Getting the Risk Right in Risk-Based Capital

Remarks Prepared for the Federal Deposit Insurance Corporation
Risk Management Conference

July 31, 2002
New York, New York

Karen Shaw Petrou
Managing Partner
Federal Financial Analytics, Inc.

The Basel Committee is now working desperately to finalize a wholesale revision of the risk-based capital rules that have governed international banking since 1988. The revisions are desirable in concept because flaws in the current rules have created anticipated, but apparently unavoidable, incentives for banks to take credit risk. However, the scope of the current project raises the risk that the revised rules may create more risks than they curtail, largely due to the fact that banking is an even more indistinct financial services sector than it was more than a decade ago. Regulatory standards that distort economic incentives — instead of appropriately disciplining them — create macroeconomic and systemic risks. As a result, the Basel Committee should move with care, first making only those revisions to the capital rules clearly warranted by widely-acknowledged problems in the current system. These changes should be based on consensus credit risk concepts, not regulatory-derived innovations, and the final rules also should not include a new capital charge for operational risk.

Credit Risk Rules and Procyclicality

One key concern — as evidenced by this panel’s main topic — is the degree to which regulatory action can exacerbate economic cycles, thus creating costly booms and busts instead of encouraging stable economic growth. To the degree that banks regulated by the Basel Accord are a diminishing segment of the total financial services industry — at least in the U.S. — this risk diminishes, but it remains a real and serious one. However, even if macroeconomic risk drops as capital rules drive risk outside the banking system, systemic risk clearly increases. Banks — in sharp contrast to other financial intermediaries — are subject to ongoing supervision and more or less meaningful capital standards. To the degree that the capital rules and other supervisory policies drive risk-taking into non-banking organizations, a truly perverse systemic result will ensue.

EU regulators have less to worry about in this regard, so they are freer to pursue capital initiatives that meet EU competitiveness goals and reflect the EU’s emphasis on capital, not bank examination. Under the EU’s Capital Adequacy Directive, the Basel rules will be applied not only to banks, but also to non-bank financial services firms. Thus, to take an example, asset management will come under the Basel operational risk rules whether conducted in a bank or outside one. In the U.S., however, the Federal Reserve does not have the authority to impose the capital rules on financial services firms that eschew the financial holding company charter. This makes the systemic risk issue a very real concern in the U.S., since banks facing uneconomic capital requirements could well abandon their charters to remain competitive.

The fact that banks in the U.S. can move some businesses outside the banking system may reduce the procyclicality concern, but it does not eliminate it. Several changes to the credit risk rules would alleviate the potential procyclical and systemic risks in the pending Basel proposal:

First, careful segregation of expected and unexpected losses in the capital framework, with reserves handling expected losses, would diminish the procyclical effect of the rule. To the degree that reserves — not capital — bear expected credit losses, bank earnings and capital are more stable through economic cycles, especially if reserve policy is improved to make reserve allocation more forward-looking. Doing this in the U.S. would require a final agreement between bank regulators and the SEC with regard to loan loss reserves. Further, U.S. policy should be changed to make reserves a tax-deductible expense, as is the case in most other G-10 countries.

Second, current efforts to keep the total impact of the revised rule within the same amount of capital now imposed means that the new rule will likely under-estimate capital for both very low and very high credit risks. This could significantly distort credit markets under stress, since banks will simultaneously flee to low-risk assets to protect capital and high-risk ones to increase return. Politics aside, the revised rule should attempt to allocate standardized and IRB capital to real risk, not try to fit the new rule into a spread-sheet with 8% at the bottom.

Finally, the credit risk indicators used in all aspects of the Basel rule should rely not only on external ratings, but also on other acknowledged indicators of credit risk. Sole reliance on ratings will exacerbate the procyclical problem in the rule, since a well-recognized failing of most ratings is their inability to anticipate risk. Ratings tend to be backward-looking — as Enron’s descent to junk status upon its bankruptcy again made clear. The Basel rules do not, for example, recognize loan-to-value ratios in setting capital for mortgage assets, despite the long-recognized relationship between LTV and mortgage credit risk. Inclusion of loss-given-default in the internal ratings-based (IRB) capital options is an effort to incorporate factors like LTV, but this approach — in contrast to explicit recognition of LTV — could increase procyclicality. Loss at default estimates will likely rise and fall with economic cycles as banks change their assessment of collateral value. As conditions worsen, potential losses increase — forcing a capital hike as the cycles trend down. Reliance on LTV at the outset, however, minimizes this risk because high-LTV loans will bear higher capital even during economic upswings.

Perverse Incentives

One big risk in the risk-based capital rules is regulatory creation of perverse incentives. These already exist in the credit rules, which — unless or until corrected — encourage banks to take more credit risk. Now, the Basel Committee is considering adding a regulatory capital charge for operational risk. This proposal could have serious and adverse consequences, and it should quickly be dropped. Regulators should devote the resources now distracted by the complex quantitative exercises necessary to develop the operational risk-based capital rules to improving supervisory standards and bank operational risk management and mitigation.

