A quarterly magazine of the IMF

December 2000, Volume 37, Number 4

The Basel Committee's new capital framework proposals will have important implications for developed and developing countries alike. Although many details remain to be worked out, it is not too early for countries to start preparing for the proposals' implementation.

Since its launch in 1988, the Basel Capital Accord has become the global standard by which the financial soundness of banks is assessed. The outcome of an agreement among the members of the Basel Committee on Banking Supervision, comprising bank regulators from the Group of 10 countries, the Accord was originally intended to apply only to internationally active banks headquartered in those countries. It is now applied, however, in most countries—industrial, emerging, and developing—and to most banks, including many that operate only domestically. Hence, recent proposals from the Basel Committee to update its now 12-year-old agreement raise significant issues that affect many countries besides the small group of countries represented on the Committee. The Committee recognizes this and has therefore been consulting widely on its proposals.

The Accord's original aims were to stem the decline in bank capital observed for much of the twentieth century (see Chart 1 for an illustration drawn from Canada, the United Kingdom, and the United States) and to level the playing field for internationally active banks. To achieve these objectives, the Basel Committee developed a simple risk-measurement framework that assigned all bank assets to one of four risk-weighting categories, ranging from zero to 100 percent, depending on the credit risk of the borrower. Thus, a bank loan to a commercial company is 100 percent weighted, whereas a domestic bank loan made to the central government in national currency (a loan that is effectively riskless) is zero weighted. Interbank lending generally attracts a 20 percent risk weighting. The Basel methodology requires banks to maintain a minimum ratio of capital to total risk-adjusted assets—that is, the total for all of a bank's assets, after the amount of each asset has been multiplied by the relevant risk weighting—of 8 percent. Although its methodology is rather crude—for example, it makes no allowance for the effect of portfolio diversification—the Accord has a number of advantages: it is relatively simple to apply, and produces an easily comparable and verifiable measure of bank soundness. Moreover, by the early 1990s, its implementation had first halted, and then reversed, the decline in banks' capital ratios in most Group of 10 countries (Chart 2).

The Basel Committee has decided to revise the Accord now for a number of reasons. One of the most important is that the last 12 years have seen the rapid development of new risk-management techniques that have left the Accord looking increasingly outdated. Many leading banks have argued that their internal risk-management systems provide better evaluations of risk than the Basel Committee's framework, which, they argue, provides insufficient differentiation of bank assets by broad risk categories.

A second important reason is the effects of a decade of financial innovation, often undertaken to circumvent capital rules. Thus, recent innovations like securitization and credit derivatives have been driven, in part, by the Basel Committee's rules, and the emergence of such innovations has reduced the Accord's effectiveness. Although the Accord initially forced banks from the Group of 10 countries to raise additional capital, more than a decade of financial innovation has created risks that are not encompassed by its measurement framework. In effect, the Accord has become progressively less binding.

Options for reform

In July 1999, the Basel Committee issued a consultative document setting out a number of options for reforming the Accord. This document introduced three pillars—improving the framework for calculating capital adequacy, developing a process of supervisory review, and strengthening market discipline.

The three pillars are intended to be mutually supportive. The overall aim is to produce a capital-adequacy standard that incorporates some of the refinements of modern risk-management practice while maintaining the concept of a regulatory minimum capital requirement.

Improved calculation of capital adequacy. Most of the attention paid to the Basel Committee's proposals to date has been concentrated on the first pillar. The Committee's aim has been to introduce greater refinement into the existing system of risk weighting, to relate its categories more accurately to the economic risks faced by banks. While attaining this goal is clearly desirable, doing so has proved to be much more difficult than expected. The primary methodology proposed by the Committee—making regulatory use of ratings assigned by the credit-rating agencies—has attracted a good deal of criticism. Much of the debate has focused on the accuracy of ratings: it has been argued that credit ratings are poor predictors of default, a contention supported by the claim that the rating agencies "missed" the Asian currency crises before they occurred in 1997-98. A related concern is that the use of ratings could boost both capital inflows and outflows, thus making a repeat of the Asian financial crises more, rather than less, likely. Some developing and emerging market countries have criticized this proposal on the grounds that it would place too much power in the hands of the rating agencies; they are also concerned at the prospect of unsolicited sovereign ratings. The rating agencies themselves have also been less than enthusiastic about being placed in the position of performing a quasi-public function.

