Contrarian investment.

For over 30 years, the efficient markets hypothesis (EMH) has remained the central proposition of financial economics. The EMH states that, as an empirical matter, prices at which securities trade in liquid financial markets are equal to their fundamental values, given by the expected present values of the cash flows accruing to these securities. In other words, the stock market prices securities at their fair values. Although the EMH flies in the face of the conventional wisdom that astute analysts can beat the market, it has withstood many empirical challenges for decades, becoming a textbook wisdom for most economists. In particular, the implication of EMH that investment strategies based on public information, including those practiced by mutual funds, cannot beat the market, has survived hundreds of tests.

In recent years, a new set of challenges to the EMH has appeared, based on some very old ideas about contrarian investment. These ideas, dating back to Graham and Dodd,(1) state that investing in value stocks - defined as stocks with low prices relative to some measures of their current fundamentals, such as earnings or dividends - is more attractive than investing in growth stocks, those with high prices relative to measures of fundamentals. Although several papers in the 1980s supported the superior returns from contrarian investment strategies based only on publicly available information, the most celebrated study came from the University of Chicago, the cradle of the EMH.

In 1992, Eugene F. Fama and Kenneth R. French reported that, between 1963 and 1990, stocks of companies with high ratios of book values of assets to market price earned higher returns than stocks of companies with low book-to-market (BM) ratios.(2) They found that the spread in returns between portfolios of stocks with high (top 10 percent) and low (bottom 10 percent) BM ratios was on the order of 10 percent per year. Despite finding this enormous benefit to investing in high BM stocks, Fama and French did not interpret the evidence as contradicting the EMH. Rather, they argued that the high BM stocks in some special ways might be riskier than the low BM stocks. According to them, this difference in risks, as measured by the difference in BM ratios, explains the difference in average returns.

These empirical findings have stimulated a great deal of further work, including our research with Josef Lakonishok and Rafael La Porta. With Lakonishok, we have looked at a variety of contrarian strategies from 1968 to 1990, and found that there is nothing special about BM as a way to identify value and growth stocks.(3) Many other ratios of price to a measure of fundamentals, such as the price-to-cash-flow ratio, and the price-to-earnings ratio, are also good predictors of returns, generating superior performance of nearly 10 percent per year for value relative to growth stocks. Moreover, all of these ratios reflect the fact that the value stocks tend to be those of companies with poor past performance, whereas the growth stocks are those of firms with good past performance. Thus we confirmed the contrarian wisdom that investing in value stocks - however measured - on average has produced superior returns for investors.

We then went on to propose an interpretation of this evidence that is not consistent with the EMH. We argued that value stocks are out of favor with investors, who extrapolate poor past performance too far into the future. Similarly, the excellent past performance of growth stocks is extrapolated too far into the future by investors. Contrary to such extreme expectations of investors, future performance of value and growth stocks is not nearly as different as it was in the past. In fact, we presented some evidence showing that the earnings performance of value stocks tends to improve, and of growth stocks tends to deteriorate; this is consistent with the view that the market over-reacts to past performance and mis-prices these securities. Thus we attributed the superior performance of contrarian investment strategies not to risk, but to inaccurate expectations of market participants, and the resulting inefficient pricing. In addition, we found no evidence that value stocks are especially risky; these stocks do especially well relative to growth stocks when the market goes down, as well as in recessions.

The empirical evidence bearing on the question of why contrarian investment strategies pay off continues to pour in. La Porta directly tested the idea that inaccurate expectations, rather than risk, explain the evidence.(4) By looking at analysts' forecasts of long-term earnings growth of different stocks, he divided firms into those with extremely bullish forecasts and those with extremely bearish forecasts. La Porta found that firms with bearish forecasts earn higher future returns than firms with bullish forecasts. Analysts subsequently revise downward their growth forecasts for firms they are bullish about, and revise upward their growth forecasts for firms they are bearish on. Not surprisingly, stocks with bullish forecasts tend to be the growth stocks (using the earlier definitions), and stocks with bearish forecasts tend to be the value stocks. La Porta thus provides direct evidence that extreme - and unjustified - pessimism among some investors can explain the current underpricing, and therefore the superior future returns, of value stocks.

With La Porta and Lakonishok, we propose yet another test of the alternative explanations of the returns to contrarian investment.(5) Specifically, if the EMH interpretation of the evidence is correct, and the market on average does not make mistakes about value and growth stocks, then the market should not be surprised, on average, when these companies announce their earnings. In contrast, if the market overreacts to the past performance of value and growth stocks, then it should be disappointed by the earnings of glamour stocks when they are announced, and pleasantly surprised by the earnings of value stocks. A look at the earnings announcements of different companies then perhaps can distinguish the two interpretations. The evidence is not favorable to the EMH: the average return of glamour stocks when their earnings are announced is negative, and that of value stocks is positive. This event study evidence thus favors the extrapolation hypothesis over the EMH.

Of course, this is not the end of the story, and much remains to be done before the dust settles. Although some studies have confirmed the superior performance of value stocks in the United States prior to the Fama-French period, and others found similar results in other countries, doubts about data snooping continue to linger. Many questions remain unanswered on exactly how investor expectations are being formed, and what information causes investors to revise their expectations. Nor is it entirely clear why more investors have not shifted into value strategies, although it is possible that a typical institutional investor does not have a long enough horizon to wait for value to pay off with a high probability. All this research, we hope, will shed new light on the validity of the EMH.

But it can do more than that as well. For example, this research has begun to illustrate the many anomalies concerning the poor performance of professional money managers, who tend not only to fail to beat the market, but actually underperform it, on average, by a significant amount.(6) We conjecture that one possible reason for the poor performance of institutional investors is their preference for holding growth stocks in their portfolios, stemming from the need to impress their investors with a good-looking portfolio. Of course, holding growth stocks can cost dearly in returns. In some of our ongoing research, as well as that by others, indeed we are finding that some of the poor performance of professional investors can be tied to their overexposure to the growth stocks. In this way, our research may illuminate the process by which the pension wealth is managed, as well as the controversies about the EMH.

1 B. Graham and D. L. Dodd, Security Analysis, New York: McGraw-Hill, 1934.

2 E. F. Fama and K. R. French, "The Cross-Section of Expected Stock Returns, "Journal of Finance 47 (1992), pp. 427-465.

3 J. Lakonishok, A. Shleifer, and R. W. Vishny, "Contrarian Investment, Extrapolation, and Risk," NBER Working Paper No. 4360, May 1993, and Journal of Finance 49 (1994), pp. 1541-1578.

4 R. La Porta, "Expectations and the Cross-Section of Expected Stock Returns, "NBER Working Paper, forthcoming.

5 R. La Porta, J. Lakonishok, A. Shleifer, and R. W. Vishny, "Good News for Value Stocks: Further Evidence on Market Efficiency," NBER Working Paper, forthcoming.

6 J. Lakonishok, A. Shleifer, and R. W. Vishny, "The Structure and Performance of the Money-Management Industry, "Brookings Papers on Economic Activity: Microeconomics (1992), pp. 339-391.