Calculating expected credit losses (ECLs) under IFRS 9 is complicated. For example, it requires banks to use historical data and assumptions about the future to form forward-looking estimates using complex modelling. In developing their ECL estimates, each bank needs to use data, assumptions, models and methods that are specific to its facts and circumstances. These complexities can make the estimates challenging for banks to explain and for financial statement users to understand. This is true during the most benign economic periods, and even more so now due to the uncertainty caused by COVID-19.

ECL disclosure requirements are largely principles-based, not prescriptive, and this could lead to diversity in disclosures. Not surprisingly, investors, bank regulators and analysts have already begun to compare disclosures and express expectations around their robustness (for example, forward-looking information, and information about scenario weighting). We expect such interest in robust disclosures that ‘tell the story’ will continue during this period of uncertainty, and that appropriate disclosures can be achieved by focusing on the following three key elements.

Interim financial statements are intended to provide an update of information included in the previous annual financial statements. IAS 34 presumes readers will have access to those annual financial statements, and so it allows entities not to repeat disclosures previously provided. Rather, IAS 34 requires entities’ disclosures to focus on new activities, events and circumstances that are significant to understanding the entity’s financial position and performance since the end of the previous annual reporting period. All of these are reasons to revisit existing disclosures (for example, granularity of information, differentiation of allowances between performing and non-performing, and changes in ECLs including write-offs) and to pay careful attention to what has changed in the current environment (such as policies related to how client relief programmes and government relief are treated, take-up rates, and significant changes in processes related to the quantification of ECLs).

In almost all cases, the economic effects of the global pandemic will represent significant changes, and the related uncertainties might be substantial. Additionally, entities may need to reconsider what disclosures are material. On the latter, it’s worth noting IFRS has a new definition applicable this year which defines information as being material “… if omitting, misstating or obscuring it could reasonably be expected to influence the decisions that the primary users of general purpose financial statements make on the basis of those financial statements, which provide financial information about a specific reporting entity”.

In order to understand a forward-looking estimate, information about the inputs used and the judgements made about the future are key. This might include the use of updated forecasts of key macroeconomic variables under multiple economic scenarios for the calculation of the ECL, probability weights that have been assigned to each scenario, and what the resulting ECLs for each scenario would be before probability weighting. While it’s certain to be imperfect, sensitivity analysis of the ECL under multiple scenarios might also be meaningful, because it enables users of the financial statements to fundamentally understand the inherent imprecision of the estimate and it can provide a reasonable range of possible outcomes. In many cases, banks provide this information in their annual reporting, but disclosure in interim financial statements is likely to be warranted, given current conditions.

Discussing the interaction between the modelling used for impairment and market inputs (such as those used to disclose fair values of the underlying loans) would also be appropriate – for instance, whether the widening of credit spreads in the market and related probabilities of default translate into the ECL for the related exposures (either listed or by proxy). The consistency of assumptions with market data and other areas of estimation in the financial statements (for example, fair value) is an important area for both the entity and users to understand.

In addition to forward-looking information and judgements, a key source of uncertainty in ECL estimates is model performance, particularly the need for model adjustments. At a high level, models have been developed using historical data that looks different from the world today, and so adjustments either to models themselves or to their outputs are often necessary. Where those adjustments are needed, it is important for disclosures to explain why that’s the case, and to articulate how the adjustments were made and what uncertainties remain. Whether quantitative, qualitative or both, detailed disclosures about the performance of models and adjustments thereto give users important insights into management’s estimate and the levels of judgement and uncertainty implicit therein.

Estimating ECLs can be challenging, and even more so in current conditions. This makes informative disclosure even more relevant, including disclosures regarding the sources of estimation uncertainty and management judgements.