HOW DISTORTED INCENTIVES INDUCE EXCESSIVE RISK:

A PROPOSAL TO REFORM BANKING REGULATIONS

A LOVE STORY
Presentation by Moshe Y. Admon on the Emory Law Journal Article
REGULATING RISK AND GOVERNANCE IN BANKS:
A CONTRACTARIAN PERSPECTIVE
Simone M. Sepe, 62 Emory L.J. 327 (2012)
Most information in this presentation is Copyright owned by Simon M. Sepe

THE MAN
THE MYTH
THE LEGEND
 What failure caused systemic, excessive risk
taking by all U.S. Banks?
• The “in-vogue” answer: There was a failure of bank governance.
CEO’s misappropriated bank governance mechanisms.
• Forget trendy, lets get real! This answer doesn’t take into account
the causal relationship between external and internal governance

BUT WHAT IS EXTERNAL AND INTERNAL GOVERNANCE?

• External Governance: Oversight and influence that external
stakeholders (debtholders and shareholders) exercise over corp.
• Internal Governance: Institutional arrangements that discipline the
corporate decision-making process.
 Why is the causal relationship between
external and internal governance important?
• Implicit shareholder–debtholder (External Control) negotiation over a
firm’s risk choices (Internal Control) would lead shareholders to seek
internal governance arrangements that can distinguish their firms as
safe investments.
• With the introduction of safety nets (deposit insurance and bailouts)
debtholders become less sensitive to risk taking. This causes
rational passivity of bank shareholders, who find themselves
benefitting from such arrangements.
WAIT, THERE’S A SAFETY NET? YOLO!
 How to fix the problem?
• Regulators presently focus on capital requirements to constrain risk taking
in the belief that higher equity levels can make riskier projects
unprofitable. This fails to give bank shareholders incentives to move away
from CEO-centric, and risk-prone, governance models.
• Regulators should make banks’ regulatory costs—capital requirements and
deposit insurance premiums—sensitive to banks’ organizational features.
Ex.: Banks opting for an advocacy-based governance model could be held
to the 8% minimum capital ratio, while CEO-centric banks have 10% ratio
• Article suggests that bank regulators should adopt a contractarian
approach - they would bargain for the same governance concessions from
shareholders that debtholders would otherwise demand.
 How can bank regulators implement this
suggested contractarian approach?
• Article suggests that in an ideal contracting world banks’ governance
arrangements would be based on the “advocacy model” - an
organizational model based on antagonistic information gathering
and contrarian thinking.
• Banks could replicate a similar decision-making structure through two
basic organizational features:
1. Appointment of a “representative” of the debtholders’ interests
2. A fully independent board of directors
But why implement this approach?
 Because of the negative effects of Moral
Hazard and Liquidity Production, duh!
• Banks have access to deposit funding and are able to carry far more
leverage than non-banking organizations. With such high leverage
shareholders are incentivized to take more risk because the
downside is disproportionately borne by debtholders.
• Incentives to take risk are exacerbated by widespread use of equity-
based schemes to compensate bank managers, high interbank
correlation, and the increased opportunities for risk taking created by
financial innovation.
 Risk preferences of the parties
• Bank Debtholders: Concave payoff schedule. Highly sensitive to
declines in asset value.
• Bank Shareholders: Convex payoff schedule. Highly sensitive to
increases in equity returns. Because of limited liability shareholders
are indifferent to losses beyond value of capital contribution.
• Bank Managers: Invested in their corporation (time, effort,
reputation) unlike diversified investors. More risk averse than
shareholders. BUT, when managers are given incentives relating to
corporate performance (stocks, options, performance bonuses they
take more risks. This is referred to as “Tail Risk.”
Debtholders & Shareholders Risk Preferences

Profit

Share Value
Managerial Risk Preferences
 The problem with increased risk taking
Interbank Correlation - Banks are correlated among two dimensions:
1. Counterparty risk dimension – arises from the high volume of
interbank transactions. Ex., Bank A borrows from Bank B. If Bank A
fails, it may cause Bank B to have solvency issues – a domino effect.
2. Asset dimension – Capital structures and assets are largely
homogeneous. Trouble in one bank may cause investors to think
that other banks have the same problems.

