Investor (Mis)Reaction, Biased Beliefs, and the Mispricing Cycle

Abstract

We construct a new measure that captures the disparity between the market reaction to earnings information and the earnings surprise (“Return-Earnings Gap”, “REG”). High REG positively predicts analyst forecast errors and firm mispricing (overvaluation). Analyst forecast errors are particularly increased when REG provides confirming information. In turn, REG is positively predicted by analyst forecast errors and higher mispricing, leading to a continuation of firm overvaluation over several quarters. Overall, our results reveal how the market’s (mis)reaction feeds back into the belief formation of analysts and investors, leading to a slow correction of firm mispricing. A simple structural model explains the predictive power of REG for analyst forecast errors and the build-up of mispricing by incorporating the dynamic expectation formation between
different agents.

Keywords: Investor beliefs, analysts, mispricing, misreaction

JEL Classification: G00, G12, G14, G40, G41