Current Expected Credit Loss (CECL) Model and Analyst Forecasts

We investigate whether the adoption of the Current Expected Credit Loss (CECL) standard by U.S. banks affects three properties of financial analysts’ loan loss provision forecasts: accuracy, dispersion, and coverage. We find that CECL adoption is associated with reduced accuracy and coverage, and increased dispersion, of analyst provision forecasts, indicating a deterioration in analyst forecast properties. We do not observe similar changes to the properties of pre-provision earnings forecasts, and we show that the main results vary cross-sectionally with proxies for the importance of the new standard to the bank, strengthening the link between CECL adoption and changes in analyst forecast properties. Finally, we provide evidence of muted stock market reactions to analysts’ provision forecast revisions after CECL adoption, suggesting that investors perceive these forecasts to be less informative post-adoption. Collectively, our results contribute to existing research on the consequences of expected credit loss standards