Many of the criticisms of the rating agencies have been overstated. As the IMF's 1999 International Capital Markets report pointed out, the reasons why the rating agencies "missed" the Asian financial crises were not that dissimilar to the reasons why many other observers, including the international financial institutions, were taken by surprise. Allegations that the use of credit ratings as measures of risk has made capital flows more cyclical are undermined by experience: ratings have been stable over time, and the response of the rating agencies to the Asian crises was much less volatile than that of financial market participants.

The proper selection of rating agencies, however, is a key prerequisite for the successful use of external credit ratings. The Basel Committee has proposed a number of criteria for the regulatory recognition of ratings agencies, but striking the right balance between having appropriate regulatory criteria and avoiding the creation of high barriers to entry into the financial services industry is a difficult matter. The experience of securities regulators in approving ratings for their own, somewhat different regulatory purposes may offer some useful guidance. Securities regulators build on the important observation that a rating agency's reputation is its most important capital asset, as well as the chief entry barrier to the market for credit ratings; accordingly, the U.S. Securities and Exchange Commission relies heavily on the opinions of market makers and participants in judging whether or not to give regulatory approval to the use of an agency's ratings. This suggests the desirability of an approval process in which rating-agency selection criteria maximize the market's input into the approval process, especially in developed country markets.

Even with substantial market input, it is likely that a decentralized approvals process would lead to inconsistencies that might compromise the Accord's application across countries. In consequence, the approvals process for credit rating agencies should be centralized, perhaps by creating a forum involving a representative group of local regulatory agencies and with the participation of multilateral financial institutions. This would ensure consistent application of the selection criteria.

Criticism of an approach based on external ratings has led the Basel Committee to place greater emphasis on an alternative it has also outlined in the consultative document. This is to permit banks to make extensive use of their own internal ratings systems to determine the amount of regulatory capital they need to hold. The change might be characterized as one from rules-based to process-oriented regulation. Whereas the original Basel Accord prescribed rules for calculating capital, which the external ratings approach seeks to refine further, the internal ratings approach would shift the regulatory emphasis to assessing the quality of banks' own risk-management processes. The shift in emphasis parallels the 1996 Market Risk Amendment to the Basel Accord, which permitted banks to use their internal value-at-risk (VaR) models as an alternative to the standardized approach for calculating capital requirements for market risk, subject to regulatory approval of these models. The new capital framework proposals seek to extend this philosophy to encompass credit risk as well.

Credit-risk models, which use many of the same statistical techniques as VaR models, are currently at the cutting edge of risk-management practice. They seek to generate a portfolio-wide simulation of the probability of default based on historic credit-loss data. One problem with such an approach, however, is that, in the Basel Committee's view, credit-risk models are not yet sufficiently refined to be used for regulatory purposes. Fundamental problems concerning the validation and back-testing of credit-risk models remain. The available data on credit defaults, for example, are not nearly as comprehensive as those on movements in the market prices of securities, which make it practical to use VaR models to assess market risk. Until these issues can be resolved, it will not be possible for credit-risk models to play the same role in assessing banks' capital adequacy that VaR models have played since 1996 in assessing market risk.

The Basel Committee consequently proposes to make use of banks' own internal rating systems. In theory, these should have some advantages over external ratings because they incorporate proprietary data about a bank's clients that are usually unavailable to the public at large. A number of studies of banks' internal rating systems show, however, that their methodologies vary widely, ranging from statistics-based systems to judgment-based ones. These systems are also used for quite different purposes and functions: some banks use them to identify deterioration in the quality of loans, while others use them to compute internal profitability—and, hence, also the remuneration of loan officers. Clearly, if banks' internal systems are to be used for regulatory purposes, greater standardization will be necessary, and the Basel Committee's current interest in them represents only the start of a long process of further development and refinement.

The potential shift toward more process-oriented regulation raises some important considerations for bank regulators in countries that are not members of the Group of 10. Most of their banks lack the sophisticated measures of internal risk that are the basis for the Basel Committee's proposed alternative approach to risk measurement. In consequence, most bank regulators will probably continue to rely on the rules-based standardized approach for the foreseeable future. In fact, only a few banks in the world currently have internal risk-management systems sophisticated enough to meet any tests that the Basel Committee is likely to set. The majority of banks—in both developed and developing countries—can therefore be expected to continue to use the standardized approach, incorporating whatever refinements the Basel Committee ultimately decides on.

