When are expected credit losses decision-useful and new to investors? Evidence from CECL adoption

Abstract

The Financial Accounting Standards Board replaced the “incurred loss” (IL) model with the “current expected credit loss” (CECL) model, which requires entities to recognize lifetime expected credit losses upon loan origination. We investigate four controversies surrounding CECL to provide evidence on when expected credit losses are decision-useful and new for investors. We examine whether (1) CECL allowances are only decision-useful for larger banks, (2) CECL allowances provide new credit loss information, (3) credit losses expected to emerge beyond one year are relevant for investors, and (4) separately reporting IL and CECL allowance amounts has more information content than reporting a single CECL allowance amount. Contrary to the views of community banks and evidence in prior literature, we find that CECL allowances are decision-useful for both small and large banks. However, CECL allowances are new information only for smaller banks. We also find that CECL allowances are decision-useful regardless of the loss emergence period, despite later-emerging losses being unrecognized under IFRS 9, CECL’s international counterpart. Finally, we find that reporting a single CECL allowance amount is relatively less decision-useful than an alternative accounting regime that reports incurred and expected future credit losses separately.

Keywords: CECL, expected credit losses, incurred loss, FASB, standard setting, IFRS 9

JEL Classification: M40, M41, M48, G21, G28