EBA欧洲银行-EBA-RTS-2015-01-and-EBA-ITS-201-501-Final-Draft-RTS-and-ITS-on-Benchmarking-Exercise_87页_1mb

EBA Final Draft Regulatory and Implementing Technical Standards on Benchmarking under Article 78 of CRD IV

Core Content

The European Banking Authority (EBA) has developed Final Draft Regulatory Technical Standards (RTS) and Final Draft Implementing Technical Standards (ITS) under Article 78 of Directive 2013/36/EU (CRD IV). These standards aim to improve the consistency and comparability of risk-weighted assets (RWAs) and capital requirements calculated by banks using internal models for credit risk (IRBA) and market risk (including VaR, SVaR, IRC, and Correlation Trading models). The EBA is also required to produce a report to assist competent authorities in their assessments.

Main Objectives

• Reduce uncertainty and differences in the calculation of capital requirements.
• Benchmark internal models to ensure consistency across institutions.
• Enable comparison between competent authorities and institutions.
• Identify and address variability drivers, including both regulatory and non-regulatory factors.

Key Information

1. Benchmarking Tool and Annual Assessments

• Competent authorities must assess the consistency and comparability of RWAs at least annually.
• The benchmarking tool allows for comparison of model outcomes.
• The EBA has introduced a rotation approach for benchmarking portfolios to manage workload.

2. Scope of the ITS and RTS

• ITS cover:

  • Templates, definitions, and IT solutions for benchmarking portfolio reporting.
  • Benchmark portfolios for credit risk (IRBA) and market risk (including VaR, SVaR, IRC, and Correlation Trading models), counterparty risk, and CVA risk.

• RTS specify:

  • Procedures for sharing assessments between competent authorities and the EBA.
  • Standards for the assessment of internal models used to calculate capital requirements.

3. Market Risk Benchmarking Portfolios

• The EBA provides individual and aggregated portfolios.
• Individual portfolios are categorized by risk types: Interest Rate, Equity, FX, Commodities, and Credit.
• Correlation trading portfolios are a separate category and not included in aggregated portfolios.
• For the initial exercise (Q4 2015), the EBA uses pre-existing Basel portfolios.
• Hypothetical portfolios are also provided for future exercises to test specific risk factors.

4. Credit Risk Benchmarking Portfolios

• Low Default Portfolios (LDP): include central governments, institutions, and large corporate exposures.
• High Default Portfolios (HDP): include corporate, SMEs, and residential mortgages.
• Real LDP cluster portfolios are based on key dimensions like rating grade, collateral type, and geography.
• Hypothetical transactions are used to assess maturity, CCF (EAD), and LGD estimates for LDP exposures.

5. Variability Drivers

• Regulatory Choices: include methods for VaR (lookback period, weighting schemes, rescaling), IRC (PD sources, transition matrices, liquidity horizons), and IRBA (data sources, rating grades, open workout procedures).
• Non-Regulatory Choices: such as simulation engines, volatility and correlation assumptions, and risk factors considered.
• Supervisory Actions: like regulatory add-ons and diversification limits, which can increase capital variability between banks and jurisdictions.

6. Assessment Methodology

• Competent authorities must assess variability and capital levels.
• Thresholds for acceptable variability are defined in the report.
• Extreme values and peer group results are key elements in the assessment.
• Back-testing is required for high default portfolios to compare estimated and historical risk parameters.

7. Reporting Templates

• Templates include detailed model parameters (e.g., PD, LGD, EAD) for credit risk.
• For market risk, historical P&L data is requested for banks using Historical Simulation.
• The EBA reuses existing COREP definitions and extends them for benchmarking purposes.

8. Data Sources and IT Solutions

• The Data Point Model (DPM) is used as a basis for benchmarking data requirements.
• Institutions must submit results of internal models applied to EBA benchmark portfolios.
• The EBA will provide clustering and analysis of model outcomes to identify variability sources.

9. Sharing Procedures

• Competent authorities must share their assessment conclusions with the EBA.
• The EBA will aggregate and analyze this feedback to identify common conclusions and provide guidance.
• Results are also shared in supervisory colleges to enhance coordination.

Summary of Key Points

• The EBA aims to improve consistency and comparability of capital requirements.
• Benchmarking portfolios are designed to capture different risk factors.
• Variability is expected due to both regulatory and non-regulatory choices.
• Supervisory actions such as add-ons and diversification limits also contribute to variability.
• Data collection is based on Q4 2015 observations.
• Templates and IT solutions are aligned with existing supervisory reporting standards.
• Assessment methodology includes checking for extreme values, model quality, and calibration features.
• Back-testing is used to evaluate the accuracy of internal models.
• The reporting process is flexible but structured to allow for future changes and analysis.

Conclusion

The EBA’s standards are designed to ensure transparent, consistent, and comparable capital calculations while allowing for evolving risk management practices. The use of benchmarking portfolios, combined with detailed reporting and assessment procedures, provides a robust framework for supervisory oversight and model validation.