Contemporary Issues in Risk Management and Banking Regulation*
William C. Hunter**
Introduction
I am pleased to have the opportunity to speak to this very impressive group of scholars
and policy makers that have come together to discuss issues related to risk management
and banking regulation. I will focus my remarks on the lessons that banking and
financial markets regulators and policy makers have learned regarding the practice of
risk management in banking and the development and implementation of risk
management techniques over the past couple of decades. Let me note that much of what
I have to say today derives from discussions held at the Federal Reserve Bank of
Chicago at our annual conference devoted to international financial policy issues. The
first conference, held in 1996 and cosponsored with the American Enterprise Institute,
examined financial derivatives and public policy. The second, cosponsored with the
World Bank, was devoted to global banking crises. The third was cosponsored with the
International Monetary Fund and focused on the Asian Financial crisis. The fourth
meeting was cosponsored with the Bank for International Settlements and examined the
lessons learned from recent global financial crises. The fifth meeting was cosponsored
with the Journal of Banking and Finance and featured academic work on quantitative
risk management modeling. This year’s conference was in April and was cosponsored
with the World Bank. The topic of the meeting was asset price bubbles and their
regulatory, monetary, and international policy implications. The proceeding of this
conference are being published by the MIT Press and the book will be available in late
fall.
Looking back at our experiences with risk management over the past couple of decades,
policy makers have learned some important lessons that are directly applicable to how
we supervise financial institutions and make monetary and regulatory policy. In
particular, I believe that some of these lessons are especially relevant to regulatory
reforms that are currently being advocated in the risk management and assessment areas.
Risk Management and Measurement
Perhaps the most basic lesson we have learned from our experience with the financial
engineering and risk management activities of banking companies over the past two
decades is that what is important is the underlying risk characteristics of financial
instruments, not what these instruments are called. This is important because some
elements of our legal and regulatory framework are undoubtedly outdated and may have
unexpected consequences for such things as legal and regulatory compliance risks. For
example, in the U.S. an instrument's name (totally divorced from its underlying risk
characteristics) may actually determine which regulatory agency has oversight
jurisdiction for that instrument and how that instrument is regulated. However, it is
*
Paper presented on 2002 Forum on Risk Management and Banking Regulation, July 4, 2002, Taipei,
Taiwan, Republic of China. The views expressed in this paper are those of the author and not
necessarily those of the Federal Reserve Bank of Chicago or the Board of Governors of the Federal
Reserve System.
**
Senior Vice President and Director of Research, Federal Reserve Bank of Chicago, USA.
exactly these risks, legal and regulatory compliance, along with market risk, credit risk,
operational risk, and reputational risk, to name a few, that are important to prudential
supervisors.
As is often noted, the use of financial derivatives does not typically involve the creation
or introduction of new types of risk. However, some financial derivatives do combine
or separate out different types of risk in new ways that require users to develop more
advanced risk-management capabilities. And as several dealers in the over-the-counter
derivatives market have learned in recent years, the marketing of new, complex
financial instruments may entail special reputational risks that demand special attention.
A second fundamental lesson learned with respect to the management and measurement
of risks over the past two decades is that risk must be measured and managed on a
portfolio basis rather than on an instrument-by-instrument or asset-by-asset basis.
Clearly, the consolidated firm has the incentive to manage its risk on an aggregate basis
whenever there are non-negligible diversification benefits deriving from its different
units/assets/or instruments. Although this is arguably the first principle of modern
finance theory and is widely practiced by financial market participants, putting this
principle into practice in prudential regulation has not been so easy. For example, past
banking crises have in part reflected a failure to recognize or to prudently limit
concentrations of risk within portfolios (most often in real estate). In recent years,
financial institutions and their supervisors have placed increased emphasis on the
importance of consolidated risk management. Sometimes called integrated risk
management, consolidated risk management refers to a coordinated process for
managing risk on a firm-wide basis. Interest in consolidated risk management has
arisen for a variety of reasons including: advances in information technology and
financial engineering, the increase in the number domestic and cross-border banking
mergers, and the increase in mergers between banks and non-bank financial institutions.
These developments have resulted in significant consolidation in the financial services
industry as well as in large, more complex financial institutions. These developments
have also made the process of risk management and risk measurement more
interdependent.
