The Use of Earnings Forecasts in Stock Recommendations: Are Accurate Analysts More Consistent?

Abstract

We examine how analysts’ incentives to build their reputation through accurate forecasting changes the relative association between analyst recommendations and rigorous valuation models versus valuation heuristics. Controlling for the firm-specific difficulty of valuation, we show that the recommendations of the most accurate forecasters within each industry have a significantly higher correlation with rigorous valuation models and a significantly lower correlation with valuation heuristics than their less accurate peers. Our results are robust to potentially confounding firm-specific effects using a within-firm sample design, a changes analysis, and for short-term, long-term and combined measures of forecast accuracy. Consistent with reputation building, we find that the recommendations of “All-Star” analysts and accurate forecasters have similar associations on rigorous and heuristic valuation approaches. Our results are consistent with the incentive to build a good reputation mitigating the influence of factors other than identifying mispriced securities, using fundamental analysis, on stock recommendations.

Keywords: Forecast accuracy, Fundamental valuation, Stock recommendations, Analyst reputation

JEL Classification: G12, G14, G29, M41