Analyst Forecast Revisions and Overconfidence

Abstract

This paper shows that analysts’ herding forecasts are accompanied by significant return reversals of 116 basis points per month, while anti-herding forecasts render reversals insignificant. These results are magnified among illiquid stocks and during high VIX months. Since analyst herding is specific to covered firms, we are also able to examine information spillover from covered to uncovered firms. Among uncovered firms, herding forecasts are accompanied by reversals of 118 basis points, but anti-herding forecasts render reversals insignificant. Our findings suggest that herding forecasts do not contain any private information, while anti-herding forecasts have a much higher informational content that can help reduce the information asymmetry between informed and uninformed investors.