This action might not be possible to undo. Are you sure you want to continue?
There is an interesting contrast between the academic discipline of financial economics and the coverage the field receives on the nightly newscast. Whenever financial markets make it on the news the story is always accompanied by pictures of people engaged in wild activity, either shouting bids and asks on the trading floor, or talking on three phones simultaneously in a trading room away from the floor. The image one gets from the news is that financial markets are dominated by people. In contrast, a reading of a standard finance textbook, such as the excellent one by Brealey and Myers, can create the impression that financial markets are nearly devoid of human activity. There is great attention to methods of computing important numbers such as present values and rates of return, an analysis of risk and how it is priced, much discussion of how much a firm should borrow and how much it should payout in dividends (answer: it doesn't matter), and even a primer on how to price options. But virtually no people. Very little in the text would be changed if all the people in both the corporate sector and the financial sector were replaced by automatons. This observation leads to a question. Since we know that there are indeed humans involved in financial markets, are the markets any different because of their presence? Would the financial world be materially different if investors, traders, managers, and workers were all replaced by computer programs? The chapters in this book suggest that the answer is yes. People make a difference. Why has financial economics largely ignored the people in preference to the prices? There are several reasons. A glib response (at least partly true) is that there are better data on prices than people. Thanks to the people at CRSP and COMPUSTAT, financial economists have had data that are the envy of other social scientists. Seventy years of stock price data on computer tape are hard to resist. In contrast, there are very few xv
data available on the behavior of individual agents. Even if one wanted to study what the people are doing it would be hard. However, while lack of data has certainly been an issue, most financial economists have not really felt deprived. nti] recently, it was possible to hold the view that there was no reason to study the behavior of people when the prices were behaving so nicely. The financial world as described in Eugene Fama's 1970 efficient markets survey was one that had no urgent need for people. Markets were efficient, prices were unpredictable, and financial economists did not know how to spell the word anomaly. Of course, times have changed. As Fama says in his 1991 sequel, the issues have gotten "thornier," and he devotes a long section to evidence on return predictability. Indeed, the findings of the last two decades have been startling. Small firms, firms with low price-earnings ratios or low ratios of market price to book value of assets, and losing firms all appear to earn higher returns than we would expect. There are also surprising calendar effects, most notably in January. Finally, there is growing evidence that the capital asset pricing model 13 has no predictive power. That is, after controlling for size or price to book value, stocks with high f3's earn no more than stocks with low 13'5. These research findings, and the experience of the 1987 stock market crash, have made financial economists more cautious, If 13 does not matter, and stock prices can full 20 percent in a day without any news, can we really be sure that people are irrelevant? Theory provided additional excuses for ignoring the people in financial markets. Although most financial economists freely admitted that many participan ts in f nancial markets were far from rational (e. g., brokers, chartists, and portfolio managers) this was not thought to matter. One smart trader with enough money was said to be sufficient to keep markets in line. If all prices are set by the rational arbitrageurs, then financial markets with a mixture of rational and irrational traders are identical to a textbook market with only rational traders. This simple view is no longer tenable. A series of papers has investigated the theory of markets with mixtures of agents. The conclusions that emerge are complex, but it is fair to say that the conditions for irrational agents to be rendered irrelevant are quite special. Most of the time, the behavior of both rational and less than fully rational agents matters. This means that it is no longer possible to have complete confidence in the usual approach (which is to characterize optimal behavior and then assume this behavior is universal). We need to be concerned with how real investors actually behave, that is, behavioral finance. What is behavioral finance? The best way to define a field is by exam-
pie, and at this time the papers in this book represent the best of what I would call behavioral finance. The common thread in these papers is the combination of a concern with real world problems and a willingness to consider all explanations in the search for understanding. I think of behavioral finance as simply "open-minded finance". Sometimes, in order to find the solution to an empirical puzzle, it is necessary to entertain the possibility that some of the agents in the economy behave less than fully rationally some of the time. Any financial economist willing to consider this possibility seriously is ready to take a try at behavioral finance. Naturally, as in any book of this kind, there were difficult choices to be made in selecting the set of papers to be included. In this case, the process of selection was partially a democratic one. The members of the Russell Sage/Nationa1 Bureau of Economic Research working group on behavioral finance reviewed a tentative table of contents and helped cut the volume down to a feasible size. If your favorite paper does not appear here, please be assured that I wanted to include it but was outvoted. I have organized the chapters into six sections. Section 1 has chapters about "Noise," the term popularized by Fischer Black's presidential address with that title. Black discusses the role of noise in financial markets. He uses the term in many ways, but perhaps one way to think about noise is that it is the opposite of news. Rational traders make decisions on the basis of news (facts, forecasts, etc.). Noise traders make decisions based on anything else. The chapter by Bradford De Long, Lawrence Summers, Andrei Shleifer, and Robert Waldmann uses Black's noise concept in a theoretical model. They investigate a world populated by rational traders and noise traders, the latter displaying time-varying shifts in sentiment. These shifts in sentiment are shown to create a new kind of risk. Assets that are held widely by noise traders bear the risk that the noise traders will become less optimistic about the future, causing the price of the assets to fall. Rational investors then demand a premium to be willing to bear this risk. An interesting special case of this noise trader model applies in the case of closed-end funds. Closed-end funds have long been <t puzzle to efficient market theorists, since the shares in such a fund often sell for prices that differ considerably from the value of the assets the fund owns. These financial curiosities are investigated in the chapter by Charles Lee, Shleifer, and myself. We show that closed-end funds are held primarily by individua1 investors, and that the discounts on closed-end funds are correlated with the prices of other securities held by individual investors (e.g. small firms). The second section is devoted to the topic of volatility. It begins with the classic paper by Robert Shiller that asks a provocative question. Since the (detrended) present value of dividends has been essentially constant
