EBA欧洲银行-CP32_BNPParibas_10页_132kb

BNP Paribas Summary of CEBS Guidelines on Stress Testing

Core Content

BNP Paribas has provided feedback on the CEBS Consultation Paper 32, which outlines a framework for stress testing in banking institutions. The document serves as a useful complement to Pillar 2 guidelines, but BNP Paribas highlights several areas for improvement, particularly in terms of qualitative judgment, proportionality, and the integration of stress testing with other risk management practices.

Main Comments

General Comments

1. Quantitative Bias
   The CEBS guidelines show a strong quantitative bias, potentially undermining the importance of qualitative expert judgment in stress testing. BNP Paribas emphasizes that stress testing is not merely a computational exercise and that qualitative aspects are essential, especially for top-down approaches. A stronger emphasis on proportionality rules is needed to avoid overburdening all business units and portfolios with stress testing.

2. Realistic Implementation Deadlines
   The proposed implementation date (June 30, 2010) is too tight. BNP Paribas requests a more realistic and progressive timeline for the development and implementation of stress testing methodologies and tools, particularly for multi-risk scenarios and feedback effects.

Special Comments

1. Calculation Infrastructure
   Stress testing should focus on linking existing systems rather than building new infrastructures. Dedicated teams are more important than dedicated systems, and the uniqueness of business lines should not be compromised by overly conservative global stress tests.

2. Scenario Preparation
   The preparation of complex and specific scenarios by banks limits regulatory consolidation. The severity of economic scenarios is not the only factor; their relevance to the portfolio is more important. BNP Paribas also notes that the number of scenarios is a management decision and should not be mandated.

3. Reverse Stress Tests
   Reverse stress tests should maintain their qualitative nature and not be used as a basis for capital planning scenarios. Their results should not be reused for other regulatory purposes.

4. Multi-Level Approach
   The distinction between "business unit" and "legal entity" should be clarified. Stress testing should primarily occur at the business unit level, and the legal entity level should not be overemphasized unless it provides added value.

5. Link with Insurance Activities
   The reference to insurance stress tests should be exclusive to the prudential banking business. Similar stress tests may be required by insurance-specific regulations, but they should not be integrated into the banking stress testing framework.

6. Results of Stress Exercises
   While management decisions based on stress test results should be documented, BNP Paribas questions the need for formal corrective actions for every stress test. Stress testing is part of management discipline, and the credibility of decisions can be enhanced by referencing the results without formalizing every action.

7. Supervisory Review
   A direct link between capital planning stress tests and capital cushions could undermine the purpose of stress testing. Institutions should have flexibility in their responses to stress test outcomes beyond just capital cushions.

8. Appendices
   The proportionality rule should also apply to the content of the appendices. They should reflect ongoing work on Pillar 1 reforms, ensuring coherence between different regulatory frameworks.

Technical Points

Operational Risk

• §2
  Stress assumptions for operational risk should differ from credit and market risk and be based on external events such as stock exchange crashes. The examples provided in the guidance should not be considered as references.

• §8
  The four elements of AMA (internal and external data, scenario analysis, business environment, and internal control factors) should not be combined with assumed weights. Institutions should implement stress tests based on their specific AMA.

• §9
  The paragraph linking stress testing to economic capital should be removed. It refers to an issue outside the scope of operational risk stress testing.

Liquidity Risk

• §4
  A new section (§6) is proposed to incorporate alternative liquidity scenarios in conjunction with other risk types (market, credit, reputation). Spillover effects should be analyzed within a globally consistent stress test framework.

• §12
  The definition of liquidity buffer should be revised. The current requirement for maintaining the same buffer at the end of the survival period is too extreme and could lead to an infinite compounding effect.

• §13
  Consolidated liquidity stress testing should consider legal and cross-border transfer constraints. Entities in different jurisdictions may face liquidity transfer risks, and these should be incorporated into scenarios.

• §12 (repeated)
  Stress test results should be used to improve liquidity risk management, including internal policies, limits, and contingency funding plans. This should be a continuous process rather than a one-time exercise.

Conclusion

BNP Paribas supports the overall goal of the CEBS guidelines but urges a more balanced approach that integrates both quantitative and qualitative elements. The document should emphasize proportionality, provide realistic implementation timelines, and ensure coherence with other regulatory frameworks. Technical aspects such as the treatment of operational and liquidity risks need refinement to better reflect practical realities and institutional capabilities.