On Thursday 26th March 2020 the PRA published guidance to help firms consider important ECL implementation issues caused by the COVID-19 pandemic. Its aim is to steer firms and auditors to account for the positive effects of government intervention in their financial risk measurement to “reduce the risk of firms recognising inappropriate levels of ECL” and ensure the financial system is “a source of strength for the real economy during this challenging period”.
How firms choose to categorise exposures as “forbearance”, a Significant Increase in Credit Risk (i.e. Stage 2) or “bad” (i.e. Stage 3, Default or Non-Performing) is important. The choice can lead to more intensive reporting and monitoring requirements, increased Risk Weighted Assets (i.e. capital demand), and increases in balance sheet Expected Credit Loss (ECL) and impairment stock/charge.
Typically, lenders will only grant concessions when customers are in financial difficulty and, typically, these exposures are higher risk than exposures without a concession (reflecting the underlying risk and customer difficulty that has led to the modification in the first place). Again, typically, we would expect firms to include as “bad” those forbearance treatments where credit behaviour shows a higher likelihood of future default.
However, the current circumstances are not typical. As well as forbearance schemes being open to all borrowers, whether in financial difficulty or not, firms need to consider customers experiencing liquidity rather than credit events, where there is good confidence about a borrower’s future ability to generate stable and sufficient cash flows. This needs a credit risk judgement to be made that a customer will remain good because of government support in the short term (e.g. the UK’s furlough salary scheme) and in the medium term because a customer’s regular source of income will likely recover.
The critical challenge is how firms and auditors might evolve established problem loan identification triggers to distinguish “good” from “bad” forbearance and appropriately assign some exposures with forbearance to Stage 1 and avoid excessive allocation of exposures to Stage 2 with associated increases in ECL. We propose that this could be achieved by using available credit risk and customer information to construct a framework of triggers to act as a proxy for information that might be garnered from more detailed customer conversations. These triggers will also need to be future-proofed to cope with the risk that the crisis enters a new, more prolonged phase than initially expected.
In addition to increases in Stage 2 and 3 driven by observed credit events, the overall financial impact of the pandemic depends on forward-looking expectations – firms’ views on economic scenarios and the likely impact of government relief measures – which will have a significant effect on ECL as a result of forward-looking Stage 2 allocation and ECL estimation. As well as reassessing their scenarios, firms are likely to have to reconsider the (mis)behaviour of their models and the potential need for overlays to reflect, for example, expectations of higher cure rates than observed historically.
For a more detailed analysis on classifying forbearance and problem loans, please click here to view our latest thought piece.
Related post: Applying IFRS 9 ECL at times of stress