Predicting Individual Analyst Earnings Forecasts

TLDR: In this paper, a model that predicts individual analyst forecasts of corporate earnings per share (EPS) using the change in the mean consensus forecast of other analysts since the date of the analyst's current outstanding forecast was proposed.

Abstract: In this study I propose and test a model that predicts individual analyst forecasts of corporate earnings per share (EPS) using the change in the mean consensus forecast of other analysts since the date of the analyst's current outstanding forecast; the deviation of the analyst's current forecast from the consensus forecast; and cumulative stock returns since the date of the analyst's current forecast. I find that these three variables explain about 38% of the variability in analyst forecast revisions. While there is evidence of a relation between changes in earnings expectations and price changes, virtually all of the explanatory power of my model arises from other analyst forecasts. Section 2 describes the data bases used and the sample selection process. Section 3 presents the model and method for predicting individual analyst forecasts. Section 4 reports the bias and accuracy of the predicted forecasts. Conclusions are in section 5.