Do errors in expectations explain the cross-section ofstock returns?

Abstract

Value stocks have historically outperformed growth stocks in most of the major international markets. Many researchers attribute this phenomenon to overly optimistic (pessimistic) expectations of investors for growth (value) stocks. In this paper, we use professional analysts' earnings forecasts from Japan to test this errors-in-expectations hypothesis. We compare the magnitude of the forecast errors, the proportion of optimistic and pessimistic forecasts, and the likelihood of downward forecast revisions, across growth and value stocks. In contrast to the predictions of the hypothesis, we do not find any evidence that earnings forecasts are systematically more optimistic for growth than for value stocks. Our results also suggest that the alleged correlation between book-to-market value, a common measure of growth, and forecast errors is the result of a measurement bias in computing the magnitude of the latter variable.