Dispersion in Analysts' Earnings Forecasts and the Cross Section of Stock Returns: What is the Driving Factor?

Abstract

Recent studies have documented that stocks with lower dispersion in analysts' earnings forecasts earn higher future returns. The economic interpretation of this observation is that dispersion in earnings forecasts represents differences of opinion among analysts. That is, under short-sale constraints, stocks prices reflect the valuation of optimistic investors and such optimism tends to be stronger when investors observe dispersed opinions about a stock's valuation. This interpretation is based on the economic meaning of the standard deviation of earnings forecasts in the numerator of the dispersion measure, while treating the absolute value of the mean earnings forecast in the denominator of the dispersion measure as just a scalar. However, contrary to the differences-of-opinion explanation, our results show that the predictability of dispersion in analysts' earnings forecasts on future stock returns mainly comes from the denominator effect rather than from the numerator effect of the dispersion measure. We also find that dispersion in analysts' earnings forecasts has the strongest predictive power among the most undervalued stocks. Our results appear to cast doubt on the differences-of-opinion explanation for the negative relationship between dispersion in analysts' earnings forecasts and future stock returns.