Federal Deposit
Insurance Corporation

By E-mail and Overnight Courier

Office of the Comptroller of the Currency
250 E Street, SW
Public Information Room, Mailstop 1-5
Washington DC 20219
Attention: Docket No. 03-14

Ms. Jennifer J. Johnson Secretary
Board of Governors of the Federal Reserve System
20th Street & Constitution Ave., NW
Washington, DC 20551
Attention: Docket No. R-1154

Robert E. Feldman
Executive Secretary
Attention Comments
Federal Deposit Insurance Corporation
550 17th Street, NW
Washington, DC 20429

Regulation Comments
Chief Counsel's Office
Office of Thrift Supervision
1700 G Street, NW
Washington, DC 20552
Attention: No. 2003-27

Re: Response to the Agencies' Invitation to Comment on the Advanced Notice of Proposed Rulemaking for a Proposed Framework Implementing the New Basel Capital Accord in the United States

Ladies and Gentlemen:

This letter and its enclosures are MBNA America Bank, N.A.'s response to the Office of the Comptroller of the Currency, the Board of Governors of the Federal Reserve System, the Federal Deposit Insurance Corporation, and the Office of Thrift Supervision's (together the "Agencies") invitation to comment on the advanced notice of proposed rulemaking ("ANPR") for a proposed framework implementing the New Basel Capital Accord (`Basel II" or the "New Accord") in the United States, issued August 4, 2003.¹

We appreciate the opportunity to provide comment on the ANPR and on the New Accord in general.

MBNA America Bank, N.A. is the principal subsidiary of MBNA Corporation and has two additional banking subsidiaries, MBNA Europe Bank Limited and MBNA Canada Bank (collectively herein referred to as "MBNA"). MBNA's primary business is retail lending, providing credit cards and other retail lending products to individuals. At September 30, 2003, MBNA Corporation reported assets net of securitizations totaling $58.7 billion. MBNA Corporation's managed assets, including securitized loans were approximately $141.1 billion as of September 30, 2003.

MBNA has been an active participant throughout the development process of the New Accord. We have participated in Quantitative Impact Study 3 ("QIS 3") and the operational risk loss data collection exercise in order to help the Committee measure the regulatory capital impact of Basel II. Throughout this process, we have consistently expressed serious reservations with many aspects of the New Accord, including its overall complexity, capital distortions created by the internal ratings-based ("IRB") approach for unsecured retail credit exposures, creation of a capital charge for operational risk, securitization treatment, and disclosure requirements. Other than the creation of the qualifying revolving retail exposure ("QRE") formula, which recognizes the importance of future margin income, very little has changed in areas important to active credit card issuers and even the QRE formula does not achieve an appropriate capital/risk balance. We hope that through this process, our concerns can be considered fully and that an approach can develop that addresses cost, complexity, regulatory burden, and competitive impact.

We note that since the release of the ANPR, the Basel Committee on Banking Supervision (the "Committee") announced four principal areas where significant changes to the Basel II framework are expected.² In its press release and the accompanying attachment, the Committee provided only a general description of how it now intends to have the New Accord treat expected and unexpected losses. It also invited interested parties to provide comment on these changes by December 31, 2003. Other than a general statement, no other information was provided. On October 30, the Agencies released a statement regarding the Basel Committees' October 11 request for comment. In their statement, the Agencies invited commenters to consider the Committee's changes when submitting responses to the ANPR. The Agencies also agreed to consider additional comments on the proposed treatment of UL/EL through December 31. In their statement, the Agencies declined to provide additional information regarding proposed changes. We believe that it would be helpful for the overall development effort of the New Accord for the Committee and Agencies to provide additional information that more fully specifies this change and its proposed application. Without that it will be difficult to both collect meaningful commentary on the changes and ensure that no institution or business line is unreasonably impacted .³ Although we support in general the changes announced by the Committee, without additional information as to how these changes will be applied and calibrated, we are limited in our ability to evaluate fully the new proposals and provide the kind of meaningful commentary we believe these changes deserve. ⁴ Without knowing more, we believe that the scope of the proposed changes also suggests the need for an additional quantitative impact study ("QIS") prior to adoption of the final rules.

