depreciate primarily against the yen and
therenmimbi.
It has often been said by European officials that so far the Euro has borne the main
runt of the need to reduce the US current account deficit. They seem to believe that when
the Asian currencies appreciate, the Euro will be able to depreciate somewhat against the
dollar.
Yusuke Horiguchi, formerly of the IMF, now the chief economist at the Institute for
International Finance has made the following argument. It is that the appreciation of the
Euro against the dollar will over a period of a few years help adjust that part of the US
deficit that is due to US imbalances with Europe. US imbalances with Asia have so far been
handled through Asian central bank purchases of reserves. When the Asians allow their
exchange rates to adjust, that appreciation will essentially compensate for their former
purchases of reserves, and leave the exchange rate of the Euro to a first approximation
unaffected
4. China. It is very difficult to predict when China and Japan will allow their
respective exchange rates to adjust. Given that the yen is not pegged, that is easier for Japan
to do than for China. Indeed Japan has already stopped intervening, although the exchange
rate has not moved much since then. Nonetheless it is safe to predict that Japan will resume
intervening if the yen begins to appreciate rapidly.
China is comfortable with its pegged exchange rate at present, and will delay moving
so long as inflationary forces in China remain under control. But given China's desire to
reduce overheating, it would be useful to allow some appreciation of the currency, perhaps
by adjusting the central rate, pegging to a basket, and gradually opening up a band. In the
medium term, China will need to allow for greater exchange rate flexibility if it is to become
a major international financial center - as it no doubt wishes to do
5. Hard landing for the dollar? Will the dollar collapse once it starts adjusting, a
possibility brought to mind by the Dornbusch overshooting model? The Asian central banks
have sufficient reserves to prevent an excessively rapid adjustment of the dollar. Since they
will be reluctant to see their currencies appreciate too fast, they will likely intervene to slow
any appreciation that begins to get out of hand.
216
Nonetheless, there is one circumstance in which the dollar could suffer a hard
landing. Since the early 1980s there has been a basic confidence in the stability of the
United States economy. That has meant that a depreciation of the dollar makes United
States assets more attractive to foreigners. If this basic confidence is lost - for instance
because of a belief that the fiscal deficit is too large - then the dollar could come under
enormous pressure as private investors react to an initial decline of the dollar by expecting
that it has further to go. Given the volume of foreign holdings of US assets, the potential
outflow could be very large, and would put pressure on the exchange rate that would be too
powerful for even the Asian central banks to handle.
Such a scenario would require a sharp reaction from policy, along the lines of the
Volcker monetary policy of 1979-81. That would slow the US and the global economies.
It is not the most probable outcome, but it is a possible result, dependent in large part on
whether investors continue to maintain confidence in the economic policies of the United
States.
In appraising the probability of a hard landing, we should recognize that we have
been worried about that possibility for a long time, and that it has not so far happened. In
addition, we need to recognize the increased capacity of the international financial system to
deal with exchange rate movements, in part because of the development of hedging
instruments. For instance, the Euro has appreciated about 50 percent against the dollar
over the past two years, but the consequences have been manageable.
6. Currency blocs. Asian central banks are already stabilizing their exchange rates
relative to the yen and the dollar - and with the renmimbi pegged to the dollar, also against
the renmimbi. There appears to be real political momentum behind the idea of an Asian
exchange rate arrangement. That will be difficult since in the medium and long term both
the yen and the renmimbi will be important Asian currencies.
In one scenario, all of East Asia, including Japan, could eventually agree on the use
of a common currency. Alternatively, a common currency could be used on the Asian
mainland with the yen continuing its separate existence - along the lines of the Euro and the
pound sterling. But all that is a very long way away. And by that time, we could even be
moving towards the use of one global currency.
When that happens, current account deficits will not matter, and we will not even
know what they are. for instance, we do not pay much attention to payments imbalances
among American states. And even in Euroland, it is already possible to run very large
current account deficits. For instance, Portugal's current account deficit in 2000 was 11
percent of GDP. But it will only be in the Keynesian long run, in a system without national
currencies, that we will cease to care about national balance of payments deficits.
217
COMMENT
Global Imbalances and Exchange Rates
Malcolm Knight
General Manager, Bank for International Settlements
It is a great pleasure for me to participate in this conference on the occasion of the sixtieth
anniversary of Bretton Woods. It is also an honour because the IMF, which was
established by the Bretton Woods Conference of 1944, has a long tradition of thinking
deeply about the topic of this panel discussion"global imbalances and exchange rates".
The near-term outlook for global output growth and external adjustment over the next 18
months or so currently appears relatively benign. But over the medium termthat is,
roughly the rest of this decadeI believe that global imbalances will pose a major
challenge for the achievement of solid output, employment and price performance in many
countries. Indeed, with due difference to Deputy Governor Li Ruogu of the People's Bank
of China, who is also present on this panel, I would hazard the view that the ancient
Chinese curse"may you live in interesting times"is rather appropriate to the subject of
this session. From the point of view of global imbalances, I think we certainly are living in
"interesting times".
If we look at the adjustment of external current account balances over the past forty
years of so, it is only a modest oversimplification to assert that the major swings in global
imbalances correspond roughly to each chronological decade. For example, the decade of
the 1960s was a period of relatively small current account imbalances, strong growth, and
low inflation. To put it another way, that decade was marked by a salubrious environment
for external adjustment. In sharp contrast, the decade of the 1970sat least after the huge
oil price increase of 1973-74as characterised by very large external current account
imbalances, stagflation, and weak macroeconomic growth performance.
The decade of the 1980s saw a very large swing in the real exchange rate of the
US dollar. The large real appreciation of the dollar from the beginning of the decade
through the mid-1980s was associated with a sharp increase in the US current account
deficit, and the subsequent large dollar depreciation was marked by a return to rough
balance in the US current account over the latter part of the 1980s, especially with the fall
in domestic investment that began as the US economy moved into recession towards the
end of the decade. The decade of 1990s began with a marked recession in a number of
countries, but thereafter it was generally marked by strongly improving fiscal balances in a
219
number of large countries, by monetary policies that were directed at maintaining low and
stable inflation, and by relatively small external imbalances. In other words, the past
decade was a salubrious one, a bit like the 1960s.
Currently we are experiencing a synchronized upturn in the world economy that
has gradually gathered momentum. Does this mean that the first decade of the third
millennium will also be one of solid macroeconomic performance, like the 1960s or the
1990s? While I agree that the short-term global outlook is relatively benign, I have major
concerns about the longer-term prospects for external adjustment, and for the achievement
of solid output and employment performance.
There are a number of reasons for these concerns. First, the current juncture is
marked by very large external current account imbalancesparticularly the deficit in the
United States and the surpluses in Asia, the transition economies, and Latin America.
Second, the stance of both fiscal and monetary policies in many countries is currently
much more expansionary than is likely to be sustainable over the longer term. Third, partly
because industrial production is shifting from regions where the primary input content of
industrial production (including energy inputs) is relatively low at the margin to emerging
market countries where it is higher, inflationary pressures from energy and other primary
product prices are emerging earlier than would typically be the case at this stage in a global
economic upturn.
To put things in a nutshell, external adjustment at the global level is likely, over the
current decade, to be much more challenging than it was in the 1990s. This challenge
comes from the need to adjust major external imbalances at the same time that a number of
major emerging market economies, in particular in Asia, are being integrated into the
global economy. The first taskmoving to a more sustainable pattern of payments
balancesis a macroeconomic adjustment issue which will require shifts in
macroeconomic policy. But the second taskthat of adjusting to large and rapid changes
in comparative advantages and trading patternswill require deep and painful structural
adjustments. This challenge will create frictions and structural problems that are largely
separate from those of trade imbalances per se. In particular, the increasing supply of
goods and services from major emerging economiesand I am not just speaking of China
and India, though they are by far the largest exampleswill trigger profound changes in
production, international trading patterns, and relative prices. In particular, deficit countries
will need to shift resources into the production of tradables at a time when the integration
of large low-wage economies into the global economy is creating intense competition in
international goods markets and is putting strong and sustained downward pressure on unit
labour costs all around the globe.
A Comparison of Current Account Cycles
Before I expand on these themes, let me briefly review major current account swings by
contrasting recent developments in the external payments position of the US economy which is almost one fifth of the global economy - with those in the rest of the world. These
developments are summarised in the table below.
220
CHANGES IN CURRENT ACCOUNT BALANCES: 1981-2003
Country/region
United States
In billions of US dollars
As per cent of GDP
-421
-3.4
-42
-0.9
39
1.0
Emerging Asia
130
3.0
Latin America
71
3.5
Transition economies
37
3.2
Others
58
Na
-128
-0.4
Euro area
Japan
Total
Country/region
As per cent of GDP
164
3.5
Euro area
-104
-1.7
Japan
- 16
-1.5
United States
Country/region
United States
Euro area
Japan
-166
57
79
As per cent of GDP
-3.6
1.7
3.0
221
The top panel of this table indicates that, over the six-year period 1997-2003, the
US current account position deteriorated by over 420 billion dollars, an increase in the US
external deficit of some 3.4 percentages of GDP-which brought the level of the external
current account deficit to around 5% of GDP by 2003. In dollar terms, the main identified
counterparts to this widening US current account deficit are not found in the industrial
countries, but in the changes to the current account positions of emerging Asia (plus
US$130 billion), Latin America (plus US$70 billion) and the central and eastern European
transition economies (plus US$40 billion). Note also that some 30% of the US deficit is
unaccounted for.
Many commentators agree that these changes in relative current account positions
are not sustainable over the long term. If they are indeed unsustainable, how will the
adjustment take place? What challenges will it present? And what are the policy
implications?
These are not easy questions to answer, we are naturally led to look for guidance
from past historical episodes. I think the decade of the 1980s presents one. Over 1981-87
the US real effective exchange rate appreciated by some 40%. US fiscal deficits were also
large, and US economic growth was strong. The middle panel of the table indicates that the
deterioration in the US current account deficit over this period was slightly larger as a
percentage of GDP, 3.6 percentage points, than it was over 1997-2003. This deterioration
amounted to some US$66 billion, and almost the exact counterpart in dollar terms was a
strengthening of the current account positions of just two regions, Europe and Japan.
The bottom panel of the table shows how the US external imbalance of the first half
of the 1980s was corrected in the late 1980s and early 1990s. From 1987 to 1991 the US
current account position improved by 3.5 percentage points of GDP. It is very interesting
that most of this improvement of 164 billion dollars was the counterpart of rising external
current account deficits in the same two regions: Europe (US$104 billion) and Japan
(US$16 billion). These deficits reflected the strong economic growth that marked the
European and Japanese economies at that time.
If we compare this full current account cycle of the 1980s with the uncompleted
cycle in recent years we see that during 1997-2002, the US dollar also appreciated strongly
and the US current account deteriorated by about the same percentage as had been the case
in the early 1980s. The question iswhat will happen next? Again, if we look at historical
experience in the last full US balance of payments cycle the current account improvement
that took place after 1985 was associated with a very large real depreciation of the US
dollar, which was quite orderly. Is this relatively orderly and benign adjustment likely to
prevail again over the remainder of the first decade after 2000? In order to answer this
question it helps to look into the reasons why the US current account deficit rose after
1997. If the causes were similar to those of the 1980s, then the swing in the real exchange
rate over the current external adjustment cycle should be associated with an opposite swing
in the external current account deficit as a percentage of US GDP.
Graph 2 shows that one reason for the widening US current account deficit over the
seven years to 2003 was that cumulative real domestic demand growth in the US was
above that in the Euro area by some 15 percentage points and exceeded demand growth in
222
Japan by a cumulative 25 percentage points. The role of relative demand growth in the
widening of the US current account imbalance is particularly important given that there is
also an asymmetry between US export and import elasticities. US demand for imports is
much more income-elastic than foreign demand for US exports. Hence the US current
account balance tends to deteriorate even if the US is growing at the same rate as its
trading partners.
Graph 3 looks at the widening of the US current account deficit since 1997 from a
saving - investment perspective. The rise in the US external deficit and the associated US
external borrowing needs of the past seven years can be divided into two phases. In the
first phase, over 1997-2000, the US current account deficit essentially reflected a modest
decline in net private saving as a percentage of GDP, and a high rate of net private
investment associated with a sustained rise in the rate of growth of productivity and
improved long-term growth prospects for the US economy. With significant fiscal
consolidation during this period, the external current account deficit was financed by
sharply rising capital inflows. By contrast, from 2001 onwards there was a sharp decline to
a lower level of net private investment, but at the same time tax reductions and rising
government expenditures led to a very marked shift from fiscal surplus to a large fiscal
deficit.
Graph 4 shows that net long-term private capital inflows into the United States,
which rose sharply from 1998 to 2000, financed the rising current account deficit.
Nevertheless, although net long-term private inflows remained fairly high thereafter, in the
second phase after 2001, the role of official foreign exchange purchases became much
more important in financing the continuing rise in US current account deficits.
With this historical perspective, I return to the question of what the global
adjustment process might look like going forward. More specifically, will it look like the
latter half of the 1980s? From the first quarter of 1985 to the end of 1987, the US dollar
depreciated by exactly the same amount (40%) as it had previously appreciated in the five
years to 1985. In other words, it depreciated to about its 1980 level and remained there for
the next four years. In addition to the lagged effects of the sharp depreciation of the dollar
that began in 1986, the main features of the improvement in the US external account
position were the marked strengthening of activity in Japan and Europe in the late 1980s.
As a result of these factors, the US current account began to improve late in 1987, reaching
a surplus in 1991 (see graph 5).
The four panels of graph 6 consider how the present level of the US real exchange
rate and external current account may look relative to the exchange rate/current account
cycle that took place in the 1980s. In these graphs the vertical line is the year of the peak in
the real exchange rate of the US dollar during each decade-long exchange rate/current
account cycle (1985 for the cycle of the 1980s and 2001 for the current (uncompleted) US
current account cycle). As shown in the south-west panel of this graph, the pattern of
current account deterioration in the five years preceding the exchange rate peak and in the
two years following it is about the same thus far in the current cycle as it was in the 1980s.
However, the current account imbalance is about 1.5% points of GDP worse than it was at
the peak of the previous cycle.
As the top left-hand panel of graph 6 indicates, the real effective appreciation of the
US dollar was slightly smaller over the five years preceding the exchange-rate peak this
223
time around than it was in the 1980s. The adjustment in the 1980s involved a complete
turnaround in the real effective rate of the dollar. By contrast, in the current cycle the
decline in the real effective rate of the dollar up to June 2004 has been only a fraction of
the preceding appreciation. This may give pause to those who think that, since its peak two
years ago, the US dollar has depreciated enough to adjust the current US position.
Of even more concern is the difference in the US structural fiscal position over the
two cycles. As the top right-hand panel shows, the structural fiscal position in the 1980s
basically deteriorated steadily prior to the maximum-deficit point of the cycle before
improving slightly thereafter. Thus the current account effects of the exchange rate
depreciation over 1985-87 were not strongly offset by changes in the fiscal position. In
contrast, the US fiscal position was improving until 2001. In the two years since then it has
deteriorated to a level that is almost the same as the fiscal position two years after the peak
of the previous cycle. This weak fiscal position is likely to offset the positive adjustment
effects of future US dollar depreciation, suggesting that the dollar may have to fall by more
this time around unless active measures of fiscal restraint are taken.
The lower right-hand panel looks at relative demand patterns in the two episodes of
US current account change. Whereas demand in the rest of the world was strengthening
relative to US demand after the US cyclical peak had been reached in the mid-1980s,
thereby reinforcing the strengthening of the US current account, the demand gap between
the United States and the rest of the world has been essentially zero during the current
cycle.
Adjustment Challenges Today
History never repeats itself exactly. The experience of the US external current account
cycle that took place during the 1980s suggests that the adjustment to reduce a very large
US current account deficit can take place without cataclysmic disruptions, (not least
because the United States is at the centre of the international monetary system and can
borrow in its own currency). But this experience also suggests that, especially in the
absence of fiscal consolidation, the current account adjustment is likely to require a rather
'large depreciation of the dollar's real effective exchange rate, that the adjustment process
will take a number of years, and that it will be marked by relatively lacklustre economic
growth performance.
Moreover, there are aspects of the current situation that were not present last time
and that could make the future adjustment process more challenging. A first key
consideration is the impact of emerging Asia. Future adjustment of global imbalances will
require that the key current account deficit country (the US) shifts resources into the
production of tradables in the face of intense competition in world markets for goods and
services coming from China, India and the other emerging market countries. This was a far
less important factor in the 1980s balance of payments cycle. The so-called ccNewly
Industrialised Countries" that entered the world trading system in the 1980s accounted for
less than 2% of the world's population. By contrast, India and China alone have one third
of the world's population. The result is a huge labour force, much of which is educated,
and much of which is currently underemployed. This is likely to put strong downward
224
pressure on unit labour costs in all developed countries for many years and exacerbate the
problem of shifting resources to tradables as the US economy endeavours to adjust.
It seems to me that this observation has two implications. First, for the United
States, the real effective exchange rate is likely to have to depreciate more than it did in the
1980s, in order to produce the required shift into tradables in a fiercely competitive world
goods market. It goes without saying that this correction will be even larger if fiscal
adjustment in the US is delayed. Second, for all developed countries, even if productivity
growth remains at current levels, real wage growth will tend to be slower than it has been
in the past. This in turn suggests weak consumption growth in the developed world, so that
the adjustment is likely to be characterised by larger than normal output gaps and weaker
output growth.
The second challenging aspect of today's adjustment process is that the next
adjustment could also be more stagflationary than the 1980s adjustment. I have already
noted that this time, primary commodity prices have risen more, and at an earlier stage in
the upswing, than was the case in the 1980s. This rise in costs risks creating stagflationary
effects. Indeed, I believe that if goods production is shifting to countries which, at the
margin, have a higher input of commodities per unit of output, the effects will be
stagflationary even if the secondround effects of prices on wages are much weaker than
they were in the 70s and 80s.
In the late 1980s, the reduction of the large US current account deficit was also
made easier by the fact that aggregate demand in Europe and Japan was accelerating. This
time, as I have already suggested, slower wage growth will result in weaker growth of
domestic consumption. Finally, the negative wealth effects of the global adjustment
process will be much larger than they were in the 1980s. The large US current account
deficits of the last few years mean that residents of the rest of the world are now holding a
larger stock of dollar-denominated financial assets, currently estimated at some
US$9 trillion. A depreciation of the US dollar that was of similar magnitude to that which
occurred from 1985 to 1987 would reduce the real value of these assets by some 40%. Like
the other effects, such a reduction in wealth would also tend to weaken demand and
economic activity. And these effects could be very large.
Policy Implications
Going forward, therefore, the risks in the global adjustment process lean towards a marked
weakening of growth performance. How should national economic authorities respond to
achieve smooth adjustment of global imbalances to sustainable levels? I do not think the
answer lies in monetary policy. Monetary policies are currently very expansionary in a
number of key industrial countries. In order to avoid exacerbating the inflationary side of a
possible stagflation, monetary policy in the key currency countries/regions must shift to a
more neutral stance. But in order to avoid overshooting of exchange rates among the key
currencies, the timing of this withdrawal of monetary stimulus will have to be consistent
across key currency zones. The precept for the coordination among monetary policy
makers over the next several years must be: "first, do no harm". By contrast, fiscal
adjustments could, in principle, assist the external adjustment process. Most obviously,
225
active policy measures to reduce the fiscal deficit in the US to a sustainable level would
contribute to a smoother adjustment of a global external imbalance.
But these adjustments, even if they are vigorous and in the right direction, may not
appear to yield much in terms of output and employment improvements as key emerging
economies that have low unit labour costs in an ever-broadening range of industrial sectors
are becoming more and more prominent in the international market place. Viewed in this
light, the adjustments needed for the US to achieve sustainable fiscal and current account
deficits may be large. The first decade of the twenty first century may indeed be
"interesting times."
226
Graph 1
US real effective exchange rate

In terms of relative consumer prices; January 1997 = 100
Graph 2
Real total domestic demand

1997 Q1= 100
US saving and investment
Graph 3
As a percentage of GDP
Defined as gross saving less gross investment.
Source: US Bureau of Economic Analysis (SCB table 5.1).
Graph 4
Current account and long-term capital in the United States

1997
1998
1999
Source: US Bureau of Economic Analysis.
12
Graph 5
US current account and domestic saving-investment balances

As a percentage of GDP
Source: US Bureau or tconomic Analysis.
Graph 6
Current and previous cycle in the United States
Years (cycle peak = O5)

1
1n terms of relative consumer prices; peak year =100. 2 General government cyclically adjusted financial balance, as a percentage of
potential GDP (OECD definition). 3 As a percentage of GDP. 4 US gap less rest-of-OECD gap; in real terms. 5 Defined by a peak
in the real effective exchange rate; current and previous peak: 2001 and 1985 respectively.
Sources: National data; OECD; BIS.
KEYNOTE SPEECHES
Improving the Policy Response to Financial Crisis

Remarks by Agustin Carstens,
Deputy Managing Director, International Monetary Fund
at the
Bank of Spain and IMF International Conference on
Dollars, Debt, and Deficits60 Years After Bretton Woods
June 14, 2004
It is a pleasure and an honor to have the opportunity to address such a distinguished
audience at this international conference that celebrates the 60th anniversary of the creation
of the Bretton Woods institutions. At the outset, I would like to thank the Banco de Espana
for its superb hospitality and for the very close collaboration with the IMF in putting all of
this together
During the last 60 years, the Fund and its member countries have been struggling
with the challenge of improving the policy response to financial crisis. As many of us
know quite well, when a crisis occurs, resolution is far from an easy task. Ideally, policy
measures will aim to limit disruptions, stabilize the economy, and lay the groundwork for
the resumption of long-term growth. Yet, the policy response involves difficult judgments
and unpleasant tradeoffs amid significant uncertainty. Measures may have temporary side
effects andespecially if reforms are not fully carried througha country can remain
exposed to significant post-crisis vulnerabilities.
It is striking how, over the last thirty years, crises in many countries have become
recurrent. Recent research we have undertaken at the Fund shows that this is particularly
true in Latin America. Since 1980, 22 Latin American countries have suffered either a
financial crisis or financial distress, in the latter case narrowly averting a full-blown crisis.
Of these 22, three countries have experienced four episodes each; three countries have
experienced three episodes each; and nine countries have experienced two episodes each.
For example, Argentina has suffered four crises, in 1980, 1989, 1995 and 2001. Ecuador
suffered three crises, in 1982, 1996 and 1998, and narrowly averted one in 2002. Brazil
and Bolivia each experienced two crises between 1985 and 1995, and pulled back from the
brink in 2002 and 2003, respectively.
I feel that more systematic research on recurrent crises is urgent. In trying to
advance some ideas on the issue, I would venture to say thatall too frequentlythe
compromises that are made with the intent of resolving current difficulties may actually
sow the seeds for the next crisis.
Accordingly, I would like to focus my remarks today on four issues, which I think
are critical for the international community to better understand our efforts to improve the
policy response to financial crisis:
First, the complexities involved in designing an appropriate policy

framework;
233
Second, the "ugly" tradeoffs policy makers face in formulating policy

options;
Third, the post-crisis vulnerabilities that might remain and the need for
countries to persevere in their efforts to reform and to "crisis-proof their
economies; and
Last, the role of the Fund in this context.
First, consider how difficult it is to identify the appropriate policy framework.

There is uncertainty regarding the scope and impact of a crisis. In the midst of a crisis, key
macroeconomic variables tend to display unusually high volatility. The resilience of the
economic system and the severity of the crisis are uncertain. As a matter of fact, once a
crisis has erupted and the Fund's mission starts its groundwork, it is not uncommon that
preliminary assessments need to be revised as, in a crisis environment, hidden liabilities
both public and privatecan pop up like rabbits out of a hat.
Balance sheet interlinkages across sectors of the economy can amplify weaknesses
in individual sectors and propagate a crisis across sectors, to the balance of payments, and
to the sovereign. Such interlinkages may complicate the policy response and magnify the
costs of crisis resolution. There is also uncertainty regarding the effectiveness of policies,
which depends critically on the perceived credibility of the corrective measures when the
reputation of the authorities tends to be at its lowest level. And there is uncertainty
regarding the political support for reforms. Policy makers face the additional challenge of
quickly mobilizing public support for often unpopular measures.
Against this background of economic and political uncertainty, the policy response
is often subject to "ugly" trade-offs. A prime example is the fiscal consolidation that may
be required to reduce imbalances or debt burdens. This must be sufficient to strengthen
confidence in the sustainability of public finances, but not so much as to undermine
medium-term growth prospects.
In cases where a banking crisis is part of the problemsuch as the recent crises in
Argentina, Turkey and Uruguaypublic support will likely be required to safeguard the
functioning of the domestic financial system. But this support can exacerbate debt
sustainability concerns and make the previously mentioned tradeoff even more difficult.
Moreover, in extreme situations, administrative measures may be seen as unavoidable for
quelling a banking crisisalthough these risk eroding confidence in the banking system
and triggering capital flight and financial disintermediation.
In cases where sovereign debt restructuring is needed, the benefits of alleviating the
liquidity or solvency constraint must be weighed against the implications for future access
to capital markets. In addition, policymakers may need to factor in the potential costs to the
domestic financial system if bank portfolios are significantly exposed to government debt.
Next, consider howeven if a credible adjustment quickly restores confidencea
number of vulnerabilities can linger and perhaps even increase. Public finances may
remain vulnerable to shocks. If public debt remains at high levels, gross financing
requirements continue to be large, and thus vulnerable to shocks or spells of market
drought. Banks may remain vulnerable to debt servicing difficulties of household and
corporate sectors, or because of a large exposure to sovereign debt. And in cases of
sovereign debt restructuring, a country may lose access to markets for a prolonged period
of time.
234
Perhaps most importantly, there is always a danger of "reform fatigue." As

countries move out of the critical stage in the process of crisis resolution, it is not unusual
to see some of them lose their drive towards reform. Several factors contribute to this: a)
the erosion of political capital; b) an early positive response from investors that might lead
to complacency; and c) the fact that many of the needed reforms do not induce
immediately higher growth and wellbeing of the population. But here is precisely where
the recurrent crises have their genesis. Once the most urgent measures have been
implemented, the tendency is to put aside the important ones for later. The message is
clear: it is critical that countries persevere with reforms to "crisis-proof their economies
and avoid recurrence of financial distress.
Finally, let me say a few words about the role of the IMF in crisis resolution, and
how we are working to improve it. A key role of the Fund is to work with members to
achieve a durable exit from crisis. The Fund helps members to consider the relevant
constraints and trade-offs and to design an appropriate adjustment program that addresses
underlying macroeconomic problems, as well as anchors investors' expectations about the
formulation and implementation of economic policies. In helping to design this program,
the Fund has to form a judgment about the appropriate balance between the availability and
scale of IMF financing, the amount of domestic policy adjustment, and securing the
support of other stakeholders (official and private creditors). And in forming this
judgment, the Fund must also consider the implications a crisis country may have on the
stability of the international financial system.
Based on previous experience, we are continually examining policies that could
help reduce the frequency and severity of crises. We are focusing on several issues at
present, including reinforcing the rigor of our debt sustainability analysis, especially
through a more thorough analysis of contingent claims. We are working to raise awareness
of balance-sheet vulnerabilities, to ensure that risks are properly assessed and effectively
addressed. We have taken steps to improving clarity about IMF lending decisions,
especially with regard to the situations when exceptional access to IMF financing may be
appropriate. And we are working to improve the process for restructuring sovereign debt
within the existing legal framework. This includes encouraging the use of collective action
clauses (CACs) in new sovereign bond issues, as well as supporting private sector efforts
to formulate a voluntary Code of Conduct.
Strong banking systems are a foundation for financial stability. Therefore, we have
been working also with the Basel Committee and its chairman, Governor Caruana, in
encouraging emerging market countries to move towards the adoption of the new set of
standards embodied in the Basel II Accord, ensuring that this transition occurs at their own
pace and based on their own priorities.
Ladies and Gentlemen, notwithstanding all these efforts, there is no doubt that
substantial challenges are still in front of us. Therefore, in closing, I would like to
commend the organizers of this conference for bringing together distinguished economists
and policymakers that through their interventions will shed some light on the critical issues
involved in building a more stable international financial system. Yet our main challenge
will be to put all of these ideas together, into a coherent strategy for crisis prevention and
resolution that is both effective and durable. If we can do this, we may also be able to abate
the recurrence of crisis.
235
The International Financial Architecture Where Do We Stand?

Remarks by Jean-Claude Trichet, President, European Central Bank,
at the
Bank of Spain and IMF International Conference on
Dollars, Debt and Deficits - 60 Years after Bretton Woods
June 14, 2004
It is a great pleasure to be here in Madrid at the invitation of the Banco de Espafia and
the International Monetary Fund. The 60th anniversary of the Bretton Woods
institutions is indeed an important opportunity to take stock of key events,
developments and issues that have shaped and are shaping the international financial
system. Let me thank the organisers, and especially our host Jaime Caruana, for setting
up a very interesting programme for this conference that encompasses all these key
topics. Tonight, I would like to offer some thoughts on the question 4Hhe international
financial architecture - where do we stand?" In my remarks, I will first have a short
look back at the changes to the international architecture over time, before
concentrating on reform efforts that have been undertaken in four areas, namely the
institutional setup, transparency and best practices, regulation of financial markets and
crisis prevention and resolution.
Let me briefly look at the changes to the international financial
architecture over time. The key aim of today's policy makers has not changed
compared to those at the Bretton Woods times - it has been, and still is, global
prosperity and stability - but the environment in which we are acting has changed
profoundly. The founders of the IMF and the World Bank wanted to create institutions
that prevent countries from falling back into autarky and protectionism and that help
them to raise growth and increase stability in a world of fixed exchange rates with still
a large degree of capital controls. Today we are striving for stability of the international
financial system in a world of free capital flows with a growing importance of private
flows and increasing trade and financial integration. Among the major factors that we
have to take into account, I would like to mention in particular:
The financial globalisation phenomenon: capital market liberalisation, both
domestically and internationally, technological advances and buoyant financial
innovations have contributed to set up a totally unknown degree of financial
globalisation - with great benefits, but also new risks.
The policy responsibility which still lies mainly with sovereign states; thus, the
challenge is to promote global financial stability very largely through national
actions enlightened and co-ordinated through a larger degree of intimate
international co-operation.
A very large consensus on giving the private sector and markets a central role on
the one hand, and relying upon sound public institutions to provide market
participants with the appropriate environment on the other hand. This shift from
237
direct public involvement to private activities is particularly striking when looking

at financial flows to emerging markets: in the 1980s, official flows were dominant,
reaching on average over 60% of total flows to emerging markets. By contrast, the
1990s saw a dramatic increase in private flows, which on average accounted for
around 85% (in the period from 1990 until 2003). Equally striking is the shift from
bank loans to negotiable securities as the major financing tool for the developing
countries.
The integration of the European Union, reinforced with the introduction of the euro,
has increased the economic, monetary and financial stability of a region that
constitutes today the world's largest trading partner and the second largest
economy. The EU has also been crucial in anchoring the transition process in
central and Eastern Europe, and in fostering stability and prosperity in this region.
The dynamics of today's world call for continued adjustment at a global level.
New challenges have been added to existing ones, such as poverty reduction. New
actors gained prominence on the international scene, with developing and emerging
markets becoming progressively full participants in the globalised economy. The
financial crises of the 1980s and the 1990s, characterised by large and sudden private
financial flow reversals, marked by a very powerful contagion phenomenon and
demonstrating some of the potential and actual vulnerabilities of the newly globalised
financial system, led to an ambitious reform agenda to strengthen the international
financial architecture.
Let me focus on the lessons from the crises in the 1990s and the ensuing work
on the international financial architecture.
On the basis of the experience with the Mexican crisis in 1994/95, the G7
summit in Halifax in June 1995 initiated work on improved crisis prevention and
management. It called for improved transparency, both at the level of individual
countries and at the IMF, and for strengthened IMF surveillance. The Halifax summit
also pointed to the importance of effective financial regulation, market-reinforced
prudential supervision and enhanced international co-operation among regulators and
supervisors. As for crisis management., concrete proposals were presented in the Rey
report to G10 Ministers and Governors in May 1996.
Work was stepped up in the aftermath of the Asian crises, which revealed
further vulnerabilities in national and international financial systems. But most
importantly, the crises in the later 1990s showed that the systemic changes in the
world's financial markets required systematic changes in the policy framework that
underlies the international financial system Almost a decade later, we can say that
many of these proposals have been implemented. Let me now focus on four different
areas, which I consider most important:
The first area concerns the international institutional set-up, which, in my
view, has been strengthened significantly since the 1990s. The existing international
financial institutions, in particular the IMF, the World Bank and the BIS maintained
their central role in the system. But they were subject to several changes to sharpen
their respective focus, to reinforce their policy advice and financial support, to enhance
their transparency and accountability and to strengthen their governance. The Bretton
Woods institutions, and particularly the IMF, underwent profound changes to adapt to
the new environment. In addition, new fora have been created in response to the
widening of the number of actors in the global economy and the growing importance of
international financial markets. The creation of the G20 in 1999 constituted in my view
a decisive and highly welcomed step to reflect adequately the newly globalised
238
economy. The G20 has turned into the international forum for appropriate dialogue and
consensus building between all economies that have a systemic influence, whether
industrialised, emerging, or in transition. Equally important is the Financial Stability
Forum, which is the first informal grouping to fully recognise the existence of a
globally integrated economic and financial system It is also the first forum to set the
goal of systemic optimisation of each of the subcomponents of the system, whether it is
banking surveillance, insurance surveillance, securities market control, accounting
rules, good practices of public and private sectors, functioning of the major market
places, governance of the IFIs etc. At the regional level, the European Union has
established a whole universe of arrangements for co-operation that is constantly being
adjusted to its changing needs and European institutions are becoming increasingly
involved at the international scene, for instance with the EU-US regulatory dialogue.
Overall, improving the governance of the international institutions and
optimising the work of the informal groupings will always remain a moving target
given that these entities permanently will have to adapt to a changing environment.
However, with the changes introduced in the recent years, the foundations of the
international financial system have been strengthened considerably.
The second area I would like to highlight regards the work to enhance
transparency and promote best practices, where significant progress has been
achieved in a number of fields. Indeed, a wide-spread consensus has developed,
which considers reliable and timely information on economic and financial data as a
precondition for well-functioning markets, since it facilitates better risk assessment and
management and hence strengthened market discipline. The IMF's special standard for
dissemination of economic and financial data has become a widely recognised
benchmark to which a large and increasing number of countries have subscribed. There
is now a presumption that IMF papers on Article IV consultations and on Fund
programmes are published. International codes of good practices have been agreed
upon, such as the ones on transparency in fiscal policy and on transparency in monetary
and financial polices. Moreover, countries' compliance with the 12 most important
standards and codes are regularly examined by the IMF and the World Bank in socalled ROSCs (Reports on the Observance of Standards and Codes), many of which are
made publicly available and have a positive impact on the market's assessments of the
countries concerned.11 consider that the progress made in the field of transparency after
the Asian Crisis is one of the main explanations for the absence of contagion in the
emerging world when the Argentine crisis erupted.
Transparency in the private sector is also crucial for well-functioning
international financial markets. Reliable and timely company information are one key
element to transparency, which is provided mainly through financial statements. Recent
corporate scandals have again brought to our minds the crucial role that accounting
standards play in this respect. In this context, I attach great importance to the reform of
the International Accounting Standards (IAS) and the key role of these standards in
advancing the European single market. The IAS, which will apply to all listed
companies in the EU, are expected to have a major impact on the European banking
system The banking sector will particularly be affected through the proposed valuation
rules for financial instruments and through the rules on disclosure.
!
The IMF and the World Bank have recognised 12 standards as useful for their operational work. These comprise
accounting; auditing; anti-money laundering and countering the financing of terrorism (AML/CFT); banking
supervision; corporate governance; data dissemination; fiscal transparency; insolvency and creditor rights; insurance
supervision; monetary and financial policy transparency; payments systems; and securities regulation.
239
The EGB has a strong interest in this debate mainly from its focus on
contributing to the maintenance of financial stability. Thus, the primary objective of
this reform has our full support as it aims to minimise the gap between the reported
information and the true risk profile of a company. I am also fully aware of the
complexities stemming from the interrelation of accounting standards with other
reporting schemes .for supervisory and statistical purposes. However, some proposals
have given rise to concerns also within the ECB. In particular, those proposals relating
to an extensive use of fair values raised concerns about the possible adverse
implications on the volatility of bank income and, eventually, on bank behaviour and
on financial stability. As a consequence, the ECB contributed to this debate,
highlighting the concerns and showing their relevance. The more recent proposals from
the IASB moved into the direction of limiting the use of fair values for those items that
can be reliably measured. The revised proposals should help to avoid undesirable
consequences such as an artificial increase in income volatility. At the current juncture,
the ECB is carrying out an exercise to check the likely impact of the new standard.
While these issues apply to all financial markets, the EU faces a particular
challenge relating to the advancement of the single market. The introduction of
harmonised EU rules regarding the setting-up of financial statements are considered to
be a crucial step towards the further integration of the financial markets in the euro area
and the European Union. Indeed, improved comparability of disclosed information
would facilitate cross-border investment and further market integration. Thus, in 2002
the European Parliament and Council adopted a Regulation requiring listed companies
to prepare consolidated financial statements in accordance with I AS from 1st January
2005. A specific endorsement process is in place to ensure legal certainty and
consistency with EU public policy concerns. This process already allowed to endorse
all I AS with the exception of the two standards concerning recognition, measurement
and disclosure of financial instruments. Recently, in order to deal with the remaining
controversial issues, the Commission took the initiative to establish a high-level
dialogue between all the constituencies interested in high quality accounting principles.
I remain confident about the positive impact of prudently implemented International
Accounting Standards on the stability and efficiency of financial markets in the EU.
The third area relates to the strengthening of financial regulation in
industrialised countries. Here, let me recall that recent crises exposed weaknesses in
the risk management practices on the part of creditors and investors in industrial
countries, pointing to the importance of financial market regulation and supervision.
We all are aware of the importance of effective financial regulation and
supervision to maintain financial stability and protect consumers, also in light of the
increased complexity of financial services and products. Let me say one word on one
important aspect of financial regulation, which is the reform of the Basel Capital
Accord, coined Basel II. I am convinced that our host, Mr. Caruana, who is the
chairman of the Basel Committee on Banking Supervision, could off hand fill the
evening by elaborating over the main features of this reform. From what I gather, your
efforts are bearing fruits and we may expect a final text to be hopefully endorsed by the
G10 Governors and the Heads of banking supervisory authorities in the forthcoming
weeks.
The Basel II reform is of key importance. New and bold developments in the
banking industry are the ultimate reason for engaging in this reform.
The ECB has expressed on various occasions its supportive stance to the new
framework. The ECB was also among the first to point out the possible macro-financial
implications of any banking prudential scheme, highlighting the potential procyclical
240
effects that might be induced by any framework relying on a comprehensive real time
risk analysis. These concerns have been taken into account in the final version of the
new framework which aims at being neutral over the cycle. Looking ahead, we have to
recognise that we stand at the beginning of the road. The success of this reform will
crucially hinge on a sound implementation of the new framework requiring strong coordinating efforts among the supervisory authorities on a global basis. With regard to
the EU context, the new institutional setting based on the Lamfalussy framework
comprising a two-tier structure of regulatory and supervisory committees is expected to
play an important role in ensuring a more uniform and flexible EU regulation and
consistent implementation resulting from convergence in supervisory practices.
As the fourth and last area, I would like to mention crisis prevention and
management. Of course, the various efforts I mentioned so far should be conducive to
prevent crises from happening. However, crisis prevention primarily rests with every
single country with strengthened macroeconomic policies and financial systems. In that
context, the experience of the past decade has highlighted the crucial importance of
well-functioning domestic rules, regulations and institutions namely the legal
framework, the regulatory system, the enforcement mechanisms, and authorities that
shape and permit the optimal functioning of a market economy with its financial
markets. This includes, in particular, central bank independence, rules for monetary
policy and for fiscal policies, appropriate supervisory frameworks and authorities.
There is strong evidence linking well-functioning institutions and good governance to
positive economic and social outcomes. Institutional factors appear to be as important
as productive factor endowments or any other explanations in determining crosscountry differences in the overall level of development.
I am confident that these lessons feed into improving domestic policy-making in
emerging market economies, making them more resilient to withstand shocks. The
continued efforts to strengthen IMF surveillance play also a crucial role in that respect.
It is clear that crisis prevention must remain the key area of all our efforts.
Crises are costly for the countries concerned and also for the international system.
Given the increasing economic and financial importance of emerging markets, major
events in these countries are bound to have spill-over effects to the rest of the world.
Let me underpin this argument with some figures: In the last four years, major
emerging markets contributed to about half of global real GDP growth (in PPP terms),
accounted for roughly 30% of world exports and received about 20% of global FDI.
Turning to crisis management, important lessons have been learnt. There has
been a growing recognition that more predictability is required on the side of the
official sector in order to set the right incentives for all the actors involved. Moral
hazard concerns and the limited availability of official funds also led to increasing
discussions about the appropriate involvement of the private sector in crisis
management. Of course, every single crisis is different and hence there is in each case
the need to find the appropriate balance in the triangle of domestic adjustment, private
sector involvement and official support. Therefore, crisis management in practice still
has to struggle with the inherent tension between rules and clarity on the one hand and
discretion and flexibility on the other.
However, considerable progress has been achieved. First, specific criteria and
procedures have been set up last year to make exceptional access to Fund resources
subject to rules and hence more predictable. We in Europe have been very much in
favour of setting such clear rules and clear limits to Fund financing in view of the very
large financing packages provided to countries in the 1990s. The IMF's debt
sustainability analysis will play an important role in that context, since clear limits to
241
official financing must be respected especially when a country faces an unsustainable

