As part of its activities, the European Securities and Markets Authority (ESMA) organizes a forum of enforcers from 38 different European jurisdictions, all of whom carry out monitoring and review programs similar to those carried out here by the Canadian Securities Administrators. ESMA a while ago published some extracts from its confidential database of enforcement decisions on financial statements, covering eight cases arising in the period from December 2021 to December 2023, with the aim of “strengthening supervisory convergence and providing issuers and users of financial statements with relevant information on the appropriate application of IFRS.” There’s no way of knowing whether these are purely one-off issues or more widespread, but some of them certainly have some relevance to matters discussed within Canadian entities once in a while. Here’s one:

• The issuer, a debt management company, acquired portfolios of credit-impairment financial assets (non-performing unsecured consumer loans). In accordance with paragraph 5.5.13 of IFRS 9 Financial Instruments, the issuer recognized as an impairment gain or loss the amount of the change in lifetime expected credit losses (ECL) at each reporting date. With this, the recognized loss allowance for the loan portfolios was equal to the cumulative change in lifetime ECL since initial recognition of the loans.
• For the ECL calculation, the issuer used a rolling 180-month collection forecast, based on the issuer’s own historical data and industry collection decay curves. At the end of each quarter, the issuer analyzed the forecast and actual collection for all portfolios where the actual collection over the last six months deviated from the forecast by both more than 5% and exceeded a predetermined absolute monetary threshold. This resulted in an updated 180-month forecast, which formed the basis for an updated ECL estimate.
• Neither current levels nor forecasts of macroeconomic factors such as unemployment rate, economic growth or interest level were used by the issuer as input factors for the ECL calculation. In the issuer’s opinion, such macroeconomic factors would be implicit in its ECL estimate, as future estimates are an extrapolation of certain patterns of historical collections. Furthermore, the issuer argued that there was no significant correlation between the ECL and changes in macroeconomic situation.
• While the ECL at any point in time was estimated using a single collection forecast per portfolio, the issuer argued that a collection forecast such as described, implicitly reflected a range of possible outcomes in a fair way.

The enforcer (as ESMA likes to term it) considered that the issuer’s ECL measurement methodology wasn’t in compliance with IFRS 9: in particular, the issuer didn’t consider reasonable and supportable information regarding forecasts of future economic conditions that was available without undue cost or effort, and didn’t evaluate in an adequate manner a range of possible outcomes when determining the ECL amount. As a consequence, the enforcer required the issuer to collect data on current and projected macroeconomic factors and to use this data as actual input data in estimating ECL, using more than one scenario. Here’s some of what it had to say:

• Although paragraph B5.5.52 of IFRS 9 acknowledges that the best reasonable and supportable information in some cases may be the unadjusted historical information, it requires that estimates of changes in ECL reflect and are directionally consistent with changes in related observable data (such as changes in unemployment rates, property prices, commodity prices, payment status or other factors that are indicative of credit losses).
• The enforcer noted that the issuer stated in its financial reporting that future collection would be affected by macroeconomic factors such as unemployment rate, economic growth, interest level and housing prices. The enforcer agreed with the issuer that there are arguments supporting the view that credit losses of debt management companies may be less sensitive to changes to the macroeconomic factors than for banks. However, the issuer did not provide sufficient evidence of a lack of correlation between changes in current or future economic conditions and credit losses. In addition, the fact that the issuer performed periodic reviews and updated its estimates could not compensate for not using all reasonable and supportable information required by IFRS 9.5.5.17 in the ECL estimation.

As so often in these cases, one would be intrigued to know more. For example, when we’re told that “the issuer argued that there was no significant correlation between the ECL and changes in macroeconomic situation,” one wonders how much evidence they had to support that contention (presumably more than just a hunch, albeit that it wasn’t enough to persuade the enforcer – we’re not told though whether the methodological weaknesses ultimately resulted in a material adjustment). It didn’t help in this case that the historical data went back only to 2015, and that the estimation period of 180 months was a relatively long one, all of which the enforcer cites as supporting the need for the use of multiple scenarios. Overall, it seems that even if the issuer’s analysis was essentially correct, it underestimated the work and support required to persuade a skeptically-minded observer of that…

The opinions expressed are solely those of the author.