Summary of IMF Staff Comments on Proposals of the Basel Committee on Banking Supervision for a New Capital Adequacy Framework

• IMF staff supports the general objectives stated in the Committee's proposals and the three-pillar approach. Top priority is for the Committee to develop capital adequacy standards and a supervisory process that promotes safety and soundness for the banks that make up the core of the international financial system.

• The Committee should strongly encourage all national authorities to apply the new standards uniformly to all banks (including domestic banks) as rapidly as possible, taking into account resource constraints and varying degrees of preparedness. This would further level the playing field, and strengthen domestic banks. The staff supports the setting of capital requirements above the minimum standard when warranted by risk conditions. Review of the Accord also provides an opportunity to define Tier 1 capital more rigorously.

• External ratings for determining risk weights are likely to play a smaller role than initially thought. At the same time, there are both risks (weak supervisory skills in assessing systems' quality; underestimation of risks) and benefits (better asset quality information; emphasis on banks' own responsibility) associated with the internal ratings approach and further research is required to adequately assess them. Continued work on standards for assessing banks' internal ratings and capital adequacy systems is required.

• The Committee should seriously consider a more rigorous treatment of off-balance sheet counterparty risks, in part because the proposed treatment of OTC derivatives does not seem to ensure that bank capital adequately reflects counterparty risk.

• Rapid progress is needed to provide guidance or standards on asset valuation, loan classification and loss provisioning, as well as internationally consistent capital calculation.

• Given the interdependence among the three pillars, the staff sees the second pillar as a critical, yet less developed, aspects of the Committee's proposed three-pillar approach.

• The time frame for implementation of a new Accord by developing and transition countries will need to be carefully considered. Most, if not all, of these countries do not have the prerequisites for effective risk management, supervision, and market discipline.

IMF Staff Comments on Proposals of the Basel Committee on Banking Supervision for a New Capital Adequacy Framework

In June 1999, the Basel Committee on Banking Supervision (the Committee) published a Consultative Paper on a New Capital Adequacy Framework for banks to replace the previous Capital Accord of 1988. The Chairman of the Committee asked the Managing Director for the staff's views, and invited IMF staff to participate in the Capital Subgroup of the Core Principles Liaison Group. The staff has now compiled this note containing its reactions to the proposals. The staff plans to continue to discuss the proposals, which are still work in progress, with representatives of the Committee and its Secretariat over the months to come. Many issues of methodology and implementation still need to be addressed before the proposed three pillar framework can be implemented effectively.

II.  General Principles

1.  The staff fully supports the Committee's stated objectives—to promote safety and soundness of the financial system, to maintain at least the current level of capitalization of the system, and to enhance competitive equality. The Committee's proposal for a more risk-focused approach to calculating capital requirements (first pillar)—in order to more closely align capital adequacy with credit risks—is appropriate, as is the increased role to be given to the supervisory process (second pillar) and market discipline (third pillar).

2.  The Committee should place the highest priority on designing a new capital framework that is appropriate for the global and internationally active banks that make up the core of the international financial system. At the same time, while the Committee's primary focus should be developing standards for global and internationally active banks, the new Accord should establish a new minimum standard to which all banks will be held, regardless of their country of origin or the extent of their activities. The Committee and national authorities, therefore, need to develop ways in which the new framework can be applied to banks of varying levels of complexity and sophistication, particularly those in emerging market, developing and transition countries. The staff recommends that the Committee neither dilute nor simplify capital adequacy standards with the sole purpose of broadening the applicability of standards to include less sophisticated (or less well-managed) financial institutions: doing so could undermine the effectiveness of market discipline across international markets. If high standards are to be maintained while broadening the application of the new standards, flexibility is best vested in the supervisory process which in our view should enforce these standards even more vigorously for less sophisticated and less well managed banks.

