This study examines the short- and long-run stock price reactions to non-performing loan (NPL) sales for Taiwan’s banks based on both efficiency hypothesis and agency problem. This study uses event study methods to estimate short- and long-run abnormal returns (ARs); the former are calculated by Ball-Brown (1968) traditional frame and Jensen’s α procedure; the latter ARs are computed by both the calendar-time portfolio approach and Jensen’s α. The sample is 519 NPL sale and trade announcements performing by both 14 financial holding companies and 18 banks from 2002 to 2008. The between-group tests for difference in ARs and the cross-sectional regressions are performed. The findings are as follows. For short-run ARs, the banks with frequent selling announcements and lower past performance have larger positive ARs on announcement day (t = 0). There are significant negative ARs on the day t =－1; in particular, if the traded amounts are relatively low and the buyers are equity-related to sellers, the ARs will be lower. Thus, the negative ARs on the trade announcements could be related to the large gap between the traded price and sale price. This evidence is consistent with the efficiency hypothesis. For long-run perspective, stock performance following NPL sale and trade announcements is poor, particular for small banks’ trades. Long-run cumulative ARs of banks with better past performance and more growth opportunities have less degree of decreases, supporting the agency theory. Therefore, NPL sales temporarily give investors an intuitive conception that bank asset quality will improve, but firm values do not enhance in the long term.
