MARKET REACTIONS TO CORPORATE EARNINGS SURPRISES IN THE FACE OF DIFFERING SIGNALS FROM ANALYSTS’ RECOMMENDATIONS

Abstract

In this paper, we investigate the reactions of the market to earnings surprises in the face of differing signals from analysts' recommendations. Using two proxies for analysts? recommendations namely the level and change in analysts' consensus recommendations, we examine market reactions to both the signs and magnitude of earnings surprises unconditional as well as conditional on those proxies. Consistent with Ho and Sequeira (2007) and Lopez and Rees (2001), we find the unconditional market reactions to be stronger for positive vis-à-vis negative earnings surprises. This holds true for both the sign and the magnitude of the surprises. More interestingly, we generally observe that market reactions to earnings news is at least as strong when conditioned on a confirmatory vis-à-vis contradictory signals from the level as well as change in analysts? consensus recommendations. There are, however, two instances where the signals from earnings surprises appear to dominate signals from analysts? recommendations. In the first, controlling for the type of earnings surprises, the level of consensus recommendation does not seem to make a significant difference in market reactions to either positive or negative earnings surprises. In the second, controlling for the change in analysts? consensus recommendations, the asymmetric market reaction remains biased towards positive earnings surprises. These results are consistent across all the five short event windows within three days of the official earnings announcement date. They are also robust to non-synchronous trading effect, size effect bias and biases induced by I/B/E/S rounding procedure in reporting stock split-adjusted earnings data.