Consensus achieved on Basel II proposals
11 May 2004
The Basel Committee on Banking Supervision is pleased to announce that it has achieved consensus on the remaining issues regarding the proposals for a new international capital standard. The group of central bankers and banking regulators who make up the Committee met today at the Bank for International Settlements in Basel, Switzerland, and will publish the text of the new framework, widely known as Basel II, at the end of June 2004. This text will serve as the basis for national rule-making and approval processes to continue and for banking organisations to complete their preparations for Basel II's implementation.
The Committee confirmed that the standardised and foundation approaches will be implemented from year-end 2006. The Committee feels that one further year of impact analysis/parallel running will be needed for the most advanced approaches, and these therefore will be implemented at year-end 2007. This will also provide additional time for supervisors and the industry to develop a consistent approach for implementation.
"Basel II introduces a far more comprehensive framework for regulatory capital and risk management than we have ever known," said Jaime Caruana, Chairman of the Basel Committee and Governor of the Bank of Spain. "The Committee owes this accomplishment to the tremendous commitment and technical support that banks, central banks, supervisory authorities and academics from around the world offered us. Now that level of collaboration will be invaluable to the prudent implementation of the new framework."
Basel II represents a major revision of the international standard on bank capital adequacy that was introduced in 1988. It aligns the capital measurement framework with sound contemporary practices in banking, promotes improvements in risk management, and is intended to enhance financial stability.
"The publication of the text opens a new phase for Basel II," agreed Nick Le Pan, Vice Chairman of the Committee and the Canadian Superintendent of Financial Institutions. "Regulators and bankers gain an opportunity to enhance banking supervision and risk management, although non-member countries must proceed at their own pace based on their own priorities."
Technical issues resolved
At its meeting, the Committee reached agreement on the outstanding issues. These included specifying a treatment for revolving retail exposures and resolving the measurements required for "loss-given-default" (LGD) parameters for banks that adopt one of the internal ratings-based (IRB) approaches to credit risk. Members also discussed the calibration of the framework and ways to uphold the Committee's objectives to maintain broadly the aggregate level of the minimum capital requirements, while providing incentives to adopt the more advanced risk-sensitive approaches of the new framework. Appendix I outlines the mechanics of the treatments to which the Committee has agreed.
Principles on cross-border implementation elaborated
Following up on its statements issued in January 2004 related to the application of Basel II across borders, the Committee furthermore elaborated certain principles and issues regarding the need for home and host country supervisors to coordinate and cooperate to reduce burdens on the industry and to employ supervisory resources efficiently and effectively. The Committee has detailed practical implications of these principles in Appendices II and III.
The publication of the framework
The Committee's decisions will be reflected in the text that details the new capital adequacy framework. The text will be released by the end of June on the Committee's home page on the website of the Bank for International Settlements.
Overview of Technical Issues Resolved
Parallel running and capital floors
Parallel running for banks adopting the foundation internal ratings-based (IRB) approach to credit risk will apply for one year during 2006.
Banks moving directly from the existing framework to the advanced approaches to credit and operational risk will have two years of parallel running/impact studies during 2006 and 2007.
The floors on both foundation and advanced approaches in 2008 and 2009 would be 90% and 80%, respectively. Foundation IRB banks will apply a floor of 95% in 2007.
Treatment of revolving retail exposures
At its May meeting, the Basel Committee on Banking Supervision specified its treatment for consumer credit cards and other revolving retail exposures. The mechanics of the treatment, which will be incorporated into the Committee's mid-year 2004 text, are outlined below.
The required capital charges for qualifying revolving retail exposures (QRRE) will be aligned to the results of recent empirical studies. The asset correlation for QRRE will be fixed at 4%, rather than requiring that correlation varies with the probability of default, as specified in the third consultative paper issued in April 2003.
With regard to securitised portfolios of QRRE, the capital framework will reflect more closely the economics of such transactions. Undrawn credit lines related to securitised exposures are allocated between the seller's and investor's interests. The seller's share of undrawn lines related to securitisation exposures will be included in the internal ratings-based (IRB) approach to credit risk, while the investors' share of undrawn lines related to such exposures will be addressed through a revised set of credit conversion factors (CCF) under both the IRB and standardised securitisation treatments of early amortisation provisions. The revised CCFs for non-controlled early amortisation provisions are provided below. Modest changes are similarly being made to the CCFs for controlled early amortisation provisions.
