Bank Loan Loss Accounting: Research Implications for the Post-Crisis Debate

Abstract

The IASB and the FASB have recently re-evaluated the current model underlying loan loss accounting (the ‘incurred loss’). Taking into consideration the G20‘s advice on using more forward-looking information, they introduce a new approach (the ‘expected loss’ model). This paper reviews the academic literature to shed some light on the new expected loss models when applied to the financial industry. The accounting literature discussed in this study outlines both general theoretical findings and empirical evidence that help to infer the potential impact of the new models. Given the link between loan loss impairment and accounting conservatism as well as earnings management, we explore these concepts from both a theoretical and empirical perspective. Indeed, the discretion envisaged in the new models may provide a potential tool for increased earnings management than at present under the current guidelines. Both models are more conservative than the current approach, but given the differences in the type of conservatism that the two new models entail, conditional and unconditional, we argue that the two proposals will provide different types of information and in this way lead to distinct economic consequences. Additionally, institutional factors and firm characteristics will also have a major impact on the way the models are used. Moreover, we also suggest new avenues for further research in conservatism and earnings management in an industry, the financial sector, where additional stakeholders show significant concerns regarding the economic impact of accounting information.

Keywords: Loan loss accounting, Expected loss model, Incurred loss model, Loan loss allowance, Loan loss provision, FASB/IASB

JEL Classification: G01, M41, M48