Seminar (Dept. of Finance& Accounting)

Title: Imperfect Expectations in Loan Loss Forecast

Speaker: Yao Lu, Assistant Professor, Cornell University

Time: 14:30-16:00, June 17, 2025 (Tuesday)

Venue: Room 1008, Mingde Business Building

Language: English/Chinese

ABSTRACT:

Loan loss forecasts that deviate from rational expectations can have a profound impact on banks’ loan loss provisions, lending procyclicality, and financial stability, especially under the forward-looking current expected credit losses (CECL) model. The behavioral finance literature suggests that bank forecasts can be influenced by the representativeness heuristic, which posits that estimates of future events might overreact to the circumstances that a forecaster is currently experiencing. Leveraging proprietary supervisory data (from Y14A filings), we document the presence of this overreaction in banks’ loan loss forecasts. Specifically, we find that upward (downward) revisions of loan loss forecasts are significantly associated with overestimated (underestimated) future loan losses. This overestimation (underestimation) leads to increased (reduced) loan loss provision, especially for CECL adopters. Supporting its robustness and prevalence, we observe the overreaction in loan loss forecasts across loan types and in forecasts of other performance metrics. Delving deeper into a key input for loan loss forecasts, we find that banks’ regional macroeconomic forecasts exhibit the same overreaction pattern.