Modeling Analyst Behavior

Abstract

Analyst earnings forecasts have formed the basis for many well-known portfolio management strategies. Earnings forecasts over a considerable period of time form large data bases scrutinized by asset managers and academic researchers alike. There has been almost no analysis of how these forecasts are derived. The analysis here suggests a classification of analysts into three categories-leaders, followers, and rebels - and aims to differentiate the independent opinion-makers (in theory, both leaders and rebels) from the followers. Opinion-makers may be influential for two distinct reasons: Their earnings forecast revisions may be accurate and provide significant explanation of subsequent asset returns, or the market may weight their earnings revisions highly out of respect or their celebrated status.