22 February 2022

EBA publishes annual assessment of banks’ internal approaches for the calculation of capital requirements

The European Banking Authority (EBA) published today its Reports on the annual market and credit risk benchmarking exercises. These exercises aim at monitoring the consistency of risk weighted assets (RWAs) across all EU institutions authorised to use internal approaches for the calculation of capital requirements. Regarding market risk, for the majority of participating banks, the results confirm low dispersion in the initial market valuation (IMVs) and increased dispersion in the VaR submissions. For credit risk, the variability of RWA remained rather stable, despite the pandemic and banks’ efforts to re-develop or re-calibrate their models to comply with the policies set out in the EBA internal rating-based (IRB) roadmap. 

Credit Risk exercise

This year’s Report provides an in-depth analysis of the observed and potential impact of the COVID-19 pandemic. It includes theoretical analysis on where potential impact is likely to be observed and empirical analysis on the development of average RWs, probabilities of default (PDs) and default rates (DRs) between 31 December 2019 and 31 December 2020 for the different benchmarking portfolios. The theoretical assessment concludes that heterogeneity of the impact is expected not only due to the different extent to which the underlying loans (obligors) are affected by the pandemic but also because of the institutions’ relevant processes for assigning and reviewing IRB ratings. The empirical analysis indicates that the observed decrease of average RW and PDs for high-default (HP) portfolios is mostly due to the re-estimation of PDs conducted in 2020. For qualified revolving exposures (RQRR) slight migrations of retail obligors/exposures towards better rating grades are observed, while for SME corporates the decrease of RW is likely to be related to the Capital Requirements Regulation (CRR) quick fix. Lastly, for retail SME portfolios, the observed decrease of average default rates may indicate a potential overcompensation of the expected impact of the economic crisis by public measures and moratoria.

The Report also includes an analysis of the developments of both average PDs and average default rates between end 2019 and end 2020, which may be due to the IRB roadmap implementation.

As usual, the Report contains a section on the competent authorities’ assessment of the deviations from the benchmarks. For HDP portfolios, this assessment underpins the results obtained from the empirical analysis and confirms that some deviations are related to the COVID-19 pandemic. The focus analysis is complemented with an extensive chart pack providing the analysis of the benchmarking metrics.

Market Risk exercise                         

The Report presents the results of the 2021 supervisory benchmarking and summarises the conclusions drawn from a hypothetical portfolio exercise (HPE) conducted by the EBA during 2020/21.

The 2021 exercise considered the same instruments applied in 2019 and 2020, which are mostly plain vanilla. This stability has facilitated the understanding of the benchmarking portfolio and contributed to an observed reduction in overall dispersion in the instruments booking. 

Regarding the single risk measures, the overall variability for value at risk (VaR) across all asset classes, except for credit spread, is slightly lower than the observed variability for stressed VaR (sVaR) (27% and 31% respectively, compared with 18% and 29% in 2020). More complex measures, such as incremental risk charge (IRC) show a higher level of dispersion (43% compared with 49% in 2020).

The increase of the VaR dispersion for 2021 was also analysed separately in order to explain the impact of the increased market volatility followed by the COVID-19 outbreak. 

Competent authorities also complemented a questionnaire on banks participating in the exercise to supplement the quantitative analysis. Although the majority of the causes were identified and actions put in place to reduce the unwanted variability of the hypothetical RWAs, the effectiveness of these actions can be evaluated only with ongoing analysis.

EBA

© EBA