Divergent Opinions and the Performance of Value Stocks

Divergence of opinions among investors, manifested in the dispersion of analysts' earnings forecasts, may play an important role in asset pricing. This article reports tests of whether disagreement can explain the cross-sectional return difference between value and growth (or “glamour”) stocks in the U.S. market over the 1983–2001 period. Consistent with the theoretical proposition that stocks subject to greater investor disagreement earn higher returns, the tests found value stocks to be exposed to greater investor disagreement than growth stocks. This finding suggests that the return advantage of value strategies is a reward for the greater disagreement about their future growth in earnings. Alternative multifactor asset-pricing tests supported the proposition that investor disagreement plays an important role in explaining the superior return of value stocks. Proponents of rational asset pricing and advocates of behavioral finance are engaged in an ongoing debate about the exact interpretation of the “value premium.” Rationalists argue that because value stocks are fundamentally riskier than growth (or “glamour”) stocks, the value premium is compensation for bearing risk. Behaviorists claim that value stocks produce superior returns because investors consistently overestimate the future earnings of growth stocks relative to value stocks. The essence of this argument is that investors are excessively pessimistic (optimistic) about value (growth) stocks because they tie their expectations for the future to past earnings. That is, investors make systematic errors in predicting future growth in earnings for value stocks, and investors' excessive pessimism about these stocks causes the superior performance of value stocks relative to growth stocks. This non-risk-based (behavioral) explanation of the value premium is known as the “extrapolation” or “errors-in-expectations” explanation, and it has been supported by several researchers.Recently, some researchers, using U.S. analyst earnings forecasts as a proxy for the market's expectations of future earnings, provided evidence against the errors-in-expectations view. Therefore, the observed abnormal return of value stocks on earnings announcement days is apparently not caused by surprise in the level of earnings but by some different mechanism. This mechanism—disagreement about future payoffs—is the focus of our article.Proponents of the rational explanation of the value premium have overlooked differences of opinion as a possible source of risk that could explain the value premium. Disagreement among investors is widely recognized, however, as a potential determinant of asset prices. The concept of heterogeneous beliefs about future stock payoffs has been introduced into a standard capital asset pricing model and shown to be a positively associated with future stock returns.In this framework, the superior future performance of certain stocks arises because not all investors possess completely accurate probability beliefs; heterogeneous expectations among investors matter in asset pricing because the opportunity set is partially unknown. When investors are uncertain about the true probability structure of stock return payoffs, they tend to hold different subjective opinions about the future of the stocks. When uncertainty about the future prospects of a stock is high, subjective beliefs will diverge, causing investors to demand high rates of return to invest in the stock.An alternative view is that the higher returns for stocks exposed to greater disagreement among investors arise because in imperfect capital markets, capital market equilibrium requires the simultaneous determination of asset prices and of the identity (that is, the opinions) of investors trading in each asset.Dispersion of opinion, then, may represent a unique source of risk, and its impact on prices should be compounded by the degree of disagreement. To examine this issue, we used dispersion in analysts' earnings forecasts as a proxy for investors' heterogeneous beliefs. We hypothesized that value (growth) stocks have greater (lower) exposure to dispersion in forecasts and, therefore, should earn a higher (lower) return.Our results are consistent with the investor disagreement explanation for the return differential between value and growth stocks. We found that the dispersion in analysts' earnings forecasts is considerably higher for portfolios composed of stocks with high book-to-market ratios (BV/MVs). We obtained similar results when we compared the extreme quintiles of stocks ranked on size. Small-capitalization stocks exhibited greater forecast dispersion than large-cap stocks. These results suggest that high-BV/MV stocks and small-cap stocks earn higher returns because there is greater disagreement among investors about the stocks' future payoffs.Our multifactor asset-pricing tests confirmed that investor disagreement, manifested in the dispersion of analysts' earnings forecasts, is a risk factor that is priced, together with other risk factors, in the determination of value stock and small-cap stock returns. These results are consistent with the view that investors require higher returns for stocks exposed to greater disagreement.