In simple terms, consolidated risk management refers to the overall process that a firm
(financial institution) follows to define a business strategy to identify the risks to which
it is exposed, to quantify those risks and to understand and control the nature of the risks
it faces. Thus, risk management comprises a series of business decisions, accompanied
by a set of checks and balances-- risk limits, independent risk management functions,
risk reporting and review, and oversight by senior management and the board of
directors. Thus, consolidated risk management involves the broad process of business
decision-making and of support to management in order to make informed decisions
about the extent of risk taken both by individual business lines and by the firm as a
whole. As is clear from this description, risk measurement-- which entails the
quantification of risk exposures-- is a key (but not the only) component of a
consolidated risk management system. This quantification may take a variety of forms-value-at-risk (VaR), earnings-at-risk (EaR), stress scenario analyses, duration gaps, etc.
depending on the type of risk being measured and the degree of sophistication of the
estimates.
2
The movement towards consolidated risk management means that the risk measures that
underpin the overall risk management process must better
 reflect the fact that risk is managed on a portfolio or consolidated basis rather
than on an instrument by instrument basis, and
 they must also be capable of capturing or quantifying the extreme/nonnormal risk events that are likely to be associated with large complex
financial institutions and the spillover effects that accompany financial crises
and severe market disruptions in a more integrated global financial system.
This is the third fundamental lesson learned.
To adequately deal with the portfolio approach to risk management, the notion (or
criterion) of a coherent risk measure has been put forth (Artzner, et al. 1997, "Thinking
Coherently," Risk, Vol. 10, No. 11, pp. 69-71). Essentially, a risk measurement
methodology will yield a coherent risk measure if the capital required to protect a
portfolio of two positions is no greater than the sum of the capitals required for each
position. That is, if the risk measure satisfies a sub-additivity property. Hence,
coherent risk measures capture the diversification effects associated with portfolios of
less than perfectly positively correlated assets or risks. With respect to the level of
capital required for bank credit risks, the recent Basle capital proposals it did not go as
far as to authorize the application of sophisticated portfolio credit risks models by
individual banks. However, the proposal clearly raises the bar for the level of
sophistication of credit risks models and has motivated the development and
enhancement of internal credit rating systems and other credit risk capital considerations
including operations risk, and their impact on economic capital and the pricing of credit
assets. Given that many of the traditional methodologies for quantifying credit and
portfolio risks such as VaR are not coherent (e.g., there are cases where a portfolio can
be split into two sub-portfolios such that the sum of the VaR corresponding to the subportfolios is smaller that the VaR of the total portfolio), more robust risk measures are
required. In addition, more research aimed at the related problem of developing
coherent risk measures that perform well in the face of extreme movements of the types
associated with financial crises and severe market disruptions would be welcomed.
Although risk measurement methodologies are still evolving and have some significant
limitations, they do facilitate the identification and quantification of concentrations of
risk within trading and asset portfolios. This is accomplished by analyzing derivatives
and other financial instruments and assets in terms of their effect on the sensitivity of
the institution's total portfolio to changes in a common set of underlying market risk
factors- interest rates, exchange rates, commodity prices, and equity index values,
among others. Clearly, the usefulness of these measures is highly dependent on the
degree that correlations among the common risk actors are accurately measured.
In recent years, leading commercial and investment banks have devoted increased
attention to measuring credit risk and have made important gains, both by employing
innovative and sophisticated risk modeling techniques and also by strengthening their
more traditional practices. For example, some market vendors model measures of
default risk by applying option theory to the market value of a borrower's equity share
price and calculates the probability of a negative net worth. This approach has the
3
important feature of incorporating market assessments of risk into the analysis and is
often used to validate a bank's independent view.
Other models, including many used at commercial banks, take a more direct approach to
calculating the fundamental elements of credit risk. They estimate the probability a
borrower will default, based on numerous measures of financial strength; the bank's
exposure given a default, reflecting any unused commitments of the bank to lend; and
the expected loss given default, taking into account any collateral or other lossmitigating features of the credit agreement.
Combined, these measures reveal the expected loss, which a bank must know to
underwrite and price a credit correctly, as well as to establish adequate loss reserves.
However, it is the volatility of this loss and the contribution of the credit to the volatility
of the bank's cash flow on a firm-wide, portfolio basis that is crucial to evaluating
capital adequacy. Thus, testing the sensitivity of these models’ results to perturbations
in the key factors that underpin their construction is a key ingredient required to ensure
the integrity of the consolidated risk management decisions that flow from these models.
Another lesson that our experience with risk management over the past two decades has
driven home is the critical importance of firms' internal risk controls. This, of course, is
a key component of an effective consolidated risk management program. As noted
above, given proper quantification of risks, a comprehensive set of risk limits carefully
monitored and strictly enforced, is a prerequisite for realizing the full benefits of new
financial engineering technologies and instruments while avoiding the obvious pitfalls.
