Choice of Interest Rate Term Structure

Models for Pricing and

Hedging Bermudan Swaptions -An
ALM Perspective
Zhenke Guan
Bing Gan
Aisha Khan
Ser-Huang Poon

January 15, 2008

Zhenke Guan (zhenke.guan@postgrad.manchester.ac.uk) and Ser-Huang Poon (ser-

huang.poon@mbs.ac.uk) are at Manchester Business School, Crawford House, University of
Manchester, Oxford Road, Manchester M13 9PL, UK. Tel: +44 161 275 0431, Fax: +44 161
275 4023. We would like to thank Peter van der Wal, Vincent van Bergen and Jan Remmerswaal
from ABN AMRO for manu useful suggestions and data, Michael Croucher for NAG software
support, and Dick Stapleton and Marti Subrahmanyam for many helpful advice.
1
Abstract
Asset Liability Management (ALM) departments in big nancial institutions
have the problem of choosing the right interest rate model for managing their
books, which typically consists of assets and liabilities denominated in dierent
currencies. Should a bank adopt a single term structure (TS) model or is it
appropriate to allow dierent branches use dierent models for ALM? Each TS
model is based on dierent measure of interest rates (e.g. spot, forward or swap
rates) with a dierent assumed dynamics for the key rate. Is it appropriate to use
a single TS model for ALM when the bank is subject to interest rate exposure
in very dierent types of economy? This paper aims to address these questions
by comparing , empirically, four one-factor TS models, viz. Hull-White, Black-
Karasinski, Swap Market Model and Libor Market Model, for pricing and hedging
long term Bermudan swaptions which resemble mortgage loans in banks book.
In contrast to pricing need in the front oce, a single-factor model is preferred
for ALM purpose because of its focus in long term risk management. Managing
long term interest rate risk using a multi-factor models is cumbersome and the
multi-factors could potentially introduce more input errors over long horizon.
The test involved calibrating the four TS models to European swaption prices
for EURO and USD over the period February 2005 to September 2007. The
calibrated models are then used to price and hedge 11-year Bermudan swaption
with a 1-year holding period. The empirical results show that, the calibrated
parameters of all four models are very stable and their pricing error is small. The
pricing performance of four models is indistinguishable in both currencies. The
hedging P&L from the four models is similar for the Euro market. For the USD
market, the short rate models preforms marginally better than SMM and LMM.
The performances of HW and BK are indistinguishable.
1 Introduction
In this paper, we study the choice of interest rate term structure models for asset-
liability management in a global bank. In particular, we compare the performance
of Hull-White, Black-Karasinski, Swap market model and Libor Market Model
for hedging a 11-year Bermudan swaption on an annual basis from February
2005 to September 2006. The 11-year Bermudan swaption is chosen because it
resembles a loan portfolio with early redemption feature, an important product
for most banks. Unlike the short-term pricing problem, the one-factor model
is often preferred for the longer term ALM purpose because of its simplicity.
For long horizon hedging, the multi-factor model could produce more noise as it
requires more parameter input. Also, for this ALM study, we do not consider
the smile eect for the same reason. Pricing performance measures a models
capability of capturing the current term structure and market prices of interest
rate sensitive instruments. Pricing performance can always be improved, in an
almost sure sense, by adding more explanatory variables and complexities to the
dynamics. However, pricing performance alone cannot reect the models ability
in capturing the true term structure dynamics. To assess the appropriateness of
model dynamics, one has to study model forecasting and hedging performance.
The last few decades have seen the development of a great variety of interest
rate models for estimating prices and risk sensitivities of interest rate derivatives.
These models can be broadly divided into short rate, forward rate and market
models. The class of short-rate models, among others, includes Vasicek (1977),
Hull and White (1990), and Black and Karasinski (1991). A generalized frame-
work for arbitrage free forward-rate modelling originates from the work of Heath,
Jarrow and Morton (HJM, 1992). Market models are a class of models within the
HJM framework that model the evolution of rates that are directly observable in
the market. Brace, Gatarek and Musiela (1997) rstly introduced the arbitrage
free process for forward libor rates which lead to the Libor Market Model (LMM,
also known as BGM model) under the HJM framework. Miltersen, Sandmann
and Sondermann (1997) derive a unied interest rates term structure which gives
closed form solution for caplet under the assumption of log-normally distributed
interest rates. Dierent methods of parameterization and calibration of LMM
are examined in Brigo, Mercurio and Mrini (2003). The term structure of swap
rates is rstly developed in Jamshidian (1997) which is known as Swap Mar-
ket Model (SMM). Jamshidian (1997) also proposed the concept of co-terminal
market model at the earliest. Base on the research of the above papers, Galluc-
cio, Huang, Ly and Scaillet (2007) use graph theory to classify the admissible
market models into three subclasses named co-initial, co-sliding and co-terminal.
Among other things, they show that the LMM is the only admissible model for
swaps of a co-sliding type.
All these models have their own strengths and weaknesses. Short-rate models
are tractable, easy to understand and implement but do not provide complete
2
freedom in choosing the volatility structure. The HJM framework is popular due
to its exibility in terms of the number of factors that can be used and it permits
dierent volatility structures for forward rates with dierent maturities. Despite
these attractions the key problem associated with the HJM is that instantaneous
forward rates are not directly observable in the market and hence models under
this framework are dicult to calibrate. The market models overcome these limi-
tations but are complex and computationally expensive when compared with the
short rate models. The question is whether one should use the simple model like
Gaussian HW model or the lognormal BK model, or should one use complicated
model like SMM and LMM. In this paper, we look at the dierence between the
four interest rate models for longer term asset liability management in two inter-
est rate regimes (i.e. Euro and USD). The choice of HW and BK is simple: at
the time of writing, they are the most important and popular short rate models
used by the industry; the choice of swap market model and Libor market model
is also simple because they are new or more recent and are widely used among
practitioners.
More recently, a number of previous studies have examined various term struc-
ture models for pricing and hedging interest rate derivatives. A large part of the
term structure literature has focused on the performance of term structure mod-
els for bond pricing (see, for example, Dai and Singleton 2000, Pearson and Sun
1994). Andersen and Andreasen (2001) use mean-reverting Gaussian model and
lognormal Libor Market Model for pricing Bermudan swaption. They nd that
for both models, Bermudan swaption prices change only moderately when the
number of factors in the underlying interest rate model is increased from one to
two. Pietersz and Pelsser (2005) compare single factor Markov-functional and
multi-factor market models for hedging Bermudan swaptions. They nd that on
most trade days the Bermudan swaption prices estimated from these two mod-
els are similar and co-move together. Their results also show that delta and
delta-vega hedging performances of both models are comparable. Gupta and
Subrahmanyam (2005) compare various one- and two-factor models based on the
out-of-sample pricing performance, and the models ability to delta-hedge caps
and oors. They nd that one-factor short rate models with time varying pa-
rameters and the two-factor model produce similar size pricing errors. But in
terms of hedging caps and oors, the two-factor models are more eective. For
both pricing and hedging of caps and oors, the BK model is better than HW
model. With regard to both pricing and hedging, their results are in line with
those obtained by Fan, Gupta and Ritchken (2006). Driessen, Klaassen and Me-
lenberg (2003) use a range of term structure models to price and hedge caps and
(European) swaptions. Regarding hedging, their results show that if the number
of hedge instruments is equal to the number of factors then multi-factor models
outperform one-factor models in hedging caps and swaptions. However, when
a large set of hedge instruments is used then both one-factor and multi-factor
models perform well in terms of delta hedging of European swaptions. Fan,
3
Gupta and Ritchken (2006) show that one- and two-factor models are as capa-
ble of accurately price swaptions as higher order multi-factor models. However,
regarding hedging , their results show that multi-factor models are signicantly
better than one-factor models. In addition, their results for swaptions show that
using multiple instruments within a lower order model does not improve hedging
performance. These results dier from Driessen et al (2003) and Andersen and
Andreasen (2001), who lend support for the one-factor models. This paper con-
tributes to the literature and the existing empirical test in: i) By comparing a
wide choice of models including the latest SMM, ii) by hedging the more compli-
cated Bermudan swaption (instead of caplet or oorlet) and iii) by focusing on
ALM perspective.
Specically, the objective is to select a interest rate term structure model
which best price and delta hedge a 11-year Bermudan swaption in dierent
currencies-EUR, USD. On the contract starting date, we constructed a hedged
portfolio for the 11-year Bermudan swaption; one year later, we get the new value
of the hedged portfolio. The hedging prot and loss is calculated as the dierence
between the value of the options and portfolios. The data are from Datastream
and ABN AMRO.
This remaining of the paper is organized as follows. Next section briey
describes the models we use and the implementation design. Data is described in
Section 3. Section 4 deals with calibration procedures of all models. Calibration
instruments and calibration algorithm are discussed. In Section 5, Bermudan
swaption prices are compared within four models. Hedging performance and
Prot and Loss are reported in Section 6 and nally, we conclude in Section 7.
2 Term Structure Models
We are interested in comparing the pricing and hedging performance of four
widely used models-HW Model, BK model, Swap market model and Libor Mar-
ket model. In this section, we briey describe the four models and there charac-
teristics.
There are numerous models for pricing interest rate derivatives,which, broadly
speaking, can be divided into two categories: sport rate models and forward rate
models. Both HW and BK model are short rate models which specify the behavior
of short-term interest rate, r. Short rate model, as it evolved in the literature,
can be classied into equilibrium and no-arbitrage models. Equilibrium models
are also referred to as endogenous term structure models because the term
structure of interest rates is an output of, rather than an input to, these models.
If we have the initial zero-coupon bond curve from the market, the parameters of
the equilibrium models are chosen such that the models produce a zero-coupon
bond curve as close as possible to the one observed in the market. Vasicek (1977)
is the earliest and most famous general equilibrium short rate model. Since the
4
equilibrium models cannot reproduce exactly the initial yield curve, most traders
have very little condence in using these models to price complex interest rate
derivatives. Hence, no-arbitrage models designed to exactly match the current
term structure of interest rates are more popular. It is not possible to arbitrage
using simple interest rate instruments in this type of no-arbitrage models. Two
of the most important no-arbitrage short-rate models are the Hull-White model
(1990) and the Black-Karasinski (1991) model.
2.1 The Hull-White and BK Model
Hull and White (1990) propose an extension of the Vasicek model so that it can
be consistent with both the current term structure of spot interest rates and
the current term structure of interest-rate volatilities. According to the Hull-
White model, also referred to as the extended-Vasicek model, the instantaneous
short-rate process evolves under the risk-neutral measure as follows:
dr
t
= [
t
a
t
r
t
]dt +
t
dz, (1)
where , a and are deterministic functions of time. The function
t
is chosen so
that the model ts the initial term structure of interest rates. The other two time-
varying parameters, a
t
and
t
, enable the model to be tted to the initial volatility
of all zero coupon rates and to the volatility of short rate at all future times.
1
Hull and White (1994) note, however, that while a
t
and
t
allow the model to
be tted to the volatility structure at time zero, the resulting volatility term
structure could be non-stationary in the sense that the future volatility structure
implied by the model can be quite dierent from the volatility structure today.
On the contrary, when these two parameters are kept constant, the volatility
structure stays stationary but models consistency with market prices of e.g. caps
or swaptions can suer considerably. Thus there is a trade-o between tighter t
and model stationarity.
In HW model the distribution of short rate is Gaussian. Gaussian distribution
leads to a theoretical possibility of short rate going below zero. Like the Vasicek
model the possibility of a negative interest rate is a major drawback of this model.
A model that addresses the negative interest rate issue of the Hull-White
model is the Black and Karasinski (1991) model. In this model, the risk neutral
process for logarithm of the instantaneous spot rate, ln r
t
is
d ln r
t
= [
t
a
t
ln r
t
] dt +
t
dz, (2)
where r
0
(at t = 0) is a positive constant,
t
, a
t
and
t
are deterministic functions
of time. Equation (2) shows that the instantaneous short rate evolves as the
1
The initial volatility of all rates depends on (0) and a(t). The volatility of short rate at
future times is determined by (t) (Hull and White 1996, p.9).
5
exponential of an Ornstein-Uhlenbeck process with time-dependent coecients.
The function
t
is chosen so that the model ts the initial term structure of
interest rates. Functions a
t
and
t
are chosen so that the model can be tted to
the chosen market volatility.
Both HW and BK model are implemented by constructing a recombining
trinomial lattice for the short-term interest rate.
2.2 Libor Market Model
In the next two subsections, we are going to describe the two most advanced
model in interest rates-Swap Market Model and Libor Market Model. They have
attracted a lot of interests from both academic and practitioners mostly because
of their characteristics - they are consistent with market standard - using Black
formula to price (European) swaption and caplet.
Let L
i
(t), i = 1, 2, ...N as the forward Libor rates. as the time interval
between two Libor rates. The dynamics of forward Libor rate under the T
n+1
terminal measure is by
dL
i
(t)
L
i
(t)
=
N

j=i+1

j
L
j
(t)
ij
(t)
1 +
j
L
j
(t)
dt +
i
(t) dW
Q
T
n+1
(t) (3)
for 0 t min (T
i
, T
n+1
) , i = 1, , n, , N. Note that L
n
(t) is driftless under
the T
n+1
terminal measure when B
n+1
(t) is used as the numeraire.
2.2.1 Model Implementation
Unlike caplet valuation, to implement LMM for pricing swaptions, it is neces-
sary to simulate all forward rates under one measure. This means we need the
dynamics (the drift term) for all L
n
(t) under the terminal measure.
Here, we use a Euler discretization scheme for the forward rate simulation.
While Glasserman and Zhao (1999) note that forward prices of bonds are not
arbitrage free under the Euler discretization scheme, we adopt the Euler scheme
nevertheless due to its simplicity. The approximation error is not severe as long
as the step size is small. Following Glasserman and Zhao (1999), the discretized
version of the LMM dynamics under the T
N+1
terminal measure in (3) is, for a
time step size h,
L
i
((k + 1)h) = L
i
(kh) exp
_
(
i
(kh)
1
2

i
(kh)
2
)h +
i
(kh)

h
j+1
_
,
where

i
(kh) =
N

j=i+1
L
j
(kh)
ij
(kh)
1 + L
j
(kh)
.
6
Here, we have chosen n = N and T
N+1
as the terminal measure. Note that under
the T
N+1
terminal measure, when i = N, the drift of L
i
= L
N
is zero and L
N
is
log-normally distributed.
Predictor-corrector scheme is used here for more accuracy(See Joshi and
Stancy 2005 for more details).
2.3 Swap Market Model
Similar as Libor Market Model, swap market model has the same good property.
However, it is not as intensively studied as LMM. Recently, people begin to realize
the usefulness of co-terminal swap market model. Given the tenor structure, a
co-terminal Swap Market Model refers to a model which assigns the arbitrage free
dynamic to a set of forward swap rates that have dierent swap starting date T
n
(n = 1, , M 1) but conclude on the same maturity date T
M
. One important
feature about the co-terminal swaption is that it is internally consistent with a
Bermudan swaption that gives the holder the right to enter into a swap at each
reset date during period T
1
to T
M1
with T
M
being the terminal maturity of the
underlying swap. When considering at each tenor date whether or not to exercise
the option to enter into a swap contract, the holder need to consider the forward
swap rate dynamics from that tenor date till nal maturity, which is actually
driven by the volatility prevailing at that time. The advantage of co-terminal
SMM over other market models in pricing Bermudan swaptions has already been
noted and discussed in Jamshidian (1997) and Galluccio et al (2007).
In Galluccio et al (2007) the co-terminal SMM is dened by introducing a
collection of mutually equivalent probability measures and a family of Brownian
motions such that for any the forward swap rate satises a SDE for all . Here we
follow their approach.
2.3.1 Drift of Co-terminal SMM under the Terminal Measure
Let C
n,m
(t) =

m
i=n+1
B
i
(t) represent the value of the annuity from T
n+1
to
T
m
. The co-terminal forward swap rates satisfy the following general SDE under
the terminal measure:
dS
n,m
(t) = S
n,m
(t)
n,m
(t) dW
M1,M
t
+ drift for n = 1, , M 1 (4)
where W
M1,M
t
is the Wiener process under the terminal measure Q
M1,M
.
We follow Joshi and Liesch (2006), who recommend using the cross-variation
in assessing impact of changing numeraire on a drift. The drift of a forward swap
rate S
n,M
(t), n = 1, , M 1 under the terminal measure Q
M1,M
is given by:
E
M1,M
[dS
n,M
(t)] =
n,M
=
C
M1,M
(t)
C
n,M
(t)
_
B
n
(t) B
m
(t)
C
n,M
(t)
,
C
n,M
(t)
C
M1,M
(t)
_
=
C
M1,M
(t)
C
n,M
(t)
_
S
n,M
(t) ,
C
n,M
(t)
C
M1,M
(t)
_
(5)
7
Note that when n = M 1,
M1
= 0 because S
n,M
(t) , 1 = 0.
In Joshi and Liesch (2006), the S
n,M
(t) term inside the square brackets is
simplied into independent Wiener process W
k
, which explains why their drift
term does not contain the correlation terms
n,m
. Although for one factor model,
the correlation terms
n,m
= 1 can be ignored, it cannot be omitted when the
dimension of W
k
is two or more. Expanding the cross-variation term in equation
(5), the drift becomes a complicated function of numeraires and cross-variation
of two forward swap rates. By induction, the general form of drift under terminal
measure can also be written as:

n,M
=
M2

i=n
_

i+1
C
i+1,M
C
M1,M

n,i+1
S
n,M

n,M
S
i+1,M

i+1,M
i
j=n+1
(1 +
j
S
j,M
)
_
(6)
Details of the derivation are given in the working paper by Gan et al. However,
the above equation is not an explicit function of S
n,M
(t) for n = 1, , M 1,
because the C
n,M
term is derived from a set of S
i,M
(t) for n i M 1. This
drift term is complicated in appearance but not time-consuming in computation.
Similar as Libor market model, SMM is also implemented using Monte Carlo
simulation, in the interest of computational eciency. Predictor-corrector drift
approximation is employed for accuracy.
3 Research Design and Data
Let t denote a particular month in the period from February 2005 to September
2007. The procedure for calibrating, hedging and unwinding a 10 1 Bermudan
swaption are as follows:
(i) At month t, the interest rate model is calibrated to 10 ATM co-terminal
European swaptions underlying the 10 1 ATM Bermudan swaption by
minimising the root mean square of the pricing errors.
(ii) The calibrated model from (i) is then used to price the 101 ATM Bermu-
dan swaption at t, and to calculate the hedge ratios using, as hedge instru-
ments, 1-year, 5-year and 11-year swaps, all with zero initial swap value at
time t.
(iii) A delta hedged portfolio is formed by minimising the amount of delta mis-
match.
(iv) At time t + 1 (i.e. one year later), the short rate model is calibrated to
9 co-terminal ATM European swaptions (9 1, 8 2, ...) underlying the
Bermudan swaption from (ii) which is now 9 1.
8
(v) The calibrated model from (iv) is used to price the 91 Bermudan swaption,
and the time t +1 yield curve is used to price the three swaps in (ii), which
are now 0 year, 4 years and 10 years to maturity.
(vi) The prot and loss is calculated for the delta hedged portfolio formed at t
and unwound at t + 1.
(vii) Steps (i) to (vi) are repeated every month for t + 1, t + 2, till T 1
where T is the last month of the sample period.
To perform the valuation and hedging analyses described above, the following
data sets are collected from Datastream:
(a) Monthly prices (quoted in Black implied volatility) of ATM European swap-
tions in Euro and USD. Two sets of implied volatilities were collected: from
February 2005 to September 2006, prices of co-terminal ATM European
swaptions underlying the 10 1 Bermudan swaption, and from February
2006 to September 2007, prices of co-terminal ATM European swaptions
underlying the 9 1 Bermudan swaption. These prices are quoted in Black
implied volatility. The implied volatility matrix downloaded has a number
of missing entries especially in the earlier part of the sample period. The
missing entries were lled in using log-linear interpolation following Brigo
and Morini (2005, p 9, 24 and 25).
(b) Annual yields, R
0,t
, for maturities up to 11 years are downloaded from
Datastream.
2
All other yields needed for producing the trinomial tree are
calculated using linear interpolation. These annual yields are converted to
continuously compounded yields, r
0,t
= ln(1 + R
0,t
).
(c) Monthly data of the annual yield curve for the period January 1999 to July
2007, i.e. total 103 observations are downloaded for Euro and USD. This
data was transformed into discrete forward rates (as in LMM) for use in
the principal component analysis. Also, Swap rate can be derived from the
forward libor rates.
(iv) Monthly data of 1-month yield for the period January 2000 to July 2007
was downloaded for estimating the mean-reversion parameter. In the
implementation, a time step (t) of 0.1 year is used for constructing the
trinomial tress, which means that the rates on nodes of the tree are con-
tinuously compounded t-period rates. Here we have used the one-month
yield as a proxy for the rst 0.1-year short rate.
2
The time step (t) for the trinomial tress in the C++ program is 0.1 year. The C++
program linearly interpolates all the required yields.
9
4 Model Calibration
Before we price Bermudan swaptions, we need to nd the parameters that give the
correct European swaptions to avoid any arbitrage opportunities. This section
discusses dierent calibration method for each model.
4.1 HW/BK Model Calibration
In this section, we discuss HW/BK model calibration. As we know, for both
models, we have the mean reversion rate, time-dependent volatility
t
and the
constant rate of mean reversion a.
The mean reversion rate parameter for the two models has been extracted
from historical interest rate data. Practitioners and econometricians often use
historical data for inferring the rate of mean reversion (Bertrand Candelon
and Luis A. Gil-Alana (2006)). In this study we calibrate the two models with
time-dependent short rate volatility, (t), and constant rate of mean reversion,
a.
4.1.1 Parameterizations of (t)
There can be dierent ways to parameterise the time-dependent parameter: it can
be piecewise linear, piecewise constant or some other parametric functional form
can be chosen. In this study, the volatility parameter has been parameterised
as follows: The last payo for the all the instruments that need to be priced
in this study would be at 11 years point of their life. We explicitly decided
these three points, because values of (t), t = 0, 3, 11 on these three points
can be interpreted as instantaneous, short term and long term volatility. These
three volatility parameters are estimated through the calibration process. The
volatilities for the time periods in between these points are linearly interpolated.
4.1.2 Choosing calibration instruments
A common nancial practice is to calibrate the interest rate model using the
instruments that are as similar as possible to the instrument being valued and
hedged, rather than attempting to t the models to all available market data.
In this study the problem at hand is to price and hedge 10 1 Bermudan. For
this 10 1 Bermudan swaption the most relevant calibrating instruments are
the 1 10, 2 9, 3 8, ., 10 1 co-terminal European swaptions (A n m
swaption is an n-year European option to enter into a swap lasting for m years
after option maturity.). The intuition behind this strategy is that the model when
used with the parameters that minimize the pricing error of these individual
instruments would price any related instrument correctly. Therefore these 10
European swaptions are used for calibrating the two models for pricing 10 1
10
Bermudan swaption.
3
When the models are used for pricing 9 1 Bermudan
swaption we use nine European swaptions; 1 9, 2 8, ., 9 1 for calibrating
the models. (
0
,
3
and
11
for the 10x1 Bermudan swaption and
0
,
2
and
10
for the 9 1 Bermudan swaption)
4.1.3 Goodness-of-t measure for the calibration
The models are calibrated by minimizing the sum of squared percentage pricing
errors between the model and the market prices of the co-terminal European
swaptions, i.e. the goodness-of-t measure is
min
n