Non-controlled early amortisation features for uncommitted retail exposures
Credit Conversion Factor (CCF)
133.33% of trapping point or more
less than 133.33% to 100% of trapping point
less than 100% to 75% of trapping point
less than 75% to 50% of trapping point
less than 50% of trapping point
Required measures for loss-given-default
Another agreement reached today addresses the potential for loss rates to be higher than average when borrowers default during an economic downturn. This issue was described in the Committee's third consultative paper published in April 2003, which requested banks adopting the AIRB approach to take this potential risk into account in assigning LGDs, particularly for exposures where it would make a material difference. Subsequent discussions with industry participants have indicated both that the importance of this issue varies across exposure types and that individual banks do not have highly-developed approaches to assess this risk.
The Committee believes that its framework should retain the concept of a single assigned LGD that should reflect "economic downturn" conditions where necessary to capture the relevant risk. The Committee considers that one possibility would be for banks' internal LGD processes to focus on assessing an expected LGD, while seeking to develop a broad consensus on how to achieve appropriate "economic downturn" LGDs for the various exposure categories. In this regard, it will be highly desirable to obtain additional industry input and dialogue on the approaches that can be used to ensure appropriate "economic downturn" LGDs are applied where necessary.
The calibration of Basel II
The Committee believes it is important to reiterate its objectives regarding the overall level of minimum capital requirements. These are to broadly maintain the aggregate level of such requirements, while also providing incentives to adopt the more advanced risk-sensitive approaches of the new framework. The Committee has confirmed the need to further review the calibration of the new framework prior to its implementation. Should the information available at the time of such review reveal that the Committee's objectives on overall capital would not be achieved, the Committee is prepared to take actions necessary to address the situation. In particular, and consistent with the principle that such actions should be separated from the design of the framework itself, this would entail the application of a single scaling factor - which could be either greater than or less than one - to the results of the new framework. The current best estimate of the scaling factor using QIS 3 data adjusted for the EL-UL decisions is 1.06. The final determination of any scaling factor will be based on the parallel running results, which will reflect all of the elements of the framework to be implemented.
Enhanced Cross-Border Cooperation
At its May 2004 meeting, the Committee reiterated its view that closer coordination between supervisors is essential to implement the New Accord effectively and efficiently. The Accord Implementation Group continues to discuss the practical implications of the high-level principles for the cross-border implementation of the new framework, which were published in August 2003.
Committee members and other supervisors are relying mainly on case studies based on actual banks' structures to explore ways to enhance communication and cooperation between home and host country supervisors. This effort also supports supervisors' commitment to communicate to internationally active banks the respective roles of home country supervisors, who would lead the coordination effort, and of host country supervisors. The Committee reiterated its commitment to pushing forward the development of these plans between home and major host countries for banking groups with major international operations, focusing on practical home/host cooperation for more advanced approaches.
In the light of its principles on cross-border cooperation, the Committee agreed on the following elaborations regarding coordination and cooperation between home and host supervisors.
- Home and host supervisors should consider practical ways to coordinate requests for information.
The Committee expects that those members needing detailed information about Basel II implementation and roll-out plans from foreign subsidiaries operating in their jurisdictions will ask for the information from the home country supervisors before addressing the bank. This should be interpreted in a practical way. This will not preclude host countries from discussing prudential matters with their banks directly, but will strengthen and rationalise the communication efforts among supervisory authorities. In addition, the Committee believes that home/host coordination of information requests will promote, in general, the ability of all host supervisors to exercise effective host banking supervision over foreign institutions operating in their jurisdictions.
- Supervisors should not duplicate the necessary approval and validation work for Basel II.
The Committee reiterates the principle that, wherever possible, supervisors should avoid performing redundant and uncoordinated approval and validation work relative to Basel II in order to reduce the implementation burden on the banks and to conserve supervisory resources. In this light, the Committee supports the principle that the home jurisdiction should play a leading role in the approval and validation of certain types of advanced techniques. As a practical application of this principle, the Committee expects that the initial validation work for most advanced IRB approaches for larger corporate exposures will be led by the home country with appropriate input from the host country supervisor and material reliance by host countries on the work of the home regulator.
- Practical considerations regarding the recognition of AMA capital across borders.