debt burden and hence requires a debt restructuring. All IMF shareholders now need to
stick by these rules, not least in order to provide the right signals to the markets and to
avoid the impression that the official sector suffers from time inconsistency between
the approval of policy principles and their actual implementation.
Second, following Mexico's bond issue with Collective Action Clauses (CACs)
in February 2003, several emerging markets included CACs in bonds issued under New
York law. More than 70% of new bond issues since early 2004 include CACs. As you
probably know, no discernible impact on borrowing costs could be detected. In order to
help making CACs a standard feature in sovereign bond contracts, the EU Member
States committed themselves to include CACs when issuing new bonds under foreign
jurisdiction. All this progress is very remarkable, especially when comparing it to the
rather sceptical stance many countries and many private sector representatives have
taken in the past vis-a-vis the recommendations in the Key report after the Mexican
crises. Of course, so far these clauses have not been tested in practice and it will take
some time until CACs are included in the entire stock of debt.
Finally, work is proceeding on a so-called Code of Good Conduct, which I
suggested myself at the IMF Annual Meetings in September 2002. Such a Code would
define best practices and guidelines for the behaviour of debtor countries and creditors
regarding information-sharing, dialogue and close co-operation in times of financial
distress. While the IMF and the G7 encouraged further work and the G20 is closely
following the process, at present the official sector confines itself to a catalysing role
and leaves the floor to the true stakeholders in the process, i.e. emerging market issuers
and private sector representatives. I understand that currently intensive discussions are
taking place on the main elements of such a Code. I would like to encourage all parties
to be as active and constructive as possible in working out what could be a significant
new tool to prevent and help solving potential crises.
Closing Remarks
We have the unique chance of living in a world which is full of opportunities, very
inspiring and very complex, very rewarding and very demanding, full of chances and of
risks. We have all been the witness of two incredible transformations of the global
economy over the last twenty five years. The technological surge which has permitted
to compute and to transfer information at practically no cost. The globalisation process
which aims at connecting all economies and finances of the world within the same
market-economy based framework. So that goods, services, capital, technologies,
concepts, ideas are moving very rapidly or even instantaneously all over the globe,
expanding considerably, in quality and in quantity, the domain of the Ricardian
comparative advantage. The significant surge of labour productivity in a number of
industrialised economies, the taking off of India, China and a very large number of
emerging countries, the rapid race of global growth. These are great successes of
today's economic world of which global finance, mirror-image of a global economy, is
both the emblem and the very powerful tool. But there is no economic success without
risks. We have been living permanently in a risky environment over the last twenty-five
years. Amongst many risks, we might mention: the debt crisis during the 1980s,
starting with Poland and Mexico and spreading to Latin America, Africa, the Middle
East and the Soviet Union; the stock exchange fall in 1987; the Mexican crisis in 1994;
the bond market crash in 1994; the Asian crisis starting in 1997; the LTCM and
Russian crises in 1998; the recent stock exchange fall and the collapse of the
242
technology bubble in 2000. We have surmounted all these crisis episodes. We have
learned a lot and we have improved a lot in these occasions. One of my friends used to
say: "Good management comes out from experience and experience comes out from
bad management!" I think we are pretty experienced now and I take it that thanks to the
lessons drawn we have now achieved a level of crisis prevention which is much better.
But we should never forget that the risks are still there because they are intimately
associated with the structural transformation of the global economy. This is not, in any
respect, a time for complacency.
If I had to sum up what should be our today's mottos, I would make the
following five recommendations:
Let us not forget the crucial role of the IFIs, in particular the Bretton Woods
institutions, in the management of the present global economy. The constant
improvement of their management and instruments is key;
Let us tirelessly improve transparency in all fields: it is the best recipe for avoiding
both misallocation of capital and global crisis contagion;
Let us continuously improve the flexibility of our economies through bold
structural reforms. Not only because it improves efficiency but also, all the more,
because it improves resilience in a world where shocks are to be expected;
Let us reinforce our methodology to ensure that we do not amplify "pro-cyclical"
phenomena: the best envisaged at a local or sectoral level can be the enemy of the
good at a global systemic level. In this respect such informal groupings like the G20
and the Financial Stability Forum are of the essence;
Let us join efforts to improve our scientific knowledge of the new world economy.
Still today, academics and practitioners are observers and actors within the
environment of largely uncharted territories. The more profoundly we understand
the functioning of today's global economy, the more efficient we will be to weather
stocks, to prevent crisis, and to pave the way for continental and global job creation,
steady growth and overall stability.
243
BLOCK II
International Financial
Architecture
Session 4
The Role of the IMF
A Model of the IMF as a Coinsurance Arrangement"

Ralph Chami
Sunil Sharma
Ilhyock Shim
Bank for International Settlements
"This paper has benefited from the input of many colleagues, and the comments of participants at the
conference on "Dollars, Debt, and Deficits: 60 Years After Bretton Woods" organized by the Bank of
Spain and the IMF in June 2004. In particular, the authors would like to thank Lorenzo Bini-Smaghi,
Sajjid Chinoy, Tom Cosimano, Burkhard Drees, Jeff Fischer, Connel Fullenkamp, Rex Ghosh, Herve
Hannoun, Mohsin Khan, Tim Lane, Jaewoo Lee, Leslie Lipschitz, Alan MacArthur, and Miguel
Messmacher. Tala Khartabil provided superb research assistance. This paper should not be reported as
representing the views of the IMF. The first draft of this paper was dated August 7, 2003. The current
draft is dated November 12,2004. The views expressed in this paper are those of the authors and do not
necessarily reflect the views of the IMF or IMF policy. Authors' e-mail addresses: |^ynii@iincffg;
ssharma@imf.org; ilhyock. shim@bis.org.
Introduction
Financial liberalization, capital account convertibility, and the increasing importance of
private capital flows have dramatically changed the international environment in which
the IMF operates. For many countries, access to international capital markets has
brought opportunities for loosening funding constraints and underpinning more
ambitious growth strategies, as well as developing financial institutions that can hold
their own in the international financial arena. However, these opportunities have come
with new attendant hazardsa greater exposure to international liquidity cycles,
changes in the moods and expectations of foreign investors, contagion, and external
shocks in general. And the financial crises of the last decadeMexico (1994), Thailand
(1997), Korea (1997), Indonesia (1997), Russia (1998), Brazil (1999), and Argentina
(2001-02)have shown that when there is a massive withdrawal of external financing,
the cost to the economy can be punitive, and the demand for IMF resources can be huge
by the standards of earlier decades.l
These large IMF programs have reopened the debate on the nature and role of
IMF financing. Some observers argue that recourse to IMF financing generates moral
hazard on the part of both borrowers and lendersleading to less due diligence by
private lenders, and allowing borrowers to incur larger debts and get by with weaker
policies and institutions. Hence, IMF financing, though offering a cushion in times of
financial crises, increases the likelihood of such events occurring. Others argue that the
moral hazard associated with international financial support is limited and that the
focus should be on containing the real hazards generated by the structural and policy
deficiencies of emerging markets and their pernicious interaction with the global
financial system Markets do not always work to provide appropriate discipline, the
extent of access and the lending terms may not be justified by fundamentals, and when
problems are eventually recognized, markets may impose punishments that are overly
severe.2
The rationale for IMF financing has remained the sameovercoming market
imperfections and enhancing the world's ability to provide international public goods
that would otherwise be in short supply.3 There are a number of factors that lead to
countries being rationed or excluded from international financial markets: imperfect
information about country prospects and institutions, problems related to enforcing
sovereign loan contracts, and coordination problems among lenders. On the supply of
public goods, just as openness to trade is considered an international public good worth
cultivating, a cautious openness to international financial markets also contributes to
the development of countries and to the common good. By providing funds temporarily
to deal with external payment difficulties so that countries do not adopt policies that are
destructive of national and international prosperity, the IMF supplements private
markets when necessary, and helps countries to become more open to trade and capital.
It enables countries to bear the risks associated with reforming and developing their
financial systems and economies, and opening them to achieve a more efficient global
allocation of resources.
'See, for example, Jeanne and Zettlemeyer (2001), Ghosh etal. (2002), Haldane and Taylor (2003), and
Independent Evaluation Office (2003).
2
For a discussion see Bordo and Schwartz (1999), Calomiris (1999), Mussa (1999), Meltzer et al. (2000),
Jeanne and Zettlemeyer (2001), the recent review by Haldane and Taylor (2003) and references therein.
3
See, for example, Masson and Mussa (1995) and Krueger (1998). Also, see Cordelia and Levy Yeyati
(2004) who argue that the insurance provided to countries by the presence of the IMF may encourage
long-term reforms in developing economies.
251
This paper uses a stylized framework to examine the role of an IMF-like

institution in the world financial system. First, it shows that a coinsurance arrangement
among countries can, in principle, play a useful role in helping countries bear the risks
involved in developing their economies and becoming part of the global financial
system.4'5 The moral hazard inherent in such an insurance arrangement can be reduced
by peer monitoring, but given the coordination and administrative costs of peer
monitoring, countries may decide to create an institutional monitor like the IMF. The
IMF, along with administering the common pool of funds created for making loans to
members, undertakes country surveillance to limit moral hazard in the system
Second, the paper tries to model the operation of the coinsurance arrangement
by examining the nature and timing of interventions. The question asked is: how should
the loan contract between a borrowing country and the IMF be structured and when
should the contractual details be decided to create the right incentivesencourage
countries to take prudent risks, do their best to prevent external payment imbalances
from emerging, and should they run into trouble take the policies to rectify the
situation. Should the IMF precommit to a predetermined contract or should the
contractual details be decided ex post after the country is in crisis?
To examine these issues, we use a two-period repeated moral hazard setting in a
principal-agent framework (with the IMF being the principal and the borrowing country
the agent). The problem is examined under two objectives for the IMF: safeguarding
of its resources and a concern for the borrowing country's welfare. If the country runs
into trouble, the IMF provides funding over the two periodsa tranche in each period.
After the two periods the IMF is paid back by the country. The IMF cannot observe the
policy effort but can observe the country's output performance. A higher policy effort
in the first period increases the probability of avoiding a crisis, and should the country
get into one, a higher policy effort in the second period increases the probability of
recovery.
In our model, the IMF and the member country take sequential decisions to
maximize their respective utilities. Three cases are considered for the timing of IMF
intervention: (i) the ex ante contract, where the IMF precommits to a contingent loan
contract for the two periodsthe contract specifies the first and second tranches before
the country has chosen its policy in the first period; (ii) an ex post contract, where the
IMF chooses the contract after the country has chosen its policy effort in the first period
and fallen into a crisisthe contract specifies the first and second loan tranches after
observing the outcome in the first period; and (iii) a variation on the preceding ex post
contract, where the IMF chooses the first tranche after observing that the country is in a
4
In this paper, we use the term "coinsurance" synonymously with "mutual insurance." The coinsurance
arrangement can be thought of as an emergency lender envisaged in Fischer (1999), who makes the case
that such an institution need not have the power to create money, as long as it has the resources to play a
useful role as crisis lender and manager. For a more traditional interpretation of the lender-of-last-resort
see, for example, Capie (1998).
5
The Chiang Mai Initiative among the ASEAN countries, China, Japan, and Korea can be thought of as a
coinsurance arrangement designed to alleviate temporary liquidity shortages. Member central banks can
swap their own currencies for certain international currencies for a short period of time. The size of the
"borrowing" can be some multiple of the amount committed by the member under the arrangement. Note
also that the original conception of the IMF was based on the idea that most countries would be both
creditors and debtors to the IMF over time. In the 1950s and 1960s, with the exception of Germany and
the United States, most members fit this description and at some point used IMF resources to help fix
external payment imbalances. However, by the 1980s most industrial countries had begun to rely
exclusively on private capital flows, and the IMF membership became divided into creditor and debtor
groups (see Boughton 2004).
6
See, for example, Laffont and Martimort (2002) and references therein.
252
crisis and then chooses the second tranche after observing the output of the program
country.
The size and design of the IMF loan contract turns out to depend crucially on
the objectives of the Fund and the timing of Fund intervention. If the Fund were to be
concerned only with safeguarding its resources, then it would demand full repayment
irrespective of the country's situation. However, if in addition to safeguarding its
resources, the Fund also cares about the welfare of its borrowers, the contractual
repayment scheme in the second period is in general contingent on the economic
situation of the country.
The size of the first tranche lowers the policy effort both for avoiding a crisis as
well as for overcoming one. This is because the first tranche has to deal with two
dilemmas in our setting, the Samaritan's dilemma and King Lear's dilemma.7 The first
tranche, which is given after a country chooses its effort in the first period, lowers the
incentives for preventing a crisis because the country knows that the IMF cares about
its welfare and will provide a cushion in a crisisthe Samaritan's Dilemma (Buchanan
1975). On the other hand, providing the first tranche before the country decides on
policy in the second period does not create the right incentives for a program country to
solve the crisis. Once the first tranche is delivered, the country's choice of policy in the
second period need not be optimal from the IMF's perspectiveKing Lear's dilemma
(Hirshleifer 1977). We argue that to deal with these dilemmas it is best for the IMF to
commit to an ex ante contractthat is, design and offer the contract before a crisis
arises. Such a contract specifies the first tranche and the state-contingent second
tranche, penalizing the country for low output, but rewarding it if the output is high and
the country emerges from crisis.
In the presence of information asymmetries and given the mandate of the IMF
to safeguard its resources and care about the welfare of members, the timing of a Fund
program has a critical effect on the country's effort to avoid a crisis, and on its effort to
recover from a crisis. We show that ex post contractsthat is, IMF intervention after a
country has fallen into crisisdoes not elicit the highest policy effort from a country.
In such cases, the program country is likely to reduce its effort to recover from the
crisis, knowing that if it does not recover, a suitable loan will be available from a
"caring" Fund. In contrast, deciding on the IMF program ex ante tends to result in
higher effort by the country to avoid and to recover from a crisis. However, we show
that such a contract is subject to time-inconsistency problems, as the Fund and the
country may both find it in their interest to renegotiate the Fund's ex ante contract,
once the country enters into a crisis. Hence, it may be best for the Fund to precommit to
a loan contract, raising the interesting question of how such a precommitment can be
enforced.
The rest of the paper is organized as follows: Section 2 develops a model of
coinsurance, first with exogenous risks and then with endogenous risks. It then
examines the issue of moral hazard in coinsurance and how it can be reduced through
monitoring arrangements. Section 3 deals with the operation of the coinsurance
arrangement and the optimal choice of loan contracts for the IMF and member
countries. The last section concludes. Proofs of all results are given in the Appendix.
For a discussion of these dilemmas and related issues, see for example, Becker (1974), Buchanan
(1975), Hirshleifer (1977), Bemheim, Shleifer and Summers (1985), Cox (1987), Bergstrom (1989),
Bruce and Waldman (1990), Chami (1996,1998) and Jurges (2000).
253
A Model of Coinsurance
Consider the possibility of voluntary coinsurance between two countries. Suppose each
country is subjected to a shock (for example, a crisis) with probability n and that the
shock results in a (fixed) loss of output. The probability n of being subjected to an
adverse shock can be either exogenous or a function of the country's policy efforts to
decrease vulnerability to shocksthat is, policies conducive to economic growth,
macroeconomic and political stability, and a healthy financial sector.
Exogenous Risk Case: No Moral Hazard
We consider two countries / and j that face income shocks that are exogenous and
i.i.d. A shock leads to a fixed loss in income, 8, where 6>Q. Two states of nature,
good (G) and bad (5), occur with probability (1-tf) and n, respectively, where
Q<n<\. If the good state prevails, country / receives income yt(G) = yi9 and if a
country suffers an adverse shock, country / receives income yi(B) = yi-6i. Thus,
country / 's expected income is
For simplicity, we assume that each country has a Von Neumann-Morgenstem
utility function, which is event-independent:
(1)
where u(-) is a continuously differentiable concave function with u > 0 and
u < 0. Thus, in the absence of coinsurance, country / receives the expected utility
(2)
Now, suppose the two countries coinsure each other: if country / is hit by a
shock but country j is not, then j transfers J3 to / , and the reverse happens when j
suffers a shock and / does not. If both countries are hit by shocks, then there is no net
transfer. As a result, there are four possible outcomes under this coinsurance
arrangement, and the expected utility for country / is given by:
(3)
In addition to symmetry
we make two assumptions:
(i)
that is, when one country suffers a shock and the other does not, the
former receives a positive transfer from the latter;
(ii)
, that is, under the coinsurance arrangement each country
has at least as much income in the good state compared to that in the bad state. In the
symmetric case, this implies that ft < 812.
The result below shows that coinsurance is utility enhancing.
254
Lemma 1. In the presence of income risk, a country will prefer to have a

coinsurance arrangement:
In the symmetric case, when the income risk is purely exogenous, the optimal
transfer is /T-512. Therefore, in this case, coinsurance which equalizes the income
across the good and the bad states is optimal.
Endogenous Risk Case: Moral Hazard
Now suppose that the probability n of a country suffering an income shock depends on
the policy effort or actions taken by a country. The policy effort is represented by the
symbol e, with higher policy effort reducing the probability of the bad outcome. We
assume such effort is private information, which is not fully revealed to an outside
observer (or the other country). Let the effect of policy on the probability of the bad
outcome be given by n(e], a convex function, i.e. ;r'<0, ff">Q. However, policy
actions that can lower the probability of a bad outcome are "costly" for a country, and
the disutility from undertaking such actions is denoted by v(e), where v(-) is a convex
function, i.e. v' > 0, v" > 0. For simplicity, a country's utility function is assumed to be
additive and separable, and given by
(4)
The expected utility of country / is

(5)
When we adopt the Nash assumption, each country maximizes utility taking J3
and the effort of the other country as given. For country / , the first order condition with
respect to ei is
(6)
and a similar condition holds for country j. Together, these two reaction functions
provide the optimal (Nash) policy efforts for the countries: ei =ef.[./?,ej and
By symmetry, we get the
equilibrium policy
effort
levels:
The next question to ask is whether providing coinsurance exacerbates the

moral hazard problem. The following lemma answers that affirmatively.
255
Lemma 2. If income risk is endogenous, larger income transfers under the

coinsurance arrangement reduce the policy effort to prevent the bad outcome i.e.
The change in J3 affects the expected utility through two channels: the direct
effect of J3, jand the indirect effect of /? through e. The proposition below shows that
countries value coinsurance even in the presence of moral hazard.
Proposition 1. Even if the income risk is endogenous, countries may prefer to
have a coinsurance arrangement, that is,
However, in the presence of
moral hazard, the optimal income transfer is smaller than in the absence of moral
hazard, i.e. 0</Ty2.
Interdependence and Coinsurance: Asymmetry Matters
In the above discussion, a country cared only about its own utility under the
coinsurance arrangement. Now let's posit that a country also cares about what happens
to the other country in a coinsurance arrangement. The interdependence of countries in
international trade, financial markets, and through the coinsurance arrangement may
make them concerned about the macroeconomic and financial health of the other
country. Thus, country / may care about the utility of country j and vice versa.
To capture the interdependence among countries, we specify the utility function
as
(7)
where A. and /l^. are parameters that depict the concern countries have for each other.
The natural question to ask is what effect the interdependence of country
utilities has on the extent of moral hazard.8 The following two propositions show under
what circumstances moral hazard can be mitigated.
Proposition 2. Suppose the two countries in the coinsurance arrangement are
identical in all respects, including A, = /L = A . Then, as
Hence, when countries are perfectly altruistic (A = l) and show the same
amount of concern for the other country as they do for themselves, a coinsurance
arrangement does not create moral hazard. Each country fully internalizes the
externality imposed on the other country. This result, however, crucially depends on the
symmetry of the problem.
Proposition 3. Suppose the two countries in the coinsurance arrangement are
identical in all respects, except for the size of their current income (i.e. yi > >>,). Then,
even as
See Chami and Fischer (1996) for a similar result in the context of insurance markets.
256
Thus, the externality generated by the coinsurance arrangement cannot be fully

internalized if the two countries have different sizes, even if they are identical in all
other respects, including the extent of their concern for each other.9 Propositions 2 and
3 show that the moral hazard problem can be reduced if the countries in a coinsurance
arrangement are concerned about each other's welfare, but that it is difficult to
eliminate it. Hence, absent such concern for each other's welfare, we may need other
mechanisms to mitigate the moral hazard problem in a coinsurance scheme between
countries. The following section deals with these issues. It also considers why a
multilateral institution, such as the IMF, that acts as a delegated monitor, and can
provide loans to countries facing external imbalances, may be able to achieve this goal
more efficiently.
Moral Hazard in Coinsurance Arrangements
Peer Monitoring
One way to reduce the moral-hazard problem in coinsurance arrangements is peer
monitoring (see, for example, Arnott and Stiglitz 1991). If countries monitor each
other, it is less likely that a country will be able to take advantage of the insurance that
is collectively offered. Arnott and Stiglitz (1991) also propose that an indirect
monitoring system will encourage peer monitoring through the creation of
interdependence, i.e., the dependence of one country's utility on the policy effort of the
other. If monitoring by peers is costless, then the moral hazard problem disappears and
the first-best coinsurance, i.e. full coinsurance, can be achieved.
Although peer monitoring can lessen the costs of surveillance in a coinsurance
arrangement, it is hard to eliminate such costs entirely. As the group of countries
grows, peer monitoring tends to become more costly because: (i) it is hard for each
member to simultaneously monitor all the other members, even if the countries are
located close to each other and considerably interact among themselves; (ii) as the
group expands, it is more likely that a new member that is relatively remote and
unknown to the incumbents, reducing the efficiency of peer monitoring; (iii) sharing
information about each member in the group is costly and sometimes strategic
subcoalitions may emerge; (iv) the expansion of the group of countries tends to
exacerbate the problem of free riding.
There are three basic schemes in peer monitoring.10
1. Mutual Monitoring (MU): under this system each member in the group is
simultaneously monitored by all of her peers. This structure is commonly observed in
practice. Obviously, when the number of members in a group exceeds two, duplication
of monitoring effort is inevitable.
2. Rotating Pyramid (RP): under this scheme, in each time period, a different
member of the group is placed at the top of the pyramid and bears the responsibility for
monitoring her peers. This structure saves on the fixed monitoring cost and avoids
duplication of the variable monitoring costs.
3. Circular Monitoring (GI): each member monitors only one member and is in
turn monitored by only one member, so that the duplication of monitoring effort is
avoided.
*Rajan and Zingales (2000) also make a similar point when they show that the existence of inequalities in
opportunities or endowments reduces cooperation.
1(
fcor a discussion of peer monitoring structures, see Armendariz de Aghion (1999).
257
Figure 1 shows how each structure worksassuming a group of three members

Ml,M2andM3.
When the variable cost of monitoring is small and the fixed per-period cost of
monitoring is large, members may favor the Rotating Pyramid structure over Mutual
Monitoring. While the Circular and Rotating Pyramid structures can avoid some
duplication of monitoring effort, they are more vulnerable to pair-wise collusion; under
Mutual Monitoring, pair-wise collusion is likely to be detected by another member.
Arguments in favor of peer monitoring over central monitoring stem from a
reduction in information costs. However, peer monitoring has its drawbacks. First,
there is the issue of free-riding in monitoring the members. Peer monitoring levies a
cost on a member for acquiring and analyzing information on others in the group, and
this may tempt some members to decrease the amount or intensity of monitoring.
Second, there is the possibility of weak punishment or forbearance in peer monitoring.
Under peer monitoring arrangements, members have some incentive to punish a
member if that member is taking too much risk, but this punishment may not be
stringent enough compared to the optimal level. DeMarzo, Fishman and Hagerty (2001)
show that self-regulatory organizations choose more lax enforcement policies
compared to those preferred by the public: investigations for cheating are less frequent
and penalties are lower than what a customer would choose.
Since peer monitoring is imperfect and costly, and each of the structures
examined has some drawbacks, group members may prefer to create an institutional
monitor that can reduce both the fixed and variable costs of monitoring and also
prevent inefficiencies arising from the formation of subcoalitions and collusion.
Centralized Monitoring and Provision of Resources
As argued above, the group of countries in the coinsurance arrangement may find it in
their interest to form an institution, say the IMF, that functions as a delegated monitor.
The moral hazard problem implicit in a collective insurance arrangement can be
contained by a rigorous system of surveillance of all members by the IMF. The more
intense and accurate the monitoring, the better the functioning of the coinsurance
scheme.
The IMF, besides monitoring the countries, can also provide temporary liquidity
to countries who have suffered an adverse income shock. In this sense, the IMF may
have to function as a financial intermediary. The difference is that the IMF does not
take deposits but uses resources from a fund created by contributions from member
countries. Since the fund is meant to be a revolving one, the IMF provides resources
only under adequate safeguards.
Coinsurance is based on a mutual agreement to insure each other against shocks
or crises. Hence, it is natural to ask whether an insurance fund needs to be established
ex-ante with upfront member contributions or quotas. Creating such a fund ex ante can
have certain advantages:
1. Time inconsistency or enforcement issues in forming a fund ex post: Even
though countries agree to insure each other ex ante, those countries that are not
subjected to a shock may, ex post, delay or refuse to pay the contributions they
promised. More importantly, since this coinsurance arrangement is between sovereigns,
the international community has only limited means to make a country pay its promised
contribution. Under an ex-post coinsurance scheme, the only punishment would be to
exclude a non-paying member from the coinsurance group and deprive it of insurance
258
in the future. Hence, setting up a fund ex ante, with insurance only available to
participating members, would make clear the resources available.
2. Quick response to the liquidity needs of member countries: Contagion from
countries in crises to others is a central issue in international financial markets.
Containing such contagion is important, since otherwise a localized or regional shock
may spread and lead to systemic problems in the international financial system Thus,
to prevent contagion and maintain confidence in the system, the coinsurance scheme
should be able to take prompt actions in the aftermath of a country or regional crisis. To
this end, it makes sense to have sufficient liquid funds available to intervene promptly
if it is deemed necessary.
3. Reduction in transaction costs: If there is no fund available ex ante, then
every time a crisis occurs the collection of contributions from member countries could
entail huge transaction costs. On the other hand, if countries set up a fund ex ante, when
a country runs into trouble, liquidity can be provided from this fond directly, reducing
the costs of putting together emergency financing packages.
Operating the Coinsurance Arrangement
In the previous section, we showed that countries may find it in their interest to operate
a global coinsurance scheme, with the IMF functioning both as a delegated monitor and
as a provider of temporary liquidity to countries with external payment imbalances.
Clearly, the success of this coinsurance arrangement depends critically on ensuring that
recourse to IMF funds does not allow countries to slacken their efforts in making their
economies more resistant to shocks, or should crises occur, allow policy makers to
postpone measures necessary for a speedy recovery, and hence repayment to the IMF.
This section deals with the question of designing loan programs to provide the
appropriate incentives to member countries. To answer this question, first we must be
clear about the IMF's objectives. Having specified the objective function, we examine
different lending contracts for IMF loans. Should the IMF precommit to a loan contract
ex ante? Or should the loan amount and contract be assessed and formulated ex post,
that is, when the country approaches the IMF for resources?
A Model of IMF Lending
In our model, the IMF and the member countries take sequential decisions as shown in
Figure 2.
1. The IMF offers a state-contingent lending contract
consisting of five elements: I} is the first tranche of the IMF loan made available if a
country enters a crisis; I^J"
are
ê values of the second loan tranche conditional on
whether (post-crisis) output is low (state L) or high (state //); ZL,ZH are the statedependent repayments to the IMF at the end of period 2.
2. A country chooses policy actions e} to foster growth, macroeconomic
stability and prevent crises.
3. Nature roles the dice and a country enters a crisis with probability x(e}) that
is dependent on the country's policy actions el. The probability function n(e^\ is a
decreasing convex function of ,. The income level is y? when there is no crisis, but
falls to y\ if there is a crisis and the country does not seek IMF help. It is assumed that
259
the country's policy effort is private information and not observed by the IMF, which
relies on the outcome to make a judgment of the policy actions.
4. If the country faces a crisis and seeks IMF help, it is given the first loan
tranche / I B After receiving the first tranche, the country chooses policy actions e2 to
remedy the country's economic situation.
5. Nature plays again, and the country has either a low output y% with
probability n(e^] or a high output y% with probability
After the output is
observed, the IMF releases the second tranche

if the output is low and I" if the
output is high.
6. At the end, the country pays back the IMF ZL if low output was observed
H
and Z if the output was high.12
In the above setup, given the state-contingent contract offered by the IMF, the
country makes two sequential policy decisions to maximize its utility: it chooses e}
when it is, to use the common parlance, a surveillance country, and it chooses e2, if it
suffers an income shock and becomes a program country.
Note that in this problem, since the country has to make repayment at the end,
the contract can be simplified by focusing on the net repayments. Let ZL =ZL -1% and
ZH =ZH -I" be the net repayments by the country at the end of the second period.
Hence, the IMF's state-contingent contract can be defined in terms of three variables
To
simplify matters, the following assumptions are made without loss of
generality:
1. A zero interest rate is charged on the IMF loans.
2. y% > ZL\ y" > ZH . In the second period, the country is able to the make the
net repayment only if it has sufficient income.
3. The first tranche I} is bounded below by q9 and bounded above by a
multiple of q, say nq, or the resources available to the IMF, x. We can interpret q as
the country's quota (contribution to the coinsurance fund).
4. yf y%. This is a sufficient condition for a country that experiences a
crisis to seek IMF heln
5. y" > max
This condition implies that a country
prefers not to fall into a crisis and seek IMF resources.
The IMF's Objective Function
The IMF is endowed with a fixed amount of resources, x, that it can use to make
contingent loans to the members of the coinsurance arrangement.13 The design of the
Note that for notational simplicity we use the same probability function n () for both efforts. Of
course, we could specify the functions as differing over countries and efforts, but this merely adds to
notational complexity without providing any additional insight.
12
The IMF could choose a fixed repayment scheme rather than a state-contingent one. Such a scheme
would be a special case of the contract considered.
13
Note that the first tranche cannot exceed the amount of resources available to the IMF, so
260
loan contracts and the associated conditionally will depend critically on the objectives
of the IMF.
First, the IMF being the guardian of a revolving coinsurance fund is mandated
to lend resources only under adequate safeguardsthat the borrowing country will
make appropriate use of the funds and, as a consequence, be in a position to repay the
IMF over a stipulated time period. This objective implies that the IMF's utility function
depends positively on the size of its own resources. We represent this utility function
as :
(8)
and the expected utility of the IMF under this specification is given by:
(9)
Second, one could argue that the IMF should also show direct concern for the
welfare of the borrowing country. Such direct concern of the IMF with the utility of the
borrowing country is represented by
(10)
where u(y) is the country's utility function and 0 is the relative weight the
IMF attaches to the utility of a borrowing country compared to safeguarding its own
resources. The resulting expected utility function is
(11)
Policy Strategy for the Country

Assuming that a country's utility depends only on its income, the standard
von Neumann-Morgenstern expected utility functions for a surveillance country
(country S in period 1) and program country (country P in period 2) are:
(12)
(13)
x>/ 1 >max(z / / ,z L )>0.

14
The utility function of the IMF, u () , and of the country, u () , are assumed to be strictly concave
and satisfy the Inada conditions.
261
where the v function representing the cost or disutility of the policy effort satisfies
and Under our assumptions, the IMF is able to induce a strictly positive policy effort
e\ from a program country in period 2, where e\ satisfies the following first-order
condition:
(14)
Equation (14) equates the marginal utility and the marginal cost of the policy
effort to overcome the crisis. Since for
and
we have
and
This implies
that, as long as e*>0, under a state-contingent repayment scheme the country's

income need not be equalized across the high and the low-income states, i.e.
. From (14), using the implicit function theorem we derive the
effect of each of the contract variables
on the policy effort, e\.
Proposition 4 . T h e larger t h ef i r s ttranche, t h e smaller i

larger t h e n e t repayment when t h e income i s low, t h e higher i
i
s
larger the net repayment when the income is high, the smaller is
Proposition 4 implies that a relatively higher repayment when output is low,
provides an incentive for the country to exert higher effort to get out of a crisis. On the
other hand, if output is high, requiring a higher repayment from the country, acts like a
tax and discourages effort. Another way to interpret the proposition is that an IMF loan
contract that accommodates low output but penalizes high output is not likely to
provide the right incentives for the country's policymakers. The implication for the
design of IMF conditionality is that to induce higher effort to overcome the crisis, such
conditionally should "bite" when the country's economic performance is low and
should be weaker when the performance is better.
Note that
holds, and that
implies that, at the
margin, the incentive effect of a higher repayment exceeds the disincentive effect. This
is because the size of the incentive effect depends on the marginal utility at the lower
income,
, whereas the disincentive effect depends on the marginal utility at
the higher income,
Now, given the optimal strategy e* for a program country, we examine the
optimal policy effort e* for a surveillance country. Again, assuming the optimal effort
e[ is strictly positive, the first order condition equates the marginal utility with the cost
of the policy effort to prevent a crisis:
05)
262
Equation (15) implicitly defines the optimal strategy e*
in terms of
Using the envelope theorem, we derive the effect of each of

the contract variables
on the optimal policy effort ef.
Proposition 5 . T h e larger t h ef i r s ttranche, t h e smaller i
larger t h e n e t repayment when t h e income i s low, t h e higher i s ;

larger the net repayment when the income is high, the higher is <
Proposition 5 shows that higher net repayments at the end, elicit a greater policy
effort e* from member countries to prevent crises. Note that in contrast to the results
for e*2, higher values of both ZL and ZH tend to induce a larger policy response e*
from surveillance countries.
Two corollaries follow directly from Propositions 4 and 5.
Corollary 1. The larger the promised first tranche, the higher is the probability
of a crisis situation developing, i.e
Corollary 2. The larger the first tranche, the higher is the probability of staying
in a crisis, i.e.
These results suggest that, ceteris paribus, the IMF should minimize the size of
the first tranche to provide member countries with the appropriate incentives, firstly,
for preventing crises, and secondly, should they enter one, for expending the right
amount of effort to get out of the crisis.
The IMF's Choice of the Lending Contract
Now, knowing a member country's optimal policy strategies e* and e\, we examine
the problem of designing the IMF's optimal contract. We classify a repayment scheme
as compensatory if
, that is the net repayment bv the country is lower in the
low income state than in the high income state. If
we will call the scheme
non-compensatory.
The IMF's problem of choosing the optimal loan contract can be specified as
follows:
263
e
the
max
/**.**
s.t.
> expected utility for S, ifS stays outside the coinsurance arrangement
> expected utility for P, if P does not use the IMF contract s u(y^).
The first two constraints are incentive compatibility constraints for a
surveillance country and a program country. The last two constraints are participation
constraints for a surveillance country and a program country. In order to set the
participation constraints, we need to specify the expected utility of a country that stays
outside the coinsurance arrangement and the expected utility of a member country in
crisis that does not use the IMF contract. Note that once a country is hit by a crisis, it
can either choose to be a program country and exert policy effort e\ or remain in crisis
and get $. From assumption 4 above, a surveillance country will choose to be a
program country, and the participation constraint for a program country is satisfied.
Before a country is hit by a crisis, the country can either accept the contract and
become a surveillance country or reject the contract and stay outside the IMF. Again,
from assumption 4 above, a country will prefer to be a member of the IMF and join the
coinsurance arrangement. Note that e** represents the policy effort of a country that
decides not to become an IMF member. From assumption 4 above, we have
a country will participate in the loan contract the IMF proposes.