i=1
_
P
i,n,model
P
i,n,market
1
_
2
where P
i,n,market
is the market price and P
i,n,model
is the model generated price of
the i (n i) European swaptions, with n = 11 when models are calibrated to
price 10 1 year Bermudan swaption and n = 10 when models are calibrated to
price 9 1 year Bermudan swaption. Instead of minimising the sum of squared
percentage price errors alternatively we could have minimised the sum of squared
errors in prices. However, such a minimization strategy would place more weight
on the expensive instruments. Minimization of squared percentage pricing error
is typically used as a goodness-of-t measure for similar calibrations in literature
and by practitioners.
4.1.4 NAG routine used for calibration
We need to use some optimization technique to solve the minimization problem
mentioned in the last section. Various o-the shelf implementations are available
for the commonly uses optimization algorithms. We have used an optimisation
routine provided by the Numerical Algorithm Group (NAG) C library.
The NAG routine (e04unc) solves the non-linear least-squares problems using
the sequential quadratic programming (SQP) method. The problem is assumed
to be stated in the following form:
min
xR
n
F(x) =
1
2
n

i=1
{y
i
f
i
(x)}
2
where F(x) (the objective function) is a nonlinear function which can be repre-
sented as the sum of squares of m sub-functions (y
1
f
1
(x)), (y
2
f
2
(x)), ,
(y
n
f
n
(x)). The ys are constant. The user supplies an initial estimate of the
solution, together with functions that dene f(x) = (f
1
(x) , f
2
(x) , , f
n
(x))
and as many rst partial derivatives as possible; unspecied derivatives are ap-
proximated by nite dierences.
3
Pietersz and Pelsser (2005) followed the same approach.
11
In order to use this routine, for our calibration purpose we did following:
(i) Set s to 1.
(ii) n = 10/9 , depending on the number of calibration instruments
(iii) f
i
(x) =
P
i,n,model
P
i,n,market
(iv) Set initial estimates for all the three parameters as 0.01 i.e. 1%.
(v) Set a lower bound of 0.0001 and upper bound of 1 for the three parameters
to be estimated.
Because of the complexity of our objective function partial derivatives are not
specied and therefore the routine itself approximates partial derivatives by nite
dierences.
4.1.5 Estimating mean reversion parameter (a)
As stated in the start of this section, the mean-reversion parameter has been
estimated from historical data of interest rates. To measure the presence of mean
reversion in interest rates we need large data set. This study spans over a period
of one year, and within this one-year period models are calibrated every month.
Hence it does not make sense to estimate this parameter for any economy on a
monthly basis. Therefore, using historical data, once we estimate the value of
this parameter and then use this value in all the tests.
For the rate of mean reversion parameter a rst order autocorrelation of
the 1 month interest rate series has been used. Here we present the basic idea
behind the estimation procedure used. Under the HW model, the continuous
time representation of the short rate process is
dr
t
= [
t
a r
t
]dt +
t
dz,
The discrete-time version of this process would be
r
t+1
r
t
= [
t
ar
t
] +
t+1
,
r
t+1
=
t
+ (1 a)r
t
+
t+1
46 (7)
where
t+1
is a drawing from a normal distribution. Equation (46) represents an
AR(1) process.
An autoregressive (AR) process is one, where the current values of a variable
depends only upon the values that variable took in previous periods plus an error
term. A process y
t
is autoregressive of order p if
y
t
=
0
+
1
y
t1
+
2
y
t2
+ .... +
t
,
t
N(0,
2
).
12
An ordinary least square (OLS) estimate of coecient (1 a) in equation (46)

would be
4
1 a =

=
(r
t+1
)(r
t
)

2
(r
t
)
=
a = 1
where is the correlation coecient between r
t+1
and r
t
and can be easily calcu-
lated using Excel. For the BK model we perform this regression using time series
of ln(r), where as before r, is 1M interest rate.
4.2 Libor Market Model Calibration
This section deals with the issue of Libor Market Model calibration. We observed
in the previous section that, in LMM, the joint evolution of forward Libor rates
under the pricing measure is fully determined by their instantaneous volatilities
and correlations. Therefore, a well specied parameters is essential for obtaining
correct prices and hedge ratios for exotic interest rate derivatives. Unfortunately,
in general, they are not directly observable. Therefore, model parameters must
be calibrated , i.e., they must be inferred either from time series of forward
rates or from market prices of vanilla derivative contracts, or from a combination
of both.
Here we follow Lvov(2005) the diagonal recursive calibration procedure to
the co-terminal European swaptions. The diagonal recursive calibration need to
make assumptions of volatility of forward rates.
Assumption: The instantaneous volatility term structure has the following form

k
(t) =
k
(T
k
t, ) (8)
where T
k
is the expiry time of rate k, is a function imposing a qualitative shape
on the volatility term structure, is a vector of volatility parameters and
k
is
a rate-specic multiplier. Brigo and Mercurio (2001) demonstrated that this
form has a potential to produce an economically meaningful volatility structure
while allowing for a satisfactory calibration to market data.
4.2.1 Tested Parameterizations
We could have dierent parameterizations of the volatility term structure. This
volatility parametric form was suggested by Rebonato (1999) which is tested
extensively in the literature and is proved to possess very good properties.
4
For the regression y
i
= + x
i
+
i
,, the OLS estimate of is

xy
=

xy

2
xy
13
The parametric form as in Rebonato(1999) as

i
(t) = (i)[(a + b(T
i
t))exp(c(T
i
t)) + d] (9)
In our test, we make (i) = 1. This parametric specication of the instantaneous
volatility, in contrast with a piecewise-const form, preserves the qualitative shape
of the implied volatility observed in the market, i.e., a hump at around two years.
4.3 Swap Market Model Calibration
Brigo, Mercurio and Morini(2003), Rebonato(2003), and other literatures at-
tribute a lot to the parameterisation and calibration of LMM. Based on their
research, Galluccio et al (2007) test calibration of co-terminal SMM s using sim-
ilar methodology as the one advocated by Rebonato(2003) in the case of LMM.
By calibrating to swaption and caplet ATM volatilities, they nally get dierent
set of parameters {
j
, a
j
, b
j
, c
j
, d
j
} where j = 1, , M 1 for M 1 number
of co-terminal swaptions under dierent forward measure. Since the model built
in this paper is under one measure and the volatility of dierent co-terminal
swap rate will aect the drift terms in a complicated way, it is unrealistic to use
dierent parameter set for dierent co-terminal swaption. Therefore, a similar
parameter formula is followed here, but four common parameters will be allocated
to volatilities of all co-terminal swaptions.
In order to capture the term structure of swaption volatilities which are peri-
odically deterministic, the linear-exponential formulation proposed in Brigo, Mer-
curio and Morini (2003) is adopted here as swaption volatility function. Further
more the swaption volatility is formulated as

n,m
(t) =
n,M
(T
n
t, a, b, c, d)
(T
n
t, a, b, c, d) = [a + b (T
n
t)] e
c(T
n
t)
+ d14 (10)
The term can preserve the well-known humped shape of market quoted Black
implied volatility.
The method adopted in this paper is to calibrate parameter a, b, c, and d to
the implied volatilities of a set of co-terminal swaptions with dierent maturities
but associated with same length of swaps. The parametric form is as below
v (t, M) = (t, a, b, c, d) = [a + bt] e
ct
+ d for T
0
< t < T
M
where v (t, M) is the market implied volatility, and t denotes the maturity of
swaption. Because of its simple log-linear character, calibration algorithm using
this parametric formula is quite fast and robust.
When performing calibration with DRC method, we use Sequential Quadratic
Programming algorithm from NAG C library to nd the optimal parameters.
This routine is based on the algorithm suggested in Gill et al (1986). Due to the
14
complexity of our target function, we approximate all partial derivatives using
nite dierences. This however does not appear to preclude the convergence of
the optimization routine.
4.4 Calibration Results
Calibration results are summarized in Table(1). The calibration is performed
on the last business date of each month from February 2005 to September 2006.
Table 1 , Figure 1 and Figure 2 describe the errors by the calibration from dierent
point of view. Generally, the root mean square error is quite small and we say
that the model is very well calibrated.
Insert Table 1 ,Figure 1,Figure 2 Here
We also want to have the calibrated parameters to be stationary. In Table 2
and Figure 3, we nd that given our parametric form of volatility, the parameters
are indeed stable.
Insert Table 2(a) and 2(b) and Figure 3 here
Several conclusions could be drawn at this stage. First, it has been shown
that all four models could calibrate to co-terminal European swaption implied
volatilities matrix pretty well. The calibration methods are quite stable for all
the testing dates in our data sets. This gives us rm ground to proceed with
pricing and hedging Bermudan swaptions.
5 Pricing Bermudan Swaption
We have recombining trinomial tree for pricing Bermudans swaptions with short
rate models. For Swap Market Model and Libor Market Model, we perform Monte
Carlo simulation with Longsta-Schwartz Least Square Method.As Bermudan
swaption are exotic interest rate derivative product, there is no market quoted
price for it. We compare the price calculated from HW, BK, SMM and LMM.
In theory, if calibrated appropriately, all models could give similar prices for
the same Bermudan swaption. The rst part, we are pricing a 11-year Bermudan
swaption whereas in the second part we are pricing a 10-year Bermudan swaption
(with the same strike as 1-year before). We price the same Bermudan swaption
on the date of holding the option.
Insert Table 3(a) and 3(b) and Figure 4 here
The results are summarized in Table (3) and Figure 4. As we could see
from the graph and tables, on most dates, the prices given by four models are
indistinguishable.
15
6 Hedging Bermudan Swaption
Changes in the term structure can adversely aect the value of any interest rate
based asset or liability. Therefore, protecting xed income securities from un-
favorable term structure movements or hedging is one of the most demanding
tasks for any nancial institutions and for the ALM group in particular. Ine-
cient hedging strategies can cost big prices to these institutes. In order to protect
a liability from possible future interest rate changes rst one need to generate re-
alistic scenarios and then need to know how the impacts of these scenarios can
be neutralized. Thus, two important issues to be addressed by any interest rate
risk management strategy are: (i) how to perturb the term structure to imitate
possible term structure movements (ii) how to immunise the portfolio against
these movements.
Next sections explain the methodologies applied in this study for
(i) estimating the perturbations by which the input term structure had been
bumped to simulate the possible future changes in the input term structure
(ii) selecting hedge instruments
(iii) delta-hedging the underlying Bermudan swaption using the selected instru-
ments.
(iv) calculating possible prots and losses (P&L).
In each section we have examples from literature have been referred to justify
the choices made.
6.1 Perturbing the term structure
Over the years researchers and practitioners have been using duration analysis for
interest rate risk management, i.e. they shift the entire yield curve upward and
downward in a parallel manner and then estimate how the value of their portfolio
is aected as a result of these parallel perturbations. They then hedge themselves
against these risks. Parallel shifts are unambiguously the most important kind
of yield curve shift but alone cannot explain completely explain the variations of
yield curve observed in market. Three most commonly observed term structure
shifts are: Parallel Shift where the entire curve goes up or down by same amount;
Tilt, also known as slope shift, in which short yields fall and long yields rise (or
vice versa); Curvature shift in which short and long yields rise while mid-range
yields fall (or vice versa). These three shifts together can explain almost all the
variance present in any term-structure and thus and one should not completely
rely on duration and convexity measures for estimating the risk sensitivity of a
xed income security. There are numerous examples in literature to support this
16
argument. Here, we mention a few studies that lead to this conclusion. Litterman
and Scheikman (1991) performed principal components analysis (PCA) and found
that on average three factors, referred to as level (roughly parallel shift), slope,
and curvature, can explain 98.4% of the variation on Treasury bond returns.
They suggested that by considering the eects of each of these three factors on
a portfolio, one can achieve a better hedged position than by holding a only a zero-
duration portfolio.
5
Knez, Litterman, and Scheikman (1994) investigated the
common factors in money markets and again they found that on average three-
factors can explain 86% of the total variation in most money market returns
whereas on average four factors can explain 90% of this variation. Chen and
Fu (2002) performed a PCA on yield curve and found that the rst four factors
capture over 99.99% of the yield curve variation. They too claimed that hedging
against these factors would lead to a more stable portfolio and thus superior
hedging performance.
Insert Figure4, Figure 5 and Figure 6 Here
Based on the ndings of these studies, in this study we performed PCA on
historical data for estimating more realistic term structure shifts. In this study
we performed PCA on annual changes of the forward rates and used scores of the
rst three principal components for estimating the shifts by which we bumped the
forward rate curves. There are not many examples of estimating price sensitivities
w.r.t. multiple factors (like 3 principal components here) with a one-factor term
structure model. Generally risk sensitivities are calculated by perturbing only the
model intrinsic factors i.e. for the one-factor model, only one-factor is perturbed
and so on.
PCA has been performed on annual changes of forward Libor rates. Annual
changes have been used because each hedge is maintained for one year. The
reason for using forward rates rather than yield curve for doing PCA is two folds:
rst forward Libor rates are directly observable in market. Second using forward
curves easily we can construct zero-curve and swap curve (needed for estimating
interest rate sensitivities of swaps that are used for hedging). Without going into
the mathematical details of this procedure, here, we briey describe how PCA
has been done for estimating the term structure shifts/bumps in the study:
6
Monthly observations of 11 forward Libor rates (f
0,0,1
, f
0,1,2
, f
0,2,3
, ..., f
0,10,11
) for
the period Jan 1999 to Dec 2006 have been used for PCA. Explicitly these
rates have been used because we give annual zero curve of maturity till 11
years as input to our short rat models and we want to estimate bumps for
all these maturities.
7
5
Litterman and Scheikman (1991, 54)
6
For details on PCA refer http://csnet.otago.ac.nz/cosc453/student tutorials/principal components.pdf
7
As we know that this dissertation is a part of a big project. The models to be implemented
are LMM and SMM. These maturity forward rates are also needed by these two models.
17
Using these monthly observations of the 11 forward rates we calculate annual
changes for each of the 11 forward rates as follows. Suppose we have
monthly observations of the 11 forward rates for a period of n years i.e. in to-
tal we have 12n monthly observations of forward Libor rates f
0,
j
,
j
+1
(t
i
),
where i = 1, 2, , 12n and
j
= 0, 1, 2, .., 10. When the observations are
arranged in ascending order (by date) then annual change for the forward
Libor rates can be calculated as
cL
i,j
= f
0,
j
,
j
+1
(t
i
+ 12) f
0,
j
,
j
+1
(t
i
)
now i = 1, 2, . . . , 12 (n 1). These values would form a 12(n-1) x 11
matrix i.e. we have 12 less entries. To clear this, say t
i
= Jan 1999, and