In response to the technical note issued in January 2004 on Principles for the home-host recognition of AMA operational risk capital, the Basel Committee received informal questions and comments on how supervisors intend to recognise a banking organisation's allocation of operational risk capital calculated under the "advanced measurement approach" (AMA) to activities and businesses that span more than one jurisdiction. A note outlining the Committee's current views on practical considerations relevant to this topic appears as Appendix III. It includes a leading role for home supervisors in coordinating supervisory assessment of AMA models.
The AIG will continue to monitor developments in home/host implementation of the framework and work to enhance cooperation in this regard.
The practical application of home-host principles for AMA operational risk capital
The Committee has received informal comments and questions from various industry participants on its recent publication of a paper on home-host supervisory principles for the advanced measurement approaches (AMA) for operational risk (AMA home-host paper).1 Included in the AMA home-host paper is an outline of what is described as a "hybrid" approach to a group-wide AMA. While the AMA home-host paper was not intended to be a consultative paper, the Committee believes that it would be appropriate to elaborate certain aspects of its views on implementing home-host supervision of operational risk AMAs.
The Committee chose not to define "significance" in determining which internationally active banking subsidiaries2 are ineligible to make use of an approved allocation mechanism. The Committee is aware of industry concerns about the extent to which stand-alone AMAs for subsidiaries could be required. It is not the Committee's intent that a large number of banking subsidiaries within a given banking group should be required to adopt stand-alone AMAs as opposed to using an approved allocation mechanism. The Committee recognises that only a small number of subsidiary banks in such a group may have the practical ability to calculate their own AMA capital requirements for operational risk, and that some supervisors may exercise national discretion in a manner that limits use of the AMA by banking organisations in their jurisdictions. The Committee expects that home and host supervisors will work together in implementing the New Accord to determine which internationally active subsidiaries can reasonably be deemed to be significant.
While supervisory processes for assessing and - where required - approving AMAs will evolve over time, the Committee is mindful that, in developing such processes, supervisors should consider the burden that such processes impose on internationally active banking organisations. As a general rule, where a banking organisation wishes (or is required) to adopt an AMA at both the group-wide and subsidiary levels, the Committee believes that it would be beneficial for the supervisory assessment of the AMA models to be coordinated by the home supervisor. While this is ultimately a matter for discussion among home and host supervisors of a given banking organisation, it would be desirable for the home supervisor to receive a banking organisation's AMA submission and coordinate comments from host supervisors in jurisdictions where the AMA will be applied.3 It is expected that the AMA submission would include, among other things, a description of the group-wide AMA; identification of significant subsidiaries that will use a stand-alone AMA; an explanation of how resources (information, staff, etc.) are shared between the group and subsidiaries that adopt a stand-alone AMA; identification of non-significant subsidiaries that may use an allocation mechanism from the group-wide AMA figure; and a description of the allocation mechanism and rollout plan, as applicable. Host supervisors will still need to be assured, however, that the board and senior management of a subsidiary bank understand the subsidiary's operational risk profile, including how its operational risks are managed, and approve its Pillar 1 methodology for determining its operational risk capital requirements, whether that methodology comprises a stand-alone AMA or an allocation mechanism.
The Committee is aware that questions remain about the application of the partial use provisions of the operational risk rules where a banking group and its internationally active banking subsidiaries are using different approaches (i.e. where a significant internationally active banking subsidiary adopts a simpler approach on a stand-alone basis even though the banking group adopts a group-wide AMA, or vice-versa). Consequently, the Committee hopes to provide greater clarity on the appropriate supervisory treatment of such situations both in this note and through possible changes to the operational risk partial use rules.
While a banking group may choose to adopt a group-wide AMA, significant internationally active banking subsidiaries of such banking groups will not be required under the partial use rules of the New Accord to adopt an AMA on a stand-alone basis. Depending on domestic implementation of the New Accord, a significant internationally active banking subsidiary could choose (or be required by its host supervisor) to adopt a simpler approach on a permanent basis even if its parent adopts a group-wide AMA. In this case, the parent would not be in violation of the operational risk partial use rules provided that, after a reasonable transitional period, the AMA metrics relevant to the subsidiary's operations are reflected in the group-wide AMA.4
Conversely, in some cases a significant internationally active banking subsidiary may choose (or be required by its host supervisor) to adopt a stand-alone AMA. The parent of such a subsidiary would not be in violation of the operational risk partial use rules if it chose to adopt a simpler approach on a group-wide basis, even if it did so permanently.5
The Committee expects that jurisdictions will have some flexibility in applying the partial use provisions of the New Accord. Supervisors should exercise reasoned judgement in assessing the appropriateness of the roll-out of a banking organisation's AMA, especially where partial roll-out is a result of jurisdictions either requiring or prohibiting the use of certain approaches to operational risk and is not a result of a banking organisation seeking favourable capital treatment (i.e. "cherry-picking").
