Contrast Effects and Analyst Forecasts

Abstract

Contrast effects take place when decision makers unconsciously interpret a signal by contrasting it with the signal preceding it. Using analyst forecast revisions in response to earnings announcements on consecutive days as the setting, we find evidence of contrast effects when the analyst covers the announcing firms on both days. The effects are driven by scenarios where the analyst is relatively inexperienced, where the firms announcing earnings on day t-1 are large relative to the firm announcing earnings on day t, and where more than one firms announce earnings on day t-1. Unlike investors who contrast earnings news against the prior day’s earnings announcements from bellwether firms, analysts do not benchmark against these firms if they are not in the coverage portfolio. Additional analyses suggest that our findings cannot be explained by information transfer or analyst limited attention.