Related to the need for effective monitoring and enforcement of risk limits is the need to
align financial incentives with management objectives. Stated differently, an effective
consolidated risk management system requires that compensation schemes in financial
firms reflect not only the returns generated by traders and portfolio managers but also
the risk assumed in generating the returns. Again, this requires that the firm accurately
quantify its risks. In this regard, many leading commercial and investment banks are
well aware of the need to link the compensation of individuals who are empowered to
commit the firm's resources to the risks actually undertaken.
The Importance of Financial Infrastructure
The Asian financial crisis of 1997 and the subsequent crises in Russia and Brazil in
1998 taught us much about the relationship between risk management and
macroeconomic and global financial stability. In the fall of 1998, after Russia’s debt
default and devaluation and the near collapse of the hedge fund Long Term Capital
Management, international financial markets seized-up for nearly all high-risk
borrowers including those in the United States and other developed economies. At that
time, the Federal Reserve’s Federal Open Market Committee lower its intended or
target federal funds rate three times (by a cumulative total of 75 basis points) in
response to the systemic threat posed by these developments. Global growth slowed
dramatically but systemic crisis was averted. Two things seem clear with hindsight.
First, credible signaling by a central bank of its intent to contain an incipient financial
crisis can solve problems associated with coordination failure in financial markets.
Second, the long period of remarkable economic growth and prosperity in Asia masked
4
weaknesses in risk management at many financial institutions. Many Asian banks did
not assess risk or conduct cash-flow analyses before extending loans. Instead, they lent
on the basis of their relationships with the borrowers and the availability of collateral-despite the fact that collateral was often hard to seize in the event of default (due to
poorly specified and/or unenforceable bankruptcy laws). The result was that loans-including loans by foreign banks--expanded faster than the ability of the borrowers to
repay. In addition, because many banks did not have or did not abide by limits on
concentrations of lending to individual firms or business sectors, loans to overextended
borrowers were often large relative to bank capital, so that when economic conditions
worsened, these banks were weakened the most.
The Asian crisis also illustrates the potential benefit of more sophisticated risk
management practices. Many Asian banks did not adequately assess their exposures to
exchange rate risk. Although some banks matched their foreign-currency liabilities
with foreign-currency assets, doing so merely transformed exchange-rate risk into credit
risk, because their foreign-currency borrowers did not have assured sources of foreigncurrency revenues. Similarly, foreign banks underestimated country risk in Asia.
In both cases, institutions seemed to have assumed that stability would continue in the
region and failed to consider what might happen if that were not the case. A greater
willingness and ability of banks to subject their exposures to stress testing could have
highlighted the risks and emphasized the importance of key assumptions. Had they
conducted stress tests, some lenders might have seen how exposed they were to changes
in exchange rates or to an interruption of steady economic growth.
Regarding the lessons to be taken from recent experiences with financial crises around
the world, (Asia and Russia being among the most recent), first, no one country should
be consider itself immune to financial crisis and second, the cost of resolving a crisis
once it has occurred is almost certain to be very high. For example, over the past 20
years alone, more than 125 countries, including the United States, have experienced at
least one serious financial or banking crisis. In more than half of these episodes, a
developing country’s entire banking system became insolvent. In more than 40 cases
the cost of resolving the crises averaged about 12.8 percent of GDP. In developing
countries (over all) the percentage was even higher at about 14.3 percent.
The details of the more recent crises are probably familiar to everyone here today, so I
will not describe them in detail. Just let me say that, similar to the Asian countries,
countries suffering financial crises over the past three decades typically exhibited many
of the same characteristics:




government directed and connected lending,
poor supervision of the financial system,
inadequate legal infrastructure,
absence of a credit culture in which lenders and investors make judgements
based on independent credit assessments and sound financial analysis,
 underdeveloped bond and long term capital markets,
 lack of adequate accounting disclosure and transparency,
 ineffective systems of corporate governance, and
5
 excessive and imprudent risk-taking by financial institutions operating with
explicit or implicit deposit insurance schemes or governmental guarantees,
among others.
However, at the same time these same countries also had unique elements that
contributed and in many cases, set off their particular episodes:
 weak fixed exchanges rate regimes (often pegged to the U.S. dollar),
 significant debt and asset price deflation, and in some cases
 persistent current account deficits, most often financed by short-term
unstable foreign capital inflows, among others.
Thus, another lesson we have learned from recent global financial crises is that while
most crises have much in common, they nevertheless can have unique precipitating
elements. The list of common characteristics observed in the crisis countries suggests
lessons in and of themselves. Stated differently, another (and perhaps the most
important) lesson we have learned regarding the potential for economy-wide financial
crises, is that infrastructure, broadly defined to include most of these common
characteristics, matters.
Countries with strong financial infrastructures including good operational (and not just
theoretical) systems of supervision and regulation, legal frameworks, and private
property rights (including bankruptcy laws) have tended to be more immune to financial
shocks and have tended to enjoy more stable rates of growth. Such stability in the
financial system breeds the necessary trust that the Federal Reserve considers essential
to a well functioning competitive financial system. Recent research has shown that
weakness of legal institutions for corporate governance had an important effect on the
extent of currency depreciations and stock market declines in the Asia crisis countries.
Measures of legal protections, enforceability of contracts, corruption, and judiciary
efficiency were all shown to be important in determining the depth of the crisis. Thus,
it is clear that an effective financial infrastructure, defined broadly as above, is a
prerequisite for promoting market discipline.
Based on my remarks to this point, it seems clear that an economy susceptible to
financial crisis due to weak infrastructure is analogous to a human body with a
weakened immune system. An immune system with strong fundamentals can shrug off
minor infections without harm. However, the same low-grade concerns can wreck
havoc on a weakened immune system leading to systemic crisis. In this case, as with
financial crises, an ounce of prevention is worth a pound of cure. Since we can not fully
protect against or eliminate every potential risk factor that might give rise to a crisis
(financial or medical), devoting proper attention to preventive actions seems very
appropriate. This includes coordinated actions to improve the global financial
infrastructure or architecture (examples include efforts like the BIS capital standards
and the core banking principles).
Although there seems to be a consensus regarding the value of a strong financial
infrastructure, there is less of a consensus regarding the optimal design of this
infrastructure, including among other things, the exact role to be played by the
multilateral agencies in the grand design. Continuing with my medical analogy:
6
 What is the role of the doctor? (the regulator?)
 What medicines should be prescribed? (what intervention or regulation is
required?)
 What is the correct dosage? (how extensive should the regulation be?)
 How long should the patient continue on the medication? (when should the
regulation be repealed?)
 How will the patient be monitored? and
 When is the patient cured?
For some illnesses, the answers to these questions are well known. For other, more
exotic diseases, more research is obviously required. Similarly, in the case of financial
crises, more research is needed on issues related to the optimal design and
implementation of a financial architecture.
Coordinated Supervision and Market Discipline
Banking supervisors have learned the importance of ensuring that incentives induced by
capital regulations are compatible with supervisory objectives. In this regard, efforts to
enhance public disclosures of the scale and scope, results, and risks of bank trading
activities have been motivated by a desire to bring greater market discipline to bear on
banks. In addition, supervisors have begun to attempt to build better financial
incentives into regulatory capital requirements. The recent proposed amendments to the
Basle Capital Accord are designed to provide incentives for accurate risk measurement
by allowing banks to select that approach which best reflects underlying risks and
associated capital levels.
As the complexity of institutions, instruments, and markets—highlighted by increased
interconnectedness driven by advances in technology (both, information processing
technology and financial technology or financial engineering) increases, the probability
of market surprises and the need for coordinated supervision and regulation will almost
surely increase. In this regard, another lesson we have learned is that direct measures of
risk-taking can provide misleading assessments of overall exposure in an environment
characterized by complex interconnections among policies, institutions, instruments,
and markets. These interactions almost always produce non-normally distributed
outcomes (the “so called fat tails” problem) or surprise correlations in situations where
outcomes were thought to be ex ante independent. For example, direct lending
exposure of global depository financial institutions to Thailand, Malaysia, and South
Korea were limited at the time of the crisis. However, proxy hedging of these risks in
other more liquid and deep markets—including those in Hong Kong, Australia, Brazil,
and Mexico proved costly when these markets were adversely impacted by the evolving
events and their spillover effects.
The possibility of the need for increased regulatory and supervisory coordination has
been taken by some as grounds for supporting the case for a single global financial
regulator. It is my belief that this growing desire for centralized regulation must be
tempered with caution, given the inherent difficulties associated with the design and
implementation of socially optimal regulation. As is well known, inappropriate
7
implementation of even well designed regulation tends to create more problems than it
solves. This is because regulation does not occur in a vacuum…firms and agents react
to changes in regulation in ways to affect the effectiveness of the regulation. In essence,
regulators must know exactly how economic agents will react in order to implement
optimal regulatory rules. However, given that regulators are generally looking from the
outside in, such understanding is difficult and elusive. This is why enhancing market
discipline is so important to the overall process of international harmonization of
regulation. Market discipline compliments regulators’ efforts allowing for more
effective supervision.
To be sure, regulators have learned that there is no way to ensure a fail-safe financial
system. Regulators need help and market discipline provides a partial solution.
Enhancing market discipline through dynamic incentive compatible approaches is the
next challenge confronting the Federal Reserve System and regulators more generally.
From the perspective of the Federal Reserve, we must respond to the recently passed
financial modernization legislation (the Gram Leach Bliley Act) which allows full
affiliation of banking and securities and insurance activities creating large complex
banking organizations which have to be effectively supervised and regulated. Clearly,
that regulation must be adaptive. Similarly, in the case of emerging markets and
transitioning economies, regulation must be dynamic and adaptive. This makes sense
since these markets/economies are in fact emerging and/or in transition.
The Chicago Fed has long pushed for more market discipline through increased
disclosure and transparency and through such schemes as mandatory issuance of
subordinated debt by large complex banking organizations. In our view, mandatory
subordinated debt is incentive compatible in that it aligns the incentives and risk
preferences of bondholders with those of bank supervisors. Being subordinated to other
liabilities, the debt holders would be risk sensitive and would monitor and discipline
bank behavior. They would demand higher rates from riskier banks (a direct effect) and
have stronger incentives to quickly resolve problems and avoid forbearance and its
associated costs. In addition, increases in interest rate spreads provide signals to
supervisors that risk is increasing (an indirect effect). A recent issue of the Chicago
Fed’s journal—Economic Perspectives (Spring 2000) contains a comprehensive article
on subordinated debt and its advantages as a tool for regulators in supervising large
complex banking organizations.
Derivatives and Monetary Policy
Before closing, let me share a few observations regarding derivatives, financial crises,
and the conduct of monetary policy. Regarding financial derivatives and their role in
contributing to financial crises and systemic risk, it turns out that all the talk about
financial wizardry that allows unbundling and transferring of risks and the lightening
speeds of transmission of these risks across markets was probably overblown. That is,
recent crises remind us that by and large, the mistakes of the past few years were rather
ho-hum or ordinary and not so exotic—making bad loans (typically in real estate),
failure to evaluate counterparty credit risk, or simply relying on the lofty reputations of
firm principals (a la Long Term Capital Management). These are age-old problems
(especially in the banking industry). The lesson here is that derivatives were not a
8
major villain in recent crises. This suggests that maybe the first lesson we should learn
is not to forget the old lessons!
With regards to matters monetary policy, there is much disagreement on the overall
effectiveness of monetary policy in a world of sophisticated financial derivatives. It
could be that derivatives, in addition to complicating monetary policy, might also
reduce the impact monetary policy has on the real economy. That is, a given amount of
monetary stimulus may have a smaller effect. For example, the existence of foreign
exchange derivatives may reduce the ability of central banks to influence exchange rates.
Since money should be neutral in a frictionless economy, the sources of monetary nonneutrality must lie in economic frictions such as informational imperfections and
transactions costs. By increasing the liquidity, depth, flexibility and transactional
efficiency of financial markets, derivatives increase the speed with which monetary
policy actions are transmitted throughout the financial system. Lower transaction costs
and reduced frictions resulting from derivatives activities increase the rate at which new
information, including policy actions, is impounded into market prices. Since
derivatives markets reduce these sorts of frictions, they provide a more efficient
mechanism for price discovery, speed up information transmission, and reduce
informational asymmetries. Thus, it follows that by reducing frictions, derivative
markets may actually reduce the real effects of monetary policy actions.
Furthermore, derivative markets act as a mechanism for spreading shocks across the
economy as a whole. To the extent that derivatives reduce the force of monetary policy,
monetary policy may become a weaker tool for counter cyclical stabilization policy.
However, if derivatives do provide the economy with the benefits cited by their
proponents, that is, a more efficient, self-correcting and shock resistant economy, then
there should be less of a need for counter cyclical monetary policy. Stated differently,
derivatives activity might actually reduce the incidence and severity of business cycles
themselves.
Conclusion
Before closing, we highlight another important lesson learned from recent experiences
with financial crises and risk management. This lesson is that modern finance (as a
discipline) can be very useful in the design of economic institutions and policies as
modern macroeconomics is. Understanding such financial concepts as market microstructure and option value can be just as important in forecasting and mitigating crises
as is a sound understanding of modern macroeconomics. The Asian crisis is a good
example of this given that most of the countries had good macroeconomic fundamentals
but significant imbalances on the finance-side which escaped the scrutiny of economists
and policy makers. Indeed, such notions as agency costs, moral hazard, signaling,
complete and incomplete markets, bankruptcy and strategic default, and other micro
level incentive related concepts proved critical in fully understanding the recent crises.
The bottom line is that finance is important and the kind of work that many of the
researchers attending this conference are doing can prove valuable for the design of
macroeconomic and regulatory policy.
9