Statement No. 238
For Information Contact:
Richard J. Herring
Edward J. Kane
For the last eight years the Basel Committee on Banking Supervision (Basel Committee) has struggled to replace the original Accord on Capital Adequacy (Basel I) with a new Accord (Basel II). At the outset of the process the Basel Committee stated that the new framework was intended “to provide approaches which are both more comprehensive and more sensitive to risks than the 1988 Accord, while maintaining the overall level of regulatory capital.
The Shadow Financial Regulatory Committee has commented regularly on the evolving Basel II proposal. In addition, the Shadow Committee has developed an alternative proposal2 to harness market discipline by using subordinated debt to make capital requirements more risk sensitive, but without the extensive compliance costs and negotiating efforts associated with implementing Basel II. The Shadow Committee continues to believe that including an appropriate subordinated debt requirement in Tier 1 capital is a superior approach.
Capital adequacy has to be assessed in a framework that operationally defines capital, risk and a specific ratio between the two that ensures an appropriate, minimum degree of safety. Basel I attempted to address each of these three elements. It defined two kinds of regulatory capital: Tier 1 and Tier 2; set out procedures for risk-weighting on- and off-balance-sheet positions to reflect bank exposures to credit risk and market risk; and established minimum ratios for Tier 1 capital and the combination of Tier 1 and Tier 2 capital relative to this measure of risk-weighted assets. Despite efforts to increase market discipline and to enhance supervisory enforcement across countries, Basel II does little to align regulatory capital requirements with the need for economic capital. The definition of regulatory capital remains the same, and does not correspond to the concept of economic capital. And, as in Basel I, it fails to develop any economic rationale for the required minimum ratios. It sought only to develop more risk-sensitive measures of exposures to credit risk and to impose capital charges on operational risks. The Shadow Committee addressed the shortcomings in the proposed treatment of operational risk in Statement No. 179 (May 6, 2002). The current statement focuses only on the treatment of credit risk.
Basel II set out three ways to risk-weight exposures to credit risk. Two of these methods make use of banks’ own internal credit risk ratings to generate risk weights: the Foundation Internal Ratings Based Approach (FIRB) and the Advanced Internal Ratings Based Approach (AIRB). The third and simplest alternative extends the framework of Basel I and is called the Standardized Approach. It requires banks to allocate their exposures to risk buckets using regulator-imposed risk weights that to some extent reflect external credit ratings.
In 2003 the U.S. federal regulatory agencies announced their intent to limit the mandatory implementation of Basel II to the largest, most internationally active banks. They further proposed that these banks be required to adopt the AIRB approach. Because the calibration of the regulatory model was designed to provide incentives to encourage adoption of the AIRB, it will usually result in lower capital charges for qualified banks. This caused other banks to fear that they would suffer a competitive disadvantage. A series of Fed-sponsored white papers analyzing the potential competitive impacts failed to allay these fears.
Moreover the results of the Fourth Quantitative Impact Survey (QIS4), conducted during the fall and winter of 2004-2005, exacerbated these concerns. This survey of 26 of the largest banks in the U.S. (including the banks that would be required to adopt the AIRB) showed substantial reductions in required capital on average. Indeed, if these banks had chosen to reduce their capital to these AIRB minimums, almost all of them would have had leverage ratios that would be categorized as undercapitalized and triggering prompt corrective action sanctions. Such a reduction in the risk-based minimum capital requirement would be inconsistent with the Basel Committee’s stated intention of “maintaining the overall level of regulatory capital.”
In response to these results, the federal regulatory authorities affirmed their support for maintaining a regulatory limit on the leverage ratio, and set limits – transition floors – on the extent to which required, risk-based capital could be reduced during the implementation of Basel II. To alleviate growing concerns about competitive inequities for smaller institutions, they also produced an Advanced Notice of Proposed Rule Making for Basel IA that would reduce capital charges for banks not subject to Basel II. More fundamentally, QIS4 showed worrisome disparities in capital charges for banks with apparently similar positions. These disparities raise doubts about the reliability of the statistical models and the data used to calibrate them even at the country’s most sophisticated banks. Some banks lack obligor-specific ratings and few, if any, have compiled over complete business cycles the kind of data needed to estimate such models with precision.
Based on these results and the inherent difficulties of estimating the probability of low-frequency outcomes, the Committee believes that it is premature at best to allow minimum risk-based capital requirements to be determined by these methods. This leads the Committee to conclude, with great reluctance, that the Standardized Approach is the best of an unappealing set of options. The Shadow Committee Statement No. 169 (February 26, 2001) underscored numerous shortcomings in the Standardized Approach. Nonetheless, relative to either internal-ratings-based approach, the Standardized Approach has three virtues: (1) superior transparency; (2) lower implementation costs; and (3) less susceptibility to manipulation by distressed banks. Furthermore, when combined with the leverage-ratio triggers for prompt corrective action, it reduces opportunities for supervisory authorities to claim justification for granting capital forbearance to troubled institutions. In any event, we remain confident that our subordinated debt proposal, described in Statement No. 160, would provide a more risk-sensitive measure of capital adequacy with much lower compliance costs than any version of Basel II.