We support the primary goal of increasing risk sensitivity and of creating a process for better differentiating risk and assigning appropriate capital to those exposures. We remain concerned, however, that the approach endorsed by the Agencies will result in a highly prescriptive set of rules which will be costly to implement and comply with and may not achieve the desired results of a risk-sensitive framework with appropriate capital requirements across product types. As a consequence we would suggest the Agencies consider an approach that more closely follows the framework of the standardized approaches of Basel II.

Our concerns about the New Accord are centered in four general areas: (1) application of the New Accord to U.S. Banks, (2) the treatment of unsecured retail credit, (3) the conservative assumptions and treatment of uncommitted credit lines affecting originators in asset securitization, and (4) inclusion of a specific capital charge for operational risk.

Application of Basel II U.S. Banks -

• Bifurcation - We are very concerned with the Agencies decision to create a bifurcated approach in assessing regulatory capital. By requiring "core banks" to adopt the most complex IRB approaches and leaving all other banks under the 1988 Capital Accord (the "Current Accord"), the Agencies are creating a framework where U.S. regulated institutions will operate in two separate spheres. As a matter of public policy we question whether there should be two entirely different capital frameworks and standards when determining capital adequacy for U.S. institutions. Moreover, from an international comity perspective, we believe that it is shortsighted to conclude that most U.S. banks should remain under the Current Accord. We believe to effect a competition-neutral result, the standardized approaches (with appropriate validation of calibration) must be applicable to all banks.

• Mandatory A-IRB & AMA Application for Core Banks - If the Agencies ignore our recommendation for a single regulatory framework, the Agencies' decision to require the application of the advanced approaches for certain large banks creates additional challenges. We believe that this approach stands on its head a basic premise of the New Accord - that banks must be permitted to choose the appropriate methodology for calculating regulatory capital. Mandatory application of the advanced approaches puts "core banks" at a disadvantage when compared to their competitors in foreign markets. These competitors (unlike their U.S. counterparts) will have the ability to choose the Basel II approach that makes the most economic sense. The Agencies should work closely to ensure that U.S. banks are not competitively disadvantaged against foreign banks and U.S. non-banks. If the costs of the New Accord (from either a compliance or capital perspective) are significant, there may be incentives to either abandon certain businesses or "de-bank" altogether. We believe that all banks should be permitted to select the framework that is the most appropriate for their business and not be governed by certain arbitrary size thresholds.

Treatment of Retail Credit -
• The A-IRB approaches will significantly impact institutions with material unsecured retail exposures. The conservative capital treatment for unsecured retail exposures should not be used by the Committee to offset lower regulatory capital requirements for other asset types without understanding their relevant risks and business models. The seemingly arbitrary approach to unsecured retail lending may cause significant competitive harm. Before the New Accord is finalized, it is critical to undertake an additional QIS to ensure that the risks for unsecured retail lending are captured accurately and an appropriate capital treatment is applied that correctly measures the underlying risks of unsecured retail lending.

• The Agencies have evidently ignored the substantial differences between revolving retail credit portfolios and corporate credit portfolios. Applying a corporate credit model (which is based on single credit exposures) to retail credit portfolios (which are managed on pools of individual exposures) has not been sufficiently tested or validated. Any credit model that is ultimately adopted for retail lending must be sound and more than simply a modified version of the corporate credit model. The unique attributes of the retail framework (definition of default, portfolio segmentation, predictable expected losses, loans priced to cover expected losses, uncommitted/undrawn lines, asset value correlation, etc.) carry a level of complexity that merits further review and study.