The IMF's contract design problem can be solved by backward induction. First,
we solve the expected utility maximization problem for a program country given
The solution yields the strategy
. Next, the expected
utility maximization problem for a surveillance country is solved, given that it will
adopt the strategy e*, if it suffers a shock and becomes a program country. Finally, we
find the optimal contract
that the IMF should offer the countries in the
coinsurance arrangement. Below, we will examine the IMF's problem of choosing the
optimal contract, under each specification of its mandate.
264
IMF Objective: Safeguarding Resources

In this section, we assume that the IMF's objective function is given by (9). We denote
the optimal contract for this objective of the IMF as A* = {/^jZ^****}- The following
proposition shows that if the IMF's sole concern is with safeguarding its resources,
then it chooses a scheme requiring the country to pay it back in full.
Proposition 6. If the IMF's only objective is to safeguard its resources, then the
optimal contract A* satisfies: q<I*A = ZLA = ZHA <mm(y^,nq).
In this case, the size of the loan is bounded above either by the country's output
in the worst case scenario15 or by the maximum possible amount of the initial tranche.
As long as I*A = ZLA = z"*, the IMF is not affected by the policy efforts of the country
since in both, the high and low income states, the optimal level of the IMF's utility is
u(x). Also, note that the problem has multiple solutions, and I*A can take any value
between q and mn(y%,nq).
It is interesting to see the effect of the optimal contract on the policy effort
levels of a surveillance country and a program country. First, from the special structure
of the utility function of the IMF where the effort of a country only affects the
probability of crisis, if a country will repay in full with certainty, the IMF does not care
about policy effort. Then, the next question is whether the surveillance country or the
program country will exert a positive level of effort or zero effort. Full repayments in
both states imply that the contract variables (I},zL,zH) do not appear in the expected
utility of both the surveillance country and the program country. Thus, we can show
that even under full repayment, the surveillance country and the program country
generally make a positive level of effort.
IMF Objective: Balancing Country Welfare and Safeguarding of Resources
For the objective function given in (11), the first order conditions for the IMF's utility
maximization with respect to the contract variables
are:
(16)
15
Note that for simplicity we are assuming that the country's entire output is pledgeable to pay back the
IMF's debt. The model can be easily modified to distinguish between tradable and nontradable output,
with only the tradable output being used to pay off the external debt.
265
(17)
(18)
Equations (16), (17) and (18) are used to solve for the optimal contract
B* = {/IB,ZB*JZ**} In this general model, depending on the values of the parameters,
the optimal repayment can be either compensatory or noncompensatory. The following
proposition shows that, depending on x, the amount of resources available to the IMF,
and 0, the extent of concern the IMF has for the country's utility, the optimal
repayment scheme can vary from zero repayment in both the high and low income
states in the second period to full repayment in both states.
Proposition 7. If the IMF's objective function is given by (11), then the
following holds: (i) If x is small, say,
, and 6 0, then the IMF's optimal
contract
s a t i s f i e s . (ii) If x is large, say,
and
, then the IMF's optimal contract
satisfies
and /*B = min (nq, x) -nq.

With explicit concern for the country's utility, the IMF faces a trade-off
between safeguarding its resources and enhancing a country's utility. If 6 is close to
0, then, the problem reduces to that solved in the previous section; the IMF focuses on
safeguarding resources and the optimal repayment scheme requires full repayment. On
the other hand, if 6 is large, the IMF puts more emphasis on a country's utility. Thus,
the IMF lends as much as possible to the country, and even no repayment by the
country may be optimal.
Now, the amount of resources available to the IMF also matters. The intuition is
straightforward. Suppose that the IMF's resources are small compared to the size of the
country. Then, given the concavity of the utility function, the IMF's utility increases by
a large amount when it receives repayment from the country, while the country's utility
decreases by a relatively small amount. Thus, the IMF's utility increases overall by
demanding higher repayment. Moreover, the higher repayment increases the policy
effort, which will generally increase the utility of the IMF. Therefore, it is optimal for
the IMF to get as large a repayment as possible.
Suppose, on the other hand, the resources of the IMF are "large" compared to
the income of a recipient country. Then, the IMF's utility decreases by a small amount
when it gives up receiving payment from the program country. On the other hand, the
country's utility increases substantially due to the higher second-period income as a
result of the IMF's net transfer. Since the IMF's utility is in turn affected by that of the
borrowing country, depending on the value of 6, some degree of debt forgiveness may
266
16
'
be optimal. Note that for the IMF's utility to increase overall, the utility increase from
caring about the country has to be larger than the decrease in the IMF's utility
stemming from the lower policy effort due to debt forgiveness.
An interesting question is whether the IMF's mandate has any implications for
the policy effort of countries. Is it the case that an IMF which only cares about
safeguarding its resources induces higher policy effort than an IMF which also cares
about country welfare? The following proposition shows that it is not necessarily so.
Proposition 8. Assume that the optimal policy actions, e* and e\, are interior
solutions. Then, if
the optimal policy actions satisfy e? > (<)e*B. Further, if
we have
The first part of the proposition states that, if the difference of the utility levels
corresponding to the good and the bad outcomes gets smaller when the IMF starts to
consider country welfare, then the program country exerts less effort to overcome the
crisis. The second part states that, if the expected utility of the program country
becomes larger when the IMF objective changes from only safeguarding resources to
considering both safeguarding of resources and country welfare, then the surveillance
country exerts less effort to avoid the crisis.
Consider the following examples based on three different sets of optimal loan
contracts.
Example 1.
This is a case where the repayment of the IMF is non-compensatory. Now
Thus,
The intuition is that a
subsidy in the good state increases the level of effort to avoid the bad outcome. In this
case, depending on the relative size of
and
we can get either
or
Example 2.
This is a case when the repayment scheme of the IMF is compensatory. Now
Thus, eÂ >e*B. Here, the forgiveness in the
bad state lowers the level of effort to avoid the bad outcome. Again in this case,
16
It is important to point out that in our framework, contracts have only a single dimensionthe size of
the loanand hence the only incentivizing device is "debt forgiveness." Introducing richer contracts
with more dimensions, such as conditionally, would give the IMF more instruments to achieve its
objectives without having to resort to "debt forgiveness."
267
depending on the relative size of EU*PA and EUp , we can get either e*A >e$B or
ee*} A <
<ee*} B .
Example 3.
and
This is the case where the resources of the IMF are very large, as in Proposition
7
(ii).
Due
to
the
strict
concavity
of
the
utility
function,
Thus,
This result shows that when
a country receives a large subsidy or debt forgiveness irrespective of whether it
emerges from a crisis or not, it exerts less effort to sort through its problems. Also, as
long as nq>y" -y\, we have
(19)
and thus e*A > e^B holds.

Thus, Proposition 8 shows that mandating the IMF to care about country
welfare in addition to safeguarding its resources, does not necessarily imply that
member countries will spend less effort in preventing imbalances from arising, or if
problems should arise, spend less effort for their resolution.
Timing of Loan Contracts
Throughout this subsection, we will assume that the IMF cares both about safeguarding
its resources and the borrowing country's utility. As pointed out in Khan and Sharma
(2003), the IMF, given such a mandate, faces the Samaritan's dilemma. This dilemma
arises whenever the availability or granting of assistance leads to making it more likely
that the conditions that evoke such aid will hold. Faced with underperformance and a
weak economy, countries know that the IMF will provide assistance because it is
concerned with the borrowing country's welfare. The knowledge that the IMF will
come to their assistance if they run into trouble, may make countries to be lax in
correcting policy imbalances and exert less effort for crisis prevention. Under these
conditions, should the IMF precommit to a contingent contract before a country
experiences a crisis, or should it formulate a loan contract after a country is in crisis?
We consider three timings for the IMF loan contract:
Case 1. Ex-ante contract.
The IMF precommits to a contingent loan contract
before the
country chooses policy effort e}. This case was analyzed in the previous section.
Case 2. Ex-post contract.
The IMF chooses contract
after e} has been chosen (and the
country enters a crisis) but before the country chooses e2.
Case 3. A variation of Ex-post contract.
The IMF chooses contract
where the IMF chooses IID after
e} has been chosen an
after the country chooses e2.

268
For Case 2, the IMF's expected utility function can be written as:
(20)
Note that (20) differs from the expected utility function in the previous section
where the IMF could precommit to a contract before a crisis occurred. The optimal
contract C* is obtained from the following first-order conditions:
0=
(21)
0=
(22)
0=
(23)
The IMF offers contract

to a country after it has entered a
crisis. Also, since the IMF's objective is to balance the safeguarding of its resources
while being concerned about country welfare, we expect the optimal contract to be
compensatory. The following proposition formalizes this intuition.
Proposition 9. If the IMF cares about both safeguarding its resources and
country welfare, under an ex-post contract, the repayment scheme is compensatory,
with the program country repaying more if the country recovers from the crisis and less
if it does not recover.
Now let's consider Case 3. Note that in this case there is no remaining
uncertainty, and an IMF that cares about country welfare finds it optimal to provide full
insurance to a country. Moreover, the repayment scheme is compensatory for strictly
positive values of 0, i.e.,
Next we compare, how the three contracts B, C, and D affect a country's
policy effort and the extent to which the IMF provides debt forgiveness. To this end,
define
as the net transfer (or the amount of debt forgiveness) to the
program country if it recovers from the crisis and
as the net transfer if the
country does not emerge from the crisis.
The following lemma shows that, given a lower bound on the probability of
remaining in a crisis (once it occurs), the effect of TL on the policy effort to avoid a
crisis is greater than that of TH.
Lemma 3. If
269
The above lemma is used in proving the following proposition on crisis

prevention and recovery effort and net transfers under the three contracts.
Proposition 10. Assume that:
(1) The expected utility function of the IMF, EUMF, is twice continuously
differentiable with respect to TH and TL.
(2)
are interior solutions.
(3)
(4)
(5)
'
"
'
"
"
Then, under assumptions (l)-(4), the three optimal contracts fl*, C", and D*
satisfy:
"(a]
a>;
Further, under assumptions (l)-(5), the optimal policy efforts elicited by the
contracts satisfy
(c)
Proposition 10 shows that when the policy efforts of the country to prevent and
overcome a crisis are not fully observed by the IMF, under some conditions a
precommitment to a lending contract, B* 9 elicits a greater amount of crisis prevention
effort and crisis-overcoming effort from countries, than the offering of the contract C*
after the crisis has occurred. It also shows that some commitment is better than none at
all, since contract C* induces higher efforts than contract D*.
In terms of debt forgiveness, precommitment leads to the least amount of
forgiveness if the program country cannot get out of the crisis. On the other hand, if the
program country is successful in getting out of the crisis, contract C* provides the
largest amount of debt forgiveness to the program country. The intuition for this is that
under contract D* the IMF decides on the amount of forgiveness after it observes
whether the program country is successful in getting out of the crisis or not. Once the
IMF knows that the country is out of the crisis, as it does under contract D*, it is
optimal for the IMF to reduce the amount of debt forgiveness compared to the situation
under contract C*, where the terms are decided before such information is available to
the IMF.
From Propositions 9 and 10, Corollary 3 follows immediately.
Corollary 3.
It implies that, compared to contract D*, under contract C* debt forgiveness by
the IMF is lower when the program country remains in crisis, and higher when it is able
to get out of a crisis.
270
Preconunitment and Time-consistency

The idea of precommitment entails an ex-ante agreement on the contingent loan
contract between the IMF and a member country. But commitment to an ex ante
contract (that is a contract agreed to before problems arise) suffers from the timeinconsistency problem (Kydland and Prescott 1977). If the country knows that ex post
(that is, after a country's situation deteriorates) the IMF will be willing to renegotiate
the contract, ex ante country ownership of the contract is less likely.17
The problem of time inconsistency arises even in an environment of complete
and perfect information. For an altruistic lender, there is the additional dimension that
the borrower knows that it will be optimal for the lender to renege on penalties agreed
to ex antethe economics of fait accompli (Bernheim and Stark 1988), Lindbeck and
Weibull 1988). The presence of informational asymmetries exacerbates the problem
even furtherthe lender cannot easily verify the policy effort and hence cannot
contract on such effort; it also changes the prescription for dealing with the dilemmas
posed by altruism and sequential decision making. This paper shows that, for an
altruistic lender facing information asymmetries, it may be best to precommit to an ex
ante loan contract rather than define the contract ex post.18
Suppose, for the sake of simplicity, that the menu of contracts consists of two
contractsthe ex ante contract B* and the ex post contract C* defined in the previous
section. Then, Proposition 10 shows that the IMF will choose contract B* before a
crisis occurs. Now consider what happens when a country runs into trouble. Once e* is
chosen and the country falls into a crisis, the optimal contract for the IMF is C*. Since
B* and C* need not be the same, the IMF has an incentive to change its contract once
the country descends into a crisis. The intuition is as follows: When the IMF offers
contract B* to a surveillance country, its objective is to provide the country with the
incentive to both, prevent a crisis, and should a crisis occur, to expend effort to change
course. Thus, the contract is designed to achieve these goals, while giving consideration
to the country's welfare. However, once a crisis erupts, the IMF's focus is to encourage
the country to exert adequate effort to get out of the crisis. Under the new
circumstances, a different contract, C*, is optimal. Hence, the IMF is willing to change
the contract over time.19
17
See, Drazen and Fischer (1997) and Khan and Sharma (2003).
Aside from altruism, the IMF faces additional hurdles in its operation compared to a traditional lenderof-last-resort (Tirole 2002 and Khan and Sharma 2003). First, lending takes place on the promise that
country authorities will implement policies to rectify imbalances, and it is difficult, if not impossible, to
establish the value of such "collateral." Second, the information asymmetries and hence the moral hazard
is likely to be more pronounced. The IMF faces what in agency theory is called "moral hazard in
teams"while program negotiations are conducted with certain representatives of the government
(central bank, finance ministry), the success of the program depends on the acceptance and effort of
many other stakeholders in society (other ministries, political parties, trade unions, professional
associations, civic groups, NGOs) (Holmstrom 1982). Third, the enforcement mechanism for ensuring
that borrowing countries live up to their obligations essentially amounts to some combination of moral
suasion, maintenance of the borrower's reputation, peer pressure, and the threat of being shut out of
international capital markets.
19
To go further, once the country chooses -e^, then depending on whether the country gets out of the
18
crisis or not, the IMF may find that yet another contract D*, different from C*, is optimal.
271
Now consider the problem the country faces over time. A country in crisis may
prefer the contract C* to going through with B* which may have been preferred before
the crisis was encountered. If the program country obtains higher expected utility from
C", renegotiation by the IMF and the country may make both of them better off.
Again, time inconsistency stems from the fact that the IMF may prefer to change its
contract as the situation changes.
This begs the question whether such contract renegotiation should be permitted
and whether it is in the global interest.20 The community of nations may well prefer that
contract 5* be offered and enforced, since it leads to higher country effort to prevent
problems from emerging, higher country effort to escape crises, and a smaller transfer
of resources by the IMF.
The model presented in the paper has analyzed a single interaction between the
IMF and a country. In a setting where there is repeated lending, it may be in the IMF's
interest to precommit to a contract such as U*3 and build a reputation for enforcing the
contract. Over time, this could lead to the emergence of an international norm under
which the IMF offers and enforces a "standard" contract. Renegotiation would be
allowed only under exceptional circumstances, for example, when it is perceived that
the country in crisis poses a systemic threat to the world economy.
Concluding Remarks
This paper argues that in the presence of information asymmetries and given the
mandate of the IMF to safeguard its resources and care about the welfare of borrowing
countries, the IMF should precommit to a lending contract.21 Such a precommitment
elicits the right policy effort from countries to prevent crises and to recover from them.
However, a contract agreed to before a crisis erupts is subject to time-inconsistency
problems, since the Fund and the country may both find it in their interest to
renegotiate the ex ante contract, once the country enters into a crisis. Hence, a country,
knowing that the IMF will renegotiate the contract if it experiences a crisis, is less
likely to own the program ex ante and give credence to statements that additional funds
or concessions will not be made.
Our results can be taken as a defense of existing IMF procedures that define
annual and overall access limits to resources for program countries. Limits on IMF
lending and rewards for good housekeeping were also favored by the independent task
force sponsored by the Council on Foreign Relations (1999). The Meltzer Commission
also suggested that the IMF should provide resources up to specific limits, but only to
prequalified countries and at penalty interest rates.22 A key objection to the
20
In the context of our model, the parameter 6 can be given a different interpretation to capture the "toobig-to-fail" issue. One could think of 6 as the importance the IMF attaches to a country, with the size of
6 depending on the consequences a crisis in that country would have for the international financial
system. Large systemically important countries would be assigned larger 6 s and for these countries the
time inconsistency problem would be more severe. Knowing this, such countries are more likely to be
successful in renegotiating their contracts and obtaining weaker programs.
21
Note that precommitment to a lending contract for all members is quite different from prequalifying
members for access as was done for the IMF's Contingent Credit Line (CCL) facility. The CCL facility
had to contend with the concern that signing up for it may be taken as a signal of weakness, and that
ineligibility at a future date may have a negative fallout.
22
See Meltzer et al. (2000). Prequalification was to be based on four factors: (i) free entry of foreign
financial institutions; (ii) regular and timely publication of the maturity structure of sovereign and
government guaranteed debt; (iii) adequate capitalization of commercial banks; and (iv) a fiscal
272
prequalification requirement was that it would exclude a large number of member

countries, and hence would be fundamentally inconsistent with the rights of all
members to access IMF resources under the Articles of Agreement.23
In many recent programs, normal IMF access limits (100 percent of a country's
quota annually and 300 percent of quota cumulatively) have been breached by wide
margins. This paper does not address the difficult question of how large the IMF should
be for effectively performing its role as a coinsurance arrangement and crisis manager.
In an era of capital mobility, country quota levels and access limits may need to be
recalibrated. Among other things, the size of the IMF will depend on the size
distribution and health of member countries; the extent to which countries have access
to private capital markets; the volatility of international finance (or more generally the
real hazards that have to be dealt with); the effectiveness of IMF surveillance; and the
catalyzing role of IMF lending.24
The design of an ex ante loan contract will also involve specifying the interest
rate charged, and the maturity of the loan. If the IMF is limited in its ability to charge
different credit spreads across countries, to safeguard its resources the IMF could still
attach different policy conditionality to the loan contract depending on a country's
characteristics and the imbalances the country is facing. Countries could be allowed to
choose loans from a predefined set of maturities, and the rate of charge could increase
with maturity to create an incentive for countries to deal with the situation quickly and
repay the IMF.
An IMF commitment to a loan contract that ex ante stipulates access limits, size
of tranches, interest rates, and maturities may have other advantages. First, it would
make it easier for countries to decide how much self-insurance they should buy.25
Second, it would make clear to private creditors the extent of IMF resources a country
could tap if it ran into liquidity problems, and hence contribute to limiting creditor
moral hazard. As a country accumulates debt in international and domestic markets,
private lenders, knowing the limits of IMF support, may be quicker to react to signs of
emerging imbalances than they would if the extent of IMF support was not specified
(Haldane and Kruger 2001). Third, an international coinsurance arrangement among
countries is essentially a self-regulatory club. It can be argued that when members do
not live up to their obligations, like other self-regulated organizations, the IMF may not
impose the discipline the international tax payers would deem appropriate. Hence, an
international norm that restricts the access of countries to IMF resources through
prespecified limits and terms, except when there is a systemic threat, may be a useful
commitment mechanism
requirement.
23
For more on the debate see Eichengreen (1999), Goldstein (2000), and U.S. Treasury (2000).
24
For a discussion on the nature and size of the IMF, see Jeanne and Wyplosz (2001). For the catalyzing
role of IMF lending see the recent survey by Hovaguimian (2003) and the references therein.
25
For a discussion of foreign exchange reserve accumulation and self-insurance, see Lee (2004).
273
Appendix: Proofs
Proof for Lemma 1.
Since
diminishing
marginal
Therefore,
holding for
utility
gives
with equality
Q.E.D.
Proof for Lemma 2.
Let
and u. = u.
From equation
(6),
where
by thej convexity of n and v.
Thus,
Q.E.D.
Proof for Proposition 1.
Let y,, = y, St=S, and u,r = u.
jf
(i)
Evaluating (i) at ft = 0, the second term in (i) disappears, and we get
Evaluating (i) at ft = \, the first term in (i) disappears, and we get
Q.E.D.
274
The first-order condition with respect to ei is
(ii)
As we did in Lemma 2, by symmetry, we can use
and
From equation (ii),
where
for all A > 0. Thus,
'(Hi)
We know from the first order condition with respect to/? that, in the symmetric
equilibrium, J3* = $2 should hold to make the marginal utilities equal. Plugging this
into (iii), as A - 1, we get - -> 0. Alternatively, as A -> 1, and the countries tend to
attach the same relative weight to each other's utility. Therefore, their marginal utilities
in
each
state
will
tend
to
equality
and
the
term
will converge to
zero.
Q E.D.

Without loss of generality, we assume that yi > y^. Then, proceeding in the same way
as in the proof of Proposition 2, we get the following:
Note that no we* * e*. due to the asymmetry y. > y^.
275
As 2 -> 1, unless
holds,
Q.E.D.
From (14),
Then,
Q.E.D.
From (15), using the envelope theorem, we get
Hence,
276
where
Q.E.D.
Proof for Corollary 1.
Q.E.D.
Proof for Corollary 2.
Q.E.D.
From the expected utility function of the IMF under contract A and the assumption
I} >max(zH9zL), we can see that the IMF's utility is maximized when I*A = ZL* = z**,
and the corresponding level of the IMF's utility is u(x) for any value of e, and e2.
Then, from the assumption that
we get
cmin(wtf,^).
Q.E.D.
Note that the sum of the right-hand side of (16), (17) and (18) is zero. Thus, if the righthand side of (17) and (18) is negative, then the right-hand side of (16) must be positive.
Suppose x oo and 6 0. Then, the second and the third terms on the righthand side of (17) and (18) converge to zero while the first term of (17) and (18) is
negative and not close to 0. Thus, the right-hand side of (17) and (18) becomes
negative, which means z"* = z* = 0 and I*B = nq.
Next, consider the case where x is small (i.e. close to 0). We know that now
the first and second terms on the right-hand side of (17) and (18) are positive and the
first terms in (17) and (18) are close to oo. Thus, the right-hand side of (17) and (18)
becomes positive, which means
as long as
Q.E.D.
277

Let
and note that
The results in Proposition 8
follow directly from (14), (15) and the fact that ~ is a strictly increasing function of
effort.
Q.E.D.
The proof is by contradiction. Suppose that z * < ZL* . Then, the following holds:
Now, from (22), we get
Also, from (23), we get
But, from the first order condition with respect to e2, u'(y2 +7, -z) >uf(y" +7, -z").
Thus, it follows that
which implies
This is a contradiction.
Q.E.D
Proof for Proposition 10 and Lemma 3.

Throughout the proof, we denote the general expected utility of the IMF by EUIMF, and
make the following assumptions:
Assumption 1.
H
(i.e. twice continuously differentiate) with respect
to T , T and K, where K is defined in (PART 1) below.

This assumption is necessary for well-behaved second-order conditions.
Assumption 2.
ire interior solutions.
Assumption 3.
278
This assumption assures that the second-order condition for a maximum holds.
It also guarantees that the expected utility of the IMF is more sensitive to its own
wealth compared to that of the country.
The proof has two parts: the first part shows that
the second part shows that
and
and T"* > Tg* ,
(PARTI)
First, we show that
When we compare the IMF's expected utility functions under contracts B and
C, we find that the first order conditions, after normalization, differ only in that each
of the first order conditions under contract B has an additional term
where / = / 19 z L ,z // . Thus, we can represent the solution of the problem under both the
contracts in one system of equations using the following term
where i = I},zL,zH . For ^ = 0, we get the appropriate conditions for contract C, and
for K = 1, we get them for contract B.
To show that e^B > e*c, first we derive the change of ZH , ZL , and I} when we
move from contract B to contract C, then we calculate the change of e* due to the
changes in ZH , ZL , and I} between the two contracts, and finally we combine these
two effects to get the total effect.
In particular, define
Then, we need to solve the following system of equations to get
where
and so on.
279
Thus, we need to show that
holds.
Note that F + G + // = 0, \/I},zL,zh ,K . Thus, from Assumption 2, one of the
terms F9 G or H is redundant. Therefore, we can redefine the problem as a 2-equation2-variable problem using the following change of variables: TH =I}-zH and
TL = 7, -Z L , where TH,TL>0 from the assumption that /, - max(z H ,Z L ) > 0. Now we
can rewrite the IMF's expected utility functions under contracts B and C using
TH,TL, calculate the derivatives e*TH9e* L9e*TH,e*TL, and redefineFand G as follows:
Note that
Thus, from Assumption
1, we have
and
To show that
we need to show that
holds, where the first term on the left side of the inequality represents the increase in e\
through
when we move from contract C to B, and the second term represents the
decrease in
through
when we move from contract C to B. To this end, it is
sufficient to show that
To derive
get the following:
where ~
and
we differentiate F and G with respect to
and so on.
280
and
Since F}, F2, F3, G13 G2, G3 are continuous, by Young's theorem, F2=G}.
We can also show that F3 < 0 and G3 < 0.
From Assumption 3, we get
Then,
. Denote
and
It foliows that T* > Tg* and T? > T^.

To prove Lemma 3, first note that
Since we already know, from the first-order condition of P with respect to e2,
that
once we have
holds.
Therefore, Lemma 3 is true.

By Lemma
3,
Using Assumption
and hence
4,
/<J=G 2 ,
have
, Then, together with
we have from the first-order condition of P, we get
Second, given
we
and
, Therefore,
we have to show that
Now denote the variables in the case of contract C by ~ . For example,

and
Solving for the surveillance country's choice of e} under contract C, given

we get
Rearranging terms,
281
Proceeding similarly for contract B, we show
Note v and v are positive unless e*'s are corner solutions; and from the assumption
that
max
it follows t h a t a n d a r e indeed
positive.
Let
and note that
Thus,/is strictly
increasing in e}, which implies that, if -
, then
is
equivalent
But Assumption 5,
to
Hence,
(PART II)
First, consider the choice of "T"* and 7^* by the IMF under contract C.
The first-order conditions with respect to T" and T are:
We know from Proposition 9 that T<f * < 7^*. From the above first-order conditions, we
get the following inequalities:
(iv)
(v)
Next, consider the choice of T and T by the IMF under contract D. The first-order
conditions with respect to T and T^ are:
(vi)
282
(vii)
Comparing (iv) and (vi), we see that T* < T . Also, a comparison of (v) and (vii)
reveals that T** > T* . Therefore, we have
(viii)
Now, we need to show
Given (7^*,7^*) already chosen by the IMF, the
program country, P, chooses
that satisfies
Rearranging terms, we have
Similarly, given (7^*,7*), the program country chooses
From (viii), we have
that satisfies
and
Thus,
which is equivalent to
Let
, and note that
Thus, g is strictly increasing in e2,
which implies that

Finally, we need to show
given
and
A
Denote the variables under contract D by . Remember the variables under contract C
are denoted by ~.
Given
D case, we get
, solving for the surveillance country's choice of e} under contract
283
Rearranging terms, gives
Similarly, for contract C, we get
Now let
Thus,
and note thai
is strictly increasing in e}, which implies that, if
But Assumption
5,
then
is
equivalent
to
Hence,
Q.E.D.
284
References
Armendariz de Aghion, Beatriz, 1999, "On the Design of a Credit Agreement with Peer
Monitoring," Journal of Development Economics, Vol. 60, pp. 79-104.
Arnott, Richard, and Joseph E. Stiglitz, 1991, "Moral Hazard and Nonmarket
Institutions: Dysfunctional Crowding Out or Peer Monitoring?" American
Economic Review, Vol. 81, pp. 179-90.
Becker, Gary S., 1974, "A Theory of Social Interaction," Jo urnal of Political Economy,
Vol. 82, pp. 1063-93.
Bergstrom, Theodore C.3 1989, "A Fresh Look at the Rotten Kid Theoremand Other
Household Mysteries," Journal of Political Economy, Vol. 97, no. 5, pp. 11381159.
Bernheim, Douglas, Andrei Shleifer and Lawrence E. Summers, 1985, "The Strategic
Bequest Motive" Journal of Political Economy, Vol. 93, no. 6, pp. 1045-1076.
Bemheim, Douglas, and Oded Stark, 1988, "Altruism within the Family Reconsidered:
Do Nice Guys Finish Last?" American Economic Review, Vol. 78, (December),
pp. 1034-1045.
Bordo, Michael D., and Anna J. Schwartz, 1999, "Under What Circumstances, Past and
Present, Have International Rescues of Countries in Financial Distress Been
Successful?" Journal of International Money and Finance, Vol. 18, pp. 683708.
Boughton, James M., 2004, "The IMF and the Force of History: Ten Events and Ten
Ideas that have Shaped the Institution," Working Paper 04/75 (Washington:
International Monetary Fund).
Bruce, Neil, and Michael Waldman, 1990, "The Rotten Kid Theorem Meets the
Samaritan's Dilemma," Quarterly Journal of Economics, Vol. 105, pp. 155165.
Buchanan, James M., 1975, 'The Samaritan's Dilemma," in Altruism, Morality and
Economic Theory, ed. by Edmund S. Phelps (New York: Russel Sage
Foundation), pp. 71-85.
Calomiris, Charles W., 1999, "How to Invent a New IMF," (Washington: American
Enterprise Institute).
Capie, Forrest, 1998, "Can There be an International Lender-of-Last-Resort?"
International Finance, Vol. 1, No. 2, pp. 311-325.
Chami, Ralph, 1996, "King Lear's Dilemma: Precommitment versus the Last Word,"
Economics Letters, Vol. 52, pp. 171-76.
, 1998, "Private Income Transfers and Market Incentives," Economica, Vol. 65,
pp. 557-80.
and Jeffrey H. Fischer, 1996, "Altruism, Matching and Nonmarket Insurance,"
Economic Inquiry, Vol. 34, pp. 630-47.
Cordelia, Tito, and Eduardo Levy-Yeyati, 2004, "Country Insurance," Working Paper
04/148, (Washington: International Monetary Fund).
Cox, Donald, 1987, "Motives for Private Income Transfers," Journal of Political
Economy, Vol. 95, pp. 508-46.
Council on Foreign Relations, 1999, Safeguarding Prosperity in a Global Financial
System: The Future International Financial Architecture, Report of an
Independent Task Force, (New York: Council on Foreign Relations).
DeMarzo, Peter, Michael Fishman and Kathleen Hagerty, 2001, "The Enforcement
Policy of a Self-Regulatory Organization," Working paper, Stanford University.
285
Drazen, Allan, and Stanley Fischer, 1997, "Conditionality and Selectivity in Lending
by International Financial Institutions," mimeo, (Washington: International
Monetary Fund).
Eichengreen, Barry, 1999, Towards a New International Financial Architecture
(Washington: Institute for International Economics).
Fischer, Stanley, 1999, "On the Need for an International Lender of Last Resort,"
Journal of Economic Perspective, Vol. 13, No. 4, pp. 85-104.
Goldstein, Morris, 2000, "Strengthening the International Financial Architecture:
Where Do We Stand?" (Washington: Institute for International Economics).
Ghosh, Atish, Timothy Lane, Marianne Schulze-Ghattas, Ales Bulir, Javier Hamann,
and Alex Mourmouras, 2002, IMF-Supported Programs in Capital Account
Crises, Occasional Paper 210 (Washington: International Monetary Fund).
Haldane, Andrew, and Mark Kruger, 2001, "The Resolution of International Financial
Crises," Financial Stability Review, Bank of England, (December), pp. 193202.
Haldane, Andrew, and Ashley Taylor, 2003, 6eMoral Hazard: How Does IMF Lending
Affect Debtor and Creditor Incentives?" Financial Stability Review, Bank of
England, (June) pp. 122-133.
Hirshleifer, Jack, 1977, "Shakespeare vs. Becker on Altruism: The Importance of
Having the Last Word," Journal of Economic Literature, Vol. 15, pp. 500-502.
Holmstrom, Bengt, 1982, "Moral Hazard in Teams," Bell Journal of Economics, Vol.
13, pp. 324-40.
Hovaguimian, Catherine, 2003, "The Catalytic Effect of IMF Lending," Financial
Stability Review, Bank of England, (December) pp. 160-169.
Independent Evaluation Office,- 2003, The IMF and Recent Capital Account Crises:
Indonesia, Korea and Brazil (Washington: International Monetary Fund).
Jeanne, Olivier, and Charles Wyplosz, 2001, "The International Lender of Last Resort:
How Large is Large Enough?" Working Paper 01/76, (Washington:
International Monetary Fund).
Jeanne, Olivier, and Jeromin Zettlemeyer, 2001, "International Bailouts, Moral Hazard
and Conditionality," Economic Policy, (October) pp. 409-432.
Jurges, Hendrik, 2000, "Of Rotten Kids and Rawlsian Parents: The Optimal Timing of
Intergenerational Transfers," Journal of Population Economics, Vol. 13, pp.
147-157.
Khan, Mohsin S., and Sunil Sharma, 2003, "IMF Conditionality and Country
Ownership of Programs," World Bank Research Observer, Vol.18, No. 2 (Fall),
pp. 227-248.
Krueger, Anne O., 1998, "Whither the World Bank and the IMF," Journal of Economic
Literature, Vol. 36, No. 4, pp. 1983-2020.
Kydland, Finn E., and Edward C. Prescott, 1977, "Rules Rather than Discretion: The
Inconsistency of Optimal Plans," Journal of Political Economy, Vol. 85, pp.
473-92.
Laffont, Jean-Jacques, and David Martimort, 2002, The Theory of Incentives: The
Principal-Agent Model (Princeton: Princeton University Press).
Lee, Jaewoo, 2004, "Insurance Value of International Reserves: An Option Pricing
Approach," mimeo, (Washington: International Monetary Fund).
Lindbeck, Assar, and Jorgen W. Weibull, 1988, "Altruism and Time Consistency: The
Economics of Fait Accompli," Journal of Political Economy, Vol. 96, No. 6,
pp. 1165-1182.
286
Masson, Paul R., and Michael Mussa, 1995, 'The Role of the IMF: Financing and Its
Interactions with Adjustment and Surveillance," IMF Pamphlet Series, no. 50,
(Washington: International Monetary Fund).
Meltzer, A, C. F. Bergsten, C. W. Calomiris, T. Campbell, E. J. Feulner, W. L.
Hoskins, R. L. Huber, M. H. Johnson, J. I. Levinson, J. D. Sachs, and E. E.
Torres, 2000, Results and Recommendations of the International Financial
Institutions Advisory Commission, Final report to the U.S. Congress and
Department of Treasury, Washington, D.C.
Mussa, Michael, 1999, "Reforming the International Financial Architecture: Limiting
Moral Hazard and Containing Real Hazard," Reserve Bank of Australia,
Conference on Capital Flows and the International Financial System.
Rajan, Raghuram G., and Luigi Zingales, 2000, "The Tyranny of Inequality," Journal
of Public Economics, Vol. 76, pp. 521-558.
Tirole, Jean, 2002, Financial Crises, Liquidity, and the International Monetary System
(Princeton: Princeton University Press).
U.S. Treasury, 2000, Response to the Report of the International Financial Institution
Advisory Commission, (Washington: U.S. Treasury).
287
Figure 1. Monitoring Structures
Figure 2. A Model of IMF Lending
The IMF in a World of Private Capital Markets"
Barry Eichengreen
University of California, Berkeley
Kenneth Kletzer
University of California, Santa Cruz
Ashoka Mody
* The views expressed are those of the authors and should not be attributed to the IMF or any other
organization. We are grateful to Enrica Detragiache, Raghuram Rajan, and an anonymous referee for helpful
comments and to Adrian de la Garza for carefully assembling a complex data set. This revised version of this
paper is dated February 21, 2005.
Introduction
Catalyzing private capital flows to emerging markets has been an objective of the
International Monetary Fund since at least the 1990s.1 The Fund provides public
monitoring services and negotiates programs that enable borrowers to reveal their
commitment to sound macroeconomic policies. In addition, its own lending may stabilize
capital flows by providing bridge finance for creditworthy countries experiencing liquidity
crises, the resolution of which may be difficult to coordinate for atomistic lenders.
In this paper we seek to better understand the roles of IMF monitoring and lending
and provide new evidence of their effects. We analyze the impact of IMF-supported
programs on market access and the cost of funds, building on three insights.
First, if banks engage in monitoring as part of their normal operation, then

IMF monitoring should have a relatively limited impact when bank
syndicates do the lending.
Second, private capital flows should be particularly sensitive to the
magnitude of IMF financial commitments when the likelihood of debt
restructuring is high.
Third, precautionary programs are a mechanism through which
governments can use their relationship with the IMF to signal their
commitment to strong policies. Differences in the impact of precautionary
and regular IMF programs should therefore be usefiil for distinguishing
between the Fund's monitoring and lending roles.
Our analysis is based on more than 6,700 loan transactions between emerging
market borrowers and international bank syndicates and some 3,500 new bond issues
placed between 1991 and 2002. We analyze the frequency of transactions and the spreads
charged. Among the explanatory variables are (a) a measure of repeat borrowing that
proxies for creditor learning about borrower characteristics and (b) the existence and size
of IMF programs. Because we analyze individual transactions rather than aggregate
capital flows or other macroeconomic conditions, our findings are less susceptible to
causality running from the outcome to the decision to initiate an IMF program.2
Important differences between bank loans and bond issues have been documented
in the domestic context.3 Banks act as delegated monitors on behalf of investors who
cannot easily observe and discipline borrowers (Diamond 1984). The information they
acquire can be used to limit the use of funds and in pricing loans. In contrast, individual
bondholders lack the incentive to incur the costs of securing expensive private information
about borrowers. Instead, public informationfor example, the information assembled by
credit rating agenciesdominates the market for debt securities.
See, for example, IMF (1999).