At present, it is unclear whether the Basel Committee intends to keep the 8 percent minimum ratio or replace it with another requirement. The stated intention is that its proposals should primarily affect the allocation of capital between banks' different activities and not the total amount of capital they are required to hold. Even in countries that make widespread use of the revised standardized approach, however, it will be important to ensure that the 8 percent ratio—or its successor—is explicitly recognized as being a minimum requirement. Countries with volatile or riskier macroeconomic environments should require their banks to maintain considerably higher capital ratios—as, in practice, many of them already do. At the same time, regulators in emerging market countries will need to continue working to improve standards and practices in loan and collateral valuation, loan classification, and provisioning for possible loan losses, without which capital ratios become unreliable indicators of banks' soundness.

Upgraded supervisory review. The second pillar of the Basel Committee's approach is supervisory review. This is intended to emphasize that bank supervision is not simply a matter of adhering to a few simple quantitative ratios but also involves making qualitative judgments on such matters as the caliber of a bank's management, the strength of its systems and controls, the viability of its business strategy, and its earnings potential. One of the shortcomings of the Accord is that it has obscured the important fact that a single ratio—like that of a bank's capital to its risk-adjusted assets—is meaningful only in the context of a broader understanding of the risks confronting a bank. All too often, a bank that has appeared to be well capitalized, on the basis of this measure, has proved to be quite fragile shortly thereafter. Indeed, the existence of an officially sanctioned minimum capital ratio may give bankers, markets, and regulators alike a false sense of security—as happened during the Asian financial crises. By giving supervisory review the status of a second pillar of bank capital adequacy, the Basel Committee intends to ensure that the qualitative dimension of banking supervision is not overlooked.

For many emerging and developing countries, placing greater emphasis on supervisory review will require substantial upgrading of regulatory skills and the mobilization of more and higher-quality staff. Emerging markets and other developing countries' markets need to improve their ability to supervise banks and other financial institutions. The key elements of this approach, which should allow for external technical assistance where necessary, should include a diagnostic assessment of the country's overall approach to supervision, the implementation of a risk-based approach to on-site examinations, the design of effective early warning systems for off-site supervision, a rigorous follow-up on weaknesses identified by off-site and on-site examinations, and the adoption of a comprehensive training program for bank supervision. Obviously, this strategy cannot be completed in just a few years, and emerging and developing country regulators will need to start planning now to build the supervisory capacity necessary to make the second pillar a functioning reality.

Increased use of market discipline. The third and final pillar is currently the least developed of the three. It envisages making greater use of market discipline an adjunct to the supervisory process: this might take the form of supervisors drawing more heavily on market information in their work, but it could also eventually result in market discipline being used as a partial substitute for official sector supervision. One of the most widely debated proposals in this category is that banks should be required to regularly issue subordinated debt. The objective of this proposal is to create a class of investors whose incentives are aligned with those of banking supervisors and deposit-protection agencies, and who actively monitor, analyze, and exert discipline on banks. Most studies on the use of subordinated debt as an instrument of market discipline have focused on its potential use in the developed world, especially the United States. But before issuance of subordinated debt can be adopted as part of an international banking standard, its applicability to developing countries needs to be considered. For example, an important question is whether various developing countries' securities markets are deep enough to support the required levels of bond issuance and liquid enough to generate prices from which meaningful information can be extracted. Similarly, the proposal presupposes the existence of a class of nonbank financial institutions that would invest in these bonds, but such institutions are lacking in many emerging market economies. While these matters merit further research, mandatory subordinated-debt proposals should not be seen as a panacea in creating market discipline, but rather as one possible element in a set of policies designed to instill market discipline. These policies should include high accounting and disclosure standards, incentive-compatible (risk-based) financial safety nets, strict no-bailout policies toward insolvent banks, and openness to foreign ownership and competition.

Conclusion

The Basel Committee's proposals will undoubtedly have important ramifications for a wide range of countries—not only for capital flows but also for the nature of the supervisory regimes countries will need to operate. Although many of the details of the proposals remain to be worked out, their broad principles are already clear and countries need to begin preparing to implement them.