for the last century, and stock prices are supposed to be a forecast of the present value of future dividend payments, why have stock prices varied so much? This paper has created a major controversy in the financial economics community." Suffice it to sa!' that when all the econometric smoke clears, the week of October 19th, 1987 makes Shiller's conclusions look very good. This point is reinforced by David Cutler, James Poterba, and Lawrence Summers-who ask, what moves stock prices? They show that many big price moves occur in the absence of important news, and conversely, economic news cannot explain a large portion of stock price changes. Of course, the question of excessive volatility is directly related to the issue of whether stock prices are equal to their "true" or fundamental value.f The chapter in this section by Lawrence Summers makes it clear how difficult it can be to test this proposition. Robert Shiller's second chapter in this section introduces the idea that fashions and fads may influence financial markets. It is an interesting commentary on the state of economics that Shiller finds it necessary to justify the idea that fashions are important determinants of behavior in other domains. Though economists themselves are not known to be particularly fashion oriented, one might have thought that they would have noticed that other people's behavior seems to change over time: hemlines go up and down, ties and lapels get wider and narrower, disco music comes and (thankfully) goes Shiller argues forcefully that fashions also influence financial markets. (This chapter also contains the most provocative sentence in the book-look for it.) The final chapter of this section, by Kenneth French and Richard Roll, is concerned with the question of whether stock markets might create their own volatility. They find that, not surprisingly, stock prices vary much more (per hour) during periods when the markets are open than when they are closed. Of course, this result might be explained by the fact that the markets are open during normal business hours when other news is being generated, so French and Roll use a controlled experiment inadvertently conducted by the New York Stock Exchange in 1968. In the second half of that year, the exchange was having trouble keeping up with the volume of trading and decided to close the markets for a series of Wednesdays. Since these Wednesdays were normal business days, French and Roll could examine whether the change in price from Tuesday's close to Thursday's opening
1. For an update. see Stephen LeRov. "Efficient Capital Markets and Martingales." Journal of Economic Literature. 2i. December 1989,151l3-1621. as well as Shiller's book. Marke! Volatilit!! (MIT Press, 1989). 2. This issue is also addressed in Lee, Shleifer, and Thaler's chapter on closed-end funds.
was less volatile when markets were closed, It was, suggesting, at least to this reader, that traders react to each other as well as to news, Section 3 is devoted to the topic of overreaction and underreaction. The first chapter, by Werner De Bandt and myself. was the result of an attempt to predict a stock market anomaly from the psychology of decision making. De Bondt and I were familiar with the work of Daniel Kahneman and Amos Tversky which showed that people have a tendency to make predictions that are not sufficiently regressive. That is, rather than being proper Bayesian decision makers, people tend to overweight recent information and underweight long-term tendencies (prior odds), We made the prediction, at the time outlandish, that stocks with extreme performance over some extended time period would display mean reversion. Indeed, we found that extreme losers outperformed the market, and extreme winners underperformed the market, though less so. Like Shiller's paper, this paper has proven to be controversial, with critics claiming that the results are explained by one of two factors: the small size of the losing firms, or their high degree of risk. These issues are taken up by Navin Chopra, Josef Lakonishok, and Jay Ritter in the next chapter. They find that even after controlling carefully for risk and size, losing firms earn excess returns, especially in the smaller deeiles. In Victor Bernard's chapter, a seemingly contrary set of results is discussed, that is, the apparent underreaction of stock prices to earnings announcements. The phenomenon, labeled post-earnings-announcement drift, is that when a firm makes an earnings announcement that is surprisingly good or bad the stock price tends to go up or down in the days following the next three quarterly earnings announcements. Bernard tries to make sense of these two surprising sets of results, In the final chapter in this section, Jeremy Stein takes the overreaction idea to another domain: options markets. He observes that stock price volatility is strongly mean reverting. This implies that the prices of short-term options should behave differently than the prices of long-term options. Specifically, when volatility increases, short-term option prices should increase more than their long-cerm counterparts. The evidence Stein presents is inconsistent with this implication of rationality. He finds that long-term options "overreact",
Section 4 contains a pair of papers that share an international perspective. The first, by Kenneth Froot and Jeffrey Frankel, addresses a wellknown anomaly in currency markets called the forward-discount bias. There is a simple proposition about exchange rates that implies that the difference between the spot rate and the forward rate is a prediction about the future movement of currency rates. The forward-discount bias refers to the fact that this prediction fails to hold, and, surprisingly,
currencies seem to move in the opposite direction from that predicted by the theory. In this chapter, Froot and Frankel investigate whether the bias can be attributed to risk. They make a convincing argument that it cannot, and suggest that the bias lies in expectations. In the other chapter, Kenneth French and James Poterba discuss a less-well-known anomaly that might be called the "invest-at-home bias". Basic diversification theory implies that investors can decrease risk while holding returns constant by investing in assets that are not highly correlated with their existing portfolio. One way to do so is to invest abroad. The anomaly is that even in small countries people tend to invest very heavily in their own country. A plausible behavioral interpretation is that investors are more confident about investing domestically, even if both foreign and domestic markets are efficient and investors are no better at picking winning stocks at home than abroad. Section 5 is devoted to corporate finance. The first chapter, by Hersh Shefrin and Meir Starman, is concerned with the major puzzle in the field: why do firms pay dividends? We know that in a world without taxes dividend policy would be irrelevant, but when dividends are taxed at a higher rate than capital gains, stockholders should complain if a firm pays cash dividends. Instead, stockholders seem to do just the opposite-they complain when dividends are cut. