2017年-EBA欧洲银行管理局_EBA_REPORT_RESULTS_FROM_THE_2016_HIGH_DEFAULT_PORTFOLIO_28HDP2920EXERCISE_69页_5mb

EBA Report: 2016 High Default Portfolios (HDP) Exercise Summary

Core Content

This report presents the results of the 2016 supervisory benchmarking exercise conducted by the European Banking Authority (EBA) on high default portfolios (HDPs), which include residential mortgages, SME retail, SME corporate, and corporate-other portfolios. The analysis is based on data reported as of 31 December 2015 by 114 institutions across 17 EU countries, representing a significant increase in the number of participants compared to previous reports.

Main Objectives

• Assess the variability in risk-weighted assets (RWA) across HDPs.
• Identify and explain the drivers of dispersion in RWA and risk parameters.
• Provide insights for competent authorities (CAs) to evaluate the quality of internal models used for calculating own funds requirements.
• Highlight the impact of observed defaults and risk parameters on the CET1 ratio.

Key Findings

1. Overall Variability

• Global Charge (GC): The average GC increased to 75% compared to 67% in the 2013 HDP report.
• GC Variability: Ranges from 8% to 293%, indicating higher dispersion than previous studies.
• RW Average: Ranges from 7% to 129%, with a simple average of 33%.
• Explainable Variability: 82% of GC variability can be explained by key factors such as the proportion of defaulted exposures, non-EU exposures, and portfolio mix.
• Unexplained Variability: The remaining 18% is attributed to underlying credit risk, modeling assumptions, and supervisory practices.

2. Cross-Sectional Analysis

• Portfolio-Specific Variability: SME corporate and corporate-other portfolios show significant interquartile range variability.
• Country Impact: Exposures in EU countries with stressed macroeconomic conditions tend to have higher RWs.
• Regulatory Approach Differences:
  • AIRB generally results in lower LGDs and CCFs than FIRB.
  • FIRB PDs are typically lower than AIRB PDs.
• Modeling Assumptions: These differences may explain the observed variations in risk parameters.

3. Outturn (Backtesting) Analysis

• PD and LGD Estimates vs. Observed Data: Estimated PDs and LGDs are generally higher than observed default and loss rates, suggesting a conservative approach by banks.
• Outliers: Some banks show observed values above estimated values, requiring further analysis.
• Country and Counterparty Influence: The country of the reporting bank and counterparty is a major driver of RW variability.

4. CAs' Assessments

• Justified Deviations: For most banks, deviations from EU benchmarks were deemed justified and not significant.
• Focus on Residential Mortgages: Identified as one of the most important portfolios to monitor due to their impact on RWA.
• Underestimations: Corporate-other and SME corporate portfolios showed the highest potential for underestimations, with no clear justifications.
• Internal Validations: Most banks did not identify significant underestimations in their internal validations.
• Supervisory Actions: CAs noted that some underestimations were identified in advance and corrective actions were taken.

5. Impact on CET1 Ratio

• Estimated Impact: Replacing PD estimates with observed default and loss rates would reduce the average CET1 ratio by only 17 bps.
• Data Quality Considerations: The analysis should be interpreted cautiously due to data quality issues and the fact that higher RWAs were not designed to estimate impacts.

Methodology and Data

Dataset

• Participants: 114 institutions from 17 EU countries.
• Data Scope: Data was reported at the highest level of consolidation.
• Excluded Data: No information on SA-rated exposures or non-HDP portfolios was collected.

Data Quality

• Constraints: Larger sample size and new technical standards introduced some data quality issues.
• Interpretation: Findings should be interpreted with caution due to potential inconsistencies in data interpretation across banks.

Assessment Approaches

• Top-Down Approach: Examines GC variability by controlling for key factors such as default status, country, and portfolio mix.
• Cross-Sectional Approach: Analyzes risk parameters across different portfolios and countries, identifying outliers and extreme values.
• Outturn (Backtesting) Approach: Compares observed values with estimated values for PDs and LGDs, highlighting potential underestimations.

Conclusion and Future Work

• The report serves as an update on RWA variability monitoring and aims to understand the drivers behind such variability.
• It provides a basis for CAs to improve their assessments and take appropriate supervisory actions.
• Future work will involve regular benchmarking exercises starting from 2018, following the new framework established by the EBA through ITS and RTS.

Key Information

• HDPs: High default portfolios include residential mortgages, SME retail, SME corporate, and corporate-other.
• Regulatory Framework: Based on Article 78 of the Capital Requirements Directive (CRD) and the Capital Requirements Regulation (CRR).
• Technical Standards: ITS and RTS from March 2015 were used to structure the benchmarking exercise.
• Data Sources: Data collected under ITS and supplemented with COREP data when necessary.
• Qualitative Insights: Interviews with 10 banks provided deeper understanding of internal modeling practices and assumptions.