j
= 0. Then f
0,0,1
(t
i
)is value of f
0,0,1
on Jan 1999; f
0,0,1
(t
i
+12)is the value
of f
0,0,1
on Jan 2000 and cL is the one-year change in the market observed
value of f
0,0,1
.
The matrix cL
i,j
is given as input to the NAG routine (g03aac) that performs
principal component analysis on the input data matrix and returns prin-
cipal component loadings and the principal component scores. The other
important statistics of the principal component analysis reported by the
routine are : the eigen values associated with each of the principal com-
ponents included in analysis and the proportion of variation explained by
each principal component.
We used the scores of rst three factors to calculate the three types of shifts for
the 11 forward Libor rates using following regression:
f
0,
j
,
j
+1
= +

j
,1
P
1
+

j
,2
P
2
+

j
,2
P
3
+
Where
j
= 0, 1, 2, .., 10 and P
k
is the vector of scores for the k
th
factor
k = 1, 2, 3. Nag routine (g02dac) has been employed to perform this re-
gression. The routine computes parameter estimates, the standard errors of
the parameter estimates, the variancecovariance matrix of the parameter
estimates and the residual sum of squares.
After performing the regressions specied above, the 11 forward rates are bumped
by shocks corresponding to the rst three factors as:
f

0,
j
,
j
+1
= f
0,
j
,
j
+1

j,k
P
k
where
j
= 0, 1, 2, .., 10, and k = 1, 2, 3.
6.2 Choosing the hedge instruments
Selecting appropriate hedge instruments is a critical part of a successful hedg-
ing strategy. In literature there are evidences of two hedging strategies, factor
18
hedging and bucket hedging. For factor hedging, in a K-factor model, K dierent
instruments (together with the money market account) are used to hedge any
derivative. The choice of hedge instruments is independent of the derivative to
be hedged i.e., the same K hedging instruments can be used for hedging any
derivative in a K-factor model, and depend only on the number of factors in the
model. For bucket hedging the choice of hedge instruments depend on the instru-
ment to be hedged and not on the factors in the model. In this hedging strategy
number of hedge instruments is equal to the number of total payos provided by
the instrument. The hedge instruments are chosen so their maturities correspond
to dierent payment or decision dates of the underlying derivative.
On using any other criteria for selecting the hedge instruments, the num-
ber of hedge instruments will lie between the numbers of hedge instruments for
these two hedging strategies. Before discussing the instruments that are used to
delta-hedge the Bermudan swaption in this study, in this section rst we briey
discuss a few examples from literature. Driessen, Klaassen and Melenberg (2002)
used (delta-) hedging of caps and swaptions as criteria for comparing the hedge
performance of HJM class models and Libor market models. They used zero
coupon bonds as hedge instruments. For each model, they considered factor and
bucket hedging strategies. DKM show that when bucket strategies are used for
hedging, the performance of the one-factor models improves signicantly Fan,
Gupta and Ritchken (2006) also used eectiveness of delta neutral hedges (for
swaptions) as criteria for comparing the hedging performance of single factor and
multi-factor factor term structure models. They used discount bonds to delta-
hedge swaptions. In this study they rst applied factor hedging for choosing the
hedge instruments and found that in context of hedging performance, multifac-
tor models outperform single factor models. Next in light of the DKM results,
they repeated their experiments using additional hedging instruments. They
found that for the one-factor and two factor models adding more instruments
did not result into better hedge results. Pietersz and Pelsser (2005) compared
joint delta-vega hedging performance (for 10x1 Bermudan swaption) of single
factor Markov-functional and multi-factor market models. They used the bucket
hedging strategy and set up hedge portfolios using 11 discount bonds, one dis-
count bond for each tenor time associated with the deal. They found that joint
delta-vega hedging performance of both models is comparable
The general implications of these examples are:
(i) Eectiveness of delta neutral hedges is often used to evaluate the hedging
performance of term structure models.
(ii) Using multiple instruments can improve the hedge performance of one factor
models. Practitioners also favour this practice
Therefore, in this study, we decide to use two dierent swaps of maturities 5
and 11years as hedge instruments. Maturity of 11-year swap coincides with the
19
maturity of the co-terminal Bermudan swaption to be hedged and the length of
other swap is almost half way the life of this Bermudan swaption. We could have
used discount bonds to hedge this Bermudan swaption
8
but use of swap is more
in line with the general practitioners practice. Studies suggest that large banks
tend to use interest rate swaps more intensively for hedging.
6.3 Constructing delta hedged portfolio
Delta hedging is the process of keeping the delta of a portfolio equal to or as
close as possible to zero. Since delta measures the exposure of a derivative to
changes in the value of the underlying, the overall value of a portfolio remains
unchanged for small changes in the price of its underlying instrument. A delta
hedged portfolio is established by buying or selling an amount of the underlier
that corresponds to the delta of the portfolio.
For interest rate derivatives,if the entire initial term structure is perturbed
by same amount say (parallel shift), then the risk sensitivity of a xed income
security w.r.t. this perturbation can be estimated as
V () V

where V is the value of the derivative calculated using initial term structure and
V () is the value of the derivative after the initial term structure is perturbed by
. If we rst increase the entire initial term structure by and then next decrease
it by , then the risk sensitivity can estimated as
V (
+
) V (

)
2
(11)
where V (
+
) is the value of the derivative calculated after initial term structure
has been shifted up by and V (

) is the value of the derivative after the initial

term structure has been shifted down by . In our case we have bumped the
initial forward rate curve by three factors. Also for each factor, we have bumped
the forward rate curve both up and down. Using the idea presented in equation
(11), we estimate the sensitivity (delta) of the Bermudan swaption and the two
swaps w.r.t the three factors as follows

B
k
=
B
P
k
=
B
+
k
B

k
P
+
k
P

k
=
B
+
k
B

k
2P
k

S
5
k
=
S
5
P
k
=
(S
5
)
+
k
B

k
P
+
k
P

k
=
(S
5
)
+
k
B

k
2P
k

S
11
k
=
S
11
P
k
=
(S
11
)
+
k
B

k
P
+
k
P

k
=
(S
11
)
+
k
B

k
2P
k
8
Pietersz and Pelsser (2005) used discount bonds to hedge Bermudan swaption
20
for k = 1, 2, 3, and (.)
+
k
and (.)

k
respectively are the prices of derivative after the
initial forward curve has been bumped up and down by the k
th
factor. Now if
we consider a portfolio, consisting of one 10 1 Bermudan swaption, x
11
units
of 11-year swap and x
5
units of 5-year swap, then total delta mismatch of this
portfolio w.r.t. k
th
factor,
k
is

k
=
B
k
x
11

S
11
k
x
5

S
5
k
(12)
x
11
and x
5
need not be whole numbers. This is the usual assumption of perfectly
divisible securities. From equation (12), we can see that when we use more than
one hedging instrument in a one-factor model, the hedge ratios would not be
unique, and some rule must be applied for constructing the hedge portfolio using
the chosen hedge instruments. Here to obtain the hedge ratios, x
11
and x
5
, we
use the basic idea behind the delta hedging. The two hedge ratios x
11
and x
5
are
obtained by minimising the total delta-mismatch of the portfolio w.r.t. the rst
three PCA factors i.e.
min
x
11
,x
5

