Divergence of Opinion and Equity Returns under Different States of Earnings Expectations

In this paper, we show that divergence of opinion trades at a discount when analysts' earnings forecasts are optimistic and at a premium when analysts' earnings forecasts are pessimistic. Our results suggest that investors tend to exaggerate the quality of their foresight and invest in low dispersion stocks when earnings expectations are optimistic (i.e., sure winners) and avoid low dispersion stocks when earnings expectations are pessimistic (i.e., sure losers). In sharp contrast with Miller's view that high divergence of opinion leads to overvaluation, we find that overvaluation occurs when divergence of opinion is low and analysts' earnings predictions are optimistic. This is also confirmed during the 1998-2000 period, which is generally regarded as a period of extreme optimism about future stock return payoffs. These findings suggest that when pessimists are out of the market, the lower the disagreement over a stock's prospects, the greater is the upward bias in its price. That is, low dispersion in analysts' forecasts along with optimistic earnings forecasts is a salient stock characteristic that fosters overvaluation. When analysts' forecasts are pessimistic low dispersion in analysts' forecasts reverses this valuation pattern. Our results are robust to the book-to-market and leverage effects, and to the use of alternative short sale constraints.