Thus, greater risk taking leads to greater chance of systemic failure.

 Intensified problems arising from aggregation of
incentivized risk & modern liquidity production
Banks don’t hold loans on their balance sheets. Instead, they distribute
them to investors, including other banks, through securitization
transactions and other financing arrangements that can transform
illiquid assets into marketable securities. This leads to two problems:

1. Securitization vastly underestimates the risk of moral hazard.

Seller can externalize the cost of its own risk taking to the buyer,
which, in turn, will act as seller in another transaction, replicating
the same moral hazard. Banks were insensitive to the quality of
securitized assets and focused on increasing securitization volumes.
Intensified problems arising from aggregation of
incentivized risk & modern liquidity production
2. Investment banks began to rely heavily on repurchase agreements
(repos) – short term financing - to compete with universal banks,
which enjoyed the comparative advantage of access to deposit
funding. This allowed the rapid growth of banks’ balance sheets
and induced more risk taking. The use of short-term debt
disproportionately increased banks’ exposure to funding liquidity
risk and with it the likelihood of sudden requests of withdrawal by
short-term creditors, which is exactly what happened in the
banking crisis.
 Oh you silly goose: The current framework
governing bank regulation
• The current theoretical framework governing regulation failed to fully
consider the relationship between external and internal governance.
• In the real world, under the protection of safety nets (i.e., deposit
insurance and bailouts), bank debtholders become almost insensitive
to risk taking and opportunistically avoid monitoring banks. Without
the constraint of debt discipline, bank shareholders, shielded by
limited liability, benefitted from apparently weak governance
arrangements that incentivized managerial risk taking.
• There is an eloquent Italian word to describe this type of oversight:
“Stupido” [stù‧pi‧do]
Even more stupido than cow racing
The prevailing theoretical approaches to bank
governance of risk taking
1. Policy Analysis Approach: According to this narrative, captured, and
often inexpert, board members joined by complacent risk managers
allowed CEOs to gain undisputed dominance over banking
organizations. THE FLAW WITH THIS APPROACH is that it focuses
exclusively on internal governance arrangements.