High-frequency data also allow us to capture the timing of programs more precisely than is possible in
aggregate studies using annual data to analyze the influence of IMF programs.
3
This difference between bank and capital markets has been well documented in the domestic US context
(see, for example, Fama 1985 and James 1987).
2
293
Securitized debt instruments, on the other hand, have superior risk-sharing

characteristics. Credit risk can be diversified away, in part, by spreading individual loans
across investors and enabling them to hold diversified portfolios. Banks cannot engage in
this practice to the same extent without eroding their incentive to invest in dedicated
monitoring technologies. This tradeoff is a way of understanding why lending takes place
through both banks and bond markets.
Banks can also more easily coordinate in response to default and restructuring.
They are relatively few in number and contractual arrangements such as sharing clauses
reduce the incentive to hold out. The advantages of creditor coordination may make it even
more profitable for banks to monitor borrowers, as we explain below. Thus, it is not
necessary to assume that banks have intrinsically superior monitoring ability; they may
simply have more incentive to invest in gathering and using relevant information.
Eichengreen and Mody (1998) find that spreads on syndicated loans fall with the
number of loans extended to a borrower. An interpretation is that contact through repeat
borrowing informs creditors about borrower characteristics, reducing uncertainty and risk
premia. That earlier paper did not also consider repeat borrowing in bond markets. We do
so here, hypothesizing that this effect is stronger for bank loans than bonds because
coordination allows banks to make better use of any information thereby gleaned.
The other potential monitor is the IMF.4 By putting a program in place, the Fund
may be able to acquire information not also available to the private sector or acquire it at
lower cost. Indeed, the Fund may convey information to the markets when it does not have
superior monitoring technology. Negotiating an IMF program may simply be a way for a
government to signal its type.5 Imagine that the standard conditions attached to Fund
programs are easier to satisfy for either economic or political reasons by governments truly
committed to strong policies and that violating that conditionality has significant costs.
Then a country with strong policies will be more likely to sign up for a program, signaling
its type and lowering its spreads.
A special case in point is when an IMF lending arrangement is converted into a
precautionary program.6 A country then volunteers to not draw on IMF resources while
still allowing itself to be subjected to Fund monitoring and conditionality.7 The Fund's
monitoring should be particularly important for bond markets not inhabited by a small
number of large investors (banks) prepared to individually invest in ascertaining the
government's type. At the same time, IMF lending, by reducing the probability of default,
could nullify the creditor coordination advantage of banks.
Consistent with these hypotheses, we find that repeat borrowing is more important
in reducing the costs of borrowing from bank syndicates than bond markets. In contrast,
4
As posited by Tirole (2002), Mody and Saravia (2003) and Bordo, Mody, and Oomes (2004).
Bordo, Mody, and Oomes (2004) have argued that the IMF's monitoring role does not imply that the Fund
has better information than the market. As such, the Fund adds value not through the mere signaling of new
information. Rather, the Fund can monitor commitment to a policy program (see also Mody and Saravia
2003). In practice it is difficult to distinguish if it is content of the program or the monitoring that is relevant.
However, because we do observe that programs (with widely varying conditionality) reduce bond interest
rate spreads, it is possible to argue that the monitoring that accompanies the core conditionality in all IMF
programs helps creditors gain confidence in the likelihood of reduced policy variability.
6
For more discussion of the channels through which IMF programs can influence international capital flows,
see Cottarelli and Giannini (2002) and Bordo, Mody, and Oomes (2004).
7
Although the financial support can still become available should the need arise.
5
294
public monitoring through IMF programs has a larger impact on spreads in markets
dominated by bonds than bank loans, again consistent with our priors. But the IMF's
presence and lending have different effects on countries in different situations. For
countries with external debt/GDP ratios below 60 per cent range, it is the IMF's presence,
as distinct from its lending, that matters for bond market access. We interpret this as
consistent with arguments emphasizing the Fund's monitoring and signaling roles. As debt
rises from there, IMF presence is still associated with lower spreads but to a diminishing
extent. The impact of IMF presence disappears when debt reaches 70 percent of GDP.
Moreover, there is little evidence in this high debt range that additional IMF lending
reduces spreads and enhances market access. For countries in this range, neither IMF
presence nor IMF lending significantly enhances market access. Evidently, countries with
such high debts have deep structural problems that must be solved before IMF intervention
can catalyze external finance. Only programs that turn precautionary - that is, where the
outlook improves sufficiently that the country can voluntarily choose to stop drawing on
Fund resources - have a significant negative impact on borrowing costs at high debt levels.
This finding is again consistent with our arguments regarding country signaling and IMF
monitoring.
The next section develops the theoretical background to these issues. The two
sections following that then provide evidence on differences in international lending
through bank loans and bond markets. We then analyze the factors that go into the decision
to borrow and the further choice between loans and bonds. The results confirm that IMF
programs do more to facilitate bond issuance than bank lending. Finally, we document the
importance for the pricing of loans and bonds of private monitoring in bank lending and of
public monitoring through IMF programs in bond markets.
Theoretical Background
Our theoretical discussion focuses on sovereign default, renegotiation, and endogenous
problems of liquidity in highlighting the IMF's monitoring and creditor-coordination roles.
Our point of departure, following Eaton and Gersovitz (1981), Bulow and Rogoff (1989),
and Kletzer and Wright (2000), is that lenders and borrowers take into account the risk of
default when agreeing to the terms of a debt contract. Changes in this risk will therefore be
reflected in the volume of debt and interest rates. Our discussion is also informed by
Tirole's (2002) exposition of dual and common agency problems in the context of
international financial contracts. As Tirole notes, the government may become an agent
even when the debt contract is between private borrowers and lenders, since government
actions bear on a private debtor's ability to repay. Private debt can have sovereign risk as a
result of explicit or implicit government guarantees and/or a debtor's recourse to domestic
legal protection. We therefore assume that the envelope of resources and government
policies determines the ability and willingness of the government and private creditors to
repay.8
The logic of our argument can be summarized as follows. First, country
fundamentals and government policies determine the ability and willingness of the
government and private creditors to service their debts. This implies that the probability of
8
Even if a private borrower derives no protection from its home government's sovereignty, the analytics
apply to any debtor that faces bounded penalties for defaulting.
295
default rises only after a threshold level of debt is reached, at which point access to credit
markets weakens and spreads on new loans begin to rise.9 Second, in the presence of
asymmetric information about borrower preferences, the ability of creditors to monitor
debtors and make use of information will influence market behavior and outcomes.
Creditor monitoring matters when debt restructuring can occurin other words, it matters
for countries with high debt levels and low credit ratings. A monitoring advantage can
arise not just when some creditors are better informed than others but also when some
creditors more readily respond to common information.10
An implication is that creditors with a monitoring advantage will tend to lend in
markets where the return to monitoring is high, and conversely. When banks have a
monitoring advantage, information about the debtor's policy preferences will affect terms
on bank loans and bonds. IMF programs can also reveal and provide confidence in debtor
government policies, reducing the common agency problem, but providing information can
also reduce the advantage of private monitoring. Monitoring differences will also affect
returns to learning about the debtor.
To further develop these points, we utilize a simple framework in which the
debtor's resources are stochastic and all debt claims have the same maturity and priority.
The debtor is willing to repay a maximal amount, V(y), in expected present value, in
equilibrium. V(y) is the value of repaying in a forward-looking equilibrium that takes
account of opportunities to renegotiate debt in the future. It is increasing in the
fundamental, y. For strong fundamentals, y, or low levels of debt, V(y) exceeds D, and the
debtor will repay. If, however, the outstanding debt, D, exceeds V(y), then the debtor is
unwilling to meet its obligations and will seek to renegotiate.
When borrowing and repayment are repeated over time, the debtor's
willingness to pay can be written as
(i)
where w(yt) indicates the debtor's equilibrium willingness to service debt today. An
interpretation is that w(yt) represents the debtor's liquid resources and VM measures
solvency. Under perfect information, current debt service obligations that cannot be met by
the debtor (Dt >V(yt}) will be rolled over into new loans, while debts that will not be
repaid in present value will be renegotiated.
The expected net return to creditors is given by
(2)
For low debt levels, an increase in debt may indicate improving fundamentals and thus result in narrower
spreads, as suggested by Pattillo et al. (2003).
10
For example, banks can have a comparative advantage in creditor coordination in the context of debt
renegotiation (when there are advantages to getting all creditors to take the same position). They may also be
in a relatively favorable position to arrange concerted lending and thus control strategic uncertainty that can
cause liquidity crises.
296
where r is the return on alternative investments. The interest rate spread will be the
difference between ERt IDt and (1+r). This spread is increasing with the level of debt for
positive probabilities that V(yt] is less than Dt. When the level of debt is low, this
probability can be zero, in which case the spread does not rise with indebtedness. But as
indebtedness rises further, the probability of default becomes positive, as does the risk
premium. Models of debt renegotiation with perfect information thus imply that spreads
will not increase with debt at low debt-to-GDP ratios but that they will start rising at an
accelerating rate after the debt-to-GDP ratio passes a critical threshold. This is
corroborated by our empirical work, below.
To motivate the role of monitoring, it is necessary to introduce information
asymmetries. Assume that the debtor's willingness to pay is known by others with
uncertainty. Specifically, suppose that lenders only know the distribution, of the debtor's
willingness to pay,
within an interval,
For simplicity, the
distribution can be taken as uniform around a mean equal toV(yt). The debtor can offer
repayment,
less than its true willingness to repay. Consistent with standard
analyses, the equilibrium offer accepted by lenders yields r e p a y m e n t , e q u a l to the

debtor's actual willingness to pay when this equals its minimum value
. For larger
realizations of
, the debtor will transfer less than its true willingness to pay and
realize a positive surplus given by the difference
The debtor pays in full if
. Because actual repayments are less than the
debtor's true capacity, the probability of default is higher and creditors' expected returns
are lower when information is asymmetric.
Creditors can extract more surplus if monitoring helps them to become better
informed about the debtor's future policy actions. Monitoring increases willingness to pay,
raising returns in the event of renegotiation and reducing the probability that renegotiation
occurs. If lenders learn about the characteristics of borrowers from repeat lending, as
appears to be the case from the evidence reported below, then spreads should fall with
repeat transactions. Similarly, if the IMF has an advantage in monitoring the policy actions
of the debtor, then agreement to establish an IMF program should reduce spreads and
increase debt issuance for a debtor with a positive probability of having to renegotiate.
Our empirical analysis in Section 5 below points to differences in the impact of
repeat lending and IMF programs on bank loans and bond spreads.11 An explanation
consistent with our findings is that banks and bondholders have different monitoring
abilities. Banks will cater to smaller, less well established borrowers, since they
presumably possess a superior monitoring technology. Bondholders will focus on large,
well-known borrowers.12 The private information revealed by clients to their banks will
then make more precise the bankers' views of their capacity to repay. On the other hand, if
banks have a monitoring advantage over bondholders, then an improvement in public
information resulting from an IMF program could reduce (or in the limit remove) that
1]
See the section on the pricing of bonds and loans.

See Petersen and Rajan (1994, 1995) for U.S. evidence consistent with this pattern.
12
297
informational advantage, reducing bond spreads and encouraging bond issuance relative to
bank loans.
With asymmetric information, the adoption of an IMF program can reveal
information to capital markets about country policies and willingness to pay. This does not
hinge on the assumption that the IMF has superior ability to collect or interpret data; the
Fund may simply have the ability to commit to actions to which private investors cannot or
will not commit.13 To the extent that the IMF has a superior ability to commit or objectives
that differ from those of private investors, adoption of a Fund program can also reveal
information about the debtor country. In turn this allows the government to signal its
intentions.14 For example, the conditionality associated with an IMF loan might be less
onerous for governments for which policy reforms are less costly, thus making it incentive
compatible for such governments to sign up with the Fund in order to signal its type.
Countries with stronger policies or a greater will to enact policy reform are thus able to
reveal this information by negotiating a Fund agreement. Of course, adopting a program
may also reveal poor fundamentals, and not just a superior capacity to enact policy
reforms, resulting in an overall ambiguous impact on spreads.
IMF programs sometimes turn precautionary: borrowing stops, but the government
continues paying a commitment fee that gives it the option to resume borrowing. By
turning a program precautionary, the debtor country government can thus reveal that it has
a diminished need for official finance but a continuing commitment to prudent policies.
This good news should be reflected in a reduction in spreads on both bank loans and
bonds.
Debt restructuring can also give rise to differences between banks and bondholders
if the members of a bank syndicate can more easily coordinate their actions. Recall that
equation (1) separates current willingness to pay into the sum of current resources
available for repaymentw(y t ) and discounted expected future willingness to pay. If
coordination failures prevent bondholders from restructuring debts quickly, then banks can
move first and secure a larger share of the pie. They can do so even when all creditors
have identical information and learn at the same time that the debt is unsustainable. Recall
that
(3)
where Bt and Lt are outstanding bonds and bank debt, respectively. Banks can reschedule
their loans and avoid immediate default by reducing repayments currently due while at the
same time increasing future repayments by rolling over their loans at higher interest rates.
Subsequent renegotiations incorporating equal sharing between bondholders and bank
lenders will then divide the settlement amount between banks and bondholders on the basis
13
It should be possible to model the IMF as endogenously gaining a monitoring advantage through its ability
to commit to lend only in a crisis in a repeated game. The approach of self-enforcing equilibrium taken by
Kletzer and Wright (2000) in the sovereign debt context could be used to model de facto IMF seniority and
why countries might meet IMF conditionality.
14
Marchesi and Thomas (1999) offer a model in which Fund conditionality serves as a screening device.
298
of the new bank share of the total debt. To illustrate, let the banks reschedule an amount
ALt of current debt repayments so that
(4)
where
and
are interest payments due on bonds and loans, respectively. The
banks then increase loans in period t+1, LM, by an amount r'AL,. The banks' share of
future repayments rises to
(5)
Since the increase in the value of bank claims comes out of the expected returns to
bondholders in the event that current total debt is unsustainable, the interest rater1 can then
be chosen to maximize the increase in expected returns,
(6)
If the banks can reschedule a sufficient share of their debt, they can eliminate their
current expected loss at the expense of bondholders. This strategic advantage contrasts
with a simple principal-agent model in which improved monitoring by banks raises the
probability of repayment and returns to banks and bondholders alike.15
The banks' strategic advantage can be reduced or eliminated by the presence of a
more senior official-sector creditor. Since the first-mover advantage arises from the
prospect of default, it can be reduced by availability of official support under an IMF
program, assuming that such funding reduces the risk of renegotiation. Absent differences
in bank and bond markets, the basic model of sovereign debt renegotiation with
asymmetric information would imply that IMF monitoring and financial resources lead to
equivalent reductions in bond and bank loan interest spreads.16 Similarly, if IMF
conditionally improves fundamentals and growth prospects, then both bond and bank
lending should increase. However, if banks have a monitoring advantage and can better
manage creditor coordination and debt restructuring problems, as assumed here, then IMF
monitoring that reveals debtor characteristics and IMF lending that reduces the likelihood
of default will benefit bondholders more than banks.
Finally, we consider the role of liquidity crises, adapting the model of Morris and
Shin (2003).1? In their model, the fundamental has a distribution that is public knowledge,
15
The sharing of negotiated repayments here contrasts with the assumptions of Bolton and Jeanne (2003) that
bonds are not renegotiable but bank loans are and separate penalties apply in selective defaults.
16
Gai and Vause (2003) present a model in which the IMF acts as a delegated monitor motivated by private
creditor coordination failures. Our emphasis on asymmetric private abilities to coordinate is different.
17
Similar models by Rochet and Vives (2001) and Corsetti, Dasgupta, Morris and Shin (2001) also take a
global games approach to catalytic finance. Chui, Gai and Haldane (2002) also discuss the policy
implications of sovereign liquidity crises.
299
but each lender in a continuum receives a privately observed noisy signal of its realization
in the current period.18 In this setting, private information can generate coordination
failures and produce liquidity crises even when debt is sustainable.
We reinterpret their model by distinguishing between banks and bondholders,
assuming that banks coordinate whereas bondholders do not. If V(yt) exceeds total debt
but the country's current liquid resources fall short of current net payments due, then a
liquidity crisis is possible. When debtor liquidity falls below a critical level, bondholders
facing uncertainty about one another's actions will be unwilling to purchase new bond
issues to replace amortizing debt.19 This may give rise to an incipient crisis. That crisis
may be prevented, however, if bank syndicates replace the maturing bonds with additional
bank loans.
Banks are able to do this, in principle, because they can coordinate among
themselves. Suppose that banks observe both a private signal drawn from the same
distribution as that of bondholders and the failure of the debtor to place new bond issues.
They can then halt a liquidity-driven crisis by replacing the maturing bonds with new
loans. They can move after bondholders exit and have an incentive to do so in order to
avoid unnecessary defaults on their long-term loans. Such models thus imply that bank
loans and bond issues should be negatively correlated if crises are caused by illiquidity.
Two further implications follow. First, a deterioration in market liquidity or
increased uncertainty that reduces bond issuance can be mitigated by the presence of bank
lending, since banks have an incentive to fill the gap. Second, the IMF, as lender in a
liquidity crisis, can help to avoid costly default and renegotiation.20 Assume that potential
purchasers of bonds are as poorly informed about what banks will do as they are about
what other bondholders will do. Banks move on the basis of private information and the
reluctance of bondholders to reenter the market. But both bondholders and banks should be
able to anticipate the IMF's strategy when a program is in place. Then the existence of an
IMF program should raise bond issuance relative to bank lending for countries susceptible
to liquidity-driven crises. We examine this proposition empirically below.
The Setting
Although international lending through bond markets was prominent in the late 19th and
early 20th centuries, from the 1960s through the 1980s private credit flows to developing
and emerging economies took place mainly through banks. Lending via bond markets was
about 10 percent of bank lending in the 1970s and early 1980s (Edwards 1986). This
changed following the debt crises of the 1980: between 1991 and 2002, credit obtained via
banks and bonds was of about the same order of magnitude, just under $700 billion
through each channel (Table 1).
01
18
Morris and Shin also distinguish short-term debt that amortizes in the current period from other debt.
Morris and Shin (2003) detail the determination of the critical level of liquidity. For our interpretation, we
leave out additive debtor effort in their model.
20
Jeanne (2001) among others discusses the lender of last resort role of the IMF.
21
While we include all bonds issued in our analysis, we restrict the sample of loans to those that were priced
based on Libor. These form the vast majority of international syndicated loans, both in terms of numbers and
in the amount borrowed. By limiting the loans to those priced off Libor, we believe that more precise
estimates of loan pricing become possible.
19
300
Differences persisted, however, in the characteristics of the typical bank loan and
bond. To show this, for each loan and bond in our data set we extracted the initial price,
the initial maturity, the amount, and the currency of denomination. Borrowers are
distinguished as sovereign, non-sovereign but public sector, and private sector.22 On
average, bank loans are more numerous and smaller. Between 1991 and 2002, Loanware
reports 6,747 Libor-based syndicated loan transactions; during the same period, Bondware
reports the issuance of just over 3700 bonds.23 On average, a bond issue was about 70
percent larger than a loan transaction.
We construct a measure of repeat borrowing, /?, separately for bank and bond
borrowing. Starting with January 1, 1991, the measure takes the value 1 the first time a
borrower enters into an international debt contract. With each subsequent instance of
borrowing we then increment the value of R by one. The results show that repeat
borrowing is more common in bond markets, where the median number of borrowings
over the period 1991 to 2002 is 3 (the 75th percentile is 8 and the 90th percentile is 27); for
banks, the median is 2 (the 75th percentile is 4 and the 90th percentile is 8). Thus, compared
with banks, which allow a diverse set of clients to episodically borrow, the bond market
caters to borrowers with name recognition who return frequently.
Relative to bank loans, bonds were more likely to be issued when the issuing
country had an IMF-supported program. About 22 percent of loans were contracted when a
country had a Fund program in place (Table 2). In contrast, just over a third of bonds were
issued in the presence of a program. To put the point another way, when countries were
under an IMF program they were about as likely to borrow through a loan or a bond, but a
loan was more than twice as likely when there was no program.
While IMF programs appear to shift borrowing toward bonds, this shift does not
occur uniformly. Table 2 shows that countries with external-debt-to-GDP ratios below 30
percent had few bond or loan transactions while under IMF programs. When the debt-ratio
was between 30 and 40 percent, more borrowing occurred under IMF programs, especially
through bonds; however, the number of credit contracts was still higher in countries
without, rather than with, IMF programs. Countries with debt/GDP ratios in the 40-60 per
cent range play an important role in our analysis. In this category, the distribution of credit
contracts between program and no program is more even: indeed, more bonds are issued
under a program than when there is no program. Finally, when external debt exceeds 60
percent of GDP, countries once again limit their international borrowing. When they do
borrow, loans and bonds are equally prevalent.
Patterns of Borrowing
In this section, we analyze the borrowing decision and the choice between bank loans and
bonds. The first probit equation (Table 3) estimates the correlates of borrowing by
sovereign, non-sovereign/public, and private entities in each country-quarter. The second
22
We use these distinctions to also construct an estimate of the numbers that did not borrow. Thus, for a given
country in a given quarter, the absence of borrowing by the sovereign implied that the sovereign had either
forgone the opportunity to borrow or had not had access to international funds. Similarly, we identify
country-quarters where no public (non-sovereign) and private borrowing occurred. For more on this, see
below.
23
Of which spreads are available for about 3 500
301
equation reports the likelihood of bond issuance rather than a bank transaction. Throughout
we report the change in the probability for an infinitesimal change in each independent,
continuous variable at its mean and the discrete change in the probability for dummy
variables. Standard errors are adjusted for clustering since the number of borrowing
transactions varies from country to country.24 Explanatory variables include issuer
characteristics (in this regression, the borrower type, with sovereign as the omitted
category), global variables (U.S. growth, the swap rate, EMBI volatility), and a vector of
country characteristics.25
Among the global variables, U.S. growth appears to facilitate borrowing, especially
by bond issuers in the medium-debt range (with a debt/GDP ratio between 40 and 60
percent). An interpretation is that global growth acts as collateral that supports additional
borrowing. If the average monthly growth of U.S. industrial production rises from its mean
of 0.3 percent to 0.4 percent, the probability of borrowing increases by just over one
percent.26
Higher volatility of J.P. Morgan's Emerging Market Bond Index, reflecting greater
uncertainty about pricing, is associated with reduced borrowing. If daily volatility
increases from its monthly mean of 2 per cent to 3 per cent, the probability of borrowing
declines by 1V4 to 2 percent.
Higher bond-market volatility also lowers the frequency of bond issuance relative
to bank loans by borrowers from countries with debt/GDP ratios below 60 percent.27 A
one percentage point increase in daily volatility reduces the likelihood of bond issuance
relative to a bank transaction by 2V-4 percent. An interpretation is that short-run liquidity
concerns and financial market disorder are more likely to generate strategic uncertainty
among bondholders, who may then withdraw to the sidelines on the fear that others are
doing the same. In contrast, banks, which are better able to coordinate among themselves,
are more likely to continue lending.28
Improved credit quality (proxied by Institutional Investor credit ratings, which run
from a low of 0 to a high of 100) allows for more borrowing both from banks and on bond
markets. The importance of the credit rating increases when the external-debt/GDP ratio
exceeds 40 percent. An increase in rating by 10 points from a mean of 52 strongly raises
the likelihood of borrowing with no apparent shift in its composition.29 An interpretation is
that whereas ratings influence the willingness of lenders to lend, a country's demand for
foreign exchange determines how much it wishes to borrow. Thus, a higher ratio of debt
service to exports increases the demand for external resources, thereby raising the
likelihood of international borrowing, provided that the debt/GDP ratio is below 60
percent. Interestingly, as the debt/GDP ratio rises, the demand for external borrowing is
increasingly met through loans. Similarly, when countries face higher export volatility,
24
This same correction for clustering is made throughout.

More detail on variable definitions and sources can be found in the appendix, below.
26
The measure of US growth used in the regressions is the average of monthly growth rates in the quarter in
which the transaction occurred.
27
Where debt ratios are higher, such compositional shifts are not statistically significant.
28
The Korean crisis in 1997-98, and other similar episodes, remind us that there may be limits to such
coordination. But an important fact about the Korean crisis is that, in the end, the banks did roll over their
loans, albeit at high interest rates. See, for example, Goldstein (1998).
29
The likelihood of borrowing rises by between 16 and 25 percent
25
302
they are less likely to borrow abroad; in particular, they are especially prone to reduce their
borrowing on bond markets.
Bond issues tend also to be larger and longer in term. Whereas the average maturity
of loans in our sample is 4V years (the median is just over 3 years), that for bonds is 6%
years (with a median of 5 years).30
IMF programs have limited influence on aggregate borrowing by countries at low
debt levels, as already suggested by Table 2. Presumably structural problems that limit the
ability to borrow also cause countries to seek Fund assistance. Table 3 suggests, however,
that such borrowers are more likely to issue bonds than borrow from banks. In the
medium-debt range, a Fund program raises the probability of borrowing by 14 per cent. At
high debt levels, the influence of IMF programs remains positive, although the effect is not
statistically significant.
We also distinguish precautionary programs. A first case is where IMF programs
are designated as precautionary at outset Country authorities declare that they do not
intend to draw on resources made available.31 Borrowing via both loans and bonds is lower
in such cases, but mainly for countries in the intermediate debt range. There is thus some
suggestion in the data that countries choosing to approach the Fund for precautionary
reasons also behave conservatively in their borrowing from banks and on bond markets.
A second case is when programs turn precautionary. In this instance the member
stops drawing on resources available through the program but continues to pay the
commitment fee to retain access. Aggregate borrowing does not appear to be affected by
such arrangements.32
The Pricing of Loans and Bonds
To analyze pricing, we use the model developed by Eichengreen and Mody (2000, 2001)
and extended by Mody and Saravia (2003). The spreads equation is linear of the form:
log (spread) =
(1)
where the dependent variable is the logarithm of the spread; Xis a vector of issue, issuer,
and period characteristics; and uj is a random error. X contains a dummy variable for the
existence of an IMF program, program characteristics if any, and interactions between the
program and country characteristics.33 Since the spread will be observed only when there is
a decision to borrow and lend, we correct for sample selection. Assume that spreads are
observed when a latent variable B crosses a threshold Bf defined by:
(2)
where Z is the vector of variables that determines the desire of borrowers to borrow and the
willingness of lenders to lend (and will also contain the IMF program variables and their
30
A borrower wishing to increase the length of maturity from the average from the average bank loan to the
average bond maturity is about 3.5 percent (1.75*0.02) more likely to issue a bond.
31
This declaration is not binding, as noted above.
32
Although borrowers from countries with high debt/GDP ratios appear to be less likely to issue bonds.
33
As discussed below.
303
interactions). 2/2 is a second error term. We assume that: uj ~ N(0,o), u^ ~ N(0,I), and corr
(uj, 1*2) = />. This is a sample selection model a la Heckman (1979). Equations (1) and (2)
can be estimated simultaneously by maximum likelihood.
Estimating the determinants of market access requires information on
nonborrowers. As noted above, for each country we consider three categories of issuers:
sovereign, other public, and private. For each quarter and country where one of these
issuers did not come to the market, we record a zero, and where they did we record a one.34
We use our measure of repeat borrowing, R, to proxy for private monitoring. It is
likely that the incremental information declines as R rises. Moreover, since R is correlated
with the number of debt obligations outstanding, a larger value of R may also create greater
coordination problems in the event of restructuring.35
The IMF dummy appears in both the selection and spreads equations. In contrast,
R appears only in the spreads equation. Other variables in the selection equation are the
global and country variables from Table 3. In addition, transaction-specific variables such
as the maturity and amount of the credit transaction and dummy variables for the currency
of issue and production sector of issuer (not shown to conserve space) are included in the
spreads equation.36 Results are in Table 4.
U.S. growth is associated with lower spreads and raises the likelihood of borrowing
through banks and bond markets. This is again consistent with the idea that stronger global
growth and export opportunities act as collateral for emerging markets. These effects are
especially important for the middle debt group: an increase in monthly growth rate of 0.1
percent (a 1.2 percent increase in annual growth) reduces loan spreads in the mid-debt
range by 2 per cent and bond spreads in that same range by about 4 per cent. Increases in
issuance probabilities are somewhat smaller.
Among the global variables, an increase in EMBI volatility has a particularly
important quantitative effect on bond issuance when a country's debt-to-GDP ratio is
below 60 per cent. If daily volatility rises by one per cent (at the daily mean of 2 per cent),
bond issuances fall by between 5 and 7 per cent (in that same debt range). Improved credit
ratings raise the probability of borrowing while lowering spreads, consistent with the idea
that their main effect is to increase investors' willingness to lend. A 10-point improvement
in the Institutional Investor rating has a large impact on spreads (with the largest effect in
the mid-debt range, 32 per cent for loans and 48 per cent for bonds). For borrowers from
countries with debt/GDP ratios below 60 per cent, improved credit ratings have a relatively
small impact on bank lending, suggesting that public rating information, while relevant to
access in both markets, is less valuable for bank decision making under normal
circumstances.
34
Leung and Yu (1996) note that the estimation does not require the variables in the selection equation and
the spread equation to be different but rather that the variables not be concentrated in a small range and that
truncated observations (no bond issuance) not dominate. We do include in the selection equation (the probit),
the ratio of debt service to exports, which appears to influence the issuance decision but not the
determination of spreads.
35
In the regressions, we use the log of/?, which has a distribution that is much closer to normal than the
(skewed) distribution of/?. We also allow all coefficientsand not just the variables of immediate interest,
R and the IMF program dummyto vary by debt category.
36
For a more extended discussion of the joint interpretation of the selection and spreads equation, see
Eichengreen and Mody (2000).
304
Our main result is that repeat borrowing reduces spreads on syndicated loans, while
IMF programs reduce spreads in bond markets. The coefficient on the log of repeat bank
borrowing is negative, significant and larger than the corresponding coefficient for bond
markets. This is true for each of the three debt/GDP categories. The effects in the loan
market are large. A second loan reduces spreads by about 10 per cent.37 A third loan has a
spread about 6 per cent lower than the second loan, after which the impact declines to low
levels. In bond markets, in contrast, only lightly indebted countries gain from repeat
borrowing.
IMF programs, on the other hand, reduce spreads and enhance access mainly in
bond markets. This effect is most evident in medium-debt countries with debt/GDP ratios
in the 40-60 per cent range. Bond issuance by countries in this category is about 13 per
cent higher when there is a Fund program and spreads are 40 per cent lower. Evidently,
bondholders become more willing to lend to such countries following the negotiation of a
Fund program. IMF programs also facilitate bank borrowing by countries in this mediumdebt range, but the impact on spreads is insignificant.
Finally, as noted in Table 2, in the low-debt range (especially when the debt/GDP
ratio is below 30 per cent), countries with IMF programs borrow little. Countries with
modest debts that nonetheless negotiate IMF programs appear to have unobserved
characteristics that raise rather than lower spreads.39
In sum, repeat transactions have a significant effect mainly on bank borrowing,
while IMF programs improve the terms of access to a greater extent for bonds.
no
Extensions
We now explore further the robustness of these results, varying the cutoff points,
considering the size of IMF programs, and distinguishing private and public borrowers.
We first ask whether the results are sensitive to cut-off points for the debt/GDP
ratio. Table 5 reports results for overlapping debt/GDP ratios, starting with the 10 to 30 per
cent range and then raising the end points by 10 percentage points over 6 intervals.40 Panel
A, for loans, confirms the value of repeat borrowing which is significant in all 6 intervals.
Comparison with the corresponding coefficients in Panel B shows that the value of repeat
borrowing is greater for loans than for bonds in every debt category. Panel A also
confirms that IMF programs do not reduce spreads significantly and are associated with
higher spreads until the debt/GDP ratio is between 40 and 50 per cent. However, once the
debt/GDP ratio exceeds 50 per cent, IMF programs are associated with a higher frequency
of borrowing from banks with no apparent adverse effect on spreads.
Panel B confirms that repeated bond issuance lowers spreads only in the 10-30 per
cent debt/GDP range and has limited value thereafter, in fact raising spreads as if a
multiplicity of bonds creates coordination problems. The contrasting importance of IMF
programs is also evident. At the low end of the debt/GDP range, there is a tendency for
37
A coefficient on the log of repeated borrowing of 0.14 times the difference between log 2 and log 1,0.69.
This finding of a strong impact of Fund programs for bond market access is also a central result in Mody
and Saravia (2003).
39
Even more for loans than bonds.
40
Ending with the 60 to 80 per cent range. We exclude the low and high ends of the debt/GDP spectrum
where outliers tend to drive the results. Thus, for example, some of the transition countries had very low
levels of debt in the mid-1990s, which may not have been an accurate reflection of their external obligations.
38
305
Fund programs to be neutral or to reduce spreads modestly, but the effect strengthens
noticeably as the debt/GDP ratio approaches 40-60 per cent. Beyond that, the influence of
IMF programs on spreads falls. Fund programs are also associated with more bond
issuance. This effect is strongest when indebtedness is between 20 and 60 per cent of
GDP.41
In Table 6 we examine the influence of the size of IMF programs.42 We interact the
IMF program dummy with the country's debt/GDP ratio and normalize the amount of IMF
lending by the country's external debt. For bonds, all the action is in the intermediate debt
category where, as above, IMF programs have their major impact on spreads. The results
in Table 6 thus reinforce the earlier finding that higher debt/GDP levels reduce the impact
of IMF programs on bond markets. At the same time, the amount of lending does not
influence spreads. These results are consistent with the Fund's value as a monitor rather
than a provider of liquidity that prevents the occurrence of a financial crisis on account of
strategic uncertainty among creditors.
In the market for bank loans, the larger is IMF assistance the higher are spreads in
the two low-debt categories at least. Thus, while availability of additional IMF resources
allows for additional borrowing, it is as if the creditor coordination advantage is
eliminated.43
In Table 7 we again consider precautionary programs. For bank loans and to a
lesser extent for bonds, programs that are precautionary at outset reduce both issuance and
spreads, as if countries entering such programs are more cautious in seeking access to
private markets.44 Spreads show a tendency to decline, as if lenders wish to acquire more
of their debt because their credit quality is perceived favorably.
But programs that turn precautionary tend not to have an impact on the frequency
of either bank loans or bond issuance. However, they do have a spread-reducing effect.
This is largest for countries in the high-debt zone. In this range borrowers both from banks
and on the bond market enjoy lower spreads, although the impact is larger in bond markets.
Thus, when a country is coming off a period during which it has relied on official finance,
a continued precautionary relationship with the Fund appears to enhance market access.
That the relationship rather than the amount lent is what matters supports the idea of a
Fund monitoring/country signaling function.45
Finally, Table 8 considers whether the market access of private borrowers is
differentially affected by the existence of an IMF program. In fact, repeat borrowing
reduces spreads more strongly for bank loans than bonds irrespective of whether the
borrower is a private- or public-sect or entity. But the effect is larger for private sector
borrowers.46 Less is publicly known about private borrowers. Their repeat borrowing
41
These results support those obtained by Mody and Saravia (2003).

Based on the findings reported in Tables 4 and 5 we again allow for the effect of programs and repeat
borrowing to vary by the level of indebtedness. But to avoid excessively detailed results, we return to
presenting results by three debt categories.
43
However, in the medium-debt range, the adverse effects of increasing debt levels from 40 to 60 percent of
GDP are mitigated by the presence of an IMF program.
44
Recall that this was what was suggested by our earlier analysis.
45
That this function is important also to bank lenders when a country is in the high-debt range suggests that
bank monitoring may not be enough when there is a high risk of insolvency.
46
Thus, a second loan reduces the spreads charged private bank borrowers by about 13 percent, while public
borrowers achieve, on average, a 7 percent spread reduction.
42
306
therefore provides particularly valuable information in the syndicated loan market. In the
bond market, in contrast, better-known private borrowers gain little from repeat borrowing.
In fact, public borrowers face rising spreads as they borrow more, presumably reflecting
the dominance of coordination effects over information gains.
The stronger influence of IMF programs when borrowing occurs through the bond
market survives splitting the sample. Again, private borrowers gain the most. The principal
action is still in the intermediate debt category. In addition, the effects for private
borrowers are substantially stronger than those for public borrowers. A Fund program
reduces bond spreads for private borrowers from countries in this intermediate debt zone
by 47 per cent while raising the probability of bond issuance by 27 per cent.47
Conclusions
Bank loans and bonds are alternative ways of transferring capital to emerging markets. The
growth of global bond markets is of course one of the signal features of the last 15 years of
international financial history. Transacting through bond markets has advantages for
investors, notably greater scope for diversifying country risk. Given the advance of
securitization across a broad front, it is therefore important to observe that bank finance
continues to play an important role in international financial markets. Bank loans are
easier to access for borrowers new to such markets, since banks have a comparative
advantage in bridging information asymmetries. Banks' intermediation technologies are
also better suited to providing small loans.
We show in this paper how the ability of banks to bridge information asymmetries
is supported by repeat borrowing. As borrowers return for credit, they reveal information
about themselves, reducing uncertainty and incurring a lower risk premium on their loans.
Since the issuers of bonds are better known, the value of information obtained through
repeat issuance is less. Indeed, to the extent that it results in a proliferation of separate
bond issues, repeat borrowing may in fact increase the risk premium, reflecting the greater
difficulty of coordinating the holders of different issues in the event of debt-servicing
difficulties.
These observations have obvious relevance to arguments about IMF monitoring
and surveillance. Our results suggest that IMF monitoring and surveillance matter more in
bond markets. This role for the IMF has the largest impact when debts reach 40 per cent of
GDP and countries are therefore vulnerable to liquidity shocks. However, as debts
continue rising from there, the impact of monitoring declines. There being relatively little
uncertainty about the nature of the problem, lenders now care mainly about whether the
IMF is providing real resources that help to keep debt service current. But as debt and
insolvency risk grow still higher, even significant amounts of additional official finance
may not make a difference. At that point, what matters most is when programs turn
precautionary, signaling that conditions have improved sufficiently that the country no
longer requires financial assistance.
The approach taken here points to the importance of distinguishing international
capital flows by instrument and intermediary. Macroeconomic analyses lumping together
bank loans and bonds tend to neglect important differences between these market segments
47
The direction of influence is the same for public issuers, but the size and statistical significance of the
outcome is weaker.
307
stemming from the nature of the information environment, the monitoring technology, and
the scope for creditor coordination. We have shown in this paper that these distinctions are
important for understanding the impact of IMF programs. We would conjecture that they
are equally important for understanding a variety of other issues in international finance.
308
References
Bolton, Patrick and Olivier Jeanne, 2003, "Structuring and Restructuring Sovereign Debt:
The Role of Seniority," Unpublished.
Bordo, Michael, Ashoka Mody, and Nienke Oomes, 2004, "Keeping Capital Flowing: The
Role of the IMF," Unpublished.
Bulow, Jeremy and Kenneth RogofF, 1989, "A Constant Recontracting Model of
Sovereign Debt," Journal of Political Economy 97, pp. 15 5-178.
Chui, Michael, Prasanna Gai and Andrew Haldane, 2002, "Sovereign Liquidity Crises:
Analytics and Implications for Public Policy," Journal of Banking and Finance,
26, pp. 519-546.
Corsetti, G., A. Dasgupta, S. Morris and H.S. Shin, 2001, "Does One Soros Make a
Difference? A Theory of Currency Crises with Large and Small Traders," Review
of Economic Studies, forthcoming.
Cottarelli, Carlo, and Curzio Giannini, 2002, "Bedfellows, Hostages, or Perfect Strangers?
Global Capital Markets and the Catalytic Effect of IMF Crisis Lending," IMF
Working Paper No. 02/193 (Washington: International Monetary Fund).
Diamond, Douglas, 1984, "Financial Intermediation and Delegated Monitoring," Review of
Economic Studies 51: 393-414.
Eaton, Jonathan and Mark Gersovitz, 1981, "Debt with Potential Repudiation," Review of
Economic Studies 48, pp.289-309.
Edwards, Sebastian, 1986, "The Pricing of Bonds and Bank Loans in International
Markets: An Empirical Analysis of Developing Countries' Foreign Borrowing,"
European Economic Review30 (3): 565-589.
Eichengreen, Barry and Ashoka Mody, 1998, "Lending booms, reserves, and the
sustainability of short-term debt: inferences from the pricing of syndicated loans,"
Journalof Development Economics 63 (1): 5-44.
Eichengreen, Barry and Ashoka Mody, 2000, "What Explains Changing Spreads on
Emerging Market Debt?" In Sebastian Edwards, ed., Capital Flows and The
Emerging Economies: Theories, Evidence, and Controversies (Chicago: The
University of Chicago Press).
Eichengreen, Barry and Ashoka Mody, 2001, "Bail-Ins, Bail-Outs, and Borrowing Costs,"
IMF Staff Papers, Vol. 47, pp. 155-87.
Eichengreen, Barry, Kenneth Kletzer and Ashoka Mody, 2004, "Crisis Resolution: Next
Steps," Brookings Trade Forum 2003, Washington, DC: The Brookings Institution,
pp. 279-352.
Fama, Eugene, 1985, "What's Different about Banks?" Journal of Monetary Economics
15:29-36.
Gai, Prasanna and Nicholas Vause, 2003, "Sovereign Debt Workouts with the IMF as
Delegated Monitor - A Common Agency Approach," Bank of England Working
Paper no. 187.
Goldstein, Morris, 1998, The Asian Financial Crisis, Washington, D.C.: Institute of
International Economics.
Heckman, James, 1979, "Sample Selection Bias as a Specification Error," Econometrica,
Vol.47,pp.l53-161.
309
International Monetary Fund, 1999, "Involving the Private Sector in Forestalling and
Resolving Financial Crises," (Washington: International Monetary Fund).
Available via Internet, http://imf.org.
James, Christopher, 1987, "Some Evidence on the Uniqueness of Banks," Journal of
Financial Intermediation 19:217-23 5.
Jeanne, Olivier, 2001, "Sovereign Debt Crises and the International Financial Architecture,
IMF working paper, (unpublished).
Kletzer, Kenneth, 2003, "Sovereign Bond Restructuring: Collective Action Clauses and
Official Crisis Intervention." Paper for Bank of England conference, "The Role of
the Official and Private Sectors in Resolving International Financial Crises," July
23 and 24, 2002, revised.
Kletzer, Kenneth and B. D. Wright, 2000, "Sovereign Debt as Intertemporal Barter,"
American Economic Review 90, pp. 621-639.
Leung, S. F., and S. Yu, 1996, "On the Choice between Sample Selection and Two-Part
Models" Journal of Econometrics, Vol. 72, pp. 197-229.
Marchesi, Silvia and Jonathan Thomas, 1999, "IMF Conditionally as a Screening
Device," Economic Journal, 109, pp. 111 -125.
Mody, Ashoka and Diego Saravia, 2003, "Catalyzing Capital Flows: Do IMF-Supported
Programs Work as Commitment Devices?" IMF Working Paper WP/03/100,
Washington D.C.
Morris, Stephen and Hyun Song Shin, 2003, "Catalytic Finance: When Does It Work?"
Yale University, Cowles Foundation Discussion Paper No. 1400, February.
Pattillo, Catherine, Helene Poirson and Luca Ricci, 2004, "Through What Channels Does
External Debt Affect Growth?" Brookings Trade Forum 2003, Washington, DC:
The Brookings Institution, pp. 229-277.
Penalver, Adrian, 2004, "How Can the IMF Catalyse Private Capital Flows? A Model,"
Bank of England Working Paper no. 215.
Petersen, Mitchell A. and Raghuram Raj an, 1994, "The Benefits of Lending Relationships:
Evidence from Small Business Data," Journal of Finance 49: 3-37.
Petersen, Mitchell A. and Raghuram Rajan, 1995, "The Effect of Credit Market
Competition on Lending Relationships," Quarterly Journal of Economics 110: 407443.
Reinhart, Carmen, Kenneth S. Rogoff, and Miguel A. Savastano, 2003, "Debt Intolerance,"
Brookings Papers on Economic Activity, 1 (Spring): 1-74.
Rochet, Jean-Charles and Xavier Vives, 2002, "Coordination Failures and the Lender of
Last Resort: Was Bagehot Right After All?" CEPR Discussion Paper no. 323.,
Tirole, Jean, 2002, Financial Crises, Liquidity and the International Financial System,
Princeton, NJ: Princeton University Press.
310
Table 1: Trends in International Bond and Bank Lending

