Need for banks’ participation for benchmarking credit risks

Global Credit Data’s latest IFRS 9 Report demonstrates that banks’ expected credit loss estimates vary widely.

Author
Date published
August 20, 2019 Categories

Global Credit Data (GCD) has released its second IFRS 9 benchmarking report, with the results highlighting that benchmarking credit risks is key to understanding and addressing variability.

As the first IFRS 9 statements are being released, banks, investors, auditors, regulators and other financial industry participants are attempting to understand the variability of loss projections, provision charges and Expected Credit Loss (ECL).

Results from the study demonstrated a significant degree of variability of around factor 4. This suggests that the IFRS 9 framework has yet to stabilise. However, the authors of the report, Daniela Thakkar and Richard Crecel from GCD feel that as institutions develop more precise methods to improve future credit loss estimates, regulators and auditors will push for greater consistency.

Key takeaways from the report are:

  1. The variability between banks’ estimates is observed for all segments defined by ECL drivers such as obligor type, geography, industry, rating and PD, facility type, guarantees and collateralization
  2. Bank-specific or reference macro-economic scenarios used for projections led to identical conclusions: in the current macro-economic environment, the variability between banks is mainly caused by banks’ different models and not by different macro-economic forecasts
  3. In order to “measure” the variability, we introduce a multiplier (=ECL 3rd quartile / ECL 1st quartile, calculated over all participating banks). We see that the multiplier is fairly stable over all asset classes and – on average – stands at least at a level of 4.

One size doesn’t fit all

IFRS 9 requires banks to estimate a 1-year and a life-time ECL and these measurements are expected to be responsive to macroeconomic developments and to include a forward-looking perspective. A certain amount of variability is expected, as ECL should also capture banks’ specificities, thus fostering diversity and enhanced resilience in the financial market. A “one size fits all” calibration shall not be the desired objective.

Richard Crecel, Executive Director of Global Credit Data commented: “We remain in the early stages, however, the high level of variability in ECL figures under IFRS 9 is something the industry will need to analyse and address. If banks don’t act, they may find the regulator acts for them – and imposes more restrictive standards than many would like.”

The report highlights the there are questions still unanswered on how much variability is too much and how should the industry go about standardising the approach to IFRS 9.

Daniela Thakkar, Methodology and Membership Executive, Global Credit Data added: “We are not looking for a one-size-fits-all calibration. Banks and regulators need to understand the potential effects of IFRS 9, and benchmarking will be a key tool in this regard. We hope this report will generate an industry-wide conversation and encourage financial institutions to participate in such benchmarking activities.”

Global Credit Data’s IFRS 9 Report

GCD conducted the study in the summer of 2018 and supporting 26 international banks in finalising their IFRS 9 implementation. Each bank in the study applied their own IFRS 9 framework to the same hypothetical portfolio of obligors and exposures, using the same macroeconomic forecasts and the same stress test scenarios. The effects of bankspecific scenarios on the ECL have been analysed as well.

Hence the study focuses on differences potentially driven by methodologies and neutralizes differences due to qualitative adjustments or differences due to banks’ specific risk profiles.

Click here to download the full report.

Exit mobile version