Implications of Analysts’ Consensus Forecast Error for Measurement of Market Expectation and Predictability of Stock Returns

Abstract

In recent studies, the analysts’ consensus forecasts are widely used as a proxy for the unobservable market’s consensus expectation of future earnings. As prior studies indicate, the analysts’ consensus forecasts measure the true underlying construct with errors and the errors may vary cross-sectionally. Based on these prior findings, this study examines the implications of the cross-sectional variation of the errors in measuring the market expectation of future earnings for the predictability of future stock returns. For this end, this study first uniquely provides a framework to estimate the cross-sectional variation of the errors in measuring the market expectation of future earnings embedded in the analysts’ consensus forecasts based on the absolute analysts’ consensus forecast errors and determinants. The findings of this study indicate that incorporating information about the cross-sectional variation of the errors can generate better specified empirical test models and improved predictions of future stock returns based on the forward E/P constructed with the mean analysts’ consensus forecasts. The results of this study also have the implications for practitioners (i.e., money managers). This study reports that the spread of the hedge returns from the forward E/P investment strategy is the largest when the errors are estimated to be the smallest. To conclude, the implications of the findings in this study for other accounting research areas are discussed.