Aggregate Value of Transactions
(US$ billions)
Number of Transactions
Year
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Total
Bonds
81
177
357
307
369
522
555
234
334
284
290
219
Loans
209
252
376
508
750
1,066
1,248
550
402
532
470
384
Total
290
429
733
815
1,119
1,588
1,803
784
736
816
760
603
Bonds
10
21
45
39
48
81
100
52
65
59
78
63
Loans
24
18
27
40
56
83
125
62
47
81
62
44
Total
34
39
73
79
104
164
225
114
113
141
140
107
3,729
6,747
10,476
661
669
1,331
311
Table 2: Number of Transactions, by Debt Category and IMF Program

Type of Credit
None
Bonds
Loans
Debt/GDP Range (0-30 percent)

No Program
IMF Program
389
1,301
1,244
57
2,606
99
None
Bonds
Loans

No Program
IMF Program
501
190
680
453
1375
240
None
Bonds
Loans
No Program
670
380
999
IMF Program
500
595
775
Debt/GDP Range (more than 60 percent)

None
Bonds
Loans
No Program
471
151
309
IMF Program
679
169
344
Full Sample
None
Bonds
Loans
No Program
2,949
2,455
5,289
IMF Program
1,758
1,274
1,458
312
Table 3: The Decision to Borrow and the Choice between Bonds and Loans
Lamount
Maturity
US Industrial Growth
Log of Swap Rate
EMBI volatility
Credit Rating
Debt/GDP
Debt Service/Exports
Real GDP growth
Export Volatility
Short-term/Total Debt
Reserves/Imports
Reserves/ST Debt
Private Credit/GDP
Public Issuer
Private Issuer
IMF Program
Precautionary
Turned Precautionary
Pseudo R-squared
Observations
(1)
1
(2)
Debt/GDP<=0.40
To
Bond
borrow or versus
loan
not to
borrow
0.103
[2.38]*
0.020
[2.48]*
2.242
-2.405
[1.29]
[0.69]
-0.062
-0.170
[2.10]*
[1.40]
-1.367
-3.757
[3.68]**
[4.67]**
0.005
-0.002
[3.00]**
[0.39]
0.332
-0.430
[1.44]
[1.51]
0.682
0.509
[5.17]**
[2.89]**
0.639
-5.670
[0.49]
[1.70]
-0.309
-0.663
[2.56]*
[1.90]
-0.163
-0.099
[1.17]
[0.43]
-0.011
0.035
[0.67]
[0.92]
-0.016
-0.029
[1.34]
[1.72]
0.071
0.071
[3.44]**
[2.19]*
0.218
-0.393
[4.90]**
[4.90]**
0.424
-0.457
[5.07]**
[6.80]**
0.027
0.290
[0.54]
[3.29]**
-0.069
-0.073
[1.08]
[0.32]
-0.007
-0.011
[0.06]
[0.13]
0.44
0.20
8505
6681
(3)
I
(4)
0.40<Debt/GDP<=0.60
To borrow Bond versus
loan
or not to
borrow
11.433
[2.07]*
-0.051
[0.83]
-1.478
[1.85]
0.016
[3.60]**
-1.094
[2.15]*
0.416
[2.21]*
3.174
[1.81]
-0.974
[3.20]**
0.331
[1.05]
0.009
[0.29]
-0.014
[1.44]
-0.044
[0.67]
0.104
[1.91]
0.312
[5.87]**
0.141
[3.45]**
-0.184
[1.97]*
0.032
[0.50]
0.33
3874
0.095
[5.78]**
0.021
[3.46]**
4.283
[0.82]
-0.023
[0.42]
-2.449
[2.46]*
0.002
[0.56]
-0.739
[1.57]
0.284
[2.70]**
1.330
[0.50]
-0.252
[1.33]
0.035
[0.14]
-0.027
[0.64]
-0.008
[0.54]
0.009
[0.23]
-0.316
[6.64]**
-0.514
[7.17]**
-0.024
[0.36]
0.131
[1.50]
0.153
[1.68]
0.24
2721
(5)
(6)
Debt/GDP>0.60
To borrow
Bond versus
or not to
loan
borrow
5.643
[0.87]
-0.135
[1.33]
-2.021
[1.82]
0.025
[5.31]**
-0.207
[0.70]
0.164
[0.44]
0.880
[0.26]
0.133
[0.71]
-0.387
[1.06]
0.043
[1.00]
-0.075
[2.07]*
-0.068
[1.02]
0.100
[1.42]
0.303
[3.66]**
0.084
[1.02]
-0.047
[0.23]
0.135
[0.72]
0.29
1976
0.130
[6.45]**
0.020
[3.99]**
23.516
[2.44]*
-0.303
[3.77]**
-0.984
[0.88]
-0.001
[0.28]
-0.113
[0.34]
-0.770
[3.38]**
2.058
[0.72]
0.011
[0.12]
-0.165
[0.57]
0.106
[3.81]**
-0.013
[0.36]
-0.097
[1.96]
-0.477
[5.24]**
-0.688
[4.44]**
-0.008
[0.14]
-0.061
[0.72]
-0.196
[4.44]**
0.42
965
The values reported represent the probability for an infinitesimal change in each independent, continuous
variable (at its mean) and the discrete change in the probability for dummy variables. Robust z statistics
(based on country clusters) in brackets, * significant at 5%; ** significant at 1%
313
Table 4: Pricing of Loans and Bonds

(1)
Debt/GDP range
Maturity
Log of Swap Rate
EMBI volatility
Credit Rating
Debt/GDP
Real GDP growth
Export Volatility
Reserves/Imports
Private Credit/GDP
Public Issuer
Private Issuer
IMF Program
Log of Repeat Borrowing
(3)
Loans
Medium
(4)
High
Low
Spread Equation
0.033
-0.084
-0.095
-0.105
[1.04]
[8.24]** [2.91]** [3.10]**
0.014
0.008
0.040
0.012
[2.98]**
[0.62]
[0.86]
[5.34]**
-6.521
-11.756
-8.075
-17.801
[1.23]
[0.92]
[0.63]
[2.08]*
0.246
0.263
0.005
0.258
[3.51]**
[0.06]
[4.24]**
[1.40]
-0.995
3.481
-0.202
-1.521
[0.60]
[1.63]
[0.12]
[2.73]**
-0.034
-0.032
-0.017
-0.022
[4.07]** [2.81]** [2.70]** [10.31]**
0.097
-0.821
-0.472
0.222
[0.50]
[0.84]
[1.22]
[0.19]
-10.008
-5.028
-6.479
-11.443
[3.05]**
[0.97]
[2.36]* [3.04]**
-0.218
0.137
-0.702
-0.336
[0-48]
[0.54]
[0.99]
[1.83]
0.267
-0.038
0.252
-0.214
[0.19]
[0.81]
[0.47]
[1.17]
0.018
-0.050
0.059
0.006
[0.67]
[0.75]
[0.98]
[0.25]
-0.037
0.047
0.033
-0.007
[0.84]
[0.55]
[0.16]
[0.78]
-0.095
0.197
-0.291
0.086
[0.98]
[0.58]
[0.95]
[0.42]
0.195
0.198
0.267
-0.162
[2.25]*
[0.87]
[0.63]
[0.37]
0.092
-0.041
-0.093
0.368
[1.57]
[1.12]
[0.27]
[3.38]**
-0.038
-0.149
-0.142
-0.139
[4.27]** [3.10]** [4.84]** [2.56]*
Low
Log of Amount
(2)
(5)
(6)
Bonds
Medium
High
0.000
[0.01]
0.008
[1.02]
-36.014
[3.06]**
0.263
[1.91]
7.180
[1.53]
-0.048
[5.00]**
0.675
[0.60]
-9.887
[2.99]**
0.678
[0.54]
-0.851
[1.20]
0.074
[1.85]
-0.060
[0.57]
0.247
[1.85]
0.520
[3.34]**
-0.392
[2.74]**
0.047
[1.54]
-0.001
[0.02]
0.012
[2.00]*
-3.713
[0.21]
-0.060
[0.32]
-0.303
[0.12]
-0.018
[0.92]
4.157
[2.60]**
-4.641
[1.53]
-0.161
[1.60]
0.574
[1.54]
0.038
[0.74]
-0.260
[2.23]*
0.090
[0.42]
0.599
[1.62]
-0.033
[0.34]
0.015
[0-45]
314
Table 4 (continued): Pricing of Loans and Bonds (selection equation)

(1)
(2)
(3)
Loans
Debt/GDP range
Log of Swap Rate
EMBI volatility
Credit Rating
Debt/GDP
Debt Service/Exports
Real GDP growth
Export Volatility
Reserves/Imports
Reserves/Short-term Debt
Private Credit/GDP
IMF Program
Public Issuer
Private Issuer
Lambda
No. of Transactions
Observations
Low
Medium
7.547
[2.40]*
-0.081
[1.23]
-0.467
[0.78]
0.011
[2.45]*
0.671
[1.42]
0.575
[3.35]**
2.389
[0.71]
-0.752
[2.16]*
-0.323
[1.23]
-0.018
[0.63]
-0.034
[1.91]
0.104
[1.89]
-0.077
[0.76]
0.591
[9.51]**
0.811
[11.43]**
-0.032
[0.35]
4278
14.170
[2.11]*
-0.141
[1.41]
-0.061
[0.05]
0.018
[3.09]**
-1.441
[2.17]*
0.277
[1.46]
4.716
[1.82]
-1.257
[3.08]**
0.380
[0.85]
-0.004
[0.12]
-0.025
[1.97]*
-0.017
[0.20]
0.115
[1.61]
0.414
[6.35]**
0.713
[9.77]**
0.054
[0.15]
1771
6389
3102
(5)
(4)
(6)
Bonds
High
Low
Selection Equation
2.577
2.772
[0.69]
[0.38]
-0.165
-0.066
[2.67]**
[0.80]
-1.810
-6.506
[5.81]**
[2.05]*
0.008
0.016
[2.17]*
[3.88]**
0.203
-0.056
[0.44]
[0.22]
0.157
1.543
[6.25]**
[0.47]
1.714
-0.281
[0.17]
[0.51]
-0.585
0.072
[2.14]*
[0.48]
-0.303
-0.141
[0.94]
[0.40]
-0.046
0.007
[1.01]
[0.22]
-0.044
-0.065
[1.74]
[1.76]
0.164
0.019
[5.88]**
[0.42]
0.168
0.090
[2.23]*
[1.23]
0.211
0.560
[1.76]
[8.60]**
0.365
0.670
[13.07]** [2.97]**
-0.044
0.081
[0.60]
[0.52]
2220
648
1783
4510
Medium
High
13.310
[1.58]
-0.080
[0.99]
-4.992
[3.92]**
0.017
[3.05]**
-1.470
[1.96]
0.643
[2.66]**
3.829
[1.21]
-1.097
[2.18]*
0.286
[0.79]
-0.009
[0.16]
-0.035
[1.89]
-0.016
[0.26]
0.132
[1.96]
-0.142
[1.76]
-0.010
[0.13]
-0.657
[3.20] **
899
13.009
[2.54]*
-0.240
[4.28]**
-1.425
[1.28]
0.015
[5.19]**
0.831
[2.62]**
-0.090
[0.55]
1.239
[0.50]
-0.008
[0.09]
-0.308
[1.87]
0.034
[1.34]
-0.056
[3.40]**
-0.066
[1.54]
0.041
[1.44]
-0.159
[2.50]*
-0.107
[1.38]
0.145
[0.44]
281
2351
1310
Robust z statistics, based on country clusters, in brackets, * significant at 5%; ** significant at 1%.
Note: Among issuer types, sovereign is the omitted category. The spreads equation also has dummy variables
for sector of issuer (e.g., manufacturing, services, finance) interacted with issuer type (public, private). Also
included are dummy variables for currency of issue and, for bond markets, a dummy variable for fixed rather
than a floating rate of interest. In the selection equation, the values reported represent the probability for an
infinitesimal change in each independent, continuous variable (at its mean) and the discrete change in the
probability for dummy variables.
315
Table 5A: Loans: Impact of IMF Programs and Repeat Borrowing
(1)
Debt Range
(%ofGDP)
10-30
IMF Program
Repeat Borrowing
IMF Program
(2)
20-40
(3)
30-50
Spread Equation
0.561
0.230
0.272
[3.06]** [2.25]* [3.53]**
-0.090
-0.058
-0.174
[4.89]** [3.05]** [2.77]**
Selection Equation
0.120
0.045
0.041
[0-51]
[1.30]
[0.48]
(4)
40-60
(5)
50-70
(6)
60-80
-0.041
[0.27]
-0.149
[3.10]**
-0.091
[0.70]
-0.159
[3.99]**
-0.081
[0.88]
-0.146
[5.15]**
0.115
[1.61]
0.162
[1.99]*
0.135
[1.55]
No. of Transactions
1908
2598
2426
1771
1355
571
Observations
2960
4066
3804
3102
2647
1471
Table SB: Bonds: Impact of IMF Programs and Repeat Borrowing
(i)
10-30
Debt Range
(%ofGDP)
IMF Program
Repeat Borrowing
IMF Program
(2)
20-40
(3)
30-50
Spread Equation
0.034
-0.000
-0.043
[0.26]
[0.01]
[0.55]
-0.067
-0.022
-0.004
[0.28]
[2.92]**
[1.50]
Selection Equation
0.250
0.153
0.221
[2.90]** [2.92]**
[1.78]
(4)
40-60
(5)
50-70
(6)
60-80
-0.392
[2.74]**
0.047
[1.54]
-0.252
[1.86]
0.067
[2.31]*
-0.023
[0.20]
0.013
[0.39]
0.132
[1.96]
0.045
[1.05]
0.048
[1.51]
No. of Transactions
789
1653
1539
899
580
272
Observations
1911
3227
3038
2351
1973
1212
Notes: Robust z statistics, based on country clusters, in brackets, * significant at 5%; ** significant at 1%.
Other variables included in these regressions are those listed in Table 4, including those referred to in the
footnote to that Table. In the selection equation, the values reported represent the probability for an
infinitesimal change in each independent, continuous variable (at its mean) and the discrete change in the
probability for dummy variables.
316
Table 6: Does the Amount of IMF Lending Matter?
Debt/GDP range
IMF Program
IMF*Debt/GDP
IMF Amount/Debt
IMF Program
IMF*Debt/GDP
IMF Amount/Debt
Observations
(1)
(2)
Loans
(3)
(4)
(5)
Bonds
(6)
Low
Medium
High
Low
Medium
High
-2.469
[2.04]*
4.354
[1.81]
-2.401
[0.99]
0.051
[1.76]
-1.052
[0.87]
1.393
[0.79]
1.846
[1.13]
0.018
[0.55]
0.941
[2.28]*
-2.991
[2.50]*
3.047
[1.34]
0.409
[1.04]
-0.689
[1.11]
1.205
[1.27]
899
2351
281
1483
0.689
1.623
[1.15]
[3.74]**
-3.818
-1.550
[0.86]
[3.98]**
2.941
6.922
[2.76]** [5.05]**
-0.143
-0.139
[4.48]** [3.27]**
-0.063
[0.20]
-0.106
[0.12]
0.386
[0.32]
4278
6389
0.141
[0.29]
-0.371
[0.38]
4.357
[3.03]**
1771
3102
Spread Equation
0.270
-0.122
[0.65]
[0.15]
-0.737
-0.039
[0.59]
[0.03]
1.343
1.801
[1.64]
[0.82]
-0.040
-0.146
[4.83]** [2.64]**
Spread Equation
0.226
-0.237
[0.83]
[0.48]
-0.210
1.245
[0.35]
[1.38]
0.033
0.376
[0.02]
[0.34]
648
1783
2220
4510
Robust z statistics, based on country clusters, in brackets, * significant at 5%; ** significant at 1%. Other
variables included in these regressions are those listed in Table 4, including those referred to in the footnote
to that Table. In the selection equation, the values reported represent the probability for an infinitesimal
change in each independent, continuous variable (at its mean) and the discrete change in the probability for
dummy variables.
317
Table 7A: Bank Loans: Is Precaution Valuable?
(1)
10-30
Debt Range
(%ofGDP)
IMF Program
Precautionary
Program
Program
Repeat Borrowing
0.587
[2.92]**
-0.372
[1.10]
-0.174
[4.89]**
(2)
(3)
30-50
20-40
Spread Equation
0.299
0.291
[2.95]**
[3.15]**
-0.125
-0.533
[0.85]
[2.12]*
-0.022
-0.075
[0.36]
[0.15]
-0.059
-0.091
[3.17]**
[2.72]**
(6)
60-80
(4)
40-60
(5)
50-70
0.086
[0.59]
-0.348
[2.02]*
-0.097
[0.77]
-0.164
[3.39]**
0.151
[0.94]
-0.470
[2.39]*
-0.264
[2.55]*
-0.186
[4.90]**
0.206
[2.07]*
-0.477
[2.79]**
-0.350
[2.44]*
-0.144
[5.23]**
Selection Equation
IMF Program
Precautionary
Program
Program
No. of Transactions
Observations
0.166
[1.88]
-0.341
[1.68]
0.093
[0.92]
-0.172
[1.16]
-0.117
[0.65]
0.086
[0.88]
-0.200
[1.30]
-0.055
[0.61]
0.196
[2.69]**
-0.272
[2.32]*
-0.015
[0.18]
0.250
[2.71]**
-0.310
[2.07]*
0.026
[0.18]
0.149
[2.16]*
-0.177
[1.16]
0.165
[0.84]
1908
2960
2598
4066
2426
3804
1771
3102
1355
2647
571
1471
318
Table 7B: Bonds: Is Precaution Valuable?
(1)
10-30
Debt Range
(%ofGDP)
IMF Program
Precautionary
Program
Program
Repeat Borrowing
0.053
[0.37]
-0.198
[0.96]
-0.068
[2.98]**
(2)
20-40
(3)
(4)
30-50
40-60
(5)
50-70
(6)
60-80
-0.282
[2.50]*
-0.140
[0.98]
-0.267
[1.60]
0.049
[1.74]
0.013
[0.08]
-0.283
[0.82]
-0.622
[2.01]*
0.065
[2.32]*
0.148
[1.36]
-0.372
[3.20]**
-0.331
[2.12]*
0.007
[0.24]
Spread Equation
0.014
0.043
[0.17]
[0.81]
-0.077
-0.153
[1.31]
[0.28]
-0.130
-0.131
[2.36]*
[2.33]*
-0.005
-0.023
[0.37]
[1.66]
Selection Equation
IMF Program
Precautionary
Program
Program
No. of Transactions
Observations
0.248
[2.63]**
-0.324
[2.93]**
0.238
[2.48]*
-0.290
[2.03]*
0.138
[1.52]
0.233
[2.88]**
-0.155
[1.75]
0.033
[0.32]
0.112
[1.99]*
-0.067
[0.62]
0.161
[1.25]
0.065
[0.94]
-0.115
[1.06]
0.064
[0.47]
0.065
[1.45]
-0.064
[0.65]
-0.003
[0.03]
789
1911
1653
3227
1539
3038
899
2351
580
1973
272
1212
dummy variables.
319
Table 8: Do Private Borrowers Benefit More than Public Borrowers From IMF
Programs?
A: Private Borrowers
(1)
Low
Debt/GDP range
IMF Program
IMF Program
Observations
(2)
Loans
Medium
(3)
High
(4)
(5)
(6)
Low
Bonds
Medium
High
-0.466
[2.85]**
-0.034
[1.33]
-0.031
[0.13]
-0.103
[1.50]
0.266
[2.51]*
-0.018
[0.65]
Spread Equation
-0.114
0.096
0.245
-0.116
[2.25]*
[1.19]
[1.62]
[0.56]
-0.179
-0.098
-0.179
-0.133
[4.14]** [2.93]** [7.08]** [4.52]**
Spread Equation
-0.044
0.130
0.200
0.091
[2.01]* [1.46]
[0.77]
[2.71]**
3315
1672
784
2109
890
452
B: Public Borrowers
Low
(2)
Loans
Medium
0.599
[5.12]**
-0.138
[4.38]**
-0.065
[0.50]
-0.109
[2.10]*
-0.053
[0.49]
0.054
[0.85]
3074
1430
(1)
Debt/GDP range
IMF Program
IMF Program
Observations
(3)
(4)
High
Low
Spread Equation
-0.038
0.118
[0.26]
[1.34]
-0.014
-0.075
[1.10]
[0.61]
Spread Equation
0.085
0.045
[1.19]
[1.35]
999
2401
(5)
Bonds
(6)
Medium
High
-0.153
[1.55]
0.099
[2.67]**
0.018
[0.22]
0.083
[4.82]**
0.075
[1.16]
0.072
[2.37]*
1461
858
dummy variables.
320
Data Appendix
Bond Characteristics
The bond dataset, obtained from Loanware and Bondware covers the period 1991 to 2002
and includes: (1) launch spreads over risk free rates (in basis points, where one basis point
is one-hundredth of a percentage point); (2) the amount of the issue (millions of US$); (3)
the maturity in years; (4) whether the borrower was a sovereign, other public sector entity,
or private debtor; (5) currency of issue; (6) whether the bond had a fixed or floating rate;
(7) borrower's industrial sector: manufacturing, financial services, utility or infrastructure,
other services, or government (where government, in this case, refers to subsovereign
entities and central banks, which could not be classified in the other four industrial
sectors).
Global Variables
(1) United States industrial production growth rate: average of month-month growth rate
over a quarter. (2) United States ten-year swap spread. (3) Emerging Market Bond Index:
standard deviation of difference in log of daily spreads.
Country Characteristics
Variable
(Billions)
Periodicity
Source
Series
Total external debt

(EDT)
Gross national
product (GNP,
current prices)
Gross domestic
product (GDPNC,
current prices)
Gross domestic
product (GDP90,
1990 prices)
Total debt service
(TDS)
Exports (XGS)
Exports (X)
Reserves (RESIMF)
Imports (IMP)
Domestic bank
credit (CLNTPVT)1
Short-term bank debt
(BISSHT)2
Total bank debt
(BISTOT)3
Credit rating
(CRTG)
US$
Annual
WEO
us$
Annual
WEO
NGDPD
National
Annual
WEO
NGDP
National
Annual
WEO
NGDPJl
US$
Annual
WEO
DS
us$
us$
us$
us$
Annual
Monthly
Quarterly
Quarterly
Quarterly
IPS
IPS
IPS
IPS
WEO
BX
National
us$
us$
semi-annual
BIS
semi-annual
BIS
Scale
semi-annual
Institutional
Investor
321
M#c|70 dzf
q#c| 11 dzf
q#c|71 dzf
q#c|32d_zf
Constructed Variables
Debt/GNP
EDT/GNP
Debt service/exports
TDS/XGS
GDP/growth
0.25*ln[GDP90 t/GDP90 {M}1
Standard deviation of export growth
Standard deviation of monthly growth rates of

exports (over six months)
Reserves/imports
RESIMF/IMP
Reserves/GNP
RESIMF/GNP
Reserves/short-term debt
RESIMF/BISSHT
Short-term debt/total debt
BISSHT/BISTOT
Domestic credit/GDP
CLM PVT/(GDPNC/4)
Sources: International Monetary Fund's World Economic Outlook (WEO) and International Financial
Statistics (IFS)'JMF program data from the IMF's Executive Board Documents and Staff Estimates; World
Bank's World Debt Tables (WDT) and Global Development Finance (GDF)\ Bank of International
Settlements' The Maturity, Sectoral, and Nationality Distribution of International Bank Lending. Credit
ratings were obtained from Institutional Investor's Country Credit Ratings. Missing data for some countries
was completed using the US State Department's Annual Country reports on Economic Policy and Trade
Practices (which are available on the internet from
http:www.state.gov/www/issues/economic/trade_reports/). U.S. industrial production was obtained from the
Federal Reserve and Swap rates and EMBI from Bloomberg.
1
Credit to private sector.

Cross-border bank claims in all currencies and local claims in nonlocal currencies of maturity up to and
including one year.
3
Total consolidated cross-border claims in all currencies and local claims in nonlocal currencies.
2
322
Session 5
Innovations in Private and Multilateral
Lending
Dollars, Debt, and the IFIs:

Dedollarizing Multilateral Lending*
Eduardo Levy-Yeyati
Business School
Universidad Torcuato Di Telia
*The author would like to thank Eduardo Fernandez Arias and Esteban Molfmo for fruitful discussions and
suggestions, and Daniel Chodos and Ramiro Blazquez for their excellent research assistance. This draft is
dated September 30, 2004.
Introduction
Financial dollarization (FD) has been placed increasingly at the forefront of the policy
debate in many emerging economies, driven by concerns about the currency imbalance and
the associated financial fragility that it introduces for the economy as a whole.1 As a result,
the center of the FD debate has moved from a generally passive stance aimed at
minimizing its negative implications, to a more proactive one oriented to offset the
incentives that favor dollarization while developing local currency substitutes.2
While the debate has tended to center on the propensity to save in a foreign
currency and the limitations to borrow internationally in the local currency, one of the most
important sources of FD in emerging economies is their dependence on lending from
international financial institutions (IFIs), which has been historically denominated in a
basket of hard currencies. It follows that any dedollarization strategy should in principle
encompass the particular issue of the denomination of multilateral lending, not only due to
its implications regarding the country's overall currency mismatch but also as a potential
ingredient conducive to the development of a market for local currency securities of long
duration. Indeed, the convenience of dedollarizing part of the lending granted by IFIs has
already been highlighted in recent proposals (see, e.g., Eichengreen and Hausmann 2002).
However, for a number of reasons, these ideas have been received with skepticism or
indifference by market practitioners and IFI staff.
The present paper redresses this issue, by identifying and discussing old and new
theoretical and practical arguments in favor and against IFI lending in local currency. In
particular, and in contrast with the existing proposals that stress the potential demand from
non-residents seeking a currency-diversified portfolio, this paper argues that any such
initiative should (and realistically can) rely, at least at an early stage, on the demand from
emerging market residents in search for local currency assets that minimize the volatility of
returns measured in the local consumption basket, but reluctant to take on sovereign risk.
In line with this view, the paper suggests a more limited approach to dedollarizing
multilateral lending that may overcome some of the obstacles inhibiting their practical
implementation while serving as a first step towards the more ambitious initiatives already
on the table.
Saving in the local currency faces two fundamental obstacles in developing
countries: high nominal volatility (that is, unpredictable inflation due to nominal shocks or,
most notably, to attempts to dilute the real value of local currency liabilities through
inflation), and high credit risk (that is, a high probability of default, including through the
violation of the terms and conditions of both public and private contracts under local
jurisdiction, or the imposition of confiscatory taxes on the stock of savings).
The first obstacle can be largely mitigated through the use of indexation, typically
to the CPI, which limits the incentives for debt dilution. The second obstacle is more
]
In what follows, following what has become standard in the dollarization literature, "dollar" and "foreign
currency", and "peso" and "local currency" are used interchangeably.
2
In this context, dedollarization is understood as a voluntary process, as opposed to a compulsory currency
conversion. For an overview of the financial dollarization and dedollarization debates, see the papers
presented in the lADB/World Bank Conference on Financial Dedollarization: Policy Options, December 1-2,
2003. at http://www.iadb.org/ros/DeDolarizacion/Agenda.htm.
327
difficult to tackle. Country risk encompasses not only the possibility of outright default by
a particular debtor but also a number of sovereign actions that negatively affect creditor
rights. Both the index and the terms and conditions of contracts can suffer unexpected and
undesired modifications, such as an involuntary currency conversion or debt restructuring,
due to the intervention of local institutions.
These risks, in turn, generate incentives to relocate savings in countries where
property rights are better defined and protected. Thus, even in the absence of nominal
instability (or despite the mitigating presence of indexation), residents may end up
dollarizing their savings simply because of the lack of local currency assets free from
country-specific credit risk. Under these conditions, there is a potential demand of
investment-grade securities in local currency that cannot be satisfied by non-investment
grade countries.
As a large part of the domestic pool of savings moves abroad as a result, the
country is forced to rely on foreign (and, in particular, multilateral) credit.3 IFIs can in
practice make up (at least partially) for the lack of domestic funds due to their greater
ability to enforce the contractual terms where private creditors fail, which enables them to
collect funds and issue loans at close to risk-free rates to countries facing high country risk
premiums. Thus, by means of their preferred creditor status, they are able to mitigate the
agency problems underscoring the high cost of capital in emerging economies, playing a
"risk transformation" role.4
In this light, the IFIs are natural candidates to launch the investment-grade local
currency market. By issuing debt in emerging market currencies to fund local currency
loans to emerging countries, they could dedollarize an important portion of the country's
external liabilities (converting existing IFI loans into the local currency while keeping a
balanced currency position), thereby providing the needed liquidity to start up these
missing markets.5
The main deterrent to advance with this type of initiative has been the untested
conjecture that the representative international investor would not be attracted by local
currency assets. However, the minimum liquidity needed to launch markets for investmentgrade local currency securities can be obtained from a latent demand for these securities
coming from the country's residents.
The FD literature has made, both analytically and empirically, a distinction
between residents and non-residents as potential demanders of local currency assets.
Analytically, instruments denominated in the local currency will look relatively more
attractive to risk-averse local savers (borrowers), as they mirror their stream of future
De la Torre and Schmukler (2003) discuss offshorization as a mechanism to cope with country-specific risk.
If offshorization could successfully protect from country risk, foreign borrowing could readily substitute for
the decline in domestic funds. However, the extent to which offshore claims are less exposed to countryspecific risk than onshore assets is not obvious, as witness the recent Argentine default.
4
IFIs also benefit from the implicit guarantee provided by their member countries, although the incidence of
these guarantees on the costs of their lending to emerging economies is difficult to assess in the absence of
default episodes.
5
Indeed, if we accept that financial dollarization is a source of financial fragility, the partial dedollarization of
their lending would, at best, reduce their exposure to sovereign risk. Needless to say, this argument as well as
much of the discussion below applies primarily to multilateral development banks, as opposed to institutions
like the IMF that are not funded in the market.
328
consumption (income) more closely.6 Empirically, there is evidence that shows that past
debt dedollarization processes have been largely driven by a deepening of the domestic
markets based on local demand.7
On the other hand, some observers have argued that, inasmuch as domestic
currency mismatches tend to cancel out, the negative consequences of FD are specifically
driven by the country's foreign currency position vis-a-vis non-residents, which would cast
doubt on the benefits of a strategy focused on resident demand.8 However, the evidence
indicates that currency mismatches between residents do not net out in the aggregate, and
can be as harmful as external liabilities in terms of the aggregate real exchange rate
exposure.9 Moreover, as this paper will argue, targeting the stock of foreign assets held by
residents (including local institutional investors) may lead to a substitution of local
currency domestic debt for foreign currency external debt, reducing the measured
aggregate position of the country.10
The case of pension funds is illuminating. By acquiring a credit risk-free asset
denominated in CPI units, fund managers would fulfill their role by ensuring a stable
stream of retirement benefits while avoiding the risk of confiscation. However, in order to
diversify credit risk, pension funds typically invest a fraction of their portfolio in dollarized
foreign assets. As a result, while the emerging country government borrows in dollars from
IFIs, a share of residents' retirement savings is invested abroad in dollarized investment
grade paper (such as that issued by IFIs to fund their own lending). It is immediate to see
how IFIs may intermediate these funds by selling to pension funds the bonds that finance
the country loans, and how this intermediation could be done in CPI units to the benefit of
both parties involved.
This paper argues that this type of arrangement is a natural first step to dedollarize
the external debt of developing countries. The advantages of the scheme are several. First,
it partially dedollarizes the liabilities of the country, voluntarily and with no cost for the
local investor (who acquires a risk-free asset in a unit of account that minimizes the
relevant volatility of future returns) nor for the IFIs (which manage to keep a balanced
currency position).11 Second, it starts up an international market in local CPIs that can be
6
This distinction was originally made by Thomas (1985) in a two-country setup and, more recently, in Ize
and Levy-Yeyati (2003).
7
Bordo et al. (2002), analyzing the evolution of debt denomination in four British Dominions (Canada,
Australia, New Zealand and South Africa), highlight that "the onset of World War I essentially closed the
London capital market, and the response was similar in all four Dominions. The gold convertibility of the
domestic currency was suspended (and not resumed until 1925) and governments raised funds domestically,
essentially creating a domestic bond market. Foreign capital (at least for sovereign debt) would never regain
to the same extent." Similarly, Claessens et al. (2003) find that the dollarization ratio of (domestic plus
external) government bonds is negatively related with the size of domestic financial markets. See also
Martinez and Werner (2002), Herrera and Valdez (2003), and Caballero et al. (2003), for the development of
local currency markets in Mexico, Chile and Australia, respectively.
8
See, e.g., Eichengreen at el. (2003).
9
For example, Berganza and Garcia Herrero (2004) find that the incidence of balance sheet effects on country
risk arising from domestic deposit dollarization are comparable to those related with external foreign
currency debt. Levy-Yeyati (2004) and Levy-Yeyati et al (2004) report similar results for banking fragility
and crisis propensity.
10
The stock of foreign assets held by residents is typically ignored while computing the aggregate currency
mismatch.
11
Strictly speaking, currency risk would still affect the financial income of the IFI. Arguably, this is a very
minor risk cost that can be priced in the loan.
329
used as a benchmark by domestic borrowers seeking to avoid the real exchange rate
exposure characteristic of foreign currency borrowing at a reasonable cost. Third, it could
eventually attract funds from non-residents willing to hold a speculative position in the
local currency without assuming excessive risk or, when and if exotic currency markets
develop, in search for currency diversification.
The plan of the paper is as follows. Section II, succinctly revisits the literature on
FD, highlighting the relative importance of IFI lending vis-a-vis other sources of financial
dedollarization. Section III extends the model in Ize and Levy Yeyati (2003) to derive
analytically the link between country risk, offshorization of residents' savings and FD, and
reports preliminary evidence in line with the analytic results. Section IV evaluates the
menu of options already being provided by IFIs to reduce currency mismatches, reviews
the arguments for and against previous proposals to dedollarize IFI lending, and elaborates
on alternative ways to pursue this goal. Section V concludes.
Definitions, Implications, and Measurement
In this paper, financial dollarization simply denotes the holding by residents (including the
public sector) of foreign currency-denominated financial assets and liabilities. The
phenomenon, which has received quite a lot of attention in the literature,12 has been
increasingly seen as a source of concern for a number of reasons, most notably the
incidence of the associated currency mismatch on output volatility and financial fragility.
These concerns have been validated by recent empirical work. Berganza et al.
(2003) find that the response of sovereign spreads to exchange rate changes increases with
the degree of external FD, while Berganza and Garcia Herrero (2004) show that this effect
is driven largely by exchange rate depreciations (in line with the balance sheet view) both
when dollarization is external (measured as external obligations over GDP) or domestic
(proxied by the ratio of dollarized onshore deposits over GDP). In turn, Domac and
Martinez Peria (2000) find the foreign-liabilities-to-assets ratio of local banks to be
positively correlated with the probability of a systemic banking crisis. Along these lines,
Levy Yeyati (2004) finds that the propensity to face a banking crisis after a depreciation of
the local currency increases with the degree of FD of domestic banking institutions.
Finally, the evidence suggests that FD also has important consequences for the real
economy, through its association with a higher propensity to suffer sharp capital account
reversals or "sudden stops" (Calvo et al., 2004) and slower and more volatile growth rates
(Levy Yeyati, 2004).
The definition and measurement of FD in its different varieties is still subject to
discussion (see Eichengreen et al. (2003) and Goldstein (2003)). While the literature has
tended to emphasize the country's foreign currency position vis-a-vis non-residents
(typically measured by its foreign currency-denominated external debt),13 the aggregation
argument underlying this distinction (namely, that the currency exposure of resident
12
Existing explanations of FD point at portfolio hedging considerations (Ize and Levy Yeyati, 2003), time
inconsistency problems related to the temptation to dilute peso obligations through inflation (Calvo and
Guidotti, 1989), the incidence of implicit debtor guarantees (Burnside et al., 2001), currency-blind financial
regulation (Broda and Levy Yeyati, 2003) and signaling problems (De la Torre et al., 2003), among others.
See De Nicolo et al. (2003) for a discussion and empirical testing of some of these hypotheses.
13
This focus on external debt implicitly presumes a link between bondholders' residence and debt jurisdiction
that is, at best, imperfect.
330
creditors and debtors should cancel out) ignores important aggregate effects. Even if a
financially dollarized economy is currency-balanced as a whole, it will likely be
imbalanced at a micro level, leading to capital flight, bank runs and massive bankruptcies
at the time of a real exchange rate adjustment, with important real consequences. Hence,
the significant effects of domestic FD found in the literature.
Moreover, a simple portfolio approach (as the one adopted in the next section)
suggests that the degree of domestic and external dollarization should be intimately related.
To the extent that the portfolio choice of resident savers determines the volume of peso
loanable funds in domestic markets, it will be correlated with the dependence on dollarized
foreign borrowing.
With this in mind, in this paper I look at both domestic and external sources of FD.
Domestic dollarization is captured by onshore dollar deposits, which, given the standard
prudential limits on banks' net currency position, provide a good proxy for the volume of
onshore dollar loans. External dollarization, in turn, is represented by private external loans
and holdings of external bonded debt, and by multilateral lending, within which I
distinguish IMF and non-IMF loans.14 Liability dollarization, in turn, is computed as the
ratio between total foreign currency liabilities (where onshore dollar loans are proxied by
onshore dollar deposits) over total liabilities (where, for consistency, onshore loans are
proxied by onshore deposits).
Table 1 provides a first glance at these different categories. For comparison, the
table includes emerging countries for which data on all categories are available (and
excludes offshore centers where FD is likely to be driven by factors of a different nature).
Two things are worth noting in the table. First, the degree of exposure has remained
relatively stable in recent years. Second, the median exposure to non-IMF IFI lending,
which has increased slightly, exhibits levels comparable to that of external loans and is
higher than that associated with domestic dollarization and external bonded debt (the focus
of much of the empirical FD literature). Based on median values, it accounts for more than
one-fourth of total external dollarization and one fifth of total dollarization. These numbers
by themselves indicate that a strategy aimed at reducing FD cannot ignore the role of IFIs.
Offshorization and Financial Dollarization
One aspect of the FD debate often overlooked by the literature is the interaction between
country risk and the degree dollarization for non-investment grade economies. Trivially, if
country risk drives domestic savings abroad where no local currency assets are available,
higher country risk would be, other things equal, associated with higher total dollarization
ratios (inclusive of offshore deposits), leading to a smaller volume of peso loanable funds.
In turn, this deficit would be partially compensated by a greater dependence on foreign
dollar borrowing (to the extent that it insulates investors from country risk better than
domestic assets) and, ultimately, on IFI lending. This section presents a stylized analytical
example to illustrate this intuition, and tests its empirical implications in the data.
The latter distinction is important. As already noted, the approach to IFI participation in the dedollarization
effort discussed in this paper does not apply to an institution like the IMF that is not funded in the market.
331
An Analytical Example
The link between country risk, offshorization and dollarization can be illustrated by a
simple example that extends portfolio approach used in Ize and Levy Yeyati (2003).
Consider the following scenario. A continuum of measure S of risk-averse resident
investors endowed with a unit of cash, can invest in four alternative assets: peso and dollar
debt issued in a non-investment grade emerging economy (the home economy) and peso
and dollar debt issued in an investment grade developed economy (the foreign economy).
Denoting the portfolio shares by
and
, where the superscripts F, H, CF
and CH refer to dollar and peso assets in the home and the foreign economies, respectively,
the real returns on each asset as measured by the resident investor would be given by:
(1)
where and //^, ns and //c are zero-mean disturbances to the local inflation rate, the real
(peso-dollar) exchange rate, and the home country's sovereign risk, assumed to be
distributed with variance-covariance matrix [5^], with Scs=Scjr=015 In turn, E(rj)
denotes the expected real return on the assets.
On the other hand, a continuum of measure L of risk-neutral local borrowers have
access to a production technology with known unit real returns (which, for simplicity, I
assume that exceed unit borrowing costs) that can be financed from three sources:
domestic peso and dollar debt, and foreign borrowing (whose share in the liability portfolio
is denoted by xc). The first two instruments are identical to those available to investors.
Offshore lending, on the other hand, can be interpreted at least in two ways: as foreign
lending by private investors (under the assumption that offshorization indeed provides
limited protection from country-specific risk), or by IFIs (under the assumption that the
preferred creditor status eliminates default risk). For the moment, it suffices to assume that
the access to this source entails an additional unit cost <fi(X), with (f>'(X) > 0, where X'=
xcL16
Then, interest rate arbitrage implies that
15
The qualitative results are not driven by these simplifying assumptions. In Appendix II, I present the
solution for the case in which country and real exchange rate risk are not independent.
16
If the source is private foreign investment, these costs would capture the fact that foreign contracts, while
perceived as providing better protection from confiscation risk than domestic contracts, are still subject to
sovereign default. Alternatively, if the source is IFI lending, the costs may reflect the upward sloping charge
scale typically applied to non-concessional multilateral loans, as well as costs associated with the process of
requesting the loan and complying with the attached conditionalities.
332
Assume further that investors maximize risk-adjusted real return measured in units
of the local consumption basket:
y
(2)
where
It can be shown (see Appendix II) that any solution to the portfolio problem can be
characterized by the following dollarization and offshorization ratios:
(3)
and
(4)
where
(5)
and
(6)
is the dollarization share in the absence of real return differentials (henceforth, the
underlying dollarization ratio). Note that, in this simplified setup where domestic interest
rate differentials are ruled out by arbitrage, A. = fa
Equations (3) and (4) characterize a continuum of portfolios that maximize the
investors' utility. While the availability of offshore peso assets has no incidence on the
dollarization ratio if y < AU9 it does so in the case in which y > Aw , where the excess of
offshore assets over the desired dollar assets can only be held in pesos (that is XCH > y/It,,). Therefore, if offshore peso assets were not available, dollarization would be driven
entirely by offshorization, as the latter could never be below the dollarization ratio.
More generally, it can be shown that, if the optimal ofFshorization ratio exceeds the
desired dollarization ratio, in the absence of country risk-free peso assets, the new
dollarization ratio (identified by the lower bar) is given by (see Appendix II):
(7)
333
The intuition is straightforward: if peso assets are not available abroad, capital
flight translates directly into an increase in the dollarization ratio.17 The difference A - A tt ,
which is increasing in country risk, is solely due to the absence of a country risk-free asset
in the local currency.
The previous analysis can be readily extended to the case of CPI-indexed domestic
assets, as perfect indexation could be expressed, in terms of this example, as fjn = 0.
Therefore, if domestic peso assets are indexed to the local CPI, the dollarization ratio from
(7) would be equal tc
and would be entirely driven by country risk. In other
words, country risk sets a floor to the extent to which a non-investment