3.  Fund staff endorses the application of capital requirements on a consolidated basis, and the inclusion of holding companies that are parents of banking groups. The staff also supports the requirement that each bank in a group be adequately capitalized individually, and that subconsolidation takes place for each of the banks in the group, thus "compartmentalizing" the risks as well as the capital buffer.

4.  Further efforts are required to align capital standards for the three major financial subsectors: banking, investment business, and insurance. Considerable additional work is necessary to develop methods for the supervision of banks that are part of a diverse group of corporations. Mechanisms need to be developed for coordinated action by financial sector supervisors in different countries to address problems involving an international group that is active in the three major financial subsectors, and which may have other financial interests as well.

III.  First Pillar: Minimum Capital Requirements

5.  The staff supports the objective of calibrating risk weightings more accurately to reflect the relative credit quality of particular asset classes and counterparties, and also underlines the need for rapid progress in the development of standards on asset valuation and loan loss provisioning. Capital calculations, and hence capital adequacy ratios are misleading if provisioning is not adequate. Recent experience in crisis countries demonstrates that there was considerable understatement of provisions, and accordingly, stated capital adequacy levels were highly exaggerated. When setting standards, it should be stressed that primary responsibility for proper provisioning lies with the bank and its external auditors.

6.  The practice in some countries of using much lower capital adequacy requirements for domestic banks than for internationally active banks should be changed. Domestic banks often operate in environments that are far more risky than those of internationally active banks.

7.  Also, the capital calculation method underlying the capital adequacy ratio should be reviewed, in order to develop a more strict definition of Tier I capital, and to eliminate, for instance, the inclusion of non-cumulative preferential shares, and other country practices that undermine the quality of bank capital, such as counting tax benefits against future losses as capital.

A.  Use of External Ratings

8.  Although the staff shares the Committee's view that external ratings will lead to a more accurate measure of credit quality than the present system, we also share some of the concerns of other commentators about the proposal to rely on commercial rating agencies, particularly for sovereign risk. It will be difficult to design eligibility criteria for external credit assessment institutions. Issues arise with regard to the oversight of rating agencies, quality and comparability of ratings, and possible conflicts of interest.

9.  The potentially procyclical effects of external ratings (on which research is being undertaken) could be reinforced if risk weights are linked to assessments of compliance with international standards such as the SDDS and the Basel Core Principles. These standards have not been designed to provide a quantitative pass or fail benchmark. Furthermore, to use the assessments as a basis for capital adequacy purposes would create incentives for compliance above reform. Staff believes that compliance assessments can in due course provide useful input as a factor in setting capital requirements. However, the question how these different sets of standards are incorporated into a capital adequacy framework raises many complex issues which will need further consideration.

10.  While external ratings will, no doubt, be appropriate in many cases, the staff sees benefits in using banks' internal ratings as a basis for capital adequacy charges, especially in cases where supervisors have the necessary skills and resources to properly assess the quality of the internal rating systems. It now seems likely that most banks active in cross-border lending will use this approach rather than rely on external ratings. The proposals provide an incentive to banks to develop their internal ratings.

11.  The use of internal systems would place more responsibility on banks. It would establish a closer connection between a bank's internal management information systems and its credit allocation and review process, thus helping to mitigate the perception that regulation is something to be circumvented. Internal ratings thus can help tailor capital requirements more closely to the actual risks faced by banks, which would then be less likely to structure their assets around the risk weighting system. As the proposals recognize, internal ratings potentially benefit most from complete information on the performance of credits and the background of the borrowers, and other sources of information.

12.  Nevertheless, there are several potential weaknesses associated with internal ratings systems. Recent research indicates that banks' internal ratings can be even more procyclical than external ratings. ¹ More evidence, therefore, seems necessary on the accuracy of internal rating systems in assessing risks. Furthermore, unless the assessments of banks' internal rating systems are highly standardized, capital requirements and the resulting capital adequacy ratios of individual banks in various jurisdictions may become less comparable. It is critical that progress be made rapidly in developing guidelines for supervisory assessment of internal systems. Especially in emerging market, developing and transition countries, the use of untested internal systems could entail risks.

13.  Uniform standards to help assess the quality of banks' internal systems are essential. Supervisors will need guidance in order to maintain equality and consistency of treatment and to be able to assess with sufficient confidence banks' risk management capabilities so as to avoid the danger of banks deliberately underestimating risk. Allowing banks to use their own systems without appropriate safeguards can create incentives for underestimating risk. The staff supports the Committee's efforts to develop standards for internal rating systems, as well as a supporting methodology for mapping internal ratings into standard risk weights. In this context, the staff welcomes the paper "Range of Practice in Banks' Internal Ratings Systems" as a first step in a more comprehensive study. The staff also supports the Committee's work on the development of credit risk models as a longer run component in assessing capital adequacy.

14.  Greater reliance on internal systems will demand higher skills from the banking supervisory authorities, given that they will be required to assess the adequacy of a bank's internal credit rating systems, instead of simply verifying regulatory compliance. Improving supervisory skills is a highly desirable objective, which Fund staff actively supports in its technical assistance, design of adjustment programs, and surveillance work. However, in many countries such improvement is bound to take time and will be resource intensive.

15.  Supervision of internal credit ratings would be consistent with the trend in other areas of banking supervision, which increasingly focus on supervisory assessment of the banks' own risk management systems and internal controls, rather than requiring compliance with quantitative standards, however useful these will remain as basic indicators.

16.  The Committee's proposals go some way toward more effectively recognizing the risks in off-balance-sheet activities, and we welcome the elimination of the cap on OTC credit risk weightings. However, the present system of capital charges for off-balance-sheet credit risks, and in particular for OTC derivative transactions, would largely remain unchanged. The increasing sophistication of banks in arbitraging capital requirements and the dynamic nature of contingent OTC derivative exposures is likely to widen existing gaps in the measurement of banks' overall credit exposures, and consequently in setting appropriate capital levels. Capital requirements need to recognize the linkages between credit and market risks more accurately, particularly, with regard to interest rate risks in emerging markets. The Committee should give serious consideration to ways in which capital charges could more closely reflect the significant changes (positive and negative) that occur in a bank's current and potential future credit exposures when market prices change. In that context, banks' internal credit risk systems could be required to quantify off-balance-sheet credit exposures (both current and potential) as a basis for appropriate capital charges—subject to verification through effective supervision (the second pillar).

17.  In general, the use of risk mitigation techniques should be encouraged, and the staff welcomes the recent issuance by the Committee of the paper "Industry Views on Credit Risk Mitigation." However, experience has shown that in many countries traditional techniques do not always function properly as a result of thin markets for collateral, deficient valuation methods, and difficulty in foreclosing on collateral and enforcing netting agreements, especially across borders. More sophisticated techniques should be permitted as a means of reducing the net exposure that would be subject to capital charges only when the supervisory authorities have the expertise to understand and assess the adequacy of these techniques and their residual risks.

18.  Narrowing the definition of "short-term claims," as proposed, from one year to six months is an unsatisfactory compromise solution to an admittedly difficult problem. On the one hand, increasing the risk weight on claims with maturities between 6 months and 12 months is appropriate due to the higher risks to individual banks inherent in such exposures. On the other hand, the proposed change would favor even shorter-term bank lending, which tends to be highly volatile and difficult to track, and could thus contribute to potential systemic instability. Moreover, the existing one year definition was intended to capture the bulk of money market placements and thus appears to have a more logical justification. According to some commentators, risk weighting 364 days revolving credit facilities would have serious consequences for corporate financing and could reduce liquidity in commercial paper markets. Nevertheless, the staff believes that such exposures represent a real risk and should be subject to some form of capital adequacy requirement. As a general issue, staff believes that claims on banks should be weighted on the basis of their own condition.

