EBA欧洲银行-ReportonGL10PeerReview_16页_157kb

Summary of CEBS Peer Review on Guidelines for AMA and IRB Approaches

Introduction

The CEBS peer review exercise aimed to promote supervisory convergence within the European Economic Area by assessing the implementation of the Guidelines on the implementation, validation and assessment of Advanced Measurement (AMA) and Internal Ratings Based (IRB) approaches. The review focused on the practical application of these provisions, emphasizing cooperation between home and host supervisors. It was conducted as a pilot study, with the goal of testing the peer review mechanism and learning from the process, rather than identifying non-compliance for punitive purposes.

The review was based on 38 validation cases recorded by the end of March 2008. These were the first such cases, and thus the findings may not be representative of future applications. The review concentrated on the main steps of the validation process outlined in Annex 1 of the Guidelines: Model assessment (Step 3), Decision (Step 4), and Implementation (Step 5). The initial steps (Pre-application and Formal application) and the ongoing review (Step 6) were excluded from the scope.

Level of Compliance

• The overall level of compliance with the Guidelines was very high, with only a few minor exceptions.
• Most supervisory authorities reached joint decisions within the six-month period, although some, such as Belgium, France, and the United Kingdom, experienced delays.
• The delay was often due to the need to allow host supervisors more time to comment or for institutions to provide additional information.
• Germany, for instance, retroactively sought host supervisors' agreement to formally meet the six-month target.
• Some countries, like Germany, reported a few non-joint decisions, and the Netherlands noted non-compliance with the principle of regular information exchange.

I. Main Issues

1. Six Months Period

• The CRD allows six months for the competent authorities to reach a joint decision, which is legally binding.
• The clock starts upon receipt of the complete application by the consolidating supervisor.
• However, some countries used alternative triggers, such as the dispatch of documents to host authorities or the date of national law transposition.
• There is a need for clearer guidance on the definition of "complete application" and the maximum duration for checking its completeness.
• The preconditions for stopping the clock are not sufficiently detailed in the Guidelines, leading to inconsistent practices.
• The impact of national transposition timelines on the six-month requirement was identified as a critical area needing clarification.

2. Consistency of Implementation of the Joint Decision

• In jurisdictions where the joint decision is not self-executing, there was concern about potential deviations during local transposition.
• Language barriers and differences in national administrative systems contributed to this risk.
• However, no evidence of material inconsistency was found during the review.
• It was noted that national provisions typically require the joint decision to be reflected in local transposition, ensuring consistency.

3. Involvement of Host Supervisors in Later Phases of Assessment

• Host supervisors expressed concern that they were not involved in later phases of model assessment, especially when centralized models were rolled out.
• The Guidelines suggest that host supervisors focus on local implementation, not re-open technical discussions.
• Some countries, like Austria, France, and Italy, used supplementary joint decisions for roll-out plans to ensure consistency.
• There is a need to involve host supervisors in the first assessment phase to avoid technical issues in local implementation.

4. Degree of Reliance on Pre-Validation Self-Assessment

• Supervisory authorities generally reviewed the self-assessments of institutions.
• The United Kingdom relied heavily on self-assessments in its risk-based approach.
• The degree of reliance was based on factors such as systemic risk, application type, complexity, and capital charge implications.
• There is a need for the Guidelines to specify the appropriate level of reliance on self-assessments to ensure consistency.

5. Communication Issues

• Effective communication between consolidating and host supervisors is essential for smooth cross-border validation.
• Communication issues were identified in several cases, including between Germany and France, and between the German consolidating supervisor and the Bulgarian host supervisor.
• Some supervisors did not maintain a regular exchange of information during the implementation phase, particularly regarding the basis for the joint decision.
• The Dutch Authority reported no such process in place, which may be due to the short time since approvals were granted.

II. Lessons Learnt and Way Forward

1. Peer Review Methodology

• The peer review methodology includes self-assessment and peer review based on common criteria.
• The process was designed to encourage supervisory convergence and to enhance the practical application of the Guidelines.
• The pilot nature of the exercise highlighted the need for refinement of the methodology, particularly in terms of the questionnaire design to capture more detailed information on processes and practices.

2. Adequacy of the Guidelines

• The Guidelines were found to be useful but not comprehensive.
• They lacked detailed guidance on several key areas, including:
  • Criteria for the assessment of the completeness of an application.
  • Maximum duration for checking completeness.
  • Preconditions for stopping the clock.
  • Impact of national transposition timelines on the six-month requirement.
  • Consistency between decision and local transposition.
  • Reliance on institutions' self-assessments.
  • Involvement of host supervisors in later phases of the roll-out plan.
• Review Panel members support a revision of the Guidelines to provide more practical and specific guidance on these issues.

Conclusion

The peer review exercise demonstrated a high level of compliance with the CEBS Guidelines, although several issues were identified that require further clarification and harmonization. The findings suggest that while the current framework supports convergence, it needs refinement to address practical challenges in the validation and implementation of advanced measurement approaches. A more detailed and consistent methodology, along with updated Guidelines, is recommended to ensure effective and uniform application across Member States.