3
k=1

2
k
where
k
is given by equation (12). Some optimization technique needs to be
applied to solve this minimisation problem. We use the C implementation of
Downhill Simplex Method provided by NAG to estimate x
11
and x
5
.
9
The nag
routine (e04ccc) minimises a general function F(x) of n independent variables
x = (x
1
, x
2
, .., x
n
)
T
.
6.4 Calculating P&L
Once hedge has been established on say day t, the hedge error can be evaluated
one year later on day t + 1 as follows
P&L
t+1
= (B
t
+ x
11,t
S
11,t
+ x
5,t
S
5,t
) (1 + i
0,t
)
(B
t+1
+ x
11,t
S
11,t+1
+ x
5,t
S
5,t+1
) (13)
where B
t
is the value of a 10 1 Bermudan swaption on day t; x
k,t
are units of
k-year swap in the hedge portfolio; S
k,t
is the value of k-year swap on day t; and
i
0,t:
is the forward Libor f
0,0,1
on day t.
At the point of initiation the value of any swap is zero. This means S
11,t
= 0
and also S
5,t
= 0. Therefore, equation (13) can be written as
P&L
t+1
= B
t
(1 + i
0,t
) B
t+1
x
11,t
S
11,t+1
x
5,t
S
5,t+1
. (14)
One year later, on day t+1, when hedge portfolio is unwound the 101 Bermudan
swaption is a 9 1 Bermudan swaption. On day t + 1, we recalibrate the model
9
The downhill simplex method (Spendley 1962, Nelder and Mead 1965, Press et al. 1992)
is an ecient algorithm for solving unconstrained minimisation problems.
21
and using new parameters calculate the model price of this away-from-money
9 1 Bermudan swaption B
t+1.
Strike rate for the 9 1 Bermudan swaption is
kept same as it was for the 10 1 Bermudan swaption on day t, as objective is
to nd the current value of that old swaption. We also calculate the values of
the two swaps on day t + 1 Using the example of 5-year swap we show, how the
swaps are re-evaluated one year later on day t + 1.
At the time of swap initialization, i
0,t
is known
i
0,t
=
_
1
P(0, 1)
1
_
and the swap rate k
5,t
for a 5-year swap on day t and can be calculated as
k
5,t
=
1 P(0, 5)
P(0, 1) + P(0, 2) + ... + P(0, 5)
At time t + 1, this 5-year swap is just a 4-year swap and also oating rate i
0,t+1
for the next reset is now known. The value of this swap at t + 1 would be
S
5,t+1
= (i
0,t
k
5
) + (i
0,t+1
k
5
)P(0, 1) + (
1,t+1
k
5
)P(0, 2)
+(
2,t+1
k
5
)P(0, 3) + (
3,t+1
k
5
)P(0, 4)
= (i
0,t
k
5
) +
__
1
P(0, 1)
1
_
k
5
_
P(0, 1)
+
__
P(0, 1)
P(0, 2)
1
_
k
5
_
P(0, 2) +
__
P(0, 2)
P(0, 3)
1
_
k
5
_
P(0, 3)
+
__
P(0, 3)
P(0, 4)
1
_
k
5
_
P(0, 4)
= (i
0,t
k
5
) + [1 P(0, 4)] k
5
[P(0, 1) + P(0, 2) + P(0, 3) + P(0, 4)]
= (i
0,t
k
5
) +
__
1 P(0, 4)
P(0, 1) + P(0, 2) + P(0, 3) + P(0, 4)
_
k
5
_
[P(0, 1) + P(0, 2) + P(0, 3) + P(0, 4)]
= (i
0,t
k
5
) + [k
4,t+1
k
5
] [P(0, 1) + P(0, 2) + P(0, 3) + P(0, 4)]
where k
4,t+1
would be the swap rate for any 4-year swap that would be initiated
on day t +1. When the swap is re-evaluated on day t +1 all the discount bonds
prices are computed using the yield curve observed on that day.
The hedging prot and loss is reported in Table 6 and Figure 7. As in Euro
market, the hedging P &L from the four models are very similar to each other.
In USD market, HW and BK model is marginally better than more complicated
SMM/LMM model. This would lead us to conclude that, from ALM perspective,
we actually could use simple short rate models, like HW and BK model, to
management the Bermudan swaption portfolios. The performance of HW and
BK are indistinguishable.
Insert Table 6 and Figure 7 Here
22
7 Conclusion
The goal of this paper is to provide an empirical analysis and comparison of
four one-factor interest rate term structure models from ALM perspective. In
contrast to previous research, we have compared four very dierent models and
we performed the pricing and hedging tests for Bermudan swaptions. We also
use a more rened method in delta ratio calculation during the hedging process.
For model calibration, HW and BK model use time-dependent volatility for
better calibration result and LMM and SMM model use market standard para-
metric form for volatility. Though the modelling approaches are vastly dierent,
the error from calibration is quite small for all four models and all the calibrated
parameters are quite stable.
Bermudan swaptions are priced via recombining trinomial tree with HW/BK
model while for SMM and LMM, Monte Carlo simulation with predictor-corrector
drift approximation is used. Most of the time, the prices from four models are
within the bid-ask spread. Thus we conclude that, for one factor model, we could
choose any of the four models for pricing as long as they calibrated to the same
instruments.
Minimizing the variations of hedging prot and loss is the objective for ALM.
In Euro market, the hedging P&L from the four models are very similar to each
other. In USD market, HW and BK model is marginally better than more com-
plicated SMM/LMM model. This would lead us to conclude that, from ALM
perspective, we actually could use simple short rate models, like HW and BK
model, to management the Bermudan swaption portfolios. The performance of
HW and BK are indistinguishable.
This analysis could have an immediate impact on the decision making of
the ALM department of big nancial institution on choosing the right interest
rate term structure model to manage their balance sheet. Future research could
applying the same test procedure on interest rate from a dierent economic regime
such as that of emerging market or a developing country.
References
[1] Andersen L. 2000, A simple approach to the pricng of Bermudan swaptions
in the multi-factor LMM, Journal of computational Finance, 2000.
[2] Andersen L. 2001, How many factors, really?, Journal of Finanical Econome-
tirc, 2001
[3] Andersen, L. and Andreasen, J. (2000) : Volatility Skews and Extensions of
the Libor Market Model, Applied Mathematical Finance 7(1):1-32.
[4] Andersen, L. and Andreasen, J. : Volatile volatilities, Risk 15(12), December
2002.
23
[5] Andersen, L., Brotherton-Ratclie, R. :2001, Extended Libor Market Models
with Stochastic Volatility, Working Paper, ssrn.
[6] Bertrand Candelon and Luis A. Gil-Alana (2006) Mean Reversion of Short-
run Interest Rates in Emerging Countries, Review of International Eco-
nomics, 14(1), 119135.
[7] Black, 1976, The pricing of commodity contracts, the Jounal of Financial
Econometrics
[8] Black F. and P. Karasinski (1991) Bond and option pricing when short rates
are lognormal, Financial Analysts Journal 47(4), 5259.
[9] BRACE, A., GATAREK, D., MUSIELA, M. (1997). The market model of
interest rate dynamics. Mathematical Finance, Vol. 7, 127-154.
[10] Brigo, Morini 2003, An empirically ecient analytical cascade calibration of
the LIBOR Market Model based only on directly quoted swaptions data
[11] Driessen J., P. Klaassen and B. Melenberg (2003) The Performance of Multi-
factor Term Structure Models for Pricing and Hedging Caps and Swaptions,
Journal of Financial and Quantitative Analysis, 38, 635-672.
[12] Fan R., Gupta A. and P. Ritchken (2006) On Pricing and Hedging in the
Swaption Market: How Many Factors, Really? SSRN Working paper
[13] Gill P E, Murray W, Saunders M A and Wright M H (1986) Users guide for
NPSOL (Version 4.0): a Fortran package for nonlinear programming Report
SOL 86-2 Department of Operations Research,Stanford University
[14] Glasserman P., X,Zhao, Arbitrage-free Discretizaton of LMM , Finance and
Stochastic, 1999.
[15] Gupta, A., and M. Subrahmanyam (2005) Pricing and Hedging Interest Rate
Options: Evidence from Cap-Floor Markets, Journal of Banking and Fi-
nance, 29, 701-733.
[16] Hull J. and A. White (1993), One-factor Interest Rate Models and the Valu-
ation of Interest Rate Derivative securities, Journal of Financial and Quan-
titative Analysis, 28, 235-254.
[17] Hull J. C. and White, A. (1994) Numerical Procedures for Implementing
Term Structure Models I: Single Factor Models, Journal of Derivatives, Fall
7-16.
[18] Hull J. C. and White, A. (1994), Numerical Procedures for Implementing
Term Structure Models II: Two Factor Models, Journal of Derivatives, Win-
ter 37-47.
24
[19] Hull J. C. and White, A. (1996), Using Hull-White Interest-Rate Trees,
The Journal of Derivatives, Winter Vol. 3, No. 3, 26-36.
[20] Hull, White, 2000,Forward rate volatilities, swap rate volatilities, and the
implementation,
[21] Hull J. C. and A.White, (Nov-Dec 2001), The General Hull-White Model
and Super Calibration, Financial Analysts Journal, 57, 6.
[22] Hull, John (2006) Options, Futures and Other Derivatives, London: Pear-
son/Prentice Hall, 6th edition.
[23] JAMSHIDIAN, F. (1997). Libor and swap market models and measures.
Finance and Stochas- tics, 4, pp. 293-330.
[24] JAMSHIDIAN, F., ZHU, Y. (1997). Scenario Simulation: Theory and
methodology. Finance and Stochastics, 1, pp. 43-67.
