Term structure of analyst forecast dispersion and future stock returns

Abstract

This study investigates the relationship between patterns of multiperiod analyst forecast dispersion (defined as the “dispersion term structure”) and future abnormal stock returns. I expect the slope of a firm’s dispersion term structure to be negatively related to its future abnormal stock returns because the earnings uncertainty that investors perceive for a firm may differ from the underlying uncertainty indicated by the dispersion term structure. This difference may arise for two reasons. First, investors may disregard information regarding the dispersion term structure when they believe that analysts are not credible. Second, investors’ limited attention span may prevent them from fully comprehending the implications of a firm’s dispersion term structure. However, I expect the information gap between analysts and investors to narrow as investors digest information regarding earnings uncertainty reflected by the dispersion term structure and the stock price reaction to subsequently follow. I estimate the slope of a firm’s dispersion term structure by regressing the dispersion in the earnings forecast for one quarter on the dispersion in the earnings forecast for the previous quarter. Consistent with my expectation, I find that the slope of the dispersion term structure is negatively related to future abnormal stock returns. This result indicates that investors do not fully understand the implications of the dispersion term structure for future stock returns and that they hence adjust their own positions on a delayed basis. I also observe that this negative relationship is weaker when analysts are more credible and when firms have a higher proportion of dedicated institutional investors. The results are robust to a battery of sensitivity tests.