Ability to leverage group resources
The Committee is aware that a number of banking organisations are managed on a business line basis rather than on a legal entity basis for internal economic capital allocation and other purposes. Nevertheless, just as the board and senior management of a subsidiary must satisfy themselves regarding the reasonableness of that legal entity's methodology for determining its operational risk and other capital requirements, banking supervisors have a responsibility for ensuring that specific legal entities in their jurisdictions are adequately capitalised. The Committee acknowledges the inherent friction between a business line approach to managing a global banking operation and the need to satisfy the boards and host supervisors of subsidiaries regarding the effectiveness of risk management practices and adequacy of capital on a legal entity basis. However, the Committee is not convinced that the related challenges are insurmountable or that they are unique to the hybrid approach to a group-wide AMA.
The AMA home-host paper states that subsidiaries implementing a stand-alone AMA will be permitted to leverage the resources of the group in determining their operational risk capital requirements. The Committee anticipates that this leveraging would encompass not only internal data and quantitative methodologies, but also the more qualitative elements of an approved group-wide AMA, such as the manner in which the results of risk and control self-assessments and scenario analyses are incorporated into the subsidiary's stand-alone AMA. At the same time, however, the Committee expects that the board and senior management of those subsidiaries would exercise judgement throughout this process and adjust the group-wide analyses, where appropriate, to address the unique circumstances of the subsidiary relative to the group. A subsidiary's process for leveraging group resources within its stand-alone AMA and, in particular, for adjusting the results of group-wide analyses in its process would have to be transparent to its board and host supervisor.
Some concerns have been expressed that banks managed on a business line basis at the group-wide level will not be able to satisfy the so-called "use test" in the AMA requirements and therefore will be unable to qualify to adopt an AMA. The Committee does not share the view that banks that manage themselves on a business line basis will be unable to satisfy the use test at the level of a significant internationally active subsidiary that implements a stand-alone AMA. Such subsidiaries may make use of group-wide processes and resources - even if these processes and resources function primarily on a business line basis - so long as the board and senior management of the subsidiary have reasonable assurance that the manner in which they are used results in a regulatory capital requirement that is commensurate with that subsidiary's operational risk profile.
The Committee notes that many of the issues discussed here may apply to the internal ratings-based approach for credit risk as well. Consequently, the Committee will continue working to ensure that the New Accord is implemented in a manner that is as reasonable and consistent as possible. A number of exercises are currently underway in the Committee's Accord Implementation Group, including actual case studies, which will help supervisors to identify key implementation issues and concerns. This important work, which has focused primarily on credit risk to date but which will increasingly scope in operational risk as well, will continue throughout the period leading up to implementation of the new capital framework. The Committee is committed to maintaining a dialogue with banking organisations throughout this period in order to identify and address implementation-related concerns.
1 Principles for the home-host recognition of AMA operational risk capital, January 2004 (available on the BIS website at http://www.bis.org/publ/bcbs106.htm).
2 The AMA home-host paper applies specifically to internationally active banking subsidiaries because these subsidiaries will be subject to the scope of application of the New Accord. The stand-alone treatment of non-internationally active subsidiaries is not within the scope of the New Accord and is therefore a matter of domestic supervisory discretion.
3 In accordance with the general home-host principles set forth in the Committee's August 2003 paper on High-level principles for the cross-border implementation of the New Accord, the Committee expects that home and relevant host supervisors will cooperate in both initial validation of an AMA and ongoing monitoring of a banking organisation's operational risk management.
4 A significant internationally active banking subsidiary's exposure to, and management of, operational risk must be explicitly considered in the banking group's overall AMA calculation, even if that subsidiary uses a simpler approach - on a stand-alone basis - for its own regulatory capital purposes. Subject to the approval of the banking group's home supervisor, this requirement may not apply while the banking group is rolling out the AMA across its global operations in accordance with an approved rollout plan.
5 The AMA partial use rules as currently drafted may prevent the parent from including the results of a subsidiary's AMA in the calculation of its global, consolidated capital requirements for operational risk. Changes to the rules are being considered that would permit this to occur in limited circumstances.