[25] Joshi,M., and Lorenzo Liesch (2006), Eective Implementation of Generic
Market Models. Working paper Available at Social Science Research Net-
work (SSRN), http://ssrn.com/abstract=907387
[26] Litterman R., J. Scheikman and L. Weiss (1991) Volatility and the Yield
Curve, Journal of Fixed Income 49-53.
[27] Longsta, Schwartz, A simple approach for pricing American with Simula-
tion, Jornal of nancial econometrics, 2001.
[28] Longsta, Schwartz, Throwing Billion dollars away, Jornal of nancial
econometrics, 2001.
[29] Lvov 2005, Monte Carlo methods for pricing and hedging: Application to
Bermudan swaptions and convertible bonds
[30] Miltersen, K.R., Sandmann, K., Sondermann, D., 1997. Closed form solu-
tions for term structure derivatives with log-normal interest rates. Journal
of Finance 52, 409C430.
[31] Piesz R., A.Pelsser,R Fast drift approximation in LMM, Journal of compu-
tational nance , 2001.
[32] Pietersz Raoul and Antoon A. J. Pelsser (2005) A Comparison of Single
Factor Markov-functional and Multi Factor Market Models, working paper?
[33] Pietersz, Pelsser, risk managing Bermudan swaption in the Libor BGM
model, Jounal of Derivatives, 2001
[34] Rebonato, R. (1999a) Volatility and Correlation. Wiley, Chichester.
25
[35] Rebonato, R. (1999b). On the Simultaneous Calibration of Multifactor Log-
normal Interest Rate Models to Black Volatilities and to the Correlation
Matrix, The Journal of Computational Finance, 2, 5-27.
[36] Ritchken, P., Chuang, I., 1999. Interest rate option pricing with volatility-
humps. Review of Derivatives Research 3, 237C262.
[37] Ritchken, P., Sankarasubramanian, L., 1995. Volatilitystructures of forward
rates and the dynamics of the term structure. Mathematical Finance 5,
55C72.
[38] Vasicek, O., 1977. An equilibrium characterization of the term structure.
Journal of Financial Economics
26
Table 1: Root Mean Square Errors from Model Calibration to European co-terminal swaption (11Y and 10Y) in EUR
and USD Markets from Feb 2005 to Sep 2007
EUR 11Y Date HW BK SMM LMM USD 11Y Date HW BK SMM LMM
2005-2-28 0.4% 0.4% 0.2% 1.0% 2005-2-28 0.6% 0.7% 0.5% 0.4%
2005-3-31 0.5% 0.6% 0.6% 0.7% 2005-3-31 0.5% 0.7% 0.4% 0.4%
2005-4-29 0.5% 0.7% 0.5% 0.6% 2005-4-29 0.5% 0.7% 0.5% 0.5%
2005-5-31 0.6% 0.7% 0.5% 0.4% 2005-5-31 0.6% 0.6% 0.4% 0.4%
2005-6-30 0.3% 0.3% 0.2% 0.2% 2005-6-30 0.0% 0.8% 0.3% 0.4%
2005-7-29 0.3% 0.5% 0.3% 0.3% 2005-7-29 0.9% 0.8% 0.5% 0.4%
2005-8-31 0.5% 0.8% 0.6% 0.5% 2005-8-31 0.6% 0.7% 0.5% 0.1%
2005-9-30 0.6% 0.8% 0.5% 0.5% 2005-9-30 0.6% 0.7% 0.4% 0.8%
2005-10-31 0.9% 1.1% 0.9% 0.9% 2005-10-31 0.4% 0.5% 0.4% 0.3%
2005-11-30 0.5% 0.7% 0.3% 0.2% 2005-11-30 0.6% 0.6% 0.5% 0.5%
2005-12-30 0.4% 0.5% 0.3% 0.3% 2005-12-30 0.5% 0.5% 0.2% 0.2%
2006-1-31 0.6% 0.6% 0.6% 0.4% 2006-1-31 0.5% 0.5% 0.4% 0.5%
2006-2-28 0.5% 0.5% 0.5% 0.5% 2006-2-28 0.8% 0.8% 0.2% 0.2%
2006-3-31 0.4% 0.5% 0.3% 0.3% 2006-3-31 0.7% 0.7% 0.3% 0.2%
2006-4-28 0.6% 0.7% 0.6% 0.7% 2006-4-28 0.7% 0.8% 0.3% 0.3%
2006-5-31 0.4% 0.5% 0.4% 0.3% 2006-5-31 0.9% 0.9% 0.3% 0.2%
2006-6-30 0.5% 0.5% 0.4% 0.3% 2006-6-30 0.6% 0.7% 0.1% 0.1%
2006-7-31 0.4% 0.5% 0.3% 0.3% 2006-7-31 0.8% 0.8% 0.4% 0.3%
2006-8-31 0.4% 0.6% 0.4% 0.3% 2006-8-31 0.9% 0.9% 0.3% 0.2%
2006-9-29 0.3% 0.2% 0.3% 0.3% 2006-9-29 1.0% 1.0% 0.2% 0.2%
Total 9.3% 11.9% 8.6% 9.0% Total 12.8% 14.5% 7.1% 6.6%
EUR 10Y Date HW BK SMM LMM USD 10Y Date HW BK SMM LMM
2006-2-28 0.3% 0.3% 0.6% 1.2% 2006-2-28 0.8% 0.8% 0.6% 0.8%
2006-3-31 0.2% 0.3% 1.1% 0.5% 2006-3-31 0.7% 0.7% 1.1% 0.9%
2006-4-28 0.5% 0.6% 0.9% 1.2% 2006-4-28 0.5% 0.6% 0.9% 1.0%
2006-5-31 0.3% 0.3% 0.5% 0.4% 2006-5-31 0.7% 0.9% 0.5% 0.6%
2006-6-30 0.4% 0.5% 0.7% 1.2% 2006-6-30 0.6% 0.6% 0.7% 0.8%
2006-7-31 0.3% 0.3% 0.8% 1.0% 2006-7-31 0.8% 0.9% 0.8% 0.8%
2006-8-31 0.4% 0.5% 0.6% 0.9% 2006-8-31 0.7% 0.7% 0.6% 0.6%
2006-9-29 0.2% 0.2% 0.7% 0.9% 2006-9-29 0.7% 0.8% 0.7% 0.7%
2006-10-31 0.7% 0.6% 0.6% 1.7% 2006-10-31 0.8% 0.8% 0.6% 0.6%
2006-11-30 0.4% 0.4% 0.6% 1.5% 2006-11-30 0.4% 0.4% 0.6% 0.3%
2006-12-29 0.6% 0.5% 0.7% 2.0% 2006-12-29 0.5% 0.6% 0.7% 0.5%
2007-1-30 0.4% 0.6% 2.0% 0.7% 2007-1-30 0.2% 0.3% 2.0% 0.7%
2007-2-28 0.3% 0.3% 0.5% 0.9% 2007-2-28 0.3% 0.3% 0.5% 0.6%
2007-3-31 0.3% 0.2% 1.0% 0.4% 2007-3-31 0.4% 0.0% 1.0% 1.1%
2007-4-30 0.3% 0.3% 1.0% 1.1% 2007-4-30 0.4% 0.5% 1.0% 0.8%
2007-5-31 0.3% 0.3% 0.7% 1.4% 2007-5-31 0.7% 0.8% 0.7% 0.7%
2007-6-30 0.3% 0.3% 0.4% 1.1% 2007-6-30 0.2% 0.2% 0.4% 0.9%
2007-7-29 0.4% 0.4% 1.0% 0.9% 2007-7-29 0.3% 0.3% 1.0% 0.7%
2007-8-31 0.7% 0.7% 0.3% 0.8% 2007-8-31 0.2% 0.3% 0.3% 0.3%
2007-9-28 0.6% 0.6% 1.1% 1.3% 2007-9-28 0.5% 0.6% 1.1% 1.2%
Total 8.2% 8.3% 15.8% 21.0% Total 10.6% 11.1% 15.8% 14.3%
Note: HW stands for Hull-White modle, BK stands for Black-Karasinski model, SMM stands for Swap Market Model and LMM
stands for Libor Market Model) by date. '11Y' denotes calibration results for 11Y co-terminal European swaptoins from Feb 2005 to
Sep 2006 ; '10Y' denotes calibration results for 10Y co-terminal European swaptoins from Feb 2006 to Sep 2007. 'EUR' denotes
Euro market and 'USD' denotes US-dollar market.
Date a b c d Date a b c d
2005-2-28 0.0100 0.0062 0.0062 0.0072 2005-2-28 -0.0208 -0.0082 0.1031 0.1753
2005-3-31 0.0100 0.0065 0.0061 0.0067 2005-3-31 0.0188 0.0215 0.6595 0.1351
2005-4-29 0.0100 0.0061 0.0060 0.0064 2005-4-29 0.0197 0.0143 0.5514 0.1343
2005-5-31 0.0100 0.0064 0.0064 0.0059 2005-5-31 0.0441 -0.0016 0.0919 0.1288
2005-6-30 0.0100 0.0069 0.0063 0.0064 2005-6-30 0.0465 0.0158 0.4431 0.1480
2005-7-29 0.0100 0.0063 0.0064 0.0066 2005-7-29 0.0196 0.0136 0.4466 0.1466
2005-8-31 0.0100 0.0063 0.0066 0.0065 2005-8-31 0.0190 0.0164 0.4144 0.1533
2005-9-30 0.0100 0.0063 0.0066 0.0064 2005-9-30 0.0117 0.0196 0.4089 0.1542
2005-10-31 0.0100 0.0065 0.0067 0.0071 2005-10-31 0.0019 0.0164 0.5374 0.1580
2005-11-30 0.0100 0.0066 0.0068 0.0065 2005-11-30 0.0032 0.0197 0.4492 0.1552
2005-12-30 0.0100 0.0064 0.0068 0.0064 2005-12-30 0.0060 0.0153 0.3217 0.1585
2006-1-31 0.0100 0.0061 0.0067 0.0058 2006-1-31 0.0300 0.0114 0.1652 0.1221
2006-2-28 0.0100 0.0059 0.0065 0.0059 2006-2-28 0.0181 0.0096 0.1807 0.1296
2006-3-31 0.0100 0.0061 0.0063 0.0060 2006-3-31 0.0030 0.0075 0.3124 0.1360
2006-4-28 0.0100 0.0063 0.0061 0.0065 2006-4-28 0.0045 0.0089 0.5624 0.1332
2006-5-31 0.0100 0.0063 0.0061 0.0062 2006-5-31 0.0152 0.0049 0.3026 0.1259
2006-6-30 0.0100 0.0063 0.0062 0.0062 2006-6-30 0.0171 0.0030 0.1787 0.1218
2006-7-31 0.0100 0.0062 0.0062 0.0060 2006-7-31 0.0094 0.0075 0.3131 0.1274
2006-8-31 0.0100 0.0063 0.0063 0.0060 2006-8-31 0.0041 0.0118 0.4010 0.1371
2006-9-29 0.0100 0.0064 0.0063 0.0060 2006-9-30 0.0041 0.0118 0.4010 0.1371
Date a b c d Date a b c d
2005-2-28 0.0087 0.1642 0.1328 0.1521 2005-2-28 -0.0558 0.2099 1.1154 0.1373
2005-3-31 0.0087 0.1740 0.1312 0.1405 2005-3-31 -0.0169 0.1665 0.8245 0.1352
2005-4-29 0.0087 0.1716 0.1337 0.1351 2005-4-29 0.0203 0.0997 0.6972 0.1338
2005-5-31 0.0087 0.1860 0.1465 0.1229 2005-5-31 0.1163 0.0052 0.2577 0.1271
2005-6-30 0.0087 0.2123 0.1488 0.1412 2005-6-30 0.1555 0.0856 0.5681 0.1468
2005-7-29 0.0087 0.1867 0.1530 0.1401 2005-7-29 -0.0089 0.0891 0.5949 0.1459
2005-8-31 0.0087 0.1966 0.1654 0.1406 2005-8-31 -0.0537 0.1083 0.5871 0.1542
2005-9-30 0.0087 0.1933 0.1719 0.1348 2005-9-30 -0.0958 0.1051 0.5195 0.1521
2005-10-31 0.0087 0.1856 0.1657 0.1562 2005-10-31 -0.1023 0.1017 0.6665 0.1575
2005-11-30 0.0087 0.1864 0.1708 0.1397 2005-11-30 -0.1175 0.1020 0.5525 0.1537
2005-12-30 0.0087 0.1882 0.1850 0.1425 2005-12-30 -0.0836 0.0634 0.4002 0.1547
2006-1-31 0.0087 0.1677 0.1705 0.1257 2006-1-31 -0.0031 0.0362 0.2119 0.1062
2006-2-28 0.0087 0.1627 0.1659 0.1310 2006-2-28 -0.0217 0.0319 0.2372 0.1202
2006-3-31 0.0087 0.1539 0.1503 0.1300 2006-3-31 -0.0429 0.0320 0.4028 0.1342
2006-4-28 0.0087 0.1523 0.1366 0.1354 2006-4-28 -0.0298 0.0510 0.6778 0.1329
2006-5-31 0.0087 0.1523 0.1350 0.1213 2006-5-31 0.0143 0.0240 0.4130 0.1247
2006-6-30 0.0087 0.1484 0.1373 0.1231 2006-6-30 0.0256 0.0086 0.2143 0.1165
2006-7-31 0.0087 0.1505 0.1406 0.1205 2006-7-31 -0.0182 0.0304 0.3983 0.1258
2006-8-31 0.0087 0.1597 0.1500 0.1266 2006-8-31 -0.0575 0.0536 0.4868 0.1354
2006-9-29 0.0087 0.1638 0.1552 0.1283 2006-9-29 0.0148 0.0207 0.2445 0.1228
Note: HW stands for Hull-White modle,a in HW model is the mean reversion rate, b ,c and d is sigma (0),sigma(3) and sigma(11)
respectively; BK stands for Black-Karasinski model, a in HW model is the mean reversion rate, b ,c and d is sigma0,sigma 3 and sigma 11
respectively; SMM stands for Swap Market Model and LMM stands for Libor Market Model.a,b,c,d in SMM and LMM are the parameters
of the volatility functional form as suggested by Rebonato 1999. '11Y'('10Y') denotes calibration is made to co-terminal 11-year (10-year)
European swaptoins. All calibrations are performed with both EUR and USD markets.
