The Chicago Fed’s Supervision and Regulation Department, in conjunction with DePaul University’s Center for Financial Services, sponsored its third annual Financial Institution Risk Management Conference on April 6–7, 2010. The conference concentrated on comprehensive risk management, lessons learned, and headline issues.

In addition to overviews of the risk-management landscape, this year’s conference focused on commercial real estate (CRE), financial modeling, capital planning, and risk and compensation. This Chicago Fed Letter provides a summary of the relevant research presented and discussions held by the bankers, academics, and supervisors in attendance.

Opening the conference, Ali Fatemi, DePaul University, described how risk aversion on the part of both bankers and regulators had reduced the availability and increased the cost of credit. Establishing the “new normal” requires efficient allocation of credit at the lowest cost; therefore, he called for the proper balance of private and regulatory incentives to achieve this goal. Carl R. Tannenbaum, Federal Reserve Bank of Chicago, continued by contrasting the turbulent circumstances of the previous conference (held in the immediate aftermath of the worst of the financial crisis) with the relative calm surrounding the current conference, reflecting improvements in the economy and financial system.

Policymaker perspectives

Charles L. Evans, president and CEO, Federal Reserve Bank of Chicago, offered his thoughts on some of the challenges banks and regulators are likely to face.

For example, enhancements to microprudential regulations, which focus on individual institutions, need to be supplemented by macroprudential supervision, which considers risks to the financial system as a whole. However, such macroprudential approaches also face obstacles, such as knowing exactly when to intervene in a potential asset bubble. Evans advocated a multipronged approach featuring strengthened capital requirements, a comprehensive approach to risk management, a macroprudential supervisor, and a process for effectively resolving insolvencies at large institutions.

According to Evans, central banks should play a key role in financial stability and in the supervision and regulation of financial institutions. For one thing, a central bank without supervisory responsibilities would have to confront any financial crisis using only monetary policy. In such a case, the central bank might have to act against exuberance in financial markets by tightening monetary policy more than would be indicated by macroeconomic considerations alone.

A former governor of the Federal Reserve System, Randall S. Kroszner, currently of the University of Chicago, surveyed some of the lessons about risk management learned from the global financial crisis. He first provided an overview of some of the fragilities of the current financial system. Then he listed “seven deadly sins of risk management” that risk managers should strive to avoid. These included allowing accounting values to obscure economic realities, failing to fully capture risk concentrations, ignoring risks faced by funding counterparties, being overconfident during periods of high market liquidity, and failing to adequately model and manage tail risk.¹

Kroszner concluded by suggesting reforms for risk managers and policymakers to consider. Risk-management functions should be independent, have sufficient stature in the organization, and consider the full range of risks. Central clearing of derivatives should be encouraged to enhance market resiliency and mitigate interconnectedness problems. The resolution regime for large financial institutions should also be improved. Finally, regulators should improve the monitoring of liquidity risks and reform capital requirements.

CEO and chief risk officer perspectives

The conference featured perspectives on risk management from a bank CEO and a panel of bank chief risk officers. Christopher J. Murphy III, CEO, 1st Source Corporation, said that his organization had avoided much of the recent credit meltdown. However, the company had experienced virtually every imaginable type of failure, ranging from basic credit problems to an accounting dispute and an information technology security breach. He detailed the lessons learned from these various experiences. In Murphy’s opinion, the greatest failure of all was allowing human hubris to creep into the company. Murphy emphasized building a culture based on constantly reinforced values of integrity and honesty. He also stressed that management’s incentives should not be overly influenced by the short-term stock price performance of the organization.

Richard C. Cahill, Federal Reserve Bank of New York, moderated a panel of three chief risk officers—Terry J. Bulger, BMO Financial Group; John S. Fleshood, Wintrust Financial Corporation; and Larry J. Kallembach, MB Financial Corporation. In addition to identifying what was distinctive about their firms, panelists shared the lessons they had learned from the recent financial crisis, including the need to better aggregate risks and to strengthen the culture and stature of risk management.

Two years ago, BMO embarked on a structured risk improvement plan, said Bulger. One goal of the plan was to increase risk transparency: BMO enhanced its risk reporting to its board and senior management and completed in-depth risk assessments for all its trading desks. Bulger stated that his organization’s priorities for 2010 are managing risk (including the problem loan portfolio) and simultaneously building capabilities for the future. The latter includes strengthening core risk-management practices, expanding a sound risk culture throughout the firm, and more effectively managing capital.

According to Fleshood, Wintrust is somewhat unusual in that its $12 billion in total assets is distributed among 15 individually chartered community banks. Under this structure, Wintrust’s board and subcommittees concentrate on enterprisewide risks. The subsidiary banks’ boards and their subcommittees, in turn, focus more on individual banks’ performance and adherence to corporate policies. Key challenges with such a decentralized structure are effectively aggregating risks and tailoring risk reporting to the needs of diverse audiences.

