Analyst Forecast Revision Consistency and Bias in Earnings Forecast Revisions

Abstract

We address whether analysts bias earnings forecast revisions and convey the bias using forecast revision consistency, i.e., the extent to which analyst reports with earnings forecast revisions include stock recommendation and target price revisions consistent in sign with the earnings forecast revisions, the sign of which is the sign of earnings forecast revision. Incentives to curry favor with management can constrain analysts to issue biased earnings forecast revisions. As predicted, we find that forecast revision consistency is (i) positively associated with the market reaction to the analyst report; (ii) positively associated with consensus analyst forecast error (AFE), i.e., actual earnings minus the consensus of the analysts’ forecasts; and (iii) negatively associated with earnings announcement returns, incremental to AFE. Absolute forecast revision consistency is smaller (larger) when firms eliminate (initiate) guidance. These findings are consistent with analysts using forecast revision consistency to convey information about bias in their earnings forecast revisions.