The use of post-model adjustments to capture emerging risks

Mon 16 Oct 2023

Since the Covid-19 pandemic, post-model adjustments¹, or management overlays, have become an increasingly common and accepted mechanism used by banks to manage expected credit losses (ECLs).

The number of post-Covid unprecedented events related to the war in Ukraine, energy crisis and global economic uncertainty has raised a number of questions relating to the consistency and comparability of overlays to measure ECLs.

In May 2023, the Internal Accounting Standards Board (IASB) released a Post-implementation Review of IFRS 9 relating to impairment asking stakeholders to report back on any perceived fatal flaws in measuring ECLs, as well as granularity and consistency between information banks provide on post-model adjustments.

Understanding impacts

To better understand the impact of unprecedented events on ECLs, Mazars conducted an analysis of 26 banks in 11 European countries. The May 2023 benchmark study on Financial reporting of European banks, based on the banks’ year-end 2022 financial statements, follows on from previous editions of the report since its launch in 2020.

According to our study, 25 out of 26 banks analysed disclosed having overlays, establishing it as a generally applied methodology by banks to overcome the limits of their ECL models and deal with exceptional situations. Indeed, while acknowledging that challenges remain, the European Central Bank (ECB) in a post released in May 2023 said that it now accepts overlays as a way to manage risks that lack the necessary historical data on which classical ECL provisioning models depend.

Different risks are emerging

Based on the findings of our most recent study, we can observe clear differences between UK banks that have strongly reduced the weight of overlays in their ECL allowances in YE 2022, whereas French and Italian banks reported an increase in overlay weightings. The result of such differences produced a decrease in the average weight of overlays in ECL allowances across all banks from 16% in YE 2021 to 14% in YE 2022.

This decrease can partly be explained by changes in Covid-19 overlays, which were quite significant due to the global nature of the risk that impacted every bank in the world. However, our recent study showed that Covid-19 overlays have been reduced to almost zero for most banks as the perceived ECL impact has subsided.

With banks now focusing on specific risk exposures and relying, to a certain extent, on judgement calls on the use of overlays to manage ECLs, differences in the weight of overlays may become more marked in future. Despite the call of regulators to incorporate overlays in ECL models as soon as practically possible, as new unprecedented emerging risks such as the war in Ukraine, the energy crisis, geopolitical and environmental risks become more prominent in ECL allowance calculations, we expect overlays to continue to exist and remain significant in the coming years.

Working towards more accurate overlay models

It’s clear from recent International Accounting Standards Board (IASB) and European Central Bank (ECB) releases that there is general acceptance of overlays to capture and manage emerging and unprecedented banking risks. At the same time, there is a call by stakeholders for more granularity and consistency of data that is currently difficult to analyse and compare.

As expectations from users increase in relation to overlays, understanding mechanisms and providing more granular and relevant information on the impact of overlays on ECLs will be top of mind.

^{[1] A post-model adjustment is an incremental ECL that increases (or decreases) the ECL resulting from the bank’s IFRS 9 impairment models.}