EBA欧洲银行-Instructions-on-NMRF-data-collection_45页_1mb

EBA NMRF SSRM Data Collection Exercise Summary

Core Content

The European Banking Authority (EBA) has launched a Non-modellable Risk Factor (NMRF) Stress Scenario Risk Measure (SSRM) data collection exercise, which is part of the implementation of the Fundamental Review of the Trading Book (FRTB). The exercise aims to collect data to support the calibration of the SSRM methodology for NMRFs, which are risk factors with limited observable market data and deemed unsuitable for use in Expected Shortfall (ES) models.

Main Purpose

• To apply the EBA NMRF methodology in practice.
• To gather data for the calibration of the SSRM for non-modellable risk factors.
• To support the development of regulatory technical standards (RTS) and the Basel Monitoring Exercise (BM194).

Key Information

• Scope: The exercise is voluntary and targets institutions using the Internal Model Approach (IMA) for market risk.

• Methodology: The SSRM is derived by applying an extreme future shock to non-modellable risk factors. It involves various calculation methods including:

  • Direct method
  • Return historical method
  • Return sigma method
  • Fallback method

• Data Collection Process:

  1. First submission: To NCAs by 4 September 2019.
  2. Second submission: To NCAs by 9 October 2019.
  3. Data quality checks: 9–20 September 2019 and 14–25 October 2019.
  4. Data analysis and impact assessment: 28 October to 20 November 2019.
  5. Finalisation of RTS: December 2019/February 2020.
  6. Consultation and feedback: February 2020 to June 2020.
  7. Final publication: Early 2021.

Data Submission Guidelines

• Data completeness: Institutions are required to fill in all worksheets on a best-effort basis.

• Data formatting:

  • Dates must be in YYYY-MM-DD format.
  • Decimal numbers should use a dot (.) as the decimal separator.
  • Present values (PVs) should be provided with two decimals.
  • Currencies should be in ISO 4217 format (e.g., USD, EUR).
  • Boolean flags must use "Y" or "N".

• Data integrity:

  • All Pof_IDs must be defined in the "Portfolios" tab.
  • All RF_IDs must be defined in the "Risk_factors" tab.
  • Referential integrity is essential for the data structure to facilitate SQL-based data extraction.

• Portfolio Definition:

  • A portfolio refers to a set of risk positions with combined PV dependency and SSRM contributions.
  • Required portfolios include EBA benchmarking portfolios such as:
    • EBA Benchmarking 2019 ALL_IN no CTP
    • EBA Benchmarking 2019 EQUITY Cumulative
    • EBA Benchmarking 2019 IR Cumulative
    • EBA Benchmarking 2019 FX Cumulative
    • EBA Benchmarking 2019 Commodity Cumulative
    • EBA Benchmarking 2019 Credit Spread Cumulative

• Risk Factor Classification:

  • A risk factor is classified as NMRF if it does not meet the modellability criteria.
  • Modellability is determined by:
    • At least 24 real price observations in the past year and no 90-day period with fewer than four observations.
    • Or at least 100 real price observations in the past 12 months.
  • If a risk factor is a curve, surface, or cube parameter, it should be flagged accordingly.

• Risk Factor Buckets:

  • The exercise includes regulatory buckets and institution-specific buckets.
  • Each bucket must have a representative risk factor.
  • For regulatory buckets, data must be provided on the bucket level.

• Data Waterfall Logic:

  • The total SSRM is calculated using a waterfall approach based on the number of nearest 10-day returns:
    • If N ≥ 250, use direct method if available, otherwise return historical method.
    • If 12 ≤ N < 250, use return sigma method.
    • If N < 12, use fallback method.

Summary of Key Tabs and Fields

Additional Notes

• The EBA will treat all submitted data as confidential and will not attribute it to individual institutions.
• The SSRM calculation involves multiple steps, including determining the stress scenario risk factor range, tail shape parameters, and scaling to the liquidity horizon.
• The consolidated SSRM methodology is used to calculate the SSRM, and institutions are expected to follow it closely.
• Feedback from pilot banks was instrumental in refining the templates and instructions.