Information Diffusion and the Boundary of Market Efficiency : Empirical Evidence from Earnings Announcements

This paper tests several predictions from an information diffusion framework in the quarterly earnings announcement setting. First, post-announcement drift is documented only for earnings announcements that have high information content (uncertainty), measured by high abnormal trading volume and return volatility surrounding the announcements. The results are robust after controlling for the magnitude of earnings surprise. Second, in contrast to prior findings, empirical evidence in this study suggests that the magnitude of drift for good (bad) news firms in the diffusion period is positively (negatively) related to analyst following and institutional ownership. Third, price movement after a second earnings announcement is affected by the information contained in the previous announcement. The result suggests the existence of the superposition effect of two information diffusion processes. These findings are consistent with information diffusion at work in capital markets