Projected Earnings Accuracy and the Profitability of Stock Recommendations
We find that while analyst characteristics have economically small predictive power for the accuracy of earnings forecasts, they have economically significant predictive power for the profitability of stock recommendations. Analyst characteristics can be used to identify ex ante more profitable stock recommendations and therefore they are helpful for identifying profit opportunities. Taking an investor-oriented, calendar-time approach, we project the firm-specific earnings accuracy of analysts based on their individual characteristics. We show that analysts with higher projected earnings accuracy issue significantly more profitable recommendations than analysts with lower projected earnings accuracy. The long portfolio based on the recommendations of the analysts within the highest projected accuracy quintile earns excess returns of up to 7.53% per year (before transactions costs) in the 1994 to 2007 time period. Analysts with characteristics that are related to higher earnings accuracy also issue more profitable stock recommendations.
Link al Paper