Table 2(a): Model Calibrated Parameter Values for EUR Market from
February 2005 to September 2006
HW(EUR) SMM(EUR)
BK(EUR) LMM(EUR)
Table 2(b): Model Calibrated Parameter Values
for USD Market from February 2005 to September 2006
Date a b c d Date a b c d
2005-2-28 0.0060 0.0092 0.0089 0.0062 2005-2-28 0.0166 0.0488 0.2342 0.1317
2005-3-31 0.0060 0.0102 0.0087 0.0062 2005-3-31 0.0149 0.0524 0.2897 0.1354
2005-4-29 0.0060 0.0098 0.0094 0.0064 2005-4-29 0.0211 0.0470 0.2687 0.1440
2005-5-31 0.0060 0.0094 0.0094 0.0064 2005-5-31 0.0211 0.0461 0.2394 0.1438
2005-6-30 0.0060 0.0097 0.0096 0.0065 2005-6-30 0.0284 0.0326 0.3098 0.1610
2005-7-29 0.0060 0.0096 0.0090 0.0066 2005-7-29 0.0228 0.0286 0.3125 0.1521
2005-8-31 0.0060 0.0097 0.0094 0.0070 2005-8-31 0.0219 0.0306 0.2954 0.1659
2005-9-30 0.0060 0.0097 0.0094 0.0069 2005-9-30 0.0223 0.0277 0.2931 0.1537
2005-10-31 0.0060 0.0099 0.0095 0.0067 2005-10-31 0.0174 0.0340 0.2581 0.1442
2005-11-30 0.0060 0.0101 0.0101 0.0076 2005-11-30 0.0158 0.0278 0.2398 0.1563
2005-12-30 0.0060 0.0097 0.0098 0.0083 2005-12-30 0.0169 0.0039 0.3350 0.1740
2006-1-31 0.0060 0.0092 0.0099 0.0079 2006-1-31 0.0151 0.0007 0.2796 0.1654
2006-2-28 0.0060 0.0081 0.0097 0.0079 2006-2-28 0.0152 -0.0372 0.4159 0.1716
2006-3-31 0.0060 0.0083 0.0097 0.0066 2006-3-31 0.0206 -0.0077 0.2870 0.1450
2006-4-28 0.0060 0.0081 0.0088 0.0069 2006-4-28 0.0166 -0.0082 0.3910 0.1374
2006-5-31 0.0060 0.0084 0.0095 0.0071 2006-5-31 0.0199 -0.0154 0.3859 0.1449
2006-6-30 0.0060 0.0080 0.0093 0.0077 2006-6-30 0.0137 -0.0230 0.3872 0.1463
2006-7-31 0.0060 0.0080 0.0094 0.0076 2006-7-31 0.0160 -0.0220 0.3703 0.1479
2006-8-31 0.0060 0.0077 0.0096 0.0077 2006-8-31 0.0171 -0.0322 0.3567 0.1564
2006-9-29 0.0060 0.0078 0.0096 0.0077 2006-9-30 0.0171 -0.0322 0.3567 0.1564
Date a b c d Date a b c d
2005-2-28 0.0060 0.1985 0.1737 0.1004 2005-2-28 0.0255 0.0606 0.3062 0.1223
2005-3-31 0.0060 0.2028 0.1576 0.1058 2005-3-31 0.0674 0.0612 0.3763 0.1303
2005-4-29 0.0060 0.2133 0.1854 0.1044 2005-4-29 0.0253 0.0697 0.3149 0.1300
2005-5-31 0.0060 0.2123 0.1957 0.1076 2005-5-31 -0.0080 0.0756 0.3045 0.1310
2005-6-30 0.0060 0.2239 0.2044 0.1138 2005-6-30 0.0082 0.0757 0.2864 0.1300
2005-7-29 0.0060 0.2045 0.1858 0.1144 2005-7-29 -0.0484 0.0885 0.3866 0.1454
2005-8-31 0.0060 0.2214 0.2037 0.1316 2005-8-31 -0.0370 0.0814 0.3598 0.1570
2005-9-30 0.0060 0.2053 0.1923 0.1170 2005-9-30 -0.0454 0.0802 0.3516 0.1441
2005-10-31 0.0060 0.1969 0.1819 0.1150 2005-10-31 -0.0001 0.0577 0.3029 0.1314
2005-11-30 0.0060 0.2034 0.1958 0.1371 2005-11-30 -0.0213 0.0536 0.3024 0.1478
2005-12-30 0.0060 0.2013 0.1995 0.1552 2005-12-30 -0.0858 0.0648 0.4062 0.1701
2006-1-31 0.0060 0.1872 0.1959 0.1502 2006-1-31 -0.0801 0.0512 0.3185 0.1563
2006-2-28 0.0060 0.1628 0.1940 0.1586 2006-2-28 -0.2150 0.0609 0.4785 0.1700
2006-3-31 0.0060 0.1613 0.1827 0.1182 2006-3-31 -0.1376 0.0707 0.3364 0.1361
2006-4-28 0.0060 0.1513 0.1548 0.1175 2006-4-28 -0.1236 0.0674 0.4668 0.1352
2006-5-31 0.0060 0.1559 0.1677 0.1202 2006-5-31 -0.1638 0.0786 0.4562 0.1420
2006-6-30 0.0060 0.1455 0.1645 0.1316 2006-6-30 -0.1527 0.0527 0.4492 0.1443
2006-7-31 0.0060 0.1501 0.1696 0.1297 2006-7-31 -0.1650 0.0617 0.4344 0.1453
2006-8-31 0.0060 0.1513 0.1824 0.1378 2006-8-31 -0.2088 0.0661 0.4190 0.1533
2006-9-29 0.0060 0.1574 0.1869 0.1411 2006-9-29 -0.2322 0.0762 0.4611 0.1591
Note: HW stands for Hull-White modle,a in HW model is the mean reversion rate, b ,c and d is sigma (0),sigma(3) and sigma(11)
respectively; BK stands for Black-Karasinski model, a in HW model is the mean reversion rate, b ,c and d is sigma0,sigma 3 and sigma 11
respectively; SMM stands for Swap Market Model and LMM stands for Libor Market Model.a,b,c,d in SMM and LMM are the parameters
of the volatility functional form as suggested by Rebonato 1999. '11Y'('10Y') denotes calibration is made to co-terminal 11-year (10-year)
European swaptoins. All calibrations are performed with both EUR and USD markets.
HW(USD) SMM(USD)
BK(USD) LMM(USD)
Table 3(a): 11Y and 10Y Bermudan Swaption Prices
in EUR Market from February 2005 to September 2007
DATE(11Y) HW BK SMM LMM
2005-2-28 0.0465 0.0441 0.0463 0.0460
2005-3-31 0.0467 0.0441 0.0479 0.0480
2005-4-29 0.0476 0.0447 0.0481 0.0485
2005-5-31 0.0491 0.0460 0.0499 0.0501
2005-6-30 0.0508 0.0474 0.0534 0.0549
2005-7-29 0.0485 0.0455 0.0484 0.0478
2005-8-31 0.0486 0.0455 0.0483 0.0475
2005-9-30 0.0464 0.0438 0.0458 0.0464
2005-10-31 0.0459 0.0437 0.0450 0.0453
2005-11-30 0.0452 0.0432 0.0444 0.0448
2005-12-30 0.0429 0.0414 0.0415 0.0415
2006-1-31 0.0411 0.0398 0.0395 0.0393
2006-2-28 0.0396 0.0384 0.0379 0.0376
2006-3-31 0.0386 0.0376 0.0375 0.0374
2006-4-28 0.0387 0.0377 0.0386 0.0390
2006-5-31 0.0396 0.0384 0.0397 0.0404
2006-6-30 0.0386 0.0376 0.0384 0.0387
2006-7-31 0.0387 0.0376 0.0382 0.0385
2006-8-31 0.0389 0.0379 0.0384 0.0386
2006-9-29 0.0381 0.0373 0.0379 0.0382
DATE(10Y) HW BK SMM LMM Strike
2006-2-28 0.0271 0.0272 0.0273 0.0271 0.0402
2006-3-31 0.0379 0.0367 0.0365 0.0364 0.0395
2006-4-28 0.0534 0.0510 0.0519 0.0521 0.0377
2006-5-31 0.0610 0.0582 0.0591 0.0592 0.0363
2006-6-30 0.0744 0.0720 0.0722 0.0723 0.0344
2006-7-31 0.0610 0.0584 0.0585 0.0586 0.0356
2006-8-31 0.0600 0.0573 0.0575 0.0576 0.0339
2006-9-29 0.0545 0.0522 0.0519 0.0522 0.0344
2006-10-31 0.0417 0.0404 0.0399 0.0397 0.0369
2006-11-30 0.0390 0.0378 0.0371 0.0369 0.0371
2006-12-29 0.0563 0.0543 0.0540 0.0537 0.0357
2007-1-31 0.0526 0.0508 0.0504 0.0502 0.0377
2007-2-28 0.0443 0.0427 0.0423 0.0422 0.0378
2007-3-30 0.0379 0.0332 0.0328 0.0329 0.0409
2007-4-30 0.0308 0.0301 0.0293 0.0290 0.0431
2007-5-31 0.0391 0.0380 0.0370 0.0369 0.0433
2007-6-29 0.0476 0.0461 0.0464 0.0466 0.0442
2007-7-31 0.0487 0.0471 0.0473 0.0474 0.0428
2007-8-31 0.0510 0.0492 0.0502 0.0504 0.0408
2007-9-28 0.0578 0.0557 0.0573 0.0576 0.0401
Note: HW stands for Hull-Whitemodle, BK stands for Black-Karasinski model, SMM stands for SwapMarket Model and
LMM stands for Libor Market Model.11Y Bermudan swaption are priceed as at-the-money from February 2005 to
September 2006; 10 year Bermudan swaptions are priced with the corresponding 11Y ATM strikes one year ago.For
example, on February 2006, the strike rate is 0.0402 which is ATM strike rate on February 2005. 10 year Bermudan
swaption are priced from February 2006 to September 2007.
Table 3(b): 11Y and 10Y Bermudan Swaption Prices
in USD Market from February 2005 to September 2007
DATE(11Y) HW BK SMM LMM
2005-2-28 0.0543 0.0528 0.0562 0.0576
2005-3-31 0.0526 0.0516 0.0568 0.0597
2005-4-29 0.0552 0.0539 0.0574 0.0592
2005-5-31 0.0545 0.0532 0.0555 0.0566
2005-6-30 0.0557 0.0546 0.0570 0.0582
2005-7-29 0.0512 0.0508 0.0529 0.0539
2005-8-31 0.0537 0.0531 0.0547 0.0558
2005-9-30 0.0522 0.0518 0.0532 0.0540
2005-10-31 0.0517 0.0515 0.0531 0.0542
2005-11-30 0.0543 0.0541 0.0545 0.0552
2005-12-30 0.0525 0.0526 0.0520 0.0521
2006-1-31 0.0517 0.0517 0.0500 0.0495
2006-2-28 0.0470 0.0475 0.0430 0.0415
2006-3-31 0.0487 0.0486 0.0459 0.0448
2006-4-28 0.0470 0.0465 0.0455 0.0450
2006-5-31 0.0485 0.0483 0.0463 0.0454
2006-6-30 0.0465 0.0464 0.0432 0.0420
2006-7-31 0.0480 0.0477 0.0446 0.0434
2006-8-31 0.0488 0.0484 0.0441 0.0424
2006-9-29 0.0493 0.0489 0.0448 0.0431
DATE(10Y) HW BK SMM LMM Strike
2006-2-28 0.0472 0.0468 0.0430 0.0418 0.0496
2006-3-31 0.0507 0.0497 0.0471 0.0464 0.0522
2006-4-28 0.0741 0.0713 0.0701 0.0698 0.0482
2006-5-31 0.0848 0.0817 0.0802 0.0798 0.0464
2006-6-30 0.0926 0.0899 0.0883 0.0877 0.0454
2006-7-31 0.0696 0.0670 0.0643 0.0635 0.0486
2006-8-31 0.0716 0.0688 0.0653 0.0640 0.0456
2006-9-29 0.0534 0.0519 0.0485 0.0473 0.0492
2006-10-31 0.0422 0.0419 0.0383 0.0368 0.0520
2006-11-30 0.0369 0.0372 0.0349 0.0338 0.0517
2006-12-29 0.0464 0.0453 0.0428 0.0418 0.0503
2007-1-31 0.0467 0.0455 0.0424 0.0429 0.0512
2007-2-28 0.0393 0.0388 0.0372 0.0366 0.0514
2007-3-30 0.0337 0.0336 0.0327 0.0319 0.0548
2007-4-30 0.0271 0.0277 0.0270 0.0264 0.0573
2007-5-31 0.0308 0.0311 0.0289 0.0276 0.0577
2007-6-29 0.0385 0.0381 0.0371 0.0368 0.0585
2007-7-31 0.0460 0.0453 0.0453 0.0451 0.0566
2007-8-31 0.0447 0.0437 0.0458 0.0467 0.0538
2007-9-28 0.0506 0.0491 0.0526 0.0547 0.0525
Note: HW stands for Hull-Whitemodle, BK stands for Black-Karasinski model, SMM stands for SwapMarket Model and
LMM stands for Libor Market Model. 11Y Bermudan swaptions are priced as at-the-money from February 2005 to
September 2006; 10-year Bermudan swaptions arepriced with thecorresponding 11-year ATM strikes oneyear ago. For
example, on February 2006, the strike rate is 0.0402 which is ATM strike rate for the 11-year Bermudan swaption on
February 2005. 10-year Bermudan swaptions are priced from February 2006 to September 2007.
