2014年-EBA欧洲银行管理局_20131217_Summary_report_on_comparability_and_pro-cyclicality_of_the_IRB_Approach_32页_974kb

Summary Report on Comparability and Pro-Cyclicality of Capital Requirements under the IRB Approach

Core Content

This report by the European Banking Authority (EBA) addresses the comparability and pro-cyclicality of capital requirements under the Internal Ratings Based (IRB) Approach as mandated by Article 502 of the Capital Requirements Regulation (CRR). The EBA conducted several studies to assess how capital requirements vary across institutions and jurisdictions, and whether these variations are driven by risk or non-risk factors. It also examined the potential for capital requirements to be pro-cyclical, i.e., to amplify economic cycles through their impact on lending behavior.

Main Points

1. Mandates and Objectives

• The EBA is tasked with reporting on the comparability of capital requirements and their pro-cyclicality.
• The goal is to promote convergence in capital requirements and reduce the risk of unlevel playing fields among banks.
• The report is part of a broader initiative to ensure a credible and transparent solvency framework.

2. Structure of the Report

• The report integrates findings from five individual studies:
  • Top-down interim report on the consistency of risk-weighted assets (RWA).
  • Bottom-up low default portfolio (LDP) benchmarking.
  • Bottom-up SMEs and residential mortgages (RM) benchmarking.
  • Report on supervisory rules and practices.
  • Report on pro-cyclicality of capital requirements under the IRB Approach.
• The structure is illustrated in Figure 1, showing the different studies and their focus.

3. Comparability of Capital Requirements

Key Findings:

• Significant differences in capital requirements exist across institutions, especially in retail and corporate exposures.
• These differences are driven by both risk-based and non-risk-based factors.
• The top-down analysis identified four major drivers:
  • Partial use (permanent and roll-out).
  • Standardised Approach (SA) risk weights.
  • IRB portfolio mix.
  • Share and risk weights of defaulted assets.
• These four drivers can explain around 50% of the variation in the global charge (GC).
• Non-risk-based differences may arise from:
  • Divergent modeling practices.
  • Calibration of parameters like PD and LGD.
  • Supervisory discretion and national rules.
  • Definitions of default and past due.

Policy Recommendations:

• A more harmonised disclosure of aggregated data, preferably under Pillar III reporting, would improve transparency and allow better understanding of differences in capital requirements.
• The EBA recommends further convergence in supervisory practices and regulatory frameworks to reduce non-risk-based differences.

4. Pro-Cyclicality of Capital Requirements

Key Findings:

• The evidence on pro-cyclicality is weak, and a clear causal link between capital requirements and the economic cycle could not be established.
• The pro-cyclical effect is considered a potential risk, as it could lead to amplification of economic cycles.
• The report includes empirical assessments and an econometric analysis of the relationship between macroeconomic indicators and minimum capital requirements.

Policy Recommendations:

• The EBA recommends improving transparency and documentation regarding rating philosophies, PD calculation, and back-testing methodology.
• Further benchmarking and analysis are necessary to understand the full implications of pro-cyclicality.
• The supervisory benchmarking exercise, to be conducted annually, will help assess convergence of practices.

Key Studies and Data

• Top-down interim report (26 February 2013): Analyzed the banking book using aggregated data from 89 IRB banks across 16 EU jurisdictions.
• Bottom-up LDP study (5 August 2013): Focused on low default portfolios, including sovereigns, institutions, large corporate, and other corporate portfolios.
• SMEs and RM study: Ongoing, with preliminary results published alongside this summary report. It aimed to collect historical data used in model calibration.
• Supervisory rules and practices survey: Included responses from 21 competent authorities (CAs), highlighting divergent approaches in defining and applying rules.

Data Sources and Methodology

• The EBA used global charge (GC) as a key indicator, defined as $[(RWA + 12.5*EL)/EAD]$.
• Implied benchmarks were used to assess variation in RWA and EL.
• Survey data, bank interviews, and expert judgment were combined with statistical analysis to evaluate the drivers of variation.
• Comparative analysis was conducted across different IRB approaches (FIRB, AIRB) and the SA.

Conclusion

• The IRB framework has proven its validity as a risk-sensitive approach.
• However, flexibility in the framework has led to divergent practices across institutions and jurisdictions.
• The EBA believes that harmonisation of practices and convergence of methodologies are necessary to improve comparability and reduce pro-cyclicality.
• The policy recommendations aim to enhance transparency, consistency, and credibility of the regulatory framework, while preserving the risk-sensitive nature of the IRB Approach.

Key Drivers of Variability

1. GC for defaulted assets – High variability due to different approaches to LGD and ELBE.
2. Roll-out effect – Differences in the share of SA exposures.
3. Definition of default and past due – Varies across CAs and banks.
4. PD and LGD calibration – Divergent practices in model parameters.
5. Supervisory rules and practices – National discretion in defining and applying rules.
6. Portfolio mix – Divergence in the composition of IRB portfolios.

Policy Recommendations

• Develop draft regulatory technical standards (RTS) for LGD and ELBE under the CRR.
• Promote harmonised disclosure of aggregated data.
• Conduct annual supervisory benchmarking to assess convergence.
• Improve transparency and documentation of rating philosophies and modeling practices.
• Continue benchmarking and analysis of capital requirements under the IRB Approach.

Annexes

• The report includes annexes with detailed methodologies, data sources, and results of the individual studies.
• A comprehensive survey of 21 competent authorities (CAs) is also included, highlighting the diversity in supervisory approaches.

Summary

The EBA has identified significant differences in capital requirements across EU banks under the IRB Approach, which are influenced by both risk-based and non-risk-based factors. While the IRB framework supports risk-sensitive capital requirements, the flexibility of the approach has led to divergent practices. The report outlines policy recommendations to improve comparability and transparency, with a focus on supervisory convergence and modeling consistency. The pro-cyclicality of the framework remains a concern, but the evidence is currently weak, and further research is needed.