reduce FD either through monetary policy (reducing inflation volatility
rate pass-through) or through CPI indexation.
How does the offshorization of domestic savings impact
composition of resident liabilities? To answer this question, first note
balance of funds requires that
grade country can

and the exchange
on the currency
that the domestic
(8)
and, for the peso market,

(9)
where vlB denotes the borrower's dollarization ratio.

If offshorization were not binding, an increase in country risk would not alter the
amount of available peso funds, and the resulting decline of domestic funds would be
perfectly offset by an increase in foreign borrowing xc.18
However, when the offshorization ratio is binding, the supply of peso funds is
automatically determined by the offshorization ratio as, from (8) and (9), we obtain:
00)
or
In this case, the borrower has effectively two options: domestic peso loans
(limited by the domestic supply of funds) and dollarized foreign borrowing. As country
risk mounts, the cost of the former relative to the latter increases, raising the liability
dollarization ratio (which is now due entirely to foreign borrowing). The associated
increase in peso interest rates offsets rH, on the other hand, partially offsets the
offshorization (and dollarization) of domestic peso savings as a response to higher country
risk.19
17
Note that this is also the dollarization ratio that would obtain should local dollar deposits be banned, leaving
offshorization as the only option to dollarize savings.
18
In the more general case of diminishing marginal returns to investment, to the extent that the demand for
loans respond to the higher financing costs due to the increase in country risk, the decline in total borrowing
could weaken the demand for peso loans and put downward pressure on the peso-dollar differential,
increasing asset dollarization A while and reducing liability dollarization AB.
19
It can be readily seen that
so tha
implies that dAB/dScc > 0.
334
It is immediate to see that the presence of country-risk free offshore assets restores
(9), as the borrower can now meet his financing needs by borrowing pesos abroad (at the
peso risk-free rate plus the transaction cost <p ), decoupling the choice of currency and
location. In particular, increases in x as a result of higher country risk need no longer have
an impact on AB.
In sum, high country risk is associated with a high offshorization ratio and, if the
latter is sufficiently large, with a smaller supply of peso loanable funds. In turn, to the
extent that foreign borrowing is relatively immune to country risk, higher country risk
would lead to a larger share of foreign borrowing and, in the absence of risk-free peso
assets, higher dollarization ratios.
Are Nonresidents Different?
An argument repeatedly made in the literature stresses that hedging considerations indicate
that resident investors are likely to exhibit smaller dollarization ratios than non-resident
investors.20 The previous example helps illustrate the point.
First, note that, using /4 /2e - //* , where ne denotes nominal exchange rate
shocks, underlying dollarization simplifies to
01)
the coefficient of a simple regression of the inflation rate on the nominal exchange rate,
that is, a crude measure of the exchange rate pass through.
Starting from (11), and exploiting the symmetry of this setup, it is easy to verify
that, in a stylized two-country world, the degree of underlying "pesification" (that is, the
share of foreign currency assets over total assets) of foreign residents would be equal to:
(12)
where e* denotes the dollar-peso exchange rate, and ;r*the rate of inflation in the foreign
country.
It follows that the underlying demand for peso assets from residents and nonresidents is highly asymmetric. On the one hand, non-resident demand for assets
denominated in emerging currencies is proportional to the pass-through coefficient of
changes in the exchange rate vis-a-vis the emerging currency, which is unlikely to be
statistically different from zero for developed economies (and for most emerging
economies). On the other, for any pair of countries with comparable pass-through
coefficients, any coefficient below 50% would imply that the demand for local currency
assets from residents should exceed that from non-residents. In both cases, the asymmetry
20
See Thomas (1985) for an early reference.

335
deepens as inflation volatility (and the pass-through coefficient) in the emerging economy
declines and, by extension, when the peso assets are indexed to the local inflation rate.
This example certainly oversimplifies the portfolio choice of the representative
resident and non-resident investors. In particular, it abstracts from cross-border transaction
costs that in practice introduces a source of investor heterogeneity that helps explain the
permanence of a captive pool of domestic dollar funds from small investors in the midst of
a sovereign debt crisis.
However, the exercise provides a valid intuition in relation to two points that are
critically important to assess the role of IFIs in the development of a market for peso
assets: i) the incidence that country risk may have (through the offshorization of domestic
savings) in determining FD in non-investment grade countries; and ii) the fact that local
currency assets are likely to look more appealing to resident investors than to foreigners,
particularly in those countries where inflation is relatively stable.
Offshorization and Dollarization in the Data
The previous analysis offers a number of empirical implications. First, in the absence of
risk-free instruments in exotic currencies, non-investment grade countries may see a
substantial portion of their domestic savings dollarized simply as a result of the flight of
capital to safer investments abroad. In turn, this capital flight, inasmuch as it reduces the
volume of domestic loanable funds, increases both financing costs and the country's
dependence on external borrowing, to the extent that the latter is perceived as less exposed
to country-specific credit risk.
Regarding the last point, some observers have argued that offshore assets (and, in
particular, external debt) are free from government interference with the laws governing
the financial contracts or, alternatively, with the local judiciary system in charge of
enforcing them.21 From this perspective, as country risk increases, we should see the
balance between domestic and external debt tilt towards the latter. However, while a
foreign jurisdiction may certainly protect the debt holder from direct government acts,
cases in which debt restructuring discriminates in favor of domestic assets are not unusual,
as shown by the differential treatment typically assigned to domestic bank deposits during
default episodes as well as recent cases of selective default.22
At any rate, inasmuch as offshorization remains only a partial protection towards
country risk, one would expect that a more limited access to domestic finance would make
non-investment grade countries more dependent on IFIs lending.23 This section explores
whether this intuition is consistent with the empirical evidence.
To do that, I distinguish between onshore dollarization ratio (computed as onshore
dollar deposits over total onshore deposits), and deposit dollarization (/b), computed as
the share of dollar on total deposits including resident deposits abroad. The second
variable, while less frequent in the literature, is nonetheless a more accurate measure of the
degree of dollarization of residents' portfolios as depicted in the analytical example, and
21
See, for example, De la Torre and Schmukler (2003).

The recent Argentina default is a clear example in which domestic creditors received a more benign
treatment.
23
Note that this does not refer to higher IMF assistance during crisis episodes, but rather to a more permanent
dependence on lending from multilateral development banks.
22
336
the one that more clearly reflects the incidence of both underlying dollarization and
offshorization, where the latter is measured as the ratio between offshore and total resident
deposits. The onshore supply of loanable funds is proxied by the ratio of onshore deposits
over GDP.24 I look at two sources of external dollarization, namely, non-official (private)
lending (which groups external loans and external bonded debt), and official lending
(where I distinguish between IMF and non-IMF lending). Liability dollarization, in turn, is
computed as the ratio between total foreign currency liabilities over total liabilities, where
currency composition of onshore loans is proxied by that of onshore deposits. Finally,
country risk is measured as the stripped spread of sovereign debt over comparable US
Treasuries, as captured in the EMBI Global Index compiled by J. P. Morgan.
Table 2 provides a first glance at the links between country risk, the location and
currency composition of resident savings, and the different sources of external
dollarization, by looking at the correlation of their period averages.25
The first things to note from the table are the association of country risk with a
smaller volume of onshore loanable funds and a greater offshorization ratio, on the one
hand, and the high and positive correlation between the latter and total deposit dollarization
(and, in turn, between deposit dollarization and country risk), on the other. Both findings
are consistent with the implications of the previous model.
Regarding the sources of external liability dollarization, the table reveals no clear
link between country risk and non-official external finance, suggesting that, while offshore
debt may provide some protection against country risk (hence, the weaker negative link
between these two variables), it does not offset the decline in domestic funds. This decline
is ultimately compensated by a larger dependence on IFI lending, as reflected in a larger
ratio over GDP as well as in a larger IFI-to-total external credit ratio.
These links are explored more in detail in Table 3. As the table shows, risky
countries are associated with fewer onshore deposits (columns 1-3) and greater deposit
offshorization (columns 4-6), even after controlling for onshore dollarization, and for the
presence of restrictions on dollar deposits that may potentially bias residents towards
offshore assets if capital flight were motivated by currency risk (interestingly, restrictions
appear to be positively related with onshore deposits and negatively related with
offshorization).
Thus, country risk appears to be an additional important determinant underlying a
weak demand for peso assets in non-investment grade countries. However, unlike in the
standard portfolio approach, in this case dollarization is simply a by-product of the lack of
investment-grade peso assets. This is confirmed by the fact that deposit dollarization, even
after controlling for country risk and underlying dollarization (computed from equation (6)
based on monthly inflation and real exchange rate data), is still positively associated with
the deposit offshorization ratio (columns 7 and 8). This result holds for the larger sample
obtained by dropping the country risk index (column 9), and in a dynamic setting with
country fixed effects (columns 10 and 11).
Table 4, in turn, shows that the offshorization ratio is associated with greater IFI
dependence. This positive link is verified both cross-section and over time, for either IMF
24
I chose onshore deposits instead of M2 or domestic credit to be consistent with the way the offshorization
ratio is computed. However, all three variables are highly correlated and yield virtually identical results.
25
Averaging periods vary by country, as they correspond to those for which the country risk measure is
available.
337
or non-IMF official lending. This contrasts with the lack of a significant link with other
sources of external credit reported in Table 5 (columns 1 and 2). Similarly, no link is found
when total external long-term liabilities are used as a proxy for foreign currency external
debt (columns 3 and 4).26 Finally, the table shows how liability dollarization is positively
correlated with deposit offshorization (columns 5-8), in line with a higher dependence on
IFI lending.
In sum, the evidence is consistent with the hypothesis that, in the absence of riskfree instruments in exotic currencies, non-investment grade countries may see a substantial
portion of their domestic savings dollarized simply as a result of the flight of capital to
safer investments abroad. In turn, this capital flight, inasmuch as it reduces the volume of
domestic loanable funds, increases the country's dependence on dollarized IFIs lending,
shifting the currency liability composition towards the foreign currency as a consequence.
Financial Dedollarization and the IFIs
As the previous discussion highlights, IFIs tend to substitute domestic sources of finance in
non-investment grade countries. Crucially, the role of IFIs in the context of a narrow
domestic market does not necessarily entail, as sometimes argued, a significant subsidy to
emerging economies. Indeed, recent work has revealed that the subsidy component in IMF
non-concessional lending to emerging economies is virtually null (Jeanne and Zettelmeyer
2001), as follows from the absence of default episodes in the past -a result that would also
apply to other IFIs blessed with a similar preferred creditor status.
Indeed, a key characteristic of IFIs is that, unlike private investors, and for reasons
that exceed the scope of this paper, they exhibit a surprisingly good repayment record.27
Thus, at the risk of oversimplifying, one can think of IFIs as contributing to a "sovereign
risk transformation." Specifically, they can be seen as matching the supply of private funds
in search of investment grade securities, and the demand of funds by non-investment grade
economies. By intermediating between the two, the IFIs exploit their superior enforcing
capabilities to channel these funds into lending that, through their intervention, becomes
virtually risk-free.
It is only natural, then, to exploit this advantage to foster the supply of local
currency funds that are lost due to sovereign risk considerations. This does not require the
extension of additional lending by the IFIs, but rather the issuance of investment grade
paper to meet the demand for risk-free local currency securities, and the use of the
proceeds to convert part of the outstanding stock of IFI loans so as to keep a balanced
currency position.
As noted in the introduction, while schemes along these lines have already been
proposed and have been the subject of discussion by economists and IFI staff in recent
years, little, if any, progress has been made so far in that direction. This section reviews the
existing facilities offered by IFIs to their clients to hedge their currency exposure, and the
alternative proposals related to the dedollarization of external liabilities. In particular, it
26
These data is available from the World Bank's GDF for a larger sample and a longer period. The implicit
assumption that all external debt issued by non-industrial countries is denominated in foreign currency seems
to be a reasonable approximation.
27
The reasons why IFIs can successfully enforce their preferred creditor status are certainly a fruitful research
topic that exceeds the scope of this paper.
338
describes a scheme oriented to meet the demand for local-currency investment-grade

securities by residents along the lines discussed in the previous section, addressing in the
process the main criticism faced by old and new initiatives of this type.
What's in the Menu?
While the concerns related with currency mismatches have been increasingly
acknowledged by IFIs, their supply of hedging instruments available is still rather limited.
The World Bank (WB), for example, offers the option to convert outstanding loan
obligations (or to request a swap of its foreign currency obligation) into local currency.
Since WB loans are fimded in the foreign currency, the transaction requires that the Bank
arranges a local-foreign currency swap with a third financial institution to transfer the
currency exposure.28
These local currency products are not without benefits, particularly for noninvestment grade clients that would be otherwise unable to access long-dated currency
swaps directly in the capital markets. However, they are typically limited in volume,
shorter than the loan they are intended to hedge, and granted on a case-by-case basis
subject to the existence of a liquid swap market.29
More importantly, rather than expanding the pool of local currency funds, they tap
on existing swap markets. Thus, while they are likely to benefit local borrowers through
longer duration and reduced transaction (e.g., collateral) costs, they are also likely to crowd
out the available supply of hedging instruments. At any rate, and possibly because of their
limited benefits, these relatively new products have not been in high demand.30
The WB has also launched a few issues in investment-grade exotic currencies.31
The modality is not uniform. For example, the February 2000, three-year euronote in
Mexican pesos was issued abroad and was largely placed among American investors, on
the back of strong external demand shortly after rating agencies announced that they were
considering an upgrade of the country's debt to investment grade. On the other hand, the
May 2000 Chilean CPI-indexed peso 5-year euronote was distributed mainly among
domestic institutional investors, who purchase about 75% of the total issue.
While these issues are not without positive spillovers for the development of local
currency markets, their value added in the context of a dedollarization agenda is
questionable, as the risk transformation role emphasized in this paper is bound to be less
valuable for economies that already enjoy investment-grade status. By contrast, the
analysis in the previous section suggests that the best use of the IFIs' advantage entails
external issues (to minimize the crowding out of available domestic funds) in non28
For details, see the brochure on Local Currency Financial Products posted on
vww.woridban^^
Currency swaps are also offered by the Inter-American Development
BanJk"(E>B)'
29
The emerging markets that, according to the World Bank, satisfied this condition by end-2003 included
Brazil, Chile, Colombia, the Czech Republic, Hungary, India, Indonesia, Mexico, Malaysia, Philippines,
Poland, the Slovak Republic, South Africa, South Korea and Thailand. Of these, only Colombia, India,
Indonesia and the Philippines are non-investment grade countries.
30
As of April 2004, only three countries had signed the Master Derivatives Agreement required by the WB to
request a currency swap.
31
A list of recent WB issues can be found at http://www.worldbank.org/debtsecurities/recent issues.htm.
339
investment grade countries (unable to attract domestic investors in search of low-risk

assets).
A move in this direction was the March 2004 Colombian CPI-indexed bond, issued
and placed domestically by the WB within domestic institutional investors. While it still
has the potential to crowd out existing (captive) demand for peso assets, the bond was
nonetheless welcome by the government as a way to satisfy the appetite for risk-free long
assets in the local currency from the growing private pension system, which would
otherwise have to be met by foreign assets.32 Closer to the scheme proposed here was the
May 11, 2004 eurobond in Brazilian reais issued by the Inter-American Development Bank
(IDB), which included selling restrictions in Brazil to avoid crowding out domestic
resources?*
These issues certainly reflect a welcome shift in the funding strategies of some
IFIs. However, contrary to what one would be led to believe, they have been entirely
motivated by the search of lower funding costs. Indeed, rather than used to convert
outstanding loans into the same exotic currencies, their proceeds have been immediately
swapped into dollars. Thus, for all the merit that these efforts may have, their effective
impact in terms of dedollarizing the liabilities of emerging economies has been virtually
null.
What Has Been Proposed?
Most of the discussion about the type of peso instrument best fit to substitute current dollar
assets while maximizing its hedging potential has centered around CPI indexation, an
avenue that proved to be successful in containing and undoing FD in Chile and Israel,
particularly when it comes to longer financial contracts. The local CPI, the most obvious
candidate index for domestic residents, has been confronted with several alternatives in the
same spirit, particularly when targeting foreign investors.
In general, indexation of dollar-denominated instruments to a price closely
correlated with the debtor's income could in principle attain what could be labeled
synthetic dedollarization, decoupling the real cash flows of the asset (measured in units of
the debtor's income) from the evolution of the real exchange rate. While in practice these
instruments may be more opaque for the average investor than a plain CPI-indexed local
currency bond (and, in turn, more difficult to market), they may be free from moral hazard
and thus potentially attractive for sophisticated investors. Crucial in this regard is the
exogeneity of the index of choice.34
Among the latter, two alternatives stand out: a GDP index, and a commodity
index.35 Both are similar in nature, being equivalent to a plain vanilla bond plus a short
position in the commodity or the issuing country's GDP, and both are subject to the
country's sovereign risk. While commodities are more easily priced and hedged due to the
32
The same would apply to domestic issues in investment-grade Chile, where a sustained fiscal surplus leaves
little room for the issuance of long-dated sovereign paper.
33
This was just the second international issue in a Latin American currency by an IFI. The first one, in April
2004, was a global bond denominated in Mexican pesos. To my knowledge, no other multilateral
development bank has issued debt in non-investment grade currencies.
34
Indexation to a tax revenue index, for example, offers no such advantage.
35
See Borensztein and Mauro (2002) on GDP-indexed bonds, and Caballero and Panageas (2003) on copperindexed debt for the case of Chile.
340
existence of derivative markets, their use is bound to be limited to commodity exporters,

and to the extent these exports correlate with the country's income. Moreover, much in the
same way as for the currency swap discussed above, it is not clear how such indexation
improves upon a short hedge purchased directly by the issuer in the derivative markets
(although access to these markets may be more costly for non-investment grade issuers).
On the other hand, while GDP indexation may be more suitable to smooth out
countercyclical variations in debt-to-GDP ratios and borrowing costs, it is difficult to see
how GDP risk can be stripped and hedged by potential investors, particularly in the
absence of a market for GDP indexes.36
The same caveats apply in principle to CPI-indexation as a way of luring foreign
investors. Eichengreen and Hausmann (2002) stress the attractiveness of a basket of CPIindexed exotic currencies for non-residents. Moreover, they specifically propose that IFIs
issue debt in these currencies to fund their own lending to emerging economies and
provide the needed liquidity for the index. This requires matching not only the demand and
supply of funds in each currency but also across currencies to allow for the needed
diversification strategy. As such, it involves a non-trivial coordination effort. Furthermore,
while speculative non-resident demand for specific currencies perceived as undervalued is
not unlikely (as the IDE issue in Brazilian reais attests), interest from long-run
international investors seeking a diversified portfolio with stable returns is more difficult to
envisage.37
IFIs and the Intermediation of Resident Savings
Once we shift the focus away from international investors to target the demand from
residents, the use of CPI indexation presents important advantages, including the fact that
it can be measured at daily frequencies (improving the accuracy of the indexation) by an
autonomous agency (ensuring that the index is free from government manipulation). More
importantly, unlike other indexes, the CPI enjoys a demand arising from the hedging
properties highlighted in the previous section. As such, it is a natural choice to jump start
the dedollarization process with the help of an investment grade issuer (the IFIs) that
decouples sovereign and currency risk, to attract domestic investors willing to invest in
their own currency at a reasonable level of credit risk.38
Resorting directly to the domestic market, however, may have economic (and
political) drawbacks, as it crowds out already available local-currency funds by inducing a
shift from high-risk government and corporate domestic debt to investment-grade IFI
paper. In that case, while the new issue may contribute to extend the market for local
36
To my knowledge, among emerging economies, only Bulgaria has issued a GDP-indexed bond (albeit with
a call clause that eliminates the upside from indexation). The bond was originally placed among institutional
investors and has hardly traded since. At the time of this writing, September 30, the Argentine government is
considering the use of GDP indexation in its forthcoming debt exchange offer.
37
Interestingly, Borensztein and Mauro also highlight the appeal of GDP-indexed bonds for a diversified
international investor. To their credit, emerging market debt as an asset class may have looked as distant
prior to the Brady plan as GDP-indexed or CPI-indexed bonds look today.
38
Risk decoupling is at the heart of the Eichengreen-Hausmann proposal. However, in that context, currency
risk is tolerated to the extent that it can be diversified away in a basket of exotic currencies. In the current
version, by contrast, CPI indexation eliminates currency risk from the resident's standpoint, so that no
currency diversification is required.
341
currency securities onshore by bringing in new investors previously reluctant to assume

country risk, it is likely to increase the cost of funds domestically, inducing the
government to borrow abroad, with only a minor change in the overall composition of
government liabilities. Thus, in order to maximize the beneficial composition effect, the
new debt should be issued in international markets, to cater to investors that seek stable
returns in units of the local consumption basket, but are reluctant to take on sovereign risk.
Both the literature and the recent experiences point at institutional investors as the
natural target of the first issues. Consider, for example, the case of pension funds. The
advantages of the CPI as the benchmark unit of account are apparent: By acquiring a credit
risk-free asset denominated in units of the consumption basket, they fulfill their role as
guarantors of a stable stream of income after retirement while avoiding country-specific
credit risk.39 Indeed, pension funds are typically allowed to invest a fraction of their
portfolio in investment-grade foreign assets (see Table 6), a fraction that has been
growing, particularly since the Argentine debacle sounded the alarm on excessive exposure
to sovereign risk. Thus, while the government borrows from IFIs, a share of residents'
retirement savings is being invested abroad in triple A paper such as that issued by IFIs to
fund their loans. IFIs can readily channel these fimds back into the domestic economy by
selling to the pension funds CPI-indexed bonds to finance loans denominated in the same
index.40
The demand for a long-dated investment-grade local-currency paper that may come
from institutional investors, as well as resident savings abroad, may reach important levels,
as shown in Table 7, which compares stocks of pension funds assets, resident deposits
abroad, and outstanding IFI loans, for emerging economies that have recently privatized
their social security system.41
As noted, a few successful IFI issues in exotic currencies already revealed the
existence of a demand for these securities. In this context, the redollarization of their
proceeds is particularly puzzling, and at odds with the concerns about FD repeatedly
endorsed by IFI officials and publications. Given the already high exposure of these
institutions with many of their clients, it is easy to see how the swap with a third financial
institution that followed the issuance of these bonds could have been done, alternatively,
directly with the client, partially dedollarizing outstanding obligations. While the cash
flows of the bond would typically be different from that of the loan, the swap markets
provide sufficient flexibility to match both schedules with little, if any, additional
transaction costs. On the other hand, the settlement currency of both streams of cash flows
would be immaterial in this case. In particular, even if the currency of denomination of the
original loan is preserved, his obligation would be indexed to the local currency (or the
local CPI), eliminating any currency exposure -an argument also valid for new lending.
39
Pension fund regulation should acknowledge this explicitly in order to align the incentives of pension fund
managers along these lines. See Caniquiry and Gruss (2004) for a discussion of the case of Uruguay.
40
An alternative approach that has been the subject of informal discussion among IFI staff is the use of IFI
guarantees to local currency debt to reduce the credit risk of non-investment grade issues in exotic
currencies. Halfway between a risk-free IFI bond and risky emerging market paper, this combination (if
guarantees are capped in dollars) would entail for the IFIs similar risks as those associated with existing
guarantees to dollar bonds.
41
The previous analyses does not deny the existence of non-resident demand for exotic currencies. However,
while anecdotal evidence indicates that this demand do exist, it is likely to be driven by short-term
speculative appreciation games rather than by the long-run diversified investors needed to build the market.
342
Moreover, and for the same reason, there is no obvious rationale to limit the currency
conversion to the local expenditure component of the loan (as is currently the case for
existing local currency products). Indeed, an appropriate hedging strategy would need to
match the currency composition of liabilities with that of future earnings (as opposed to
past expenses). In sum, there seems to be no obvious obstacle to onlend the funds obtained
from local currency issues to emerging market clients.
Addressing the Skeptics
Besides the mixed reviews received by markets participants, the proposals to dedollarize
IFI lending have faced criticism from within IFI circles. The present analysis would not be
complete without a brief discussion of some of them.
Raj an (2004) summarizes two of the main arguments. First, he points out that a
portfolio approach to FD should take into account the correlation between financial returns
and non-financial income. More precisely, to the extent that economic activity is
negatively correlated the real exchange rate, local savers would demand lower returns on
dollar assets that are used as a hedge against economic downturns. In principle, however,
this preference should not induce FD, as local debtors would be willing to pay the higher
returns demanded on peso assets to hedge their income stream.
The second argument is more relevant to our discussion: In the presence of myopic
behavior, one would expect emerging market borrowers to exploit the lower dollar
borrowing costs in good times, disregarding the contingent cost of the associated exposure
-likely to be borne by others.42 Note that, while the peso interest rate charged by IFIs
would be below that demanded by private lenders (due to the lower credit risk), the
conversion of outstanding IFI loans to the local currency would not save debtors the
currency risk premium that induced dollarization in the first place. In other words, if FD
were the result of asymmetric risk pricing, rather than lack of investment grade local
currency assets as argued here, opportunistic debtors would turn down the offer to insure
against future balance sheet effects at a fair price. If so, the proposed dedollarization
strategy, rather than suffering from the lack of investor interest, may be condemned by the
indifference of the very debtors that it is intended to relieve.
This agency argument looks a bit overdone in light of recent dedollarization efforts
in emerging economies.43 Nonetheless, taking the argument at face value, one can only
conclude that it would be in the interest of the IFIs to correct this imperfection by
including dedollarization within the standard conditionality set, rather than offer
misleadingly cheap dollar lending to perpetuate this perverse cycle. Ultimately, agency
problems provide yet another reason for IFIs to adopt a more proactive stance.
42
Variations on this argument have been examined in the literature in relation to market imperfections such as
implicit guarantees (Bumside at el., 2001), or currency-blind regulation (Broda and Levy Yeyati, 2003) that
are conducive to excessive dollarization.
43
Examples include, among others, the gradual dedollarization of public debt in post-Tequila Mexico and,
more recently, Brazil; the revision of the prudential framework as well as the introduction of CPI-indexed
assets in Uruguay; and the imposition of quantitative restrictions on the on-lending of onshore dollar deposits
in Argentina after the demise of the currency board.
343
Final Remarks
This paper tried to convey a simple message: to the extent that country risk induces
financial dollarization through the offshorization of domestic savings -as the analysis and
the evidence presented here seem to indicate-, IFIs can exploit their superior enforcement
ability to intermediate these savings back into the domestic economy without increasing
financial dollarization.
For IFIs, this would not require expanding credit, transferring resources or
incurring currency risk. Rather, it would involve issuing local currency bonds and using
the proceeds to gradually convert current loans into (or refinance maturing loans in) the
local currency. Far from a final solution to the dollarization problem, this initiative
represents a feasible starting point for the much needed development of local currency
markets. While successful issues of IFI debt in exotic currencies are an encouraging first
step, a coordinated effort is still needed to convince governments and the IFIs of the
benefits of using the proceeds to dedollarization multilateral lending.
This paper did not argue that the scheme described above is a sufficient condition
to reduce FD in emerging economies. Needless to say, the demand for local currency assets
(and, more generally, the achievement of financial stability) would be contingent on the
implementation of responsible economic policies consistently over time. However, while
good policies are by definition a good advice, they are not always sufficient to collect the
fiill reward. It is along that margin where the IFIs can make a contribution.
344
References
Aghion P., Bolton P. and Dewatripont, M., 2000. Contagious Bank Failures in a Free
Banking System, European Economic Review.
Berganza, J. C.,.R. Chang and A. Garcia Herrero (2003). Working Paper 306, Bank of
Spain.
Berganza, J. C. and A. Garcia Herrero (2004). What Makes Balance Sheet Effects
Detrimental for the Country Risk Premium? Mimeo. Bank of Spain.
Bordo, M., Ch. Meissner and A. Redish (2002). How "Original Sin" was Overcome: The
Evolution of External Debt Denominated in Domestic Currencies in the United
States and the British Dominions 1800-2000. Mimeo.
Borensztein, E. and P. Mauro (2002). Reviving the Case for GDP-Indexed Bonds. IMF
Policy Discussion Paper 02/10.
Broda, Ch. and E. Levy-Yeyati (2003). Endogenous Deposit Dollarization. Federal
Reserve Bank of New York Staff Papers 160.
Burnside, C., M. Eichenbaum, S. Rebelo, (2001). Hedging and Financial Fragility in Fixed
Exchange Rate Regimes. European Economic Review 45 (7).
Caballero, R. and S. Panageas (2003). Hedging Sudden Stops and Precautionary
Recessions: A Quantitative Approach. Mimeo, MIT.
Calvo, Guillermo and Pablo Guidotti (1989). Credibility and nominal debt: exploring the
role of maturity in managing inflation. IMF Working Paper 89/73.
Calvo, G., A. Izquierdo and L. F. Mejia (2004). On the Empirics of Sudden Stops: The
Relevance of Balance-Sheet Effects. Mimeo, Inter-American Development Bank.
Carriquiry, F. and B. Gruss (2004). Dolarizacion y Fragilidad Financiera del Sistema de
Pensiones: El Caso de Uruguay, forthcoming, Revista de Economia, Central Bank
of Uruguay.
Cespedes, L., R. Chang and A. Velasco (2000). Balance Sheet Effects and Exchange Rate
Policy. NBER Working Paper 7840.
Claessens, S., D. Klingebiel and S. Schmukler (2003). Government Bonds in Domestic and
Foreign Currency: The Role of Macroeconomics and Institutional Factors. Mimeo,
The World Bank.
De la Torre, A., E. Levy-Yeyati and S. Schmukler (2003). Beyond the Bipolar View: The
Rise and Fall of Argentina's Currency Broad. Economia, vol. 5 (2).
De la Torre, A. and S. Schmukler (2003). Financial Contracting in Emerging Economies.
Mimeo, The World Bank.
DeNicolo, G., P. Honohan, and A. Ize (2003). Dollarization of the Banking System: Good
or Bad? Mimeo, IMF.
Domac, I. and M. S. Martinez Peria (2000). Banking Crises and Exchange Rate Regimes:
Is There a Link? Journal of International Economics, forthcoming.
Eichengreen, B. and R. Hausmann (2002). Original Sin: the Road to Redemption,
forthcoming in Eichengreen and R. Hausmann (eds.) Debt Denomination and
Financial Instability in Emerging-Market Economies, Chicago: University of
Chicago Press.
Eichengreen, B., R. Hausmann, and U. Panizza (2003). Currency Mismatches, Debt
Intolerance and Original Sin: Why They are Not the Same and Why it Matters,
NBER Working Paper No. 10036.
345
Goldstein, M, and P. Turner (2003). Controlling for Currency Mismatches in Emerging

Economies. Mimeo, Institute for International Economics.
Herrera, L. O. and R. Valdez (2003). Dedollarization, Indexation and Nominalization:
The Chilean Experience. Mimeo, Central Bank of Chile.
Ize, A. and E. Levy-Yeyati (2003). Financial Dollarization. Journal of International
Economics, 59.
Jeanne, O. and J. Zettelmeyer (2001) "International Bailouts, Moral Hazard,
and Conditionally," Economic Policy 16, issue No 33, October.
Levy-Yeyati, E. (2004). Financial Dollarization: Evaluating the Consequences. Mimeo,
http://www.utdt.edu/-elv.
Levy-Yeyati, E., M. S. Martinez Peria, and S. Schmukler (2004). Market Discipline under
Systemic Risk: Evidence from Bank Runs in Emerging Economies. Mimeo. The
World Bank.
Martinez, L. and A. Werner (2002). Capital Markets in Mexico: Recent Developments and
Future Challenges. Mimeo, Central Bank of Mexico.
Rajan, R. (2004). How Useful Are Clever Solutions? Finance & Development, March 2004.
Thomas, L.R. (1985). Portfolio Theory and Currency Substitution. Journal of Money,
Credit, and Banking, Vol. 17.
346
Table 1. Sources of Financial Dollarization

(nonindustrial economies excluding offshore centers; as percent of GDP)
Onshore
dollar
Deposits
(a)
Mean
Median
1996 Min
Max
Obs.
Mean
Median
2001 Min
Max
Obs.
0.0825
0.0625
0
0.3390
30
0.1197
0.0823
0.0003
0.5101
30
External
loans
(b)
0.1427
0.1297
0.0271
0.5295
30
0.1286
0.1207
0.0359
0.2484
30
Dollar
bonded
external
debt
(c)
0.0406
0.0272
0.0014
0.2937
30
0.0908
0.0583
0.0019
0.3251
30
IFI
lending
IMF
lending
(d)
(e)
0.2238
0.0872
0
2.4379
30
0.1838
0.0964
0.0011
1.973382
30
0.0130
0.0055
0
0.0591
30
0.0183
0.0028
0
0.0987
30
Total external
Total
(b)+(c)+(d)+(e)
0.4201
0. 3323
0.09118
2.6977
30
0.4214
0.3479
0.1783
2.2831
30
0.5027
0.4326
0.1185
2.9492
30
0.5411
0.4422
0.1814
2.7932
30
Notes: Countries in the sample: Argentina, Bulgaria, Chile, Costa Rica, Czech Republic, Dominican
Republic, Egypt, Estonia, Guatemala, Croatia, Hungary, Indonesia, Jamaica, Kazakhstan, Lithuania,
Latvia, Moldova, Mexico, Malaysia, Nicaragua, Peru, Philippines, Poland, Romania, Slovak Republic,
Thailand, Turkey, Uruguay, Venezuela and South Africa.
347
Table!
Measures of Financial Dollarization - Correlation Matrix
(non-industrial economies excluding offshore centers; period averages)
Onshore
deposits
Deposit
Offshoriz.
Deposit
dollariz.
Dollar external
liabilities (exc. IFIs)
-0.5754
(0.0011)
29
-0.4163
(0.0540)
22
0.4209
(0.0455)
23
-0.2383
(0.2313)
27
-0.2714
(0.1709)
27
0.6157
(0.0023)
22
-0.1966
(0.3687)
23
0.3060
(0.1284)
26
0.1701
(0.4062)
26
-0.1390
(0.5588)
20
0.2052
(0.3722)
21
0.2676
(0.2408)
21
-0.4280
(0.0469)
22
-0.1778
(0.4285)
22
0.2520
(0.2048)
27
IFI lending / Total ext. liabilities
-0.1388
0.4900
27
0.0674
0.7437
26
0.2008
0.3828
21
-0.5304
0.0111
22
0.7951
0.0000
27
0.1568
0.4347
27
Country risk
-0.5646
(0.0012)
30
0.3678
(0.0496)
29
0.5047
(0.0166)
22
-0.1886
(0.3774)
24
0.5633
(0.0022)
27
0.5073
(0.0069)
27
Deposit offshorization
Deposit dollarization
Dollar external liabilities (exc. FIs)
IFI lending (exc. IMF)
IMF lending
XTM*
IFI lending/
Total ext.
liabilities
0.3896
(0.0445)
27
Note: Significance levels in parentheses. Number of observations in italics. Averages computed based on observations for which the country risk index is available. All
variables computed over GDP (with the exception of the ratio of IFI lending to total external liabilities and the country risk index).
Table 3
Country Risk, Offshorization and Deposit Dollarization

(nonindustrial economies excluding offshore centers)
Onshore deposits over GDP
OLS (period averages)
Country risk
(1)
(2)
(3)
-0.030***
(0.007)
-0.039***
(0.008)
-0.031***
(0.008)
Onshore dep. doll, ratio
Deposit offshorization ratio

(4)
(5)
(6)
0.021***
(0.008)
0.089
(0.164)
0.026*
(0.013)
0.018*
(0.009)
0.006
(0.005)
0.426***
(0.089)
0.413***
(0.066)
0.274***
(0.054)
0.455***
(0.116)
0.334***
(0.072)
0.514***
(0.069)
0.116***
(0.029)
-0.045*
(0.023)
Underlying doll, ratio

Dep. offshorization ratio
Constant
Observations
R-squared
0.560***
(0.069)
0.588***
(0.092)
30
23
0.32
0.39
0.004**
(0.002)
-0.020
(0.221)
0.045
(0.047)
Restrictions
0.020*
(0.010)
Deposit dollarization ratio

FE (annual data)
(10)
(7)
(9)
(8)
(11)
0.521***
(0.072)
24
0.35
0.272***
(0.064)
0.240***
(0.069)
0.298***
(0.069)
0.261***
(0.054)
0.154**
(0.054)
0.303***
(0.044)
0.282***
(0.027)
0.423***
(0.078)
29
22
0.20
26
0.19
21
0.68
21
0.83
78
0.52
107
0.98
584
0.96
0.14
Notes: Robust standard errors in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1%. FE regressions include year dummies.
Table 4
Offshorization, IFI Lending and IMF Lending
IFI lending over GDP (exc.
IMF)
FE (annual
OLS
data)
(averages)
1
Deposit
offshorization
(2)
(3)
(4)
(5)1
(6)
0.2
53
o***
0.56
0.050
0.147
**
0 017
0.036*
**
0.005*
*
0.047*
*
(0.14
1)
(0.02
5)
(0.06
0)
(0.020)
(0.012)
(0.002)
(0.024)
**
0.008
***
(0.00
2
)
Country risk
Observations
R-squared
FE (annual
data)
(8)
(7)
OLS (averages)
a)
(0-1
61)
Constant
IMF lending over GDP
(0.06
9)
0.504
***
(0.01
9)
0.099
***
(0.02
4)
120
815
0.08
0.96
0.24
0.0
94
(0.0
69)
o***
26
0.0
9
0.001
(0.001)
0.010
0.011*
*
0.027*
**
-0.004
(0.009)
(0.006)
(0.002)
(0.013)
125
26
120
816
125
0.98
0.03
0.07
0.92
0.82
Robust standard errors in parentheses.