Table 4: PCA Factor Loadings of Libor Rates from J anuary 2000 to September 2007
EUR USD
Rates PCA1(EURO) PCA2(EURO) PCA3(EURO) PCA1(USD) PCA2(USD) PCA3(USD)
1 0.16 -0.41 0.16 0.15 -0.25 0.45
2 0.37 -0.44 0.12 0.34 -0.30 0.42
3 0.40 -0.34 0.02 0.37 -0.20 0.22
4 0.35 -0.17 -0.02 0.34 -0.09 0.03
5 0.33 -0.04 -0.12 0.32 -0.55 -0.43
6 0.30 0.11 -0.09 0.32 0.58 0.40
7 0.23 0.14 -0.76 0.30 0.10 -0.15
8 0.33 0.44 0.60 0.29 0.16 -0.19
9 0.27 0.33 -0.02 0.29 0.19 -0.26
10 0.24 0.30 -0.06 0.28 0.25 -0.21
11 0.25 0.27 -0.05 0.26 0.14 -0.22
EUR
Percentage
variance
Cumulative
variance
USD
Percentage
varariance
Cumulative
varariance
PCA1 69.2% 69.2% PCA1 79.6% 79.6%
PCA2 16.2% 85.4% PCA2 8.7% 88.3%
PCA3 7.6% 93.0% PCA3 5.9% 94.2%
Table 5: Explainary Power of the First Three Pinciple Comp
Table 6: Hedging Profit and Loss in EUR and USD Market
EUR HW BK SMM LMM
2006-02-28 0.0064 0.0024 0.0037 0.0047
2006-03-31 0.0081 0.0069 0.0117 0.0137
2006-04-28 0.0064 0.0065 0.0080 0.0080
2006-05-31 0.0063 0.0074 0.0118 0.0133
2006-06-30 0.0018 0.0042 0.0137 0.0065
2006-07-31 0.0050 0.0075 0.0127 0.0126
2006-08-31 0.0054 0.0069 0.0109 0.0099
2006-09-29 0.0057 0.0076 0.0078 0.0068
2006-10-31 0.0109 0.0112 0.0115 0.0141
2006-11-30 0.0109 0.0104 0.0130 0.0094
2006-12-29 0.0053 0.0077 0.0059 0.0057
2007-01-31 0.0060 0.0088 0.0080 0.0039
2007-02-28 0.0162 0.0198 0.0184 0.0155
2007-03-30 0.0194 0.0260 0.0259 0.0248
2007-04-30 0.0258 0.0314 0.0381 0.0388
2007-05-31 0.0361 0.0399 0.0385 0.0357
2007-06-29 0.0266 0.0317 0.0334 0.0296
2007-07-31 0.0211 0.0261 0.0246 0.0209
2007-08-31 0.0061 0.0090 0.0118 0.0040
2007-09-28 -0.0065 -0.0037 -0.0058 -0.0069
RMSS 0.0149 0.0175 0.0189 0.0175
USD HW BK SMM LMM
2006-02-28 0.0102 0.0099 0.0159 0.0168
2006-03-31 0.0009 0.0010 0.0081 0.0237
2006-04-28 -0.0024 0.0027 0.0034 0.0185
2006-05-31 -0.0083 -0.0017 -0.0010 0.0057
2006-06-30 -0.0109 -0.0035 -0.0040 0.0045
2006-07-31 -0.0031 0.0028 0.0052 0.0148
2006-08-31 -0.0007 0.0052 0.0084 0.0159
2006-09-29 -0.0003 0.0012 0.0056 0.0156
2006-10-31 0.0104 0.0107 0.0157 0.0279
2006-11-30 0.0193 0.0193 0.0216 0.0307
2006-12-29 0.0041 0.0052 0.0064 0.0172
2007-01-31 0.0043 0.0057 0.0073 0.0140
2007-02-28 0.0094 0.0103 0.0072 0.0070
2007-03-30 0.0197 0.0200 0.0180 0.0172
2007-04-30 0.0274 0.0268 0.0265 0.0260
2007-05-31 0.0239 0.0238 0.0239 0.0242
2007-06-29 0.0149 0.0156 0.0132 0.0084
2007-07-31 0.0096 0.0106 0.0071 0.0013
2007-08-31 0.0115 0.0127 0.0055 -0.0027
2007-09-28 -0.0010 -0.0005 -0.0081 -0.0126
RMSS 0.0554 0.0548 0.0570 0.0775
Note: HW stands for Hull-Whitemodle, BK stands for Black-Karasinski model, SMM
stands for Swap Market Model and LMM stands for Libor Market Model. All numbers
are in real value. RMSS is the root mean sum of sqaure of all P&L in each period
Figure 1: Root Mean Square Prices Error by Date from Calibration to 11-year Co-
terminal European Swaptions over the period of February 2005 to September 2006
Note: HW stands for Hull-White model; BK stands for Black-Karasinski model;SMM stands for
Swap Market Model and LMM stands for Libor market Model.
EUR
0.0%
0.2%
0.4%
0.6%
0.8%
1.0%
1.2%
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Date
R
M
S
E
HW
BK
SMM
LMM
USD
0.0%
0.2%
0.4%
0.6%
0.8%
1.0%
1.2%
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Date
R
M
S
E
HW
BK
SMM
LMM
Figure 2: Root Mean Square Prices Error by Contract
in EUR and USD Market from February 2005 to September 2006
Note: HW stands for Hull-White model; BK stands for Black-Karasinski model;SMM stands for
Swap Market Model and LMM stands for Libor market Model. x*y means x-year option on y-year
swap. Calibration is made simultaneously to 11-year (x+y) co-terminal European swapition.
EUR
0.0%
0.1%
0.2%
0.3%
0.4%
0.5%
0.6%
0.7%
0.8%
0.9%
1.0%
1*10 2*9 3*8 4*7 5*6 6*5 7*4 8*3 9*2 10*1
Contract
R
M
S
E
HW
BK
SMM
LMM
USD
0.0%
0.2%
0.4%
0.6%
0.8%
1.0%
1.2%
1*10 2*9 3*8 4*7 5*6 6*5 7*4 8*3 9*2 10*1
Contract
R
M
S
E
HW
BK
SMM
LMM
Figure 3: Calibrated Parameter Values of Four TermStructure Models
in EUR and USD Markets fromFebruary 2005 to September 2006
Note: HW stands for Hull-White modle,a in HW model is the mean reversion rate, b ,c and d is sigma0, sigma 3 and sigma 11 respectively; BK stands
for Black-Karasinski model, a in HW model is the mean reversion rate, b ,c and d is sigma(0),sigma(3) and sigma(11) respectively; SMM stands for
Swap Market Model and LMM stands for Libor Market Model.a,b,c,d in SMM and LMM are the parameters of the volatility functional formas
suggested by Rebonato 1999. Wecalibrated to co-terminal 11Y European swaptoins fromFeb 2005 to Sep 2006. All calibration are performed with
both EUR and USD market.
HW Parameters(EUR)
-
0.002
0.004
0.006
0.008
0.010
0.012
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Dates
v
a
l
u
e
a
b
c
d
BK Parameters(EUR)
-
0.05
0.10
0.15
0.20
0.25
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Dates
v
a
l
u
e
a
b
c
d
HW Parameters(USD)
-
0.002
0.004
0.006
0.008
0.010
0.012
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Dates
v
a
l
u
e
a
b
c
d
BK Parameters(USD)
-
0.05
0.10
0.15
0.20
0.25
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Dates
v
a
l
u
e
a
b
c
d
SMM Parameters(USD)
-0.10
-0.05
-
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Dates
v
a
l
u
e
a
b
c
d
LMM Parameters(USD)
-0.3
-0.2
-0.1
-
0.1
0.2
0.3
0.4
0.5
0.6
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Dates
v
a
l
u
e
a
b
c
d
SMM Parameters(EUR)
-0.1
-
0.1
0.2
0.3
0.4
0.5
0.6
0.7
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Dates
v
a
l
u
e
a
b
c
d
LMM Parameters(EUR)
-0.2
-
0.2
0.4
0.6
0.8
1.0
1.2
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Dates
v
a
l
u
e
a
b
c
d
Figure 4: 11Y and 10 Year Bermudan swaption Prices in EUR and USD Markets from February 2005 to September 2007
Note: HW stands for Hull-White modle, BK stands for Black-Karasinski model, SMM stands for Swap Market Model and LMM stands for Libor Market Model.11Y
Bermudanswaptionispriceasat-the-moneyfromFeb2005toSep2006; 10year Bermudanswaptionsarepricedwiththecorresponding11Y ATM strikesoneyear ago. 10
year Bermudan swaption are priced fromFeb 2006 to Sep 2007.
11Y Bermudan Swaption Prices(EUR)
0%
1%
2%
3%
4%
5%
6%
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Dates
P
r
i
c
e
s
HW
BK
SMM
LMM
10Y Bermudann Swaption Prices (EUR)
0%
1%
2%
3%
4%
5%
6%
7%
8%
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
O
c
t
-
0
6
D
e
c
-
0
6
F
e
b
-
0
7
A
p
r
-
0
7
J
u
n
-
0
7
A
u
g
-
0
7
Dates
P
r
i
c
e
s
HW
BK
SMM
LMM
11Y Bermudan Swaption Prices (USD)
0%
1%
2%
3%
4%
5%
6%
7%
F
e
b
-
0
5
A
p
r
-
0
5
J
u
n
-
0
5
A
u
g
-
0
5
O
c
t
-
0
5
D
e
c
-
0
5
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
Date
P
r
i
c
e
s
HW
BK
SMM
LMM
10Y Bermudan Swaption Prices (USD)
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10%
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
O
c
t
-
0
6
D
e
c
-
0
6
F
e
b
-
0
7
A
p
r
-
0
7
J
u
n
-
0
7
A
u
g
-
0
7
Date
P
r
i
c
e
s
HW
BK
SMM
LMM
Figure 5: PCA factor loading for the first three principle component in EUR and USD Market
P1 P2 P3
EUR 0.005216 0.002344 0.0010106
USD 0.0082656 0.002051 0.0020943
Mean of first Three Principal Components
PCA Loadong for Factor 1
0.0000
0.0005
0.0010
0.0015
0.0020
0.0025
0.0030
0.0035
1 2 3 4 5 6 7 8 9 10 11
Rates
V
a
l
u
e
P1(Euro)
P1(USD)
PCA Loadong for Factor 2
-0.0015
-0.0010
-0.0005
0.0000
0.0005
0.0010
0.0015
1 2 3 4 5 6 7 8 9 10 11
Rates
V
a
l
u
e
P2(Euro)
P2(USD)
PCA Loadong for Factor 3
-0.0015
-0.0010
-0.0005
0.0000
0.0005
0.0010
0.0015
1 2 3 4 5 6 7 8 9 10 11
Rates
V
a
l
u
e
P3(Euro)
P3(USD)
Figure 6: Factor Loadings for the first three principle components of forward
rates term structure in the EUR and USD markets
PCA 1 Facotr Loading
0.0
0.1
0.2
0.3
0.4
0.5
1 2 3 4 5 6 7 8 9 10 11
Rates
L
o
a
d
i
n
g
s
PCA1(EURO)
PCA1(USD)
PCA 2 Facotr Loading
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1 2 3 4 5 6 7 8 9 10 11
Rates
L
o
a
d
i
n
g
s
PCA2(EURO)
PCA2(USD)
PCA 3 Facotr Loading
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1 2 3 4 5 6 7 8 9 10 11
Rates
L
o
a
d
i
n
g
s
PCA3(EURO)
PCA3(USD)
Figure 7: Hedging Profit and Loss in EUR and USD market from Feb 2006 to Sep 2007
Note: HW stands for Hull-White modle, BK stands for Black-Karasinski model, SMM stands for
Swap Market Model and LMM stands for Libor Market Model. All numbers arein real monetary
value.
EUR
-0.01
0.00
0.01
0.02
0.03
0.04
0.05
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
O
c
t
-
0
6
D
e
c
-
0
6
F
e
b
-
0
7
A
p
r
-
0
7
J
u
n
-
0
7
A
u
g
-
0
7
Date
V
a
l
u
e
HW
BK
SMM
LMM
USD
-0.02
-0.01
-0.01
0.00
0.01
0.01
0.02
0.02
0.03
0.03
0.04
F
e
b
-
0
6
A
p
r
-
0
6
J
u
n
-
0
6
A
u
g
-
0
6
O
c
t
-
0
6
D
e
c
-
0
6
F
e
b
-
0
7
A
p
r
-
0
7
J
u
n
-
0
7
A
u
g
-
0
7
Date
V
a
l
u
e
HW
BK
SMM
LMM