Post-earnings announcement drift : evidence from Turkey

We study the post-earnings announcement drift (PEAD) anomaly and its determinants in Borsa Istanbul using quarterly earnings announcements and three different surprise measures. We find evidence supportive of the existence of PEAD in the Turkish stock market. Sorting stocks each quarter into three quantiles according to an earnings surprise measure based on past earnings, we find both statistically and economically significant difference of 2.9% in average cumulative abnormal return between high earnings surprise firms and low earnings surprise firms in the 60 days following the announcement. The hedge returns obtained using analyst forecast based surprise measure as well earnings announcement return based surprise measure are also significant at 2.9% and 2.2% respectively. The positive relation between post-announcement abnormal returns and the surprise in earnings remain significant after controlling for an extensive list of explanatory variables. Among these explanatory variables, we find that firm size has a negative impact on the magnitude of this positive association between surprise and subsequent equity returns. Finally, using two separate approaches we show that multifactor models fail to explain the return differentials following earnings announcements.