* significant at 10%; ** significant at 5%; *** significant at 1%
1
Includes observations for which the country risk index is available.
350
Tables
Offshorization and Liability Dollarization
Dollar external
liabilities
over GDP
(exc. IFIs)
OLS
(avgs.)
(2)
(1)
Deposit
offshorization
Constant
Observations
R-squared
PT?
(annual
)
Total external
liabilities
over GDP
(exc. IFIs)
in?
OLS
(annual
(avgs.)
)
(4)
(3)
-0.108
0.015
-0.021
0.022
(0.099)
(0.013)
(0.030)
(0.014)
0.260**
*
0.165**
*
0.139**
*
o.ioi**
(0.048)
(0.027)
(0.019)
(0.009)
23
0.04
301
0.75
120
0.00
815
0.84
Liability
dollarization ratio
(exc. IMF)
OLS (averages)
(5)
(6)1
Liability
dollarization ratio
OLS (averages) '

(7)
0.154
**
(0.06
3)
0132
***
(0.04
2)
0.156
**
(0.06
3)
0.422
***
(0.03
4)
0.467
***
(0.02
6)
0.425
***
(0.03
4)
38
0.14
88
0.15
38
0.14
(8)
0.131***
(0.034)
0.472***
(0.022)
88
0.15
Notes: Robust standard errors in parentheses. * significant at 10%; ** significant at 5%; *** significant at
1%.
'Uses total long-term external debt as a proxy for dollar long-term external debt.
351
Table 6.
Pension Fund Investments in Foreign Assets
(emerging economies with private social security systems)
Legal framework (*)
Foreign assets
share** *J
Argentina
Up to 10% of the Fund's total asset value.
9.04
Chile
Up to 20% of the Fund's asset value.
23.89
Mexico
Although the SIEFORES Law determines that total investment in

instruments denominated in foreign currencies (U.S. Dollars, Euros,
Yens) must not exceed the 10% of the Fund's total asset value, no
restrictions have been placed on the issuer's origin.
8.77
Colombia
As regards compulsory pensions, up to 10% of the Fund's total value

can be invested in foreign assets (rule effective since September 1st
2001). On the other hand, no quantitative limits have been set for
voluntary pensions, although the law requires that the issuer be
awarded the "investment grade" status by credit rating agencies.
7.36
Peru
Up to 10% of the Fund's asset value.
8.77
No less than 10% or greater than 50% of the Fund's asset value.
n.a.
Bolivia
(*) Amongst others, it includes assets issued or backed by foreign governments and central banks or
commercial banks (both foreign and international), stocks and corporate bonds, mutual funds and foreign
stock indices.
(**) Obtained as the ratio of funds invested in foreign assets to total fund's portfolio value at December
2003. For Peru, the last available data belongs to August 2003.
Source: FIAP, national pensions regulators and supervisors and national pension funds unions.
352
Table 7.
Pension Fund Stocks and Flows, Offshore Deposits and IFI Lending
(emerging economies with private social security systems)
Pensionfund
(2003)
IFI
lending
(exc.
IMF)
(Dec.
2001)
IMF
lending
(2003)
Initial
Year
Gross
Inflows
Stocks
Offshore
Deposits
(2002)
Argentina
Bolivia
Colombia
1994
1997
1994
956
192
775
15,947
1,485
7,326
23,413
1,176
7,252
21,211
3,103
8,591
15,466
278
-
Costa Rica
2001
167
304
3,234
1,654
Chile
1981
6,206
49,691
13,242
1,751
El Salvador
1998
476
15572
1,006
2,563
Mexico
Peru
Dom. Re.
Uruguay
1997
1993
2003
1996
6,765
754
34
112
35,844
6,341
34
1,232
48,616
5,894
2,391
7,500
19,852
14,688
2,447
2,302
139
130
2,407
2000
1998
2000
13
N.A.
2,822
134
2,631
11,058
2,965
1,383
19,378
N.A.
2,148
17,810
1,183
-
19,272
133,602
137,450
98,120
19,602
LATAM
EUROPE
Bulgaria
Kazajstan
Poland
TOTAL
Source: FIAP, national pensions supervisory agencies, national pension funds unions, BIS, IMF and
GDF (World Bank). Gross inflows for Costa Rica, El Salvador, Dominican Republic and Bulgaria
obtained as the difference between the stock of assets for 2003 and 2002 (both informed by FIAP).
353
Appendix I
Variable sources and definitions
Onshore dollar (peso) deposits: Foreign (local) currency deposits with domestic
banks. Source: Levy Yeyati (2004).
Onshore deposits: Onshore dollar deposits + Onshore peso deposits.
Offshore deposits: Cross-border deposits by residents with banks domiciled in BIS

reporting countries. Source: Bank of International Settlements (BIS).
Total deposits: Offshore deposits + Onshore deposits.
Deposit offshorization ratio: Offshore deposits / Total deposits.
Deposit dollarization ratio: (Onshore dollar deposits + offshore deposits) / Total

deposits.
Onshore deposit dollarization ratio: Onshore dollar deposits / Onshore deposits.
External loans: Cross-border loans to residents from banks domiciled in BIS reporting
countries. Source: BIS.
Dollar (peso) bonded external debt: Private and public external bonds denominated
in foreign (local) currency; stocks outstanding. Source: BIS.
IFI lending: Long-term debt with official creditors. Public and publicly guaranteed
debt from official creditors includes loans from international organizations (multilateral
loans) and loans from governments (bilateral loans). Source: Global Development
Finance 2003 (GDF 2003). Units: US dollars. Scale: millions.
IMF lending: Use of IMF credit. Denotes repurchase obligations to the IMF with
respect to all uses of IMF resources, excluding those resulting from drawings in the
reserve tranche. Source: GDF.
Dollar external liabilities: External loans + Dollar bonded external debt + IFI lending.
Total external debt: Includes public and publicly guaranteed long-term debt, private
nonguaranteed long-term debt, use of IMF credit, and estimated short-term debt
outstanding. Source: GDF.
Short-term external debt: Defined as debt that has an original maturity of one year or
less. Source: GDF,
Total long-term external debt: Total external debt minus short-term external debt.
Used in some tests as an alternative measure of dollar external liabilities. Source: GDF.
Liability dollarization ratio: (Dollar external liabilities + Dollar onshore

deposits)/(Dollar external liabilities + Peso bonded debt + onshore deposits). The
currency composition of deposits is used to proxy the currency composition of
domestic loans.
354
Country risk: J.P. Morgan Bond EMBI Global index. Included in the EMBI Global
are US dollar denominated Brady bonds, Eurobonds, traded loans and local market
debt instruments issued by sovereign and quasi-sovereign entities. Source: J.P.
Morgan.
Restrictions: Index of restrictiveness of rules on resident holdings of foreign currency
deposits onshore as of beginning of 2001. Source: Levy Yeyati (2004) based on IMF,
2001 Annual Report on Exchange Arrangements and Exchange Restrictions, following
the methodology proposed by De Nicolo et al. (2003).
Underlying Dollarization Ratio: (Var(D) - Cov (D,s)) / (Var (D) + Var(s) - 2Cov
(D,s)), where D and s are the monthly inflation and real devaluation rates. Source:
IMF, International Financial Statistics (IPS).
355
Deposit Dollarization Data: Countries and Periods Covered

Country
Dollariz.
Country
Dollariz.
Country
1992-2001 Ecuador*
1990-1999 Lebanon
Albania
1995-2001 Egypt
1980-2001 Lithuania**
Angola
Antigua and Barbuda* 1979-2001 El Salvador*
1982-2001 Macedonia, FYR**
1981-2001 Estonia**
1991-2001 Malawi
Argentina*
1992-2001 Ethiopia
1998-1999 Malaysia
Armenia**
1997-2001 Finland
1996-2001 Maldives
Austria
1992-2001 Georgia**
1992-2001 Malta
Azerbaijan**
1975-2001 Ghana
1995-2000 Mauritius
Bahamas, The
1987-2001 Greece
1990-2001 Mexico*
Bangladesh
1984-1997 Grenada*
1979-1999 Moldova**
Bahrain
1975-2001 Guatemala*
1995-2002 Mongolia**
Barbados*
1992-2001 Guinea
1989-2001 Mozambique
Belarus**
1976-2001 Guinea-Bissau
1990-1996 Myanmar
Belize
1994-2001 Netherlands
1993-2001 Haiti*
Bhutan
1975-2001
Honduras*
1990-2001 Netherlands Antilles*
Bolivia*
1991-2001 New Zealand
Bosnia and Herzeg.** 1996-2001 Hong Kong
1991-2001 Hungary**
1989-2001 Nicaragua*
Bulgaria**
1995-1999 Iceland
1978-1999 Nigeria
Cape Verde
1993-2001 Indonesia
1992-2001 Norway
Cambodia
1976-2001 Israel
1981-2001 Oman
Chile*
1996-2000 Pakistan
China,P.R.: Mainland 1998-2001 Italy
1992-2001 Papua New Guinea
1990-1999 Jamaica*
Colombia*
1996-2001 Paraguay*
1998-2001 Japan
Comoros
1975-2001 Jordan
1990-1999 Peru*
Congo, Dem. Rep.
1990-2002 Kazakhstan**
1998-2001 Philippines
Costa Rica*
1995-2001 Poland**
1993-2001 Kenya
Croatia**
1990-2001 Qatar
1993-2001 Korea
Czech Republic**
1981-1999 Romania**
1991-1999 Kuwait
Cyprus
1991-2001 Kyrgyz Republic**
1995-2001 Russia**
Denmark
1988-2001
Lao
People's
Dem.
Rep.
1989-2001 Rwanda
Dominica*
1992-2001 Sao Tome & Principe
Dominican Republic 1996-2001 Latvia**
Saudi Arabia
Dollariz.
Country
1993-2001
1993-2001
1997-2001
1994-2001
1996-2001
1981-1999
1975-1984
1992-1999
1991-2002
1994-2001
1992-2001
1991-2001
1991-1999
1990-2001
1975-2001
1990-2001
1990-2001
1994-2001
1996-2000
1975-1999
1990-1998
1976-1999
1988-2001
1975-2001
1982-2001
1985-2001
1993-1999
1990-2001
1993-2001
1994-1999
1995-2001
1975-2001
Sierra Leone
1993-1999
Slovak Republic** 1993-2001
Slovenia**
1991-2001
South Africa
1991-2001
Spain
1996-2001
St. Kitts and Nevis* 1979-2001
St. Lucia*
1979-1999
St. Vincent & Grens.* 1979-2001
Sudan
1992-1998
Suriname*
1975
Sweden
1994-2001
Switzerland
1998-2001
Syrian Arab Republic 1975-1998
Tajikistan*
1996-2000
Tanzania
1993-2001
Thailand
1982-2001
Trinidad and Tobago 1993-2001
Turkey
1986-2001
1993-2000
Turkmenistan**
Tonga
1994-1999
Uzbekistan
1997-1999
Uganda
1992-2000
Ukraine**
1992-2001
United Arab Emirates 1981-2001
United Kingdom
1990-2001
Uruguay*
1981-2001
1981-1999
Vanuatu
Venezuela*
1994-2001
Vietnam
1992-2001
1990-2001
Yemen
Zambia
1994-2001
Zimbabwe
1993-1999
Note: (*) denotes Latin American countries and (**) denotes Transition countries.
356
Dollariz.
Emerging Market Bond Index Global: Countries and Periods Covered

Country
Period covered
Algeria
Argentina*
Bulgaria**
Brazil*
Chile*
China: Mainland
CoteD'Ivoire
Colombia*
Croatia**
Dominican Republic*
Ecuador*
Egypt
Hungary
Korea
Lebanon
Morocco
Mexico*
Country
1999-2003
1993-2003
1994-2003
1994-2003
1999-2003
1994-2003
1998-2003
1997-2003
1996-2003
200 1 -2003
1995-2003
2001-2003
1999-2003
1993-2003
1995-2003
1997-2003
1993-2003
Malaysia
Nigeria
Pakistan
Panama*
Peru*
Philippines
Poland
Russia**
El Salvador*
Thailand
Tunisia
Ukraine
Turkey
Uruguay*
Venezuela*
South Africa
Period covered
1996-2003
1993-2003
2001-2003
1996-2003
1997-2003
1997-2003
1994-2003
1997-2003
2002-2003
1997-2003
2002-2003
2000-2003
1996-2003
2001-2003
1993-2003
1994-2003
Notes: (*) denotes Latin American countries and (**) denotes Transition countries.
Source: JP-Morgan.
357
Appendix II
Non-binding
offshorization
First note that, from (1),
(A.1)
which implies that
(A.2)
to rule out arbitrage between a portfolio XH - xp + XCF and XCH .
Then, using
and
, we obtain
, so that the investor's problem can
now be written as:
(A.3)
where
(A.4)
and
(A.5)
with
358
and
The first order conditions with respect to x can be expressed as:
(A.6)
from which we obtain:
(A.7)
or
Interest rate arbitrage by risk-neutral borrowers implies that
(A.8)
from which the dollarization and offshorization ratios are given by:
where
denotes the underlying dollarization ratio.
359
Binding offshorization and liability dollarization

Offshorization is binding iff .
or
(A.9) '
In turn, noting that, combining (9) and (A. 8),
(A. 10)
and
we can rewrite condition (A.9) as
If the previous condition is satisfied,

in equilibrium. Thus, in the absence of
foreign peso assets, the investor's problem can now be written as:
where, using XH + XCF = 1,
from which the first and second moments of the probability distribution of portfolio real
returns can be expressed as
and
360
In turn, from the first order condition with respect to XCF ,
where
and
so that the additional dollarization induced by offshorization in the absence of a country

risk-free peso assets is increasing in country risk
Finally, substituting A, = A into the balance of funds equation (9) and rearranging,
we obtain
(A.11)
from which it is immediate to see that liability dollarization increases with country risk,
since
Positive correlation between exchange rate risk and country risk

If
361
and
and
Thus, offshorization now becomes binding (/I < y) iff
where
increases with pcs.
362
Optimal Collective Action Clause Thresholds*
Andrew G. Haldane
Adrian Penalver
Victoria Saporta
Bank of England
Hyun Song Shin
London School of Economics
"This paper represents the views and analysis of the authors and should not be thought to represent those
of the Bank of England or Monetary Policy Committee members. We would like to thank Ken Kletzer,
Charles Goodhart, Chris Salmon, Misa Tanaka, David Wall, seminar participants at the Bank of England,
attendees at the IMF/Banco de Espafia conference 'Dollars, Debts and Deficits: Sixty Years After
Bretton Woods," and an anonymous referee for helpful comments. Correspondence should be directed to
Adrian Penalver at the Bank of England, Threadneedle St, London, EC2R 8AH. Telephone: +44 20 7601
3303. Fax +44 20 7601 5080. Email: adrian.penalver@bankofengland.co.uk
Introduction
Over the past year, tangible progress has been made towards improving the resolution
of sovereign financial crises. In February 2003, Mexico made a policy decision to
include collective action clauses (CACs) in its sovereign bonds issued under New York
law, contrary to market convention.1 With the first-mover problem solved, most other
emerging markets issuing in New York have followed suit. A new market standard for
emerging market sovereign bonds appears to have been set, with CACs as its
centrepiece.
The potential advantages of CACs have of course long been recognised. They
have been standard in English law bonds since the 19th century (Buchheit and Gulati
(2002). But under New York law, unanimity clauses have until recently remained the
market convention. This is generally felt to be suboptimal. Unanimity bonds mean that
debt restructurings are potentially held hostage to the actions of recalcitrant or rogue
creditors. More specifically, they may engender rent-seeking among some creditors,
with attendant welfare externalities for the system (see Kletzer 2003 and Haldane,
Penalver, Saporta and Shin 2003).
In response to these concerns, a push to introduce CACs in New York law
bonds was first made by the official sector after the Mexican crisis, with the publication
of the Rey Report (1996) by the Group of Ten countries (see also Eichengreen and
Fortes 1995). Little action followed. A second push was made by the official sector in
2002, following crises in Turkey, Brazil and most prominently Argentina. Again under
the auspices of Group of Ten, a working group was set up to draft model CACs. These
draft clauses were published in March 2003.
The aims of the Group of Ten working group were twofold. First, to examine a
range of potential contractual clauses that could be included in sovereign bonds and
recommend which ones to include. For example, they specified a threshold voting
majority of 75% for changes in a bond's financial terms. Second, to set a new market
standard. Contractual clauses set the rules of a debt restructuring and if the rules vary
too much, restructurings become more uncertain. The precise details of the clauses
would depend on the jurisdiction of issuance but the intention was to make the
substance as similar as possible.
The bonds issued by Mexico in February 2003 followed closely the G10 model
clauses, including a 75% threshold. But some subsequent issues by other countries
Brazil, Belize, Guatemala and Venezuelaopted for higher 85% thresholds. These are
closer to the levels proposed by private sector trade associations (EMCA 2002).
Some within the official sector have taken a dim view of these developments.
First, because these higher thresholds take us closer to a 100% unanimity bond, thereby
increasing the risk of holdouts. And second, because different voting thresholds risk a
splintering of the market standard (see Fortes 2003). A combination of standard and
non-standard thresholds could create unnecessary uncertainty during a debt
restructuring (see Gray 2004). One contribution of this paper is to critically evaluate
these propositions using a theoretical model of financial crisis. In particular, we ask:
what factors might determine the choice of optimal CAC threshold for a debtor? Is a
lower threshold always better? And are there valid reasons why different issuers may
!
It was not nearly as strict a convention as generally thought. Richards and Gugiatti (2003) document
that Bulgaria, Egypt, Kazakhstan, Lebanon and Qatar had issued bonds with CACs in New York prior to
2003.
365
want to set different, but country-specific, thresholds? To our knowledge, the literature
has not yet addressed these questions within an analytical framework..
Our findings suggest that there are costs to a policy of the 'lower the better'. A
lowering of CAC thresholds provides assurance to the debtor that in the event of debt
restructuring holdout creditors can be held in check with a lower offer. But this
insurance benefit of CACs has to be weighed against the prospect of a lower offer
increasing the likelihood of creditors running for the door and, consequently, higher exante interest rates.2 The choice of threshold that strikes the optimal balance between
these costs and benefits depends on debtor risk preferences and creditworthiness. Riskneutral debtors prefer high thresholds because the ex-post costs of getting away with a
lower offer are more than outweighed by the ex-ante benefits of lower interest rates and
a lower probability for a liquidity run. Risk-averse debtors may however prefer lower
CACs. Moreover, for a given level of risk aversion, the lower the debtor's
creditworthiness the more likely they will want to issue bonds with higher CAC
thresholds. In other words, different choices of threshold between countries emerge as
an optimal debtor response to different risk preferences and creditworthiness.3
A second contribution of this paper is to develop a model which nests both
liquidity runs and debt restructuring following a solvency crisis. Typically, the two are
treated separately.4 In practice, however, it is rarely straightforward to partition crises
in this way. Liquidity crises affect prospects for solvency; and expected recovery rates
for creditors following a debt restructuring will in turn affect short-term decisions on
liquidity. These interactions mean that most crises lie in the 'grey zone' between pure
liquidity and pure insolvency.
The model presented here is one such 'grey-zone' model, which allows
behavioural interactions between short-term liquidity and debt restructuring following a
solvency crisis. The framework allows us to explore the interaction between liquidity
and solvency crisis tools. For clarity of analysis, insolvency is determined by inability
to pay and therefore would be also applicable to corporate financial crises. But this
does sidestep the additional sovereign constraint of willingness to pay (Eaton and
Gersovitz 1981). But this additional layer of uncertainty only reinforces the complexity
of 'grey-zone crises'. Taking account of these interactions is important when assessing
the design and potential benefits of CACs as a crisis resolution measure.
The paper is planned as follows. Section 2 presents the stylised grey-zone
model. Section 3 discusses the determination of interest rates in the model and Section
4 the debtor's choice of optimal CAC threshold. Section 5 concludes with some policy
implications.
The Model
A debtor who wishes to finance a risky project issues a bond with collective action
clauses. These clauses stipulate that if a proportion K or more of creditors vote to
change the financial terms of the contract they bind the rest. The debtor chooses a
2
The analysis by Eichengreen et al. includes the potential for the insurance element of CACs to induce
moral hazard on the part of debtor.
3
The paper does not assess whether a country should maintain a uniform threshold in all issues if
creditworthiness and rating circumstances change.
4
See, for example, Chang and Velasco (1999) for a model of liquidity crises and Bolton and Jeanne
(2002) for a model of sovereign solvency crises.
366
collective action clause threshold K from a set of thresholds [K^K"] where K1, for
example, could be 65% and KU 85%.
A continuum of small risk-neutral creditors buys the bond. Each creditor has a
legal cost lj>0, which is the cost that this creditor would face if creditors collectively
chose to reject a restructuring offer from the debtor. It captures the disutility arising
from the costs of taking the debtor to court and prolonging the bargaining process
The probability distribution function of legal costs across the creditors is common
knowledge. The bond is short term. Creditors face the choice of whether to roll over
their funding at an interim stage and can decide not to roll over their funding to the last
period of the game.
The project's return depends on the realisation of a fundamental, 0. We assume
that the ex-ante distribution of 0 is uniform on [0,b]. Before creditors make their
decision to foreclose or roll over they receive a private signal about the fundamentals
Xj= 0+$ where si is i.i.d. with a uniform distribution over the interval [-E ,e ]. The noise
term s; is independent of i's legal costs. Creditors who foreclose receive a fixed pay-off
of 1. Creditors who roll over get a gross return (1+r) if the project succeeds. If the
project fails, they participate in a bond restructuring.6
The bond restructuring proceeds as follows. The debtor makes an offer to write
down the debt according to the contractual conditions of the bond. The creditors vote
to accept or reject the offer. If K or more creditors vote to accept the offer, it goes
ahead. Otherwise, creditors get a pro-rata share of the residual project return less their
private legal costs.7
More precisely, the extensive form of the game is as follows:
1.
The debtor chooses a CAC threshold K from [K!,KU] to insert into the
bond contract used to finance a risky project.
2.
The bond contract is offered to multiple risk-neutral creditors who buy
an equal share of the total debt stock (normalised to one here) at a market-determined
interest rate r.
3.
Creditor i observes signal x*= 0+$, as described above.
5
There are many reasons why such costs may differ across creditors. For example, some creditors (e.g.,
bond mutual funds) may have investors with shorter investment horizons than others (e.g., pension funds
and life insurance companies). There may also be differences in balance sheet structures, in agency
problems related to compensation structure, and in accounting and regulatory rules. Equivalently, lj can
be thought of as measuring the elative degree of risk aversion of different sets of creditors, in deciding
between choosing a certain option (accepting the offer) and an uncertain one (holding out).
6
We have allowed for secondary market trading because there are no pure strategy equilibria to such a
subgame (full proof available from the authors). Vulture funds have weak incentives to bid for bond
issues when: (i) ownership of the issue is widely dispersed; (ii) each creditor owns a small proportion of
the total issue; and (iii) contractual provisions ensure that during the subsequent restructuring stage
holdouts are kept in check, so that all creditors that hold on to their bonds get the same return. In
essence, the argument is identical to the one made by Grossman and Hart (1980) but in the context of
corporate raids. Creditors (shareholder) have very little incentive to tender their bonds (shares) to a
vulture fund (raider) whose participation in the restructuring stage (in management of the company) is
expected to increase the value of the debtor's offer (the value of the stock) for all, when each creditor
(shareholder) is small enough not to affect the outcome of the vulture's (raider's) bid.
For the purpose of our analysis, this assumption can be interpreted to mean that foreign creditors are
able to attache assets with value equal to the residual project return in the jurisdiction where the bonds
were issued. In practice, there are few tangible assets that creditors can attach in foreign jurisdictions.
The important aspect of this assumption is that the higher the available project return, the greater the
payout to creditors who decide to pursue their claim in courts. Other aspects of the analysis can be
generalised as long as this assumption is maintained. See also Haldane et al. (2003).
367
4.
Creditors decide whether to flee or roll over their bond contracts. We
denote by f the proportion of creditors who flee.
5.
Fundamental 0 is realised. This in turn determines the return on the
project which is given by
where f is the deadweight damage to the project from early liquidation.8

The project succeeds if there is enough project return to pay the creditors who
stay after the fleeing creditors have been paid out. That is, if
then the debtor repays the creditors who roll over in full and keeps the residual project
return for itself. The game ends. Otherwise, the project fails, the debtor defaults and we
enter the restructuring phase of the game described below.
6.
The debtor makes an offer to write down (1 -f)(l +r).
7.
Creditors vote to accept or reject.
8.
If the offer is accepted all creditors receive a payment equal to the offer.
In the event that the offer is rejected, each creditor i gets a pro-rata share of the return
to the project less their private legal costs lj.9
9.
Final pay-offs are determined.
We proceed to solve the game.
Restructuring Subgame
Stages 6-8 of the game are a voting subgame. The debtor makes a request to its
creditors to write down the repayment on its bond. Creditors decide whether to accept
or reject the offer. Since the bond has collective action clauses, the debtor's offer will
bind any dissenting creditors as long as the incidence of accepting creditors is K or
higher. If fewer than K creditors accept the offer, claims are pursued through the courts
and the debtor remains in default. In this event, each creditor will eventually receive a
pro-rata share of gross project return, 0, net of payouts to fleeing creditors, f, and net of
the damage done by early liquidation, f, and less the legal costs spent pursuing their
claim.10 The debtor receives whatever is left from the project after it has paid its
creditors.
Implicitly, we have assumed that the marginal damage to the project of one fleeing creditor is equal to 1.
The model can be easily generalised by assuming that the damage is equal to kf with& ^ 1. All the
results that follow are still valid as long as we allow for the possibility of a liquidity crisis, i.e., as long as
k>r.
ê model does not nest the case of unanimity provisions according to which all creditors are required
to agree before an offer can go through. The reason is because it is assumed that creditors lie on a
continuum. For a welfare analysis of CACs versus unanimity provisions from an ex-post perspective, see
Kletzer (2003) and Haldane et al. (2003).
10
This restructuring subgame is very similar to the one set out in Haldane et al. (2003). The main
difference is that here we do not assume that the debtor exerts adjustment effort after learning the
outcome of the vote. For our purposes, including adjustment effort would complicate the notation
without changing the analysis.
368
So the debtor's pay-offs are:

U(6-2f-offer)
U(0)
if K or more creditor accept

otherwise
(1)
where U(.) is the utility function of the debtor. The pay-off to the creditor in the
(l-K)-th quantile of the distribution of legal costs from a given offer is:
offer
0-2f-li.K
if K or more creditors accept

Otherwise
(2)
So from the point of view of the creditor and assuming full information, it is a
weakly dominant action to vote to accept the offer provided that the debtor's offer is
greater or equal to 6-2f-li.K. From the point of view of the debtor, again under full
information, and since legal costs are positive it is optimal to make an offer which is
just large enough to persuade the (l-K)-th creditor to accept, that is to offer:
(3)
This offer would be exactly equal to the net project return (after payouts to
fleeing creditors have been met) less the legal costs of the (l-ic)-th most stringent
creditor from the pool of (1-f) creditors who participate in the restructuring.n
From (2), this offer is just large enough for the marginal (l-K)-th creditor who
votes on a restructuring to accept the offer and for the deal to go through. In this setup, CACs ensure that restructuring is not delayed and resources are not expended on
litigation; they neuter rent-seeking among bondholders.12
In the event of a restructuring, therefore, the pay-off to the debtor is
(4)
whereas the pay-off to each creditor who rolled over is equal to the offer made by the
debtor, given by (3). Comparing (3) and (4) we can see that after default has occurred, a
lowering of the CAC threshold would shift the allocation of residual project return from
the creditors to the debtor and vice versa. Hence looking at solvency crises in isolation,
debtors would always prefer a lower voting threshold. We will see below, however,
that this result is modified when crises are neither pure insolvency nor pure liquidity,
i.e., when we have 'grey-zone' crises.
Rollover Subgame
Working backwards, we now determine the proportion of creditors who flee. This
involves solving stages 3-6 of the extensive form of the game, which form a rollover
global game in the manner of Morris and Shin (1998).
11
Say the threshold K is 75%, then for the offer to go through the debtor needs only to persuade the
creditor with the highest legal cost in the first quartile of the distribution of legal costs.
12
Haldane et al. (2003) show that this outcome in no longer guaranteed when there is two-sided
information asymmetry between the debtor and its creditors about how much they stand to gain or lose if
the restructuring does not take place.
369
The aggregate strategy of creditors is a rule of action which depends on whether

the signal creditors receive is above a critical threshold x*. A creditor will flee if his
private signal is lower than x* and will roll over otherwise. More formally, the strategy
u(x*) is an indicator function which takes the value one if x<x* and takes the value of
zero, otherwise. This implies that the proportion of investors who flee given the
aggregate strategy and the uniform distribution of private signals is
(5)
The equilibrium in switching strategies, in turn, implies that there is a critical

state of fundamentals, 0*, above which the project succeeds and below which the
project fails. At the equilibrium switching point, two conditions need to be met.
First, the proportion of creditors who flee must be such that the solvency
constraint just binds. We refer to this condition as the 'solvency condition'. The
solvency condition is given by
which simplifies to
(6)
From (5) and (6), we obtain

(7)
In the limit as e - 0, we have 6* - x*.

The second condition is that, at the switching point, the marginal creditor must
be indifferent between fleeing and staying. We refer to this as the 'indifference
condition'. This says that the expected pay-off from rolling over if the debtor defaults
plus the expected pay-off from rolling over if the debtor repays must equal the
deterministic pay-off from fleeing. That is:
(8)
where we have used the fact that the pay-off from rolling over if the debtor defaults is
equal to (3) and where p(6|x) is the density of 6 conditional on x. Since the prior over 6
is uniform and the noise is also uniform with support [- ,e], the conditional density
p(0|x) is uniform with support [x*-e ,x*+e ]. So, (8) can be written as:
(9)
Using (7) to eliminate x* and (5) to eliminate f we can express (9) as a

quadratic equation in 0*. In Appendix 1, we show that (9) can be written as
(10)
370
where the coefficients Fj are simple functions of 8 and the interest rate r. Appendix 1
also shows that, in the economically meaningful case where (10) has a positive root for
6*, there is precisely one such positive root. Hence, the switching equilibrium is
unique. This root gives the trigger point of fundamentals at which a liquidity crisis will
commence. This trigger point lies between two points: a 'fundamental insolvency'
point and a 'fundamental solvency' point, denoted 01 and 0U respectively. Fundamental
insolvency occurs when the project return is insufficient to pay out creditors even if
they all roll over (f=0), that is when 0 <Ql =(l+r). Fundamental solvency occurs when
the project return is so high that there is enough to pay out all creditors even if they all
foreclose (f=l), that is 0 >0U =2. In the region in between ((l+r)<0*< 2), we have
liquidity crises that is defaults which would not have occurred were it not for the
decision of a sufficient proportion of creditors to flee.
Interest Rate Determination
Having solved for the rollover stage of the game, we now solve for the marketdetermined, actuarially fair interest rate. In equilibrium, the expected return to the risky
project must equal creditors' outside option if they invest elsewhere, which for
convenience we normalise to one.
That is
(11)
where p(0) is the unconditional density of 0 which is equal to . Appendix 2 shows
b
that in the limit as the signals about fundamentals received by all creditors become very
precise, i.e., as s -> 0 (11) can be rewritten as:
(12)
where
(13)
is the positive root of (10) and where the coefficients Fj have been evaluated in the limit
as s -> 0. Expression (12) is an equation in three variables: r, II_ K and b. So, for a given
value of the CAC threshold, K, or equivalently, for a given legal cost of the (l-K)-th
creditor, II.K, and a value for the upper support of the unconditional distribution of the
project, b, we can determine the market interest rate r. Because the market interest rate
is actuarially fair it is equal to creditors' expected loss in the event of crisis. This
expected loss, in turn, is equal to the product of the loss in the event of crisis and the
ex-ante probability of a liquidity crisis. From (3), we know that creditors' loss in the
event of crisis (recovery rate) is decreasing (increasing) in the voting threshold K. We
would therefore expect the interest rate to decrease as the CAC threshold, K, increases.
371
Chart 1 illustrates for a simple parameterisation that this is indeed the case.13 The exante probability of a liquidity crisis, on the other hand, depends on both fundamentals
and on the voting threshold. First, we would expect the probability of crisis to increase
as the debtor's fundamentals, b, deteriorate.14 Chart 1 confirms this. Second, the exante probability of crisis depends on the voting threshold chosen by the debtor.
Intuitively, other factors being equal, lower voting thresholds make creditors more
trigger-happy during the rollover game because in the event of crisis they would allow
the debtor to get away with a lower offer and still keep holdout creditors in check.15
Chart 1: Variation of interest rates with CAC threshold
It is also interesting to note that although interest rates increase linearly as the
CAC threshold decreases, they increase at an increasing rate as fundamentals
deteriorate. Therefore, given a certain CAC threshold, K, one would expect to observe a
greater premium above risk-free rates for bonds issued by poor credit quality borrowers
than for bonds issued by better quality borrowers. Chart 2 illustrates this more clearly
by plotting interest rates against fundamentals for two different values of the CAC
threshold. Lower voting thresholds raise the probability of a liquidity run, the more so
the less creditworthy the borrower because the higher interest rate itself affects the
solvency constraint.
So lower-rated sovereigns need to offer creditors greater
compensation ex ante, for a given threshold. This finding is consistent with the
empirical findings of Eichengreen and Mody (2003) who find that lower-rated
borrowers are charged a premium to issue bonds with collective action clauses, whereas
higher-rated borrowers can issue at a discount.16 The reason suggested by Eichengreen
13
For the purposes of generating this figure and all the ones that follow we assume that the distribution of
legal costs is uniform on [0,1]. Under this assumption collective action clauses of 0.65, 0.75 and 0.85
imply marginal creditor legal costs, II.K, of 0.35,0.25 and 0.15 respectively.
14
As 9 is uniform on [0,b], the ex-ante expected value of the fundamental is b/2.
15
In the next section we show this more formally.
16
Richards and Gugiatti (2003), on the other hand, find no discernible difference from the inclusion of
CACs, whatever the credit rating of the issuer.
372
and Mody, and modelled in Eichengreen, Kletzer and Mody (2003), is borrowers'
greater propensity to default strategically when they lack creditworthiness. In our setup, a premium could arise even without an additional incentive to default. Rather,
investors require extra compensation as a result of the increased fragility of short-term
creditors when creditworthiness is low.
Chart 2: Variation of interest rates with fundamentals
Debtor's Choice of CAC Threshold

Arbitrage among creditors establishes a relationship between interest rates, the
collective action voting threshold and the range of fundamentals (equation 12).
The debtor can use this relationship when it chooses the voting threshold in
bonds it issues. We now turn to the first stage of the extensive form of the game, to
determine the debtor's choice of optimal collective action threshold.
Intuitively, the debtor cares about three things: the probability of crisis (which
depends on fundamentals b and on the expected rollover behaviour of creditors); its
pay-off in the event of non-crisis (which depends on the market interest rate); and its
pay-off in the event of a crisis. As we have seen from (4), the CAC threshold directly
affects the debtor's pay-off in the event of crisis. The lower the CAC threshold, the
higher the legal costs of the marginal creditor and the higher the debtor's pay-off in the
event of crisis. But Chart 2 shows that a lower collective action clause threshold results
in a higher interest rate. What the debtor gains from a higher pay-off in the bad state, it
loses in the good state. So a crucial question for the debtor in determining the optimal
voting threshold is its impact on the probability of crisis.
373
This can be determined by differentiating 0* with respect to II.K. The analytical

expression-a function of r and II-Kis complicated so results are better illustrated
80*
pictorially. In Chart 3 we plot ~for all valid combinations of r and II_K.
dl\-K
Chart 3: Variation of crisis threshold with legal costs of marginal creditor
Given that the surface always lies in the positive region,

equivalently,
or,
Therefore, the debtor can lower the probability of crisis by
raising the CAC threshold, K. A higher voting threshold provides short-term creditors
with additional assurance, and so lowers their incentive to run and hence the probability
of crisis.
This is the essence of the 'grey-zone' dimension to the model. Decisions
regarding outcomes in the event of solvency crisis affect the probability of liquidity
crisis, and hence shape ex-ante choices about optimal debt contracts. Previous papers
have looked at the effects of solvency crisis measures on contract design (e.g., Haldane
et al. 2003 and Kletzer 2003) or the effects of liquidity crisis measures on contract
design (e.g., Dooley 2000 and Gai and Shin 2003) but none, to our knowledge, have
considered the interaction among these measures.
More formally, the debtor would choose the CAC threshold, K to maximise its
expected return on the project. In doing so the debtor takes into account the pay-off
they expect to receive in the event of a crisis, U(li-K), and the pay-off they expect to
receive if no crisis occurs,
So the debtor solves the following
maximisation problem:
17
Recall that we are conducting our analysis as 6 > 0. A consequence of this is that the proportion of
creditors who flee becomes polar 0,1.
374
(14)
where 6* is the trigger point at which short-run creditors flee, (13), and p(6), as before,
is
f
In what follows, we assume that the debtor has utility
where 2-p
is the coefficient of relative risk aversion. As a result,
(15)
Taking into account the endogeneity of the crisis threshold and interest rates,
this can be rewritten as
(16)
The maximisation problem in equation (16) does not have a closed-form
solution, but simulated results provide interesting insights.
First, for all values of creditworthiness b, a risk-neutral debtor (p=2)

will prefer a higher collective action threshold to a lower one. This is
illustrated in Chart 4. Intuitively, a risk-neutral debtor does not care
about the distribution of possible pay-offs, only their expected value.
And the expected repayment by the debtor to its creditors is equal to 1
from equation (11). What the debtor gains by increasing its ex-post
payout in a crisis by lowering the collective action threshold is offset
exactly by higher ex-ante interest charges when there is no crisis. Since
the debtor makes a surplus when the project succeeds, a risk-neutral
debtor gains by simply minimising the probability of crisis, which is
achieved by issuing bonds with high collective action thresholds.
375
Chart 4: Variation of debtor utility with CAC threshold - risk-neutral debtor
Second, this conclusion is modified if the debtor is risk-averse (p <2).

