[Research Seminar] Accounting: “Bank reporting of expected credit losses under IFRS 9: Information or confusion for financial markets?” L.DESROUSSEAUX & O.MADADIAN – IÉSEG

Speaker: Luc DESROUSSEAUX & Oveis MADADIAN
IÉSEG School of Management

Date and Location – Monday January 10th 2022 from 13:30 to 15:00
in R217 (Lille campus), PR13 (Paris campus) and on Zoom

—————————————–

ABSTRACT

The introduction of a new forward-looking principle-based Expected Credit Loss model (ECL model) under IFRS9, for recognizing and measuring impairment of financial assets, as a replacement for the existing rule-based Incurred Credit Loss model (ICL model) required by IAS 39 is known by practitioners to have been a game changer for banks. We analyse how the shift from an incurred to an expected credit loss models for bank loan portfolios impacted the information environment using analyst forecast characteristics. Using an international sample of banks reporting under IFRS, we show that the shift from ICL model to ECL model led to a decrease in analyst forecast dispersion but no impact on analyst forecast precisions. The results hold when we either use EPS or Loan Loss Provision (LLP) analyst forecasts as dependent variables. In additional tests, we show that analyst forecasts of Non-Performing Loans (NPL) are less predictive of future reported levels and change in NPL following IFRS 9. Overall, these findings suggest that the introduction of the CECL model for bank loans yields mixed results with regard to the quality of banks’ information environment.