Are financial analysts' forecasts of corporate profits rational?

This paper develops generalized method-of-moments tests for the rationality of earnings per share forecasts made by individual stock analysts. We fail to reject the hypothesis of rationality as long as we take into account two complications: (1) the correlation in a given period of analysts' forecast errors in predicting earnings for firms in the same industry and (2) discretionary asset write-downs, which affect earnings but are intentionally ignored by analysts when they make earnings forecasts. Our results challenge earlier work by De Bondt and Thaler and by Abarbanell and Bernard that found irrationality in analysts' forecasts.