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

We examine how analysts' conflicting incentives to be either accurate or optimistic affect their choice to generate stock recommendations with rigorous valuation models or growth-based heuristics. Consistent with prior research the average analyst recommendation is negatively associated with rigorous valuation models and positively associated with growth-based heuristics, we document that these associations are weakest for the most accurate analysts and strongest for the least accurate analysts. We also find evidence consistent with consistency between recommendations and valuation models underlying the positive future returns from trading on the most accurate analysts' recommendations. Our results are consistent with reputation incentives to be accurate mitigating the use of optimistic growth-based models in generating stock recommendations.