19.  There are difficulties in quantifying operational, legal, and reputational risks, and in establishing capital charges for such risks. These risks traditionally have been partially covered by unallocated capital and reserves, and can also be partially offset by specific provisions--for example, for legal risks flowing from disputed claims or obligations. This point notwithstanding, the staff supports a more precise system where banks are at least required to identify more clearly the possible types of general business risk and set aside capital for those risks. This system could, for instance, be used as a basis for higher general loan-loss provision requirements for banks with a relatively high risk profile. By contrast, banks that have exceptionally strong risk management systems could be rewarded with lower general provision requirements.

20.  The staff supports the overriding role of the supervisory review process (second pillar) in assessing banks' internal systems for risk management and asset valuation, and ensuring the appropriate capital requirement for each bank. The staff sees the second pillar as one of the most important, yet underdeveloped, parts of the Committee's proposals.

21.  The staff agrees that individual capital adequacy requirements for banks with a relatively high risk profile and/or operating in a high risk environment should be higher. The supervisory process (second pillar) needs to be strong enough to ensure that a bank's capital position not only meets the minimum levels implied by the capital standards, but is consistent with a bank's overall risk profile and strategy, and its operating environment. Supervisors should encourage banks to develop their own internal capital assessment processes that are conceptually sound and robust. Early supervisory intervention should be used if capital is deemed not to provide sufficient buffer. At the same time, supervisors should have the ability to require banks to hold capital in excess of the minimum depending on the quality of bank management, its track record in managing risks, and the overall economic environment.

22.  The Committee's proposals suggest that supervisors should take account of a bank's relative importance in the system and its potential to trigger systemic instability. The staff agrees that banks with systemic importance should be subject to more intense supervision, but doubts whether a higher minimum capital charge is the best way to achieve this objective, especially as such banks are also more likely to use adequate internal controls. The supervisory review process is more likely to be the most effective way of dealing with this kind of differentiation.

23.  In developing and transition countries, there is a particularly pressing need for enhancing the supervisory review process. In crisis countries, flawed capital calculations were compounded by weak supervisory practices and public sector willingness to tolerate or support undercapitalized banks. In such an environment, credit problems, and thus an under-estimation of the true capital position, can remain hidden for some time. This strongly indicates that sound supervisory practices are at least as important as modern quantitative methods in assessing bank soundness. For this reason, the Fund's, and the World Bank's, technical assistance efforts have been focussed to this end.

24.  However, in many countries the skills to perform these assessments need to be developed further. The authorities will also need to be given regulatory powers to evaluate banks' internal risk management systems, asset valuation procedures, and to introduce clear sanctions against banks with weak systems. Countries need to be aware that broader qualitative standards for supervisory practices have undergone significant development since 1988, and that such standards, whether or not they will be incorporated into new capital proposals, should be adopted.

25.  The staff strongly supports a greater role for market discipline in supplementing and supporting supervision. Regulators increasingly are obliging banks to disclose more information about their financial condition and risk profile, and are relying on market forces. More comprehensive disclosure on a consolidated basis could lead to a more differentiated view of the risks taken by each bank and banking group. Encouraged by market participants, banks would have, therefore, a greater incentive to hold more capital and to adopt more robust risk management methodologies.

26.  Disclosure practices differ across countries, however. In some countries, even prudential reports provided for supervisors are publicly available, while in others, banks are obliged to publish only their summary financial statements or perhaps their capital adequacy ratios and data on nonperforming loans and provisions. In some markets, disclosure and transparency are insufficiently reliable, due to deficient data quality that results from flawed valuation and accounting practices, insufficient legal mandates for financial disclosure, and unduly restrictive secrecy rules. Improvements in disclosure requirements would, therefore, require improved rules for accounting and auditing, asset classification and loss provisioning, as well as other legal changes. The staff proposes that more stringent disclosure requirements be placed particularly on the systemically more important banks.