Investors' Misreaction to Unexpected Earnings: Evidence of Simultaneous Overreaction and Underreaction

Behavioral Finance aims to explain empirical anomalies by introducing investor psychology as a determinant of asset pricing. Two kinds of anomalies, namely underreaction and overreaction, have been established by an impressive record of empirical work. While underreaction defines a slow adjustment of prices to corporate events or announcements, overreaction deals with extreme stock price reactions to previous information or past performance. Theoretical models have shown that both phenomena find potential explanations in cognitive biases, that is, investor irrationality. This study investigates current and past earnings surprises and subsequent market reaction for listed US companies over the period 1983-1999. The results confirm the existence of post-earnings announcement underreaction and provide new evidence of overreaction to highly unexpected past earnings surprises. Traditionally, unexpected earnings are defined by the difference between the last individual or consensus estimate and the actual, reported earnings per share, standardized by the share price. Other methodologies include scaling by the actual eps or the forecast made. While former is a measure of the market surprise, latter can be considered as the analyst's proportional estimation error. In order to capture the extent to which an earnings surprise is highly unexpected, some of the tests performed in this study use the standard deviation of consensus forecasts as a scaling factor. The results suggest that investors simultaneously exhibit shortterm underreaction to earnings announcements and long-term overreaction to past highly unexpected earnings. A potential explanation for the reported overreaction phenomenon is the representativeness bias. As I show, the overreaction and the later reversal is stronger for events, which exhibit a long series of similar past earnings surprises.