Summary and Conclusions

Abstract

The forecasting of earnings per share, eps, is a most important input to an investment strategy. There is a tremendous literature regarding forecasting of corporate eps and whether the forecasts are more accurate than a random walk or a random walk with drift. Much of the literature can be summarized as follows: (1) analysts’ forecasts are not statistically different from a random walk with drift model; that is, analysts’ forecasts can be approximated with a first-order exponential smoothing model forecast; (2) analysts’ forecasts are biased; analysts’ forecasts are optimistic; (3) analysts’ forecast revisions and the direction of their revisions are more highly correlated with stock returns than earnings forecasts themselves; (4) earnings forecasts are highly statistically significant in forecasting total stock returns; (5) earnings forecasts, revisions, and direction of revisions can be combined with fundamental data, such as earnings, book value, cash flow, sales, these variables relative to their histories, and price momentum strategies to identify mispriced stocks; (6) smaller capitalized stocks are more mispriced than larger capitalized stocks; and (7) international and global stocks are more mispriced than the US stocks.