The proposed new capital charge for operational risk is based on a simple percentage of gross income, despite the fact that operational risk is not linear. Further, this approach means that the more banks spend on operational risk mitigation — back-up facilities, disaster planning, insurance — the higher their actual regulatory capital burden. Thus, lower risk banks will pay higher effective regulatory capital. September 11 demonstrated all too graphically the vital importance of all forms of operational risk management — especially contingency risk planning and disaster recovery. It further highlighted the value of insurance. Look at how few commercial and investment banks experienced earnings volatility — let alone solvency or capital problems — after the tragic World Trade Center attacks, and the vital role of third-party insurance becomes clear. Why regulators are considering an operational risk-based capital rule that actively discourages risk mitigation is hard to fathom, especially given the vital need to enhance financial system risk management and mitigation in an era of new threats.

To be sure, the regulators are attempting to remedy some of these problems through an "advanced measurement approach" to operational risk-based capital. However, the sophisticated approach would have a floor based on gross income, wholly reducing the benefits such an approach could offer. It’s hard to understand why all the work is going into an operational risk capital charge when regulators have long agreed to leave interest-rate risk in the supervisory context without a specific capital charge. Interest-rate risk is far more quantifiable than operational risk — as the minute-to-minute pricing of interest-rate swaps makes clear. Thus, even the advanced approach is deeply flawed because of its assumption that operational risk can now be quantified, and the entire effort should therefore be dropped.

Converging on a Solution

One must conclude that the Basel effort to rewrite the risk-based capital rules is a necessary one, given the numerous problems in the 1988 Accord, but also one fraught with potential risks. Not only are there serious risks that the new rules will exacerbate the procyclical impact of the current standards, but the rules could also create a range of new, perverse incentives that lead banks to take more — not less — risk. To some degree, the systemic implications of the Basel rules are mitigated — at least in the U.S. — by the fact that banks subject to uneconomic regulatory standards can opt out and structure themselves as non-bank financial services firms. This is true even with respect to very traditional banking businesses — taking deposits, for example — which can be housed in savings associations or special-purpose banks not necessarily subject to the Basel standards. Arguably, these charter options make the U.S. financial system more efficient, given that companies do have choices if regulation becomes excessive or uneconomic. However, it also makes the system riskier.

Some defenders of the pending Basel rules acknowledge that the capital charges may not correspond to economic risk, but defend this by pointing to the government support afforded banks. However, it is important to remember that large companies that elect not to become banks also enjoy explicit and implicit federal support. Securities firms, for example, have discount window access — and they don’t pay for it through the cost of regulation borne by banks that, admittedly, are first in line at the window during market turmoil. However, investors know that the Fed will step in for large broker/dealers as it did in 1987 and in 2001, protecting the financial system as a whole — not just banks. Further, the Fed’s active role in rescuing Long-Term Capital Management in 1998 led many investors to conclude that even big hedge funds have an implicit Fed guarantee, as long as their balance sheet is big enough. Banks do, of course, have FDIC insurance not provided to holders of mutual funds or other retail investments, but capital is only one of the rules intended to protect taxpayers from the risk that the deposit insurance funds run dry. Further, insured depositories have paid the premiums that fund the system, albeit not necessarily under extreme situations in which taxpayer-funded support must be lent.

To the degree that the Basel rules impose uneconomic costs on banks and banks abandon their charters, overall systemic risk increases. That this may happen only in the U.S. and not the EU is cold comfort, especially to those of us here. Worthwhile efforts are underway to design a regulatory structure for a rapidly-converging financial services industry, with rules applied on a line-of-business, not charter, basis. This is a very worthwhile effort, but until it develops a meaningful capital and supervisory framework, bank regulators need to take extra care that their initiatives conform to market realities, imposing no more regulatory cost on banks than the economic risk requires. To the degree that emerging risks — operational ones, for example — concern regulators, they should turn first to their own responsibilities and revise and improve supervisory standards. It is worth remembering that several recent bank failures in the U.S. were of banks that met or exceeded the regulatory capital requirements, but they still failed — at considerable cost — because of admitted supervisory glitches and accounting gaffes.

In proposing the new Basel rules, regulators have outlined a three "pillar" approach to proper bank regulation, with capital standing next to improved supervision and expanded disclosures. However, virtually all of the effort to date has focused on the capital rules, with the proposals creating potential macroeconomic and systemic risks as outlined above. Because of the direct impact regulatory capital rules have on a bank’s return-on-equity and, therefore, profitability, the impact of improperly-drawn capital rules is far more severe than that associated with potentially burdensome supervision or confusing disclosures. Thus, the excessive focus on capital to the exclusion of the other two pillars is a high-risk regulatory strategy. Improved supervision must be a top regulatory priority, with considerable care taken as the risk-based capital rules are revised.