2. Debt Contract Approach: Under this view, the contractual

negotiation between the debtholders and the firm is the primary
instrument that serves to control risk taking. THE FLAW WITH THIS
APPROACH is that it is indifferent to internal governance.
The prevailing theoretical approaches to bank
governance of risk taking
3. Optimal Compensation Contract Approach: This approach
advocates that a manager’s payoff schedule should be tied to both
the firm’s equity and the firm’s debt. THE FLAW WITH THE
APPROACH is that it fails to take into account issues of contractual
incompleteness and, primarily, monitoring, which is, instead, an
essential governance component. Conditions may change and
compensation schemes may fail to provide managers with the right
incentives to act as perfect agents, potentially demanding the
undertaking of corrective action by the principals.
Those theoretical approaches to bank governance
of risk taking are about as useful as…
A NEW APPROACH: Regulators should govern banks
as debtholders would in a world without safety nets
A counterfactual analysis of what bank governance would be absent
safety nets:
1. Investigate the general dynamics of the interaction between
external and internal governance in corporations.
2. Apply the results of this investigation to corporations with high
leverage, since banks fit into this paradigm.
3. Contrast these results with the reality of the distortions introduced
in governance mechanisms by deposit insurance and other safety
nets.
 1. Multi-Dimensional Moral Hazard and the
Dynamics of Governance
On a general level, the governance activity of both shareholders and
debtholders has two components—monitoring and the exercise of
governance levers.
1. Monitoring involves the collection and processing of both
prospective and retrospective information about the firm.
2. The exercise of governance levers consists of the actions investors
take—based on the information they gather through monitoring —
to protect their investment expectations. These actions comprise:
1. Exit (investor selling shares in the corporation), and
2. Voice (the exercise of control rights)
 1. Multi-Dimensional Moral Hazard and the
Dynamics of Governance
The Effort Dimension
• Shareholders care most about managerial effort. But, there are
informational asymmetry problems. A useful way to mitigate
shareholder-managerial conflict is equity-based compensation but
there are also problems with this approach:
1. Manager performance is difficult to verify
2. Not effective in limiting private benefit extraction by managers – it actually
improves over time.
3. Does not limit entrenchment. Managers first concern is preserving power.
• But, Shareholders can use threat of Exit and Voice to discipline the
organization, in addition to minimizing costs and the cost of debt.
 1. Multi-Dimensional Moral Hazard and the
Dynamics of Governance
The Risk Dimension
• Scarcity of observable and verifiable information leaves debtholders
incapable of bargaining for interest rates contingent on actions taken
by the firm. To protect their investment expectations, debtholders
price debt through a pooling mechanism: they pool firms in risk
categories and price debt based on the average risk of each category.
This leads to an inefficient allocation of debt capital.
• Since managers’ most often run the negotiation with debtholders,
equity-based compensation alights managers’ risk preference with
those of the shareholders.
WAKE UP!
 2. Governance with High Leverage
With high leverage and rapid exit rights, the debtholders can promptly
gain corporate control by triggering insolvency or converting their
investment. Thus, shareholders have stronger incentives to distinguish
their firms as safe investments. But, pooling provides a problem.
• If bank shareholders are invested in Project A (low risk), but can
substitute it with Project B (high risk), pooling accounts for the
probability of substation and compensates debtholders ex ante for
the probable risk.
• If Project B fails to materialize, shareholders are still paying the higher
interest rate. Thus, they have incentive to send debtholders a signal
that they can credibly commit the bank to Project A.
 2. Governance with High Leverage
• Signaling will only be efficient for shareholders when its cost is less
than the net gains shareholders obtain from lower interest rates.
• Negotiation of control covenants is the standard signal that firms
employ to reduce the costs of debt pricing. Covenants are costly:
• Burden firms with high opportunity costs, and
• Burden debtholders with high monitoring costs
• Posting costly collateral is another signal used to contain interest
rates. But, collateralization may be inadequate to protect debtors
interest, particularly when assets are highly fungible (as in banks),
because firms can substitute assets and “hide” risks.
 2. Governance with High Leverage
Thus shareholders need a “validating signal” to differentiate their firms’