This is illustrated in Chart 5. The more risk-averse the debtor, the more
it weighs the pay-offs it receives in a crisis period over those in a noncrisis period. With strong risk-aversion, the debtor may prefer to issue
bonds with a low voting threshold, so that it reduces what it has to pay
out ex post in a debt restructuring, even though this increases the ex-ante
risk of crisis and raises the ex-ante interest rate.
Chart 5: Variation of debtor utility with CAC threshold - effect of risk

aversion
376
Third, the lower the creditworthiness, the more likely it is that a

moderately risk-averse debtor will want to issue bonds with a high
collective action threshold (Chart 6). This occurs because creditors
considering their rollover decision require increasingly more
compensation for the risk of default as creditworthiness declines (Chart
2): when fundamentals are weak, short-term creditors are more triggerhappy. This will tend to lead lower-rated debtors to choose a higher
threshold to minimise this risk.
Chart 6: Variation of debtor utility with CAC threshold - effect of
fundamentals
Policy Implications
The model of grey-zone crisis developed here suggests the following policy
implications:
In choosing the optimal CAC threshold, there are costs to a policy of "the lower
the better1. A lowering of CAC thresholds may provide some assurance to the debtor in
the event of a default, because it allows holdout creditors to be held in check with a
lower offer. But there is a down side to this CAC benefit. The prospect of a lower offer
increases the likelihood of creditors running for the door in the first place. So a lower
CAC threshold risks increasing the chances of liquidity crisis, thereby raising initial
borrowing costs for the debtor. For example, in the simple model presented here, if a
debtor is risk-neutral they would prefer a higher collective action threshold to a lower
one because the ex-post benefit of having CACs is more than outweighed by its ex-ante
cost.
In more general settings, higher thresholds may no longer be optimal.

For example, a risk-averse debtorone which places greater weight on
the adverse consequences of defaultis more likely to choose a lower
377
CAC threshold.18 This suggests there may be a role for the international
official community in promoting lower thresholds. It is tantamount to
asking the sovereign to behave in a more risk-averse way. Put
differently, if CACs are insurance, then asking sovereigns to self-insure
by choosing a lower threshold lowers the burden on the official sector as
a centralised provider of sovereign insurance.
Even when there is a demand (voluntary or induced) for CAC insurance,
the optimal degree of insurance will vary across borrowers. The more
creditworthy the debtor, the lower the tolerable CAC threshold, for a
given level of risk aversion. Therefore, although there may be benefits in
having a market standard for CAC thresholds to reduce uncertainty in
debt restructurings, such a standard may be suboptimal for some
debtors.
The most general policy lesson which emerges from the model is that
solvency crisis tools and liquidity crisis tools cannot and should not be
viewed in isolation. The interaction and spillovers between them need to
be weighed carefully when crafting both sets of policy. This has been
demonstrated here in the context of CAC design and its important
implications for liquidity crisis. But it equally applies in reverse when
considering the efficacy of liquidity crisis measures. The model
presented here provides a framework for assessing these welfare
questions in an integrated fashion. It also provides a vehicle for
exploring complementarities between different (liquidity and solvency)
tools.
18
There may be other (than risk aversion) reasons why a lower CAC threshold may have a benefit
for example, in situations where there is two-sided information asymmetry between debtors and creditors
(Haldane et al. 2003). On the other hand, others argue that mitigating the costs of crisis of the debtor by
lowering the CAC threshold may weaken the disincentives to default (Dooley 2000).
378
Appendix 1
Recall from the main text that there are two equations, (7) and (9), in two unknowns,
0*, the trigger value of fundamentals for a crisis and x*, the signal of the marginal
creditor who flees. The relationship between 0* and x* also determines the proportion
of people who flee, f(u(x*)), and final project return at the trigger value of crisis 0*f(D(x*)).
The first step in the solution is to rewrite (7) for x*. For convenience, this can
be written in two forms:
(A-l)
where
and
(A-2)
where
and
The indifference condition (9) can be written as
(A-3)
where
Expressions (A-l) and (A-2)
are two equations in two unknowns x* and 0*. Substituting (A-l) and (A-2) into (A-3)
we obtain:
which is equal to
where
and
This can then be rewritten in the form
(A-4)
379
where
and
For s
and
Provided that
F3 + FAl}_K < 0, (A-4) will have one negative root and one positive root. The positive
root will be the economically meaningful solution for 0*. Therefore,
(A-5)
380
Appendix 2
Since 9 i s distributed uniformly o
Substituting p(0) into (11) we obtain:
(B-l)
In the limit, as s -> 0, signals become fully informative and all creditors would
fleeif # < 0 * , i.e.,f=l for -0. Therefore as s ->0, (B-l) becomes
(B-2)
where 6* is given by (13).
381
References
Bolton, P and Jeanne, O (2002), 'Sovereign debt structuring and restructuring: an
incomplete contracts approach', mimeo.
Buchheit, L and Gulati, M (2002), 'Sovereign bonds and the collective will',
Georgetown University Law Center Working Paper no. 34.
Chang, R and Velasco, A (1999), 'Liquidity crises in emerging markets: theory and
policy' NBER Working Paper no. 7272.
Dooley, M (2000), 'Can output losses following international financial crises be
avoided?', NBER Working Paper no. 7531.
Eaton, J and Gersovitz, M (1981), 'Debt with potential repudiation theory and
estimation', Review of Economic Studies, Vol. 48, pages 289-309.
Eichengreen, B, Kletzer, K and Mody, A (2003), 'Crisis resolution: next steps', IMF
Working Paper 03/196.
Eichengreen, B and Mody, A (2003), 'Would collective action clauses raise borrowing
costs? An update and additional results', mimeo.
Eichengreen, B and Fortes, R (1995), Crisis what crisis? Orderly -workouts for
sovereign debtors, Centre for Economic Policy Research, London.
Emerging Market Creditors Association (2002), Model covenants for new sovereign
debt issues}.
Gai, P S and Shin, H\ S (2003), 'Debt maturity structure with preemptive creditors',
Bank of England Working Paper no. 201.
Gray, R (2004), 'Collective action clauses: the way forward', paper presented at
Georgetown University Law School Conference on Sovereign Debt.
G10 Working Group on Contractual Clauses (2003), Report of the G10 working group
on contractual clauses.
Grossman, S and Hart, O (1980), 'Takeover bids, the free rider problem and the theory
of the corporation', Bell Journal of Economics and Management Science,
Spring, pages 42-64.
Haldane, A, Penalver, A, Saporta, V and Shin, H S (2003), 'Analytics of sovereign debt
restructuring5, Bank of England Working Paper no. 203 and forthcoming in the
Journal of International Economics.
Kletzer, K (2003), 'Sovereign bond restructuring: collective action clauses and official
crisis intervention', in Haldane, A (ed), Fixing financial crises, London:
Routledge.
Morris, S and Shin, H S (1998), 'Unique equilibrium in a model of self-fulfilling
currency attacks', American Economic Review, Vol. 88 (3), pages 587-97.
Fortes, R (2003), 'Resolution of sovereign debt crises: the new old framework', mimeo.
Rey, J-J (1996), 'The resolution of sovereign liquidity crises', Report to the Ministers
and Governors of the G10.
Richards, A and Gugiatti, M (2003), 'Do collective action clauses influence bond
yields? New evidence from emerging markets', International Finance, Vol. 6
(3), pages 415-47.
382
COMMENT
The Role of the IMF
Lorenzo Bini Smaghi
Director General for International Finance,
Italian Ministry of the Economy and Finance
I will base my discussion on the following question: how do the two papers help us in
addressing the key policy issues concerning the role of the International Monetary Fund
60 years after the Bretton Woods agreements?
These two papers fit quite nicely with the two key issues relating to the main
tasks of the Fund: crisis prevention and crisis resolution.
I will use the paper by Eichengreen, Kletzer and Mody ("Monitoring
International Borrowers: The IMF's Role in Bank and Bond Markets ") as the basis for
my remarks on crisis prevention, and the one by Chami, Sharma and Shim ("A model
of the IMF as a Coinsurance Arrangement") to address the issue of crisis resolution.
I draw from the first paper (EKM) the following considerations:
In a world of imperfect information, the role of the Fund in surveillance is
essential.
Surveillance by the Fund is most valued by market participants when
countries have a programme (i.e. they borrow), including precautionary.
However, this is true only when the country has a sustainable level of debt
Let me give the authors a note of caution on their findings. When this type of
analysis is conducted, one should check for three main problems:
Endogeneity
Missing variables
Alternative explanations.
For example, the paper tests the following causal relationship: an IMF
programme means more information; more information means a lower risk premium,
and a lower risk premium means more bond issuance.
There can be an alternative explanation: an IMF programme may be interpreted
as the expectation of a bail-out, which means a lower risk premium, but also more bond
issuance. The two hypotheses imply a similar relationship between the existence of an
IMF programme and bond issuance.
The fact that the link between IMF programme and risk premium does not work
at a high level of the debt should provide reassurance against this alternative
hypothesis.
Another note of caution. Bank credit and bond issuance are not perfectly
substitutable. They depend on a series of factors, related only on the creditors' choice.
383
Third note of caution. Bank loans are not necessarily easier to restructure than
bonds, as the paper seems to suggest. This was the case in the Asian crisis. However,
the Turkish case (2002) has shown that it is increasingly difficult to coordinate bank
roll-over, especially if the central banks do not play a leading role. The Argentine and
Uruguay cases have shown that bondholders have over time found solutions to the
coordination problem. (Politically it might become even more important in the future.
Financial institutions do not vote, while bondholders do).
What are the main policy lessons that I draw from the EKM paper?
First, IMF programmes are a very important element in surveillance. This
means two things:
1.
2.
Surveillance needs to be strengthened independently of programmes.

There should be a way for all countries to have access to the kind of
surveillance and monitoring that is foreseen for countries that have an
IMF programme.
The G7 is currently working on a strategic review of the Bretton Woods

institutions. One idea is to have IMF programmes that do not necessarily entail
borrowing from the IMF, the so-called "non-borrowing programmes". Non-borrowing
programmes would be based on strong policy commitments by countries. The Fund's
task would be to provide high frequency monitoring and sending clear signals on the
extent to which the country is achieving the objectives. It would be stronger than staffmonitored programmes because it would involve the Board. The reports would also be
published.
Such programmes could also help as "exit strategies" from programmes. They
could also support the transition to accessing capital markets and could complement
precautionary programmes, with less risk to the exposure of the Fund.
Second conclusion, IMF surveillance for high debt countries should be
strengthened (at present it does not add much to available information) and be made
more accountable and more transparent, i.e. more independent.
The IEO has shown that one of the main problems in surveillance and, in
particular, Debt Sustainability Analysis, has come from the over-optimistic
assumptions made by the IMF staff. This is not surprising. If you ask the same people
who have designed a programme to assess it (in particular in terms of DS A, stress text,
sensitivity analysis), they will select hypotheses which are consistent with the
programme. There is a conflict of interest.
Surveillance (including DSA) is not sufficiently independent of programme
design.
In any financial institution, the decision to lend is separate from risk
management. The Analysis produced by risk management is provided independently to
the Board.
In the IMF, the management has access to both assessments, but DSA is
produced by country desk. The case of Argentina since 2001 is one where the
information has not been provided transparently to the decision-making body, which is
the Board. The risk profile of the IMF portfolio has to be decided by the Board.
In the Strategic Review, the G7 are examining ways to strengthen surveillance,
ensuring greater accountability.
Two ideas can be put forward:
384
1.
2.
Better integration between multilateral, bilateral and regional

surveillance;
Better separation between surveillance and programme design.
Concerning the second point, greater independence could be achieved by

making PDR responsible for DSA. This would ensure independence of programme
design.
Let me turn now to crisis resolution. Here I will use the paper by Chami,
Sharma and Shim.
The main note of caution with this paper, as recognized in the paper itself, is in
footnote 20: time inconsistency of the optimal policy by the country.
The paper does not include reference to private sector financing and its
relationship with IMF financing. This is particularly relevant in addressing recent
crises.
Another issue that the paper does not address, but which is very relevant to the
current international debate is the status of the IMF as a preferred creditor. The ongoing
discussions about Argentina underline the importance of this issue. How can an
assumption on IMF repayment be made without some specific target concerning the
primary surplus and independently of the hypothesis on haircut? How can the IMF
defend disengagement while the private sector is taking a substantial hit?
The other issue that is not addressed in the paper is the possibility of defaulting
with the IMF. The whole LIA policy is ignored.
Two counter-arguments have been put forward in the past: resources and
micromanagement.
With respect to resources, some savings can be made in other areas, especially
if bilateral and regional surveillance are more closely integrated. Concerning
micromanagement by the Board, it is often forgotten that the principles of good
Corporate Governance (e.g. those by the DECD) give responsibility to the Board to
ensure that organizations are effective and avoid conflicts of interest.
As the paper recognizes, the key to ensuring the credibility of IMF policies is
clear access limits. New procedures and criteria have been adopted recently. This has
been a major innovation in IMF policy. We now have to implement it! This is the major
challenge for the IMF. Recent research by the Banco de Espafta shows that the main
shortcomings in recent reforms has been implementation. One recent example is the
concept of "negotiations in good faith" which has been questioned in the resolution of
the Argentine crisis. Further clarification could be useful, in particular in the context of
the Code of Conduct. This is an area where further progress would be very welcome.
385
COMMENT
Innovations in Private and Multilateral Lending
John Murray
Advisor, Bank of Canada
I have been asked to discuss two thoughtful and interesting papers, and must admit at the
outset that I have very little to offer by way of useful criticism or comment. Both papers
present credible and convincing arguments for the positions that they advance, and I am in
broad agreement with most of what they say. With these apologies duly recorded, allow me
to make a few observations on what I regard as the papers' major contributions as well as
some modest suggestions concerning additional points that the authors might want to
consider in future extensions of their work.
Let me begin with the paper by Eduardo Levy-Yeyati ("Dollar, Debts and the
IFIs: Dedollarizing Multilateral Credit"). This paper focuses on the efforts of emerging
market economies (EMEs) to develop strong domestic capital markets and to minimize the
problems posed by foreign borrowing and currency mismatches. More interestingly, it
advances an intriguing proposal for dedollarizing existing sovereign debt. According to
Levy-Yeyati's proposal, IFIs would be encouraged to roll-over their existing, largely
dollar-denominated, loans for new obligationsdenominated in local currencies. LevyYeyati notes that IFI debt now represents a significant portion of all outstanding EME
liabilities, and that dedollarizing this debt would make an important contribution towards
resolving the currency mismatch problem. In addition, the newly created stock of high
grade, local currency, bonds could serve as a catalyst for the development of a more active
and extensive domestic market in private and public sector debt. The IFIs would not lose
anything through this conversion, the author suggests, since EMEs almost always repay
their IFI debts and the loans could be indexed to guard against any loss in their real value.
The IFIs, in turn, could finance these loans by tapping into the ready market created by
emerging market investors who had sent their money off-shore, but who still might value
the hedging opportunities afforded by risk free, local currency denominated, bonds.
Levy-Yeyati acknowledges the mixed reaction that similar proposals have received
in the past, as well as the recent criticisms put forward by various IFIs, but claims that
these concerns are misplaced or of limited relevance, given the unique features of his
proposal. As I suggested earlier, I have some sympathy for both Levy-Yeyati's proposal
and his response to the critics. There might be a lack of investor and debtor interest in such
instruments, as some critics have claimed, but it is not clear that anything would be lost by
testing the waters. Any problems the IFIs might encounter would probably be of an
operational nature, since the transactions costs associated with local-pay instruments might
be slightly higher than those on dollar-denominated debt, and the resulting markets would
probably be less liquid. In additional, it would be more difficult for IFIs to match their
387
currency exposures on an ongoing basis. Non-resident investors would likely show little
initial interest in these instruments, but resident investors who have shipped their money
off-shore would seem to be obvious candidates.
The two main concerns that I have with Levy-Yeyati's proposal are the following.
First, it is possible that IFI claims would attract savings away from local EME financial
institutions and capital markets, and draw towards off-shore financial centres. In other
words, the IFIs could "cherry pick," owing to their preferred creditor status, and actually
frustrate or delay the development of local capital markets by redirecting funds that would
have otherwise stayed at home. Second, from a broader perspective, it is not obvious that
much would have been accomplished from the perspective of the EME. If sovereign
obligations are (nearly) always repaid and the new loans are fully indexed, is there any
material difference between these local currency instruments and the dollarized debt that
they are replacing? There may be, in other words, a sort of "Modigliani-Miller irrelevance"
aspect to this whole initiative.
By the same token, however, one could ask, what is the harm? If the new
instruments are issued and there is limited take up or, alternatively, they simply substitute
for existing claims, no one is worse off. On the other hand, there is a chance that they might
actually do some good, and have the sort of catalytic effect that Levy-Yeyati claims.
The second paper1, co-authored by Andrew Haldane, Adrian Penalver, Victoria
Saporta, and Hyun Song Shin, addresses a different, yet related, issuethe optimal design
of collective action clauses for sovereign debt. Whereas Levy-Yeyati focuses on a specific
innovation that might reduce the exposure of sovereign borrowers to sudden exchange rate
movements and capital reversals, Haldane et al. (henceforth, HPSS) look at some of the
endogenous elements that are in play when one designs collective action clauses (or CACs).
Although CACs can both reduce the chances of a crisis emerging and facilitate the
resolution of those which do occur, inappropriate or inflexible CAC features could have
unintended and undesirable consequences.
The authors start with the observation that some bonds, recently issued by EMEs,
incorporate slightly different features in their collective action clauses than the standard
London-style instruments. More specifically, the voting thresholds required to approve a
debt restructuring have, on occasion, exceeded the 75 per cent norm that has existed for
many years on blonds issued under U.K. law. The major contribution of the paper is the
construction of an elegant model which captures the inherent endogeneity of the decision
process faced by borrowers and lenders, and which is also able to explain why some
sovereign borrowers might want to issue bonds with a somewhat higher threshold. The
logic runs as follows.
Countries which are perceived to be greater credit risks, as the authors suggest, may
find it advantageous to raise the voting threshold because the potential costs that they might
incur, in the form of a more difficult debt restructuring, are more than offset by the
reduced probability that skittish investors will race for the exits in the event of a liquidity
problem.
Once again, as I admitted earlier, I find it difficult to identify any significant flaws in
the authors' argument or the way the model is developed. Nevertheless, there are a few
thoughts that occurred to me as I read the HPSS paper.
Ôptimal Collective Action Clause Thresholds."
388
The first point that I would raise concerns the risk of giving too much significance
to recent events. While it is true that Belize, Brazil, and Guatemala have all issued bonds in
the last few months with somewhat higher voting thresholds, these seem to have been
exceptional events that might not be repeated in the future. Indeed, one of the countriesBrazilhas come to the market again, but with a more traditional covenant. Although the
reasons behind the initial decision to include a slightly higher voting threshold are not
exactly clear, it might have had more to do with increased attention that CACs were
attracting at the time, and a desire on the part of Brazil to give some recognition to the
hard-line that the bondholder committees were espousing. Is it really the case that the risks
associated with Belize, Brazil, and Guatemala were substantively different than those of the ,
Philippines and Turkey? Why then were the Philippines and Turkey able to borrow with the
traditional 75 per cent voting threshold? In addition, the results reported by researchers
such as Richards and Gugiatti suggest that creditors are often completely unaware of
CACs, let alone fine differences in their voting thresholds.
My second point is more conceptual in nature and relates to an important aspect of
CACs that seems to have been under-appreciated in the authors' model. It concerns the
benefits that both borrowers and creditors are expected to realize from the inclusion of
CACs in bond covenants. The gains that are realized from CACs, in terms of defusing the
collective action problem and the cost of restructuring, should this be necessary, accrue to
both parties. However, the model seems to assume that they reside mainly with the
borrower/While problems might arise when the threshold is set too low, one wonders if
the incentive to run and thereby trigger a liquidity crisis is materially different at a threshold
of 75 per cent as opposed to 85 per cent.
Why then, one might ask, would groups such as EMC A push for higher voting
thresholds? A cynical interpretation of events might run as follows. Certain borrowers and
creditors found it advantageous to operate under the old system, without CACs, and did
not want to introduce any arrangements that might make it easier to restructure debt in the
case of serious liquidity or solvency problems. This is not because the resultant pain was
viewed as a necessary deterrent to strategic default or moral risk on the part of borrowers,
but because the old arrangements maximized the chances of a public sector bailout.
Anything that would make debt workouts more efficient and reduce the need for official
intervention was not in their joint interest therefore. Having resisted the adoption of CACs
for so long, it would have been awkward to give way too easily and simply adopt the
London terms. Creditors, and as a consequence some borrowers, may have felt it necessary
be seen entering the new arrangements grudgingly. Having made their point, however, and
saved face, they could then revert to the London norms prescribed by the G-l0.
My final point turns the CAC issue on its head, and asks whether the authors'
hypothesis does not run counter to the Current strategy of the G-10. This strategy seems
designed to push all emerging country borrowers onto the same template or CAC model.
Does this make sense if, as the authors suggest, some variability in voting thresholds and
other CAC conditions is welfare improving? On the other hand, does the fact that the 75
per cent threshold has existed for so long, and been applied so uniformly, indicate that the
benefits from contract standardization far outweigh the benefits of tailoring the covenants
to the debtors' differing circumstances? Revealed preference would seem to indicate that
the "boiler plate approach" dominates, in which case the G-10 is correct to push for
uniformity and a common best practice. Time will tell whether the differences that HPSS
389
have identified are important. The simulations that they report, however, suggest that they
are likely to be rather small.
To conclude, HPSS have written an interesting and thought-provoking paper. The
model that they have constructed is elegant, and captures important features of the multistage game which borrowers and creditors play. Nevertheless, I have some reservations
about the practical significance of the differences that they highlight.
390
COMMENT
Future of the International Financial Architecture
Dr. Guillermo Ortiz
Governor of the Bank of Mexico
I would like to begin by expressing my appreciation to the Banco de Espafia and the IMF
for inviting me to participate in this important conference to commemorate the sixtieth
anniversary of the Bretton Woods institutions.
Since the second half of the nineties there has been an almost continuous debate
and an extensive consideration of the changes that should be made to the architecture of
the international financial system to improve its effectiveness. Let me say at the outset that
after all these years of analysis and discussion, I remain convinced that:
The broad principles of the Bretton Woods institutions as expressed in the

articles of agreement remain valid today; and
The institutional changes should be evolutionary and not revolutionary.
Most of the recent debate on the present and future role of the IMF originated from
the economic crises that have been observed since the early nineties. Notwithstanding the
many criticisms to which the Fund has been subject as a result of its involvement in these
episodes, I am of the opinion that overall the IMF did a reasonable job. To a large extent,
the challenge for this institution has been to understand the opportunities and difficulties
that result from the impact of the globalization of capital markets on emerging economies.
My remarks will focus on four subjects related to these issues, which I consider
central to the future of the international financial architecture:
What have we learned from recent crises?

Achievements and challenges in crisis prevention.
Crisis resolution.
Governance of the international financial institutions.
The Nature of Recent Crises

The economic crises observed in emerging market economies since the second half of the
1990s present a dual nature that distinguishes them from those observed in previous years.
Almost all of the crisis economies went through typical balance of payments
problems that led to large current account deficits and appreciated real exchange rates,
caused mainly by excessive spending by private or public sectors, financed by credit
expansion and substantial capital inflows in the pre crisis stage. But the imbalances were
generally not large enough to explain the virulence of the crises that followed. In fact, the
391
main common element of these crises was the presence of financial panic and herd
behavior, which translated in capital account shocks.
Beyond the role played by weak macroeconomic fundamentals and insufficient
transparency, these events have shown us that countries are more prone to crisis under the
presence of balance sheet mismatches, and weak and inefficient domestic financial
markets.
The risks of a crisis are exacerbated under the presence of implicit or explicit
government guarantees. I would point in particular to two of them.
First, during the 1990s international investors swiftly realized that implicit or
explicit deposit insurance implied that in the case of a systemic crisis, liabilities of banks
were effectively a contingent liability of the public sector.
A second, and even more important guarantee, relates to the economy's exchange
rate system. Fixed exchange rate regimes (in the context of open capital markets)
encourage short term capital flows and currency mismatches, and accentuate the risks of
capital flow reversals and self-fulfilling crises.
In fact, the merits of flexible exchange rate systems for those countries integrated to
the international financial markets have been increasingly clear after the Argentine crisis,
and as new evidence has emerged suggesting that emerging market economies should
consider flexible exchange rate regimes as they develop economically and institutionally.1
The IMF could play a useful role here by providing technical assistance to help countries
meet the conditions needed for a successful transition to a floating regime.
There is a last element that deserves emphasis. Countries that suffered crises were
also characterized by deficient internal institutions: inappropriate frameworks for financial
regulation and supervision, weak legal frameworks and judicial institutions, government
inefficiency, widespread corruption, and so on.
Achievements and Challenges in Crisis Prevention
The improvements in the understanding of the nature of economic crises has allowed
substantial progress in crisis prevention. At the domestic policy level, we are far more
aware nowadays of the challenges faced by macroeconomic policies in an environment of
balance of payments pressures dominated by the capital account, of the importance of
strong and adequately supervised financial systems, and of the risks posed by balance sheet
mismatches, among others.
The strengthening of the framework for crisis prevention has also included major
efforts at the level of the IMF:
The framework for surveillance has been improved;

A huge amount of work has been carried out in the area of standards and
codes. As of May 30, 2004, 400 ROSC modules had been published for 93
countries;
A major contribution has been made by setting in motion the Financial
Sector Assessment Programs;
See K. Rogoff, A:M Husain, A. Mody, R. Brooks, and N. (Domes, Evolution and Performance of Exchange
Rate Regimes, IMF Working Paper, December 2003.
392
Transparency has improved substantially, both at the country and at the

Fund level; etc.
Mexico has benefited a lot from these efforts. Suffice to say that we have
undergone standards and codes assessments in 8 different areas, and that a Financial Sector
Assessment Program for Mexico was carried out in 2001, an exercise which proved to be
very valuable for us.
Although crisis prevention is one of the areas in which more tangible progress has
been made during the last years, there is clearly room for improvement. I would like to
refer to four specific issues in this connection:
First, I agree with the view that crisis prevention will be made more
effective if the Bretton Woods institutions complement each other in an
efficient way2 that implies that the Fund and the Bank must focus on their
core areas of responsibility. This is not an easy task. Traditionally,
cooperation between the Bank and the Fund has been a difficult endeavor.
At least to some extent, this has been the result of the fact that there are
some areas where these institutions' involvement is fundamental and where
the division of responsibilities is not totally clear. For instance, it is
important to include the Fund in the discussion of issues related to
institution building, because these are central to assess a country's credit
risk and its financial stability.
Second, the Fund must be flexible enough when assessing risks in any
economy. It is a good idea to use a set of vulnerability indicators to support
crisis prevention efforts. However, it would be a mistake to apply them
mechanically, without taking into consideration each country's particular
situation. There is no such thing as a crisis-proof toolbox.
Third, we must be aware that effective crisis prevention will require in
many cases a joint effort on the part of the country's authorities and the
international community. Take the cases of Brazil and Turkey. Both
countries have made strenuous efforts to achieve prudent macroeconomic
policies, increase their primary surpluses, implement structural reforms, etc.
And yet, in view of the large size of their public debts, they are very
vulnerable in an environment of rising world interest rates. More generally,
those countries which are highly integrated to international financial
markets but whose domestic markets are relatively illiquid and present
heavy debt burdens, are very exposed to speculative attacks during episodes
of turbulence. Furthermore, debt levels can only be reduced gradually.
What should the IMF do to be of help in these cases? I am convinced that
the key question for crisis prevention here is to create the perception in
international markets that both the resources of the institution and its access
policy are such, that these economies will receive adequate support in case
they need it. In other words, it is fundamental to avoid the perception that
See The Bush Administration's Reform Agenda at the Bretton Woods Institutions: A Progress Report and
Next Steps; Testimony by John B. Taylor, Under Secretary of Treasury for International Affairs before the
Gommittee on Banking, Housing, and Urban Affairs, United States Senate, May 19, 2004.
393
Fund financing will be subject to rigid rules or that countries are going to be
forced into debt rescheduling. There is an additional element worth noting.
The International Monetary Fund has been insisting on incorporating debt
sustainability assessments as a major element of its crisis prevention toolkit.
However, we must be very careful in making use of these analyses. As
should be evident from my previous comments, even with the same policy
stance debt sustainability can change drastically if our economic
assumptions-for instance, the degree of risk aversionare altered.
Fourth, I am of the opinion that the IMF would have made a valuable
contribution to crisis prevention by keeping a facility like the CCL alive.
This is clearly an area to which we should be devoting closer attention in
the near future.
Crisis Resolution
Unfortunately, progress in the area of crisis resolution has been far more modest than in
the case of crisis prevention. Let me begin with the consideration of the role of the Fund in
crisis resolution through its financial support. I wish to make three comments in this
respect.
First, economic crises today, as was the case many years ago, need to be
solved through a combination of adjustment and financing. The crucial
point nowadays is that with the dual nature of crises, dominated by the
capital account, an overshooting of both financing and policy adjustment
are needed. For this reason, the Fund has nowadays a far more prominent
role in attempting to give viability to the adjustment efforts of emerging
market economies facing economic crises.
Second, the Fund has responded to the challenges posed by recent capital
account crises by helping members to meet unusually large financing needs.
This has led to one of the main concerns that arise when discussing the role
of the IMF: large financing and its potential implications for moral hazard.
In my view, its importance has been greatly overstated. I have not seen as
yet any convincing evidence to support the notion that moral hazard
represents a serious difficulty.
Third, the involvement of the Fund in crisis resolution in Argentina has put
its reputation at stake. It is paradoxical and politically very questionable to
see that Argentina is nowadays less vulnerable to increases in international
interest rates because it is in default, while other emerging markets adhering
closely to the recommendations of the Fund face a more complicated
outlook. The Fund has been supporting Argentina under its policy of
lending into arrears. The latter rests on the presumption that "the member is
pursuing appropriate policies and is making a good faith effort to reach a
collaborative agreement with creditors". Given the doubts that have
emerged, the Fund should attempt to clarify the specific meaning of these
conditions. Broad political support of the membership is of course required.
394
Let me turn now to debt restructuring. I wish to say first of all that the Fund cannot
be criticized for lack of effort on this front. In fact, the SDRM represents one of the main
initiatives supported by Fund management in recent years. However, this initiative was
doomed to fail and fortunately it did. The reasons are well known.
The contractual approach to sovereign debt restructuring, based on a more
widespread use of Collective Action Clauses, represents a more realistic alternative. The
strong support for the contractual approach from most participants in the international
financial markets, the success of the Mexican as well as other subsequent issues of bonds
with CACs, and the absence of evidence showing an interest rate spread associated to the
use of CACs, have helped move the discussion forward. But I believe that the real test for
the usefulness of CACs will come only at the time when a country having a majority of
bonds issued with CACs faces the need to restructure its external debt. Evidently, it will
take years before we face a situation like this.
Another avenue that has been getting attention recently as a mean to strengthen the
framework for crisis resolution, is the design of a Code of Conduct for Sovereign Debt
Restructuring. It has been argued that, by providing guidance on how debtors and
creditors should behave during periods of difficulties, and by promoting dialogue between
them, the drafting of a Code of Conduct would contribute to an early resolution of
solvency problems. However, given the fact that the objectives of debtors and creditors are
misaligned, to be able to reach an agreement the Code will have to be written at a high
level of generality. Under such circumstances, its usefulness as a roadmap will likely be
very limited.
Improving Governance in IFIs
Let me comment briefly now on decision making at the IFIs. While as I said before, I am
not of the view that the Bretton Woods institutions need to be reinvented, we do need
organizations that are able to adapt to the challenges posed by an evolving world economy.
In this process, it is essential that we set in place the institutional adjustments
required to ensure that these objectives are met. An adequate decision making structure at
these institutions is essential to this end. The changes that have been observed in recent
years in the world economy have been accompanied by an increasing importance of the
emerging markets. Unfortunately, these developments have not been reflected in the IFIs
voting structure. For instance, in the case of the IMF, Horst Kohler underscored recently
the existing scope for raising the voting share of emerging market countries, "whose
quotas no longer reflect their true weight in the world economy."3 I am certain that a
revision of the voting structure aimed at allowing emerging markets to have a say in
decision making at the Fund consistent with their position in the world economy, is
essential if the Fund is to adjust its role in the world economy as needed in coming years.
Allow me one last comment before closing my intervention. We have spent by
now nearly a decade debating on the international financial architecture. These discussions
have been very productive, but they cannot go on forever. In my view, the time has come
to move on to put this issue behind us and move from the consideration of the international
financial architecture to the most concrete and substantive aspects of the international
economic policy arena.
Interview with Horst KShler, IMF Survey, May 2004.
395

IMF 2005 - Dollars Debt and Deficits - Sixty Years After Bretton Woods

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

IMF 2005 - Dollars Debt and Deficits - Sixty Years After Bretton Woods

Uploaded by

Copyright:

Available Formats

Dollars, Debt, and Deficits

Sixty Years after Bretton Woods

International Monetary Fund. Not for Redistribution

2005 International Monetary Fund

Please send orders to:

International Monetary Fund. Not for Redistribution

BLOCK I: GLOBAL IMBALANCES, EXCHANGE RATE

Session 1; Global Imbalances

Session 2; Exchange Rate Issues

Session 3; Debt in Emerging Economies

International Monetary Fund. Not for Redistribution

BLOCK II: INTERNATIONAL FINANCIAL ARCHITECTURE

Session 4; The Role of the IMF

Session 5; Innovations in Private and Multilateral Lending

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

Eduardo Levy Yeyati, Professor, University Torcuato Di Telia

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

frequency monitoring without increasing IMF exposure. This non-borrowing facility

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

This page intentionally left blank

International Monetary Fund. Not for Redistribution

Opening Remarks by Rodrigo de Rato y Figaredo,

Introduction: Welcome and a Few Words on the IMF's Mission

International Monetary Fund. Not for Redistribution

globalization is a welcome development. It provides opportunities to channel large-scale

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

financial support entirely is undesirable in view of the volatile character of international

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

This page intentionally left blank

International Monetary Fund. Not for Redistribution

Opening Remarks by Pedro Solbes, Second Vice-President of the

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

Finally, technical assistance should be included as a key plank of the

First, debt sustainability analysis should be a key issue of the

International Monetary Fund. Not for Redistribution

Opening Remarks by Jaime Caruana, Governor of Banco de Espana

International Monetary Fund. Not for Redistribution

oriented macroeconomic policies and, in particular, to conduct an independent monetary

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

This page intentionally left blank

International Monetary Fund. Not for Redistribution

International Monetary Fund. Not for Redistribution

This page intentionally left blank

International Monetary Fund. Not for Redistribution

A Map to the Revived Bretton Woods End Game: