Post-Announcement Drift

Abstract

Brown, Goetzmann and Ross (1995) document that ex-post conditioning can significantly bias empirical results based on observed rates of return. These results have interesting implications for cross-sectional cumulated excess return measures [CAR’s] that are commonly used in the context of event studies (see Brown and Warner, 1981). Ball and Brown [1968] note an upward drift in cumulated excess returns subsequent to a positive earnings announcement surprise. Subsequent work by Foster [1977] and Foster et al [1984] among others has documented that this drift is related to size of the firm in question. The current state of this literature is summarized in Ball [1992].