• Venture capital firms provide an answer in how they govern start-

ups. They require start-ups’ provide for debtholders to make some
board appointments, therefore less biased decision making.
• This is known as the “advocacy system”, a decision-making structure
that combines an independent decision maker with the competing
interaction of individuals appointed to be advocates of specific
causes (Ex.: a court proceeding – plaintiff, defendant, judge).
 2. Governance with High Leverage
How would this governance structure work in banks? Would require
two defining features:
1. The appointment of a representative of the debtholders’ interest
within the bank. The CRO (chief risk officer) would be a natural
candidate and can have a payoff schedule that decreases with the
level of the bank’s risk taking.
2. A fully independent board of directors. This would guarantee the
impartiality of the board as adjudicator of the competing panel of
risk information provided by the CEO and the CRO. [We need to
watch out for these guys since they golf together.]
Are we there yet?!?!
Soon… soon…
 3. Governance and Insurance
Why safety nets—deposit insurance and bailouts—give rise to the
adoption of governance arrangements that promote rather than
constrain risk taking.
• Moral Hazard of Bank Debtholders: When debtholders are insured,
they are less sensitive to increased risk taking, which weakens the
disciplinary effects of debt. Also, since monitoring is costly, with
safety nets, debtholders may have incentives to abandon monitoring.
• Inactive Bank Shareholders: Safety nets lead to rational shareholder
passivity. with bank debtholders’ opportunistic abandonment of
monitoring, bank shareholders have no incentives to police
governance arrangements that induce risk taking (tail risk).
 And now a short, relaxing intermission.
rd
Please find the 3 derivative of the function.
REFORMING BANKS’ REGULATORY DISCIPLINE
• Safety nets are an inevitable response to bank runs and the
macroeconomic shocks from individual bank failures.
• Bank regulators should act as substitutes for bank debtholders,
exerting the same kind of discipline debtholders would bargain for if
they were not implicitly or explicitly insured.
• Regulators should expand the set of regulatory tools they use to
discipline banks, making capital requirements and deposit insurance
premiums sensitive to the risk propensity of a bank’s organizational
structure.
REFORMING BANKS’ REGULATORY DISCIPLINE
Self-Regulation and Safety Nets
Why not just eliminate safety nets?
1. Safety nets protect the banking system in uncertain times, avoiding
catastrophic spillover effects. Even a single run may have systemic
effects, either through contagion effects (i.e., “bank panics”) or
interbank correlation. Also, the risk of bank runs has increased
because of banks’ growing reliance on short-term liabilities.
2. As a corollary of (1) above, legislative commitments against the
future use of safety nets are not credible, due to the catastrophic
effect of too big to fail institutions on the economy.
REFORMING BANKS’ REGULATORY DISCIPLINE
• Prudential regulation comprises four key sets of provisions: (i)
supervision rules; (ii) activities restrictions; (iii) deposit insurance
rules; and (iv) capital adequacy requirements. One rational for
prudential regulations is to protect small, unsophisticated depositors.
Another is to prevent systemic risk and, with it, costly ex post
interventions by the government in support of failing banks.
• Prudential regulators should act as substitutes for opportunistic bank
debtholders, exerting on banks the same kind of discipline
debtholders would bargain for if they were uninsured. This will bring
about shareholders’ commitment to more conservative risk choices
through the use of safe governance arrangements.
REFORMING BANKS’ REGULATORY DISCIPLINE
• Nevertheless, current regulatory practice largely fails to consider bank
governance and focuses instead on capital-related rules—primarily,
capital requirements. This is a crude response.
 The limits of capital requirements
Why have Capital Requirements?
• A larger equity buffer reduces the expected gains from taking more
risk.
Problems with Capital Requirements:
• Capital requirements need to be very high in order to be effective.
• Capital regulation may limit liquidity production, which can
exacerbate the negative effects of economic shocks.
• Tends to be socially expensive. Banks may accept fewer deposits,
thus reducing loans. May direct shareholders to divert money to
more profitable industries.
 Contracting around bank governance
• Using risk as a bargaining chip between banks and regulators can
prove highly efficient. Ex. Bank shareholders can commit to
undertaking a less risky project by using a credible signal in return for
the bank regulator excluding additional capital requirements.
• Shareholders can optimize their regulatory cost of capital by
choosing, for example, an advocacy model over a CEO-centric model.
• Under the current regulatory framework there are only superficial
references to banks’ governance arrangements in the form of
examination ratings, such as the CAMELS ratings. More
fundamentally, governance provisions play almost no role in the
regulators’ evaluation of a bank’s appetite for risk.
 Incentivizing advocacy in banks
Regulators should resort to one-on-one negotiations only for “large,
interconnected bank holding companies” as defined in the Dodd-Frank
Act (bank holding companies “with total consolidated assets” of $50
billion). Other banks should be subject to a standardized system.
1. The Standardized System for Small and Medium-Sized Banks. Regulation
should include just two “contracts” which would limit banks’ alternatives to
either an advocacy model or a CEO-centric model.
2. The Tailored System for Systemic Banks. (1) A standardized system can never
eliminate the likelihood of misalignment between optimal private sorting and
optimal social sorting. (2) The relatively small number of large, interconnected
banks (approx. thirty) would make an assessment system based on individual
negotiations feasible.
 Rebuffing objections
1. One potential objection to the adaptive regulatory approach proposed
here is that it would be costly.
• Relative to bailouts, the cost is de minimis
• There is already a system in existence to build on. Ex. The regulatory infrastructures
provided by examination ratings; existing deposit insurance rules for incorporating
bank governance into the determination of deposit insurance premiums.
2. A more radical objection to this Article is the cost of active governance
taking into account shareholders’ coordination problem.
• This objection overlooks that shareholders can exit. They had the choice to exit
before the crisis, but they rationally chose to remain because riskier governance
arrangements were profitable.
• As well, the increase in activist hedge funds and private equity funds points to three
being great room for shareholder voice in large banks.
A special note to dissenters of this article
CUE FOR AUDIENCE TO WAKE UP OR AVERT
EYES FROM FACEBOOK