EBA欧洲银行-2014-08-19-BSG-Opinion-on-EBA-CP-2014-07_11页_264kb

EBA Banking Stakeholder Group Consultation Summary on Benchmarking Portfolios

Core Content

The EBA Banking Stakeholder Group (BSG) has provided detailed comments on the Consultation Paper EBA/CP/2014/07, which outlines draft Regulatory Technical Standards (RTS) and draft Implementing Standards on benchmarking portfolios. The BSG supports the initiative to harmonize supervisory rules across Europe, aiming to ensure fair competition, improve efficiency for cross-border groups, and facilitate data sharing between supervisors. However, they identify several issues that need clarification to prevent unintended consequences.

Main Views and Key Information

General Comments

• The BSG emphasizes the importance of involving banks early in the benchmarking process to avoid erroneous conclusions due to insufficient information.
• Reporting requirements for benchmarking portfolios are substantial and costly. The BSG supports using existing reporting systems and the Data-Point-Model (DPM) to reduce duplication and costs.
• The draft RTS does not sufficiently address the complexity and variability in internal risk modeling practices, which can affect capital requirement calculations.
• A harmonized stress period is welcomed, provided it is defined in coordination with credit institutions and allows for a sufficient implementation period.
• The BSG suggests that the stress period should start from 2008 to ensure data availability.
• The use of standardized approaches for benchmarking is questioned due to their limited risk sensitivity and potential to obscure rather than support benchmarking.

Replies to Questions

Q1: Common benchmarks for credit and market portfolios

• The BSG supports the use of common benchmarks to ensure a common approach.
• They note that differences in risk model development practices may make comparisons difficult.
• The BSG suggests that the draft RTS does not adequately address these challenges and may lead to counterproductive outcomes.

Q2: Sufficient proportionality and flexibility of benchmarks

• The BSG finds the proposed benchmarks (first and fourth quartiles) to be unclear and unstable over time.
• They propose replacing the term "extremes" with "outliers" and using Box plots to identify outliers.
• This would allow for more stable and meaningful benchmarking over time.

Q3: Limitations of proposed benchmarks

• The use of quartiles is limited in identifying the actual reasons for differences in risk estimation.
• The comparison between internal models and the standardized approach is not useful due to the latter's lack of risk sensitivity.
• The comparison between estimates and outturns requires further definition of rating philosophies and in-depth model analysis.

Q4: Appropriate benchmark for assessing underestimation

• The BSG suggests comparing model estimates with actual long-term default rates and credit losses.
• They recommend using lower quartiles (e.g., 90%, 95%, 99%) for validation due to their reduced data requirements and stability.

Q5: Appropriate market risk portfolios for initial exercise

• The BSG suggests using the TBG benchmark portfolio for institutions involved in the FRTB QIS.
• A limitation to plain vanilla instruments could improve comparability.

Q6: Appropriateness of EBA's approach for future annual exercises

• The BSG supports the general approach but stresses the need for clarity on the role and impact of the benchmarking method.
• They also emphasize the need for well-defined comparison methods across EU institutions.

Q7: Alternative proposals

• The BSG does not have any alternative proposals for future exercises.

Q8: Preferred phasing-in option

• The BSG prefers Option 2 for its cost efficiency.

Q9: Ambiguities in credit risk portfolios (Annex I)

• The BSG identifies potential ambiguities in modeling practices, such as rating philosophy and economic cycle definitions.
• Clarification is needed on how to treat counterparties as legal entities and how to handle different ratings within a banking group.

Q10: Additional credit risk portfolios

• The BSG does not have suggestions for additional credit risk portfolios.

Q11: Ambiguities in market risk portfolios (Annexes VII.a and VII.b)

• The BSG does not identify specific ambiguities in the market risk portfolios defined in these annexes.

Q12: Additional market risk portfolios

• The BSG does not have suggestions for additional market risk portfolios.

Q13: Exemptions for reporting portfolios

• The BSG agrees that banks should be allowed to refrain from reporting portfolios if they meet the conditions in Article 3.
• They suggest clarifying whether basic IRB banks need to report LGD, EAD, and maturity.

Q14: Additional exemptions from reporting

• The BSG suggests that banks should be allowed to abstain from reporting certain exposures if they are immaterial to the assessment.
• They recommend that banks only calculate risks for products with regulatory model approval to ensure comparable standards.

Conclusion

The BSG supports the harmonization of benchmarking practices but stresses the need for further clarification and flexibility in the proposed standards to ensure meaningful and accurate comparisons across institutions. They advocate for the use of outliers and Box plots for benchmarking, emphasize the importance of involving banks in the process, and suggest focusing on products with regulatory approval to avoid distortions.