Why do investors discount earnings announced late?

Holding earnings surprises constant, investors react negatively to delayed earnings announcements. One standard deviation of delay (5 days) corresponds to about 21 bps negative abnormal returns over a two-day announcement window. We show that the results are robust after further controlling for various firm characteristics, earnings characteristics, and the industry effect. We examine alternative explanations, including the earnings manipulation hypothesis proposed in the literature, which suggests that delayed earnings announcements are susceptible to manipulation and are thus discounted by investors. We find no evidence supporting the earnings manipulation hypothesis. Instead, our results are consistent with the effect of concurrent information disclosure. We show that there is a cluster of bad news for late announcements. As investors react to not only a firm’s own announcements but also concurrent earnings announcements, the negative information of concurrent announcements contributes significantly to lower stock returns. In addition, information update by the management and analysts also has a significant effect on market reactions to delayed announcements. In particular, information update by analysts may help the market to incorporate future earnings information more efficiently into stock prices. We show direct evidence that negative market reactions to delayed earnings announcements contain information of future earnings.