• Under the IRB approach, capital requirements for credit card loans are higher than both the Current Accord and the standardized approach of Basel II. We believe that this result contradicts the New Accord's stated objective that the IRB approaches would result in more effective risk measurement and, therefore, lower capital requirements than the standardized approach. Our internal analyses has determined that, from a portfolio point of view, the economic risk of the A-IRB approach should be less than the CP 3 standardized approach for unsecured retail lending. As such, substantial recalibration of the A-IRB will be necessary to correct these major differences.

• Banks should hold capital for unexpected losses only. Although the Committee has announced its intention to separate the treatment of unexpected losses and expected losses, how this change will be applied requires additional clarification by the Committee and the Agencies. We are particularly concerned with the Committee's conclusion that expected one-year losses must be measured against the loan loss reserve and that any shortfall would be taken as a deduction of 50% from Tier 1 capital and 50% from Tier 2 capital. This approach is counter to both U.S. GAAP and current U.S. regulatory policy and ignores the impact of future margin income for unsecured retail lending.

• The potential risk of additional draws from uncommitted retail credit lines that can be terminated at will by a lender does not warrant a charge for additional capital. The risk associated with undrawn, uncommitted lines for unsecured retail loans is very low, particularly when they are closely monitored and readily cancelable by the lender. In MBNA's case, for example, over 90% of available U.S. credit card lines are in accounts with expected PDs less than 2%.

• The asset value correlation ("AVC") factors are not consistent with our own experience. We suggest that each institution should be permitted to establish its own AVC factors. At the very least, the Committee should lower the range of AVC factors to 2% - 5% for QREs, with a corresponding reduction for other retail exposures.

Asset Securitization -
• The requirement that originators hold more capital than investors for similar risk exposures is overly conservative and unnecessary. We believe that originators should not be burdened with higher capital requirements compared to investors in equivalent risk positions.

• Undrawn, uncommitted credit lines related to revolving accounts included in securitization transactions should not require capital. In typical revolving securitization structures, both current drawn balances and future Customer draws, are securitized. During the revolving period, investors do not have the ability to choose whether or not to purchase newly originated loans, nor do they have the ability to purchase only low-risk receivables. Rather, investors are required to purchase receivables, on a pro-rata basis, from all accounts in the securitization vehicle. If the Agencies are trying to allocate capital for the risk of amortization, that risk is already captured through the proposed new early amortization capital requirement.

Operational Risk -
• Operational risk management is an emerging discipline; the current state-ofthe-art practices for operational risk measurement are still in their very early stages. As such, we question the wisdom of a specific capital charge for operational risk at this time. We see little harm in waiting to apply any change as an interim step since most larger banks are "well capitalized" and have an adequate cushion in place to cover operational risks. It is imperative that banks be given time to evolve their operational risk measurement practices before any capital charge for operational risk goes into effect.

• Consistent with our recommendation for credit risk and with the Committee's decision to rely solely on unexpected losses for the measurement of riskweighted assets, any application of operational risk capital charge must be limited to unexpected losses, and not include expected losses.

• Direct calculation of specific risk results to a 99.9% confidence level, with a verifiable degree of accuracy, will not be possible for most business lines given the lack of available data or will result in an extremely conservative capital charge, which would not make economic sense for the institution.

As noted in the enclosed, we continue to have major concerns about the remarkable complexity of the New Accord. Our concerns regarding regulatory burden and whether elements of the New Accord can in fact be applied beyond the "laboratory" and at the operational level relate to virtually every facet of the proposal. Moreover, our concerns about competitive harm between large and small banks, monoline and full service banks, and regulated and non-regulated financial institutions are central to our overall reservations about the draft. We urge that the Agencies study the impact that these sweeping changes will have on their regulated institutions before embarking upon this new direction and that they consider fully alternative approaches that address these major concerns.

Finally, the established timeframes are much too aggressive for development and final approval of the New Accord. Currently, the Committee expects that the final version of the New Accord will be completed by the summer of 2004. Meanwhile, the Committee and the Agencies are directing institutions that will operate under the advanced approaches to begin making the necessary investments to be ready for implementation by year end 2006. As part of that process, the Agencies expect core banks to begin collecting data and making other operational changes before a final rule is adopted. To do this, each affected institution must begin now to make major investments in systems and personnel - even before a final rule is issued. We are concerned that the Agencies may have predetermined the result of this rulemaking, calling into question the soundness of the entire process. We suggest that the more prudent approach would be for the Agencies (1) to proceed with another QIS, (2) to prepare a second ANPR (collecting commentary on the results of the QIS and any additional changes the Committee proposes), (3) upon consideration of responses to the second ANPR, to complete the formal rulemaking process with development and adoption of final regulations, and (4) to approve a final version of Basel II only after items 1- 3 have been completed. Any other approach may be problematic and open to challenge.

We appreciate the opportunity to provide these comments to the Agencies. If you have any questions regarding this submission or if we can provide further information, please contact Vernon Wright directly by telephone at 302-453-2074 or by e-mail at vernon.wright@mbna.com.

Yours truly,

¹ Enclosed with this letter at Appendix 1 is our response to each of the ANPR questions we found applicable to MBNA. For clarity, we also included both the original captioned headings in the ANPR and every specific question for which public comment was sought (appending Federal Register page numbers) regardless of whether MBNA submitted a response. We also numbered each question presented in the ANPR, to assist in cross-referencing our other responses to the Agencies' questions. Included with our responses to the specific questions are also general comments adjacent to the captioned headings that address additional issues, not raised as matters requesting comment by the Agencies. Also enclosed at Appendix 2, is MBNA's prior submission to the Basel Committee on Banking Supervision providing comments to Consultative Paper 3 (July 31, 2003). We include at Appendix 3, MBNA's methodology in producing the appropriate estimated asset value correlation and at Appendix 4, a graph showing the appropriate qualifying revolving exposures risk weights by probability of default.

² The four areas are: "[1] changing the overall treatment of expected versus unexpected credit losses; [2] simplifying the treatment of asset securitisation, including eliminating the "Supervisory Formula" and replacing it by a less complex approach; [3] revisiting the treatment of credit card commitments and related issues; and [4] revisiting the treatment of certain credit risk mitigation techniques. The Committee did not offer information concerning items 2-4, where additional changes are expected. We anticipate that further guidance will be provided for these three areas.

³ We agree with the concerns expressed by Daniel Bouton, chairman of the Institute of International Finance's regulatory capital steering committee and chief executive of French banking group Societe Generale who said that in light of the Committee's October 11 announcement it is important that the modifications ensure that recalibrations of capital requirements resulting from revision of the unexpected/expected loss framework do not cause significant disadvantage to any constituency of banks or to any business line that is an important source of financing to the economy. See IIF Says More Work Needed on Basel II, Global Risk Regulator, Breaking News Service (Oct 14, 2003).

⁴ We would welcome the opportunity to provide additional comment to the Agencies once they have bad the opportunity to consider the proposed changes and provide appropriate regulatory guidance on how they propose to apply these changes. We believe that this would be most appropriately accomplished through, an additional ANPR process.
 

Enclosures:*

Appendix 1- MBNA's Comments to Advanced Notice of Proposed Rulemaking on the Implementation of the New Basel Capital Accord

Appendix 2 - MBNA's Letter to the Basel Committee on Banking Supervision, dated July 31, 2003

Appendix 3 - MBNA America's Methodology in Producing the Appropriate Estimated AVC

Appendix 4 - Qualifying Revolving Exposures Risk Weights by Probability of Default

*Enclosures may be viewed at the FDIC Public Information Center, Room 100, 801 17th St., NW, Washington, DC, between 9 a.m. and 4:30 p.m. on business days.

C:
Financial Services Authority (United Kingdom)
Office of the Superintendent of Financial Institutions (Canada)
Irish Financial Services Regulatory Authority
Banco de Espana
European Commission
The Basel Committee on Banking Supervision, The Bank for International Settlements