Analyst bias and forecast consistency

Abstract

We contribute to the literature examining inconsistent analyst forecast revisions. While prior research suggests inconsistent analyst forecasts are less accurate, we find this result is sensitive to both the definition and direction of inconsistency. We define consistency relative to each analyst's own forecast error and find the relation between analyst characteristics and the likelihood of issuing an inconsistent forecast differs depending on the direction of the revision. Specifically, less accurate analysts are more likely to issue an inconsistent positive revision and have higher future accuracy than their consistent counterparts. Our results suggest that analysts derive a benefit (better future accuracy) from issuing upward forecast revisions in the period immediately following an earnings announcement regardless of the resulting forecast pattern.