MB Financial also has a somewhat atypical structure in that its enterprisewide risk function evolved out of the operations/technology area, which focuses on traditionally underemphasized risks, such as anti-money laundering, business continuity, and information security. Kallembach, the company’s chief information officer, also serves as its enterprise risk officer, responsible for overseeing the management-level risk committee. Other important pieces of the enterprisewide view of risk are provided by a risk-management department (under the administration area), a board-level credit committee, internal loan review (which reports to the audit committee), and the chief credit officer.

Commercial real estate outlook

Weaknesses in managing CRE concentrations were responsible for much of the current deterioration in banking conditions affecting many small and mid-sized banks. Furthermore, policymakers are concerned that continuing weaknesses in CRE could impede the economic recovery. James D. Shilling, DePaul University, moderated a panel that provided diverse perspectives on CRE. The panelists were Robert Bach, Grubb and Ellis Company; Timothy Riddiough, University of Wisconsin–Madison; and Brian D. Gordon, Federal Reserve Bank of Chicago.

Bach provided data and analysis on current trends in CRE markets, including rising vacancy rates, falling rents, sharply reduced investment volumes, and a growing pool of distressed assets. Riddiough analyzed how weaknesses in residential lending had combined with macroeconomic distress to undermine CRE markets. According to Riddiough, in the short term CRE lending faces a continuing lack of liquidity, and its longer-term recovery depends on the return of securitization markets, albeit under a stricter regulatory regime.

Gordon offered suggestions on how to estimate CRE losses. Loss estimation is critical for stress testing, analysis of capital adequacy, and evaluation of the adequacy of loan-loss provisioning (funds set aside as an allowance for bad loans). Gordon cautioned that broad loss rates should not be applied to individual portfolios, which can be highly diverse. He also emphasized that capital is meant for unexpected losses, not expected losses, which should be reflected in loan-loss reserves. Finally, Gordon highlighted the key points in the banking regulatory agencies’ guidance on prudent CRE loan workouts.² This guidance emphasizes that excessive foreclosures are in no one’s best interest and that loans should not be adversely classified solely because the value of collateral has declined. Overall, better estimation of CRE losses would have improved the adequacy of loan-loss reserves and capital at banks with high CRE concentrations and reduced the severity of the current banking crisis.

Financial modeling

Many have identified failures of (and inappropriate use of) financial models as among the key causes of the financial crisis. Tannenbaum examined the strengths and weaknesses of modeling with a panel composed of William H. Schomburg III, State Street Corporation; Michael Alix, Federal Reserve Bank of New York; and Deborah J. Lucas, Congressional Budget Office.

Schomburg said that models provide an analytic framework to assess risks and a common language to communicate these risks to others. To provide the necessary controls over models, State Street uses a highly structured four-level approach, comprising the model owners, validation and assessment groups, and a high-level risk-management committee with final responsibility. Schomburg emphasized that modeling needs to be supplemented with stress tests and expert judgment. However, a common weakness of modeling, stress tests, and expert judgment is a bias toward recent data.³

Alix provided a supervisory perspective. He stressed the need to understand the purpose of a particular model before trying to determine its effectiveness. In his view, one commonly perceived problem with risk models—that they don’t produce sufficiently distressed results—is attributable not to technical shortcomings but rather to the inability of users to “think outside the box” (i.e., to consider plausible, but never experienced, shock scenarios). Other problems are the lack of reliability of input data and the uncritical use of models’ output.

Lucas was not convinced that models themselves had failed during the recent crisis. However, there were clearly shortcomings in the implementation and use of models. It will be relatively easy to correct some of the technical problems of models and to use data that cover at least an entire deep business cycle. But it will be harder to model tail events, to effectively use the information that models provide, and to change the culture of how models are used. Specifically, decision-nmakers need to better understand models, to integrate analysis and judgment more, and to use models to inform decision-making (not only justify decisions already made), said Lucas.

Capital planning

In light of the large number of financial institutions whose capital proved inadequate during the financial crisis, internal capital planning has become a key focus of bank supervision. It was the subject of a panel moderated by Andre Reynolds, Federal Reserve Bank of Chicago. This panel featured Tanya K. Smith, Office of the Comptroller of the Currency; Ron Feldman, Federal Reserve Bank of Minneapolis; and Joseph R. Mason, Louisiana State University.

Smith outlined the “new normal” for capital planning. Capital planning encompasses identification of risks and risk tolerance, risk measurement, goal setting (for risks and capital), analysis of capital supply and demand, assessment of a range of operating requirements, and development of capital contingency plans. All of these must be incorporated into a sound governance framework. The end result of capital planning should not be just a single number. Instead, such planning should constitute a well-articulated, well-supported, and well-understood process surrounding the many facets of capital and risk.

Feldman considered capital in relation to the too-big-to-fail (TBTF) problem. TBTF refers to the provision of discretionary government support to the uninsured creditors of financial institutions perceived to pose systemic risk. He compared two options for addressing TBTF—a capital surcharge for systemically important institutions and a premium (e.g., incorporated into deposit-insurance premiums) that charges systemically important firms for the implied government support they receive. He concluded that a premium is superior because it is likely to 1) more efficiently and effectively discourage excessive risk-taking by financial institutions and 2) more transparently address the TBTF problem.

Finally, Mason applied a circular five-step risk-management cycle (develop goals, identify/quantify exposures, define philosophy, implement program, and evaluate and control) to a wide range of historical risk scenarios, both financial and nonfinancial. Failure to adhere to the cycle can lead to large unexpected losses, as in the case of Barings Bank in 1995 or Société Générale in 2008.

Risk and compensation

Alteration of incentive compensation practices is also high on the reform agenda. Keith M. Howe, DePaul University, led a panel on risk and compensation that featured Steven N. Kaplan, University of Chicago; Kevin J. Murphy, University of Southern California; and James W. Nelson, Federal Reserve Bank of Chicago.

Kaplan sought to determine whether poorly designed top executive compensation at financial firms had fueled the financial crisis. If this had been the case, we would have expected to find that top bank executives were rewarded for short-term results with large amounts of upfront cash pay; bank executives did not hold sufficiently large amounts of stock to align their interests with those of shareholders; and firms utilizing more short-term pay and less stock ownership as compensation performed worse in the crisis. Recent research that Kaplan cited did not support these propositions. Therefore, poorly designed top executive compensation does not appear to have played a significant role in the financial crisis, especially compared to other factors. Kaplan argued that more regulation of top bank-executive pay is unnecessary and would have negative unintended consequences, such as driving the most talented employees to unregulated sectors, such as hedge funds, private equity funds, and boutique firms.

Murphy was also pessimistic about increased government regulation of compensation. He said that regulation is often designed to be punitive and advance political agendas rather than to be constructive and foster creation of shareholder value. Nevertheless, Murphy argued that compensation practices in financial services could be improved. He suggested bonus deferrals and “clawback” provisions (where rewards are recovered if critical indicators on which bonuses were based are revised in the future). He also recommended basing bonuses on value creation rather than on sheer volume of transactions.

Nelson presented the Federal Reserve’s proposed guidance on sound incentive compensation policies.⁴ This guidance is based on three fundamental principles. First, incentive compensation arrangements should not provide employees with incentives to take risks beyond an organization’s ability to effectively identify and manage those risks. Second, such arrangements should be compatible with effective risk management and controls. Third, these policies should be supported by strong corporate governance, including active and effective board oversight. This proposed guidance is being supplemented by two supervisory initiatives—one for large, complex banking organizations and another for the remaining organizations.

Summing up

Eugene A. Ludwig, Promontory Financial Group, was the U.S. Comptroller of the Currency over the period 1993–98. He indicated that the “new normal” will incorporate stronger, fortresslike balance sheets and more effective chief risk officers and boards. While many of the firms that became troubled did have chief risk officers that were formally independent of business line management, often these individuals lacked sufficient influence to restrain excessive risk-taking. For chief risk officers to gain more influence, Ludwig supported producing targeted, high-quality risk reporting to boards rather than generating a massive tome (known in one firm as “the brick”) that nobody reads. In addition, risk-management models should place more emphasis on tail events, become more forward-looking, and draw on real-time, enterprisewide data.

For Ludwig, concentration levels in CRE at small and mid-sized banks were less of a problem than their business models that ignored other types of lending opportunities, such as business lending. Ludwig also recommended that banks improve the quality and diversity of their earnings by emphasizing deposit accounts, developing sustainable fee income, and looking hard for cost-saving opportunities.

Notes

¹ Technically, tail risk is a form of portfolio risk that arises when the possibility that an investment will move more than three standard deviations from the mean is greater than what is shown by a normal distribution. More broadly, the term is used to refer to the risk of large unexpected losses for the financial sector as a whole.

² For further details, see the Federal Reserve press release on prudent CRE loan workouts at www.federalreserve.gov/newsevents/press/bcreg/20091030a.htm.

³ Financial models often use recent data that do not cover past business cycles. For example, during the late 2000s, data from only the preceding few years would not have included any instances of sharp declines in asset values and thus would have produced overly optimistic modeling results. 

⁴ See the Federal Reserve press release at www.federalreserve.gov/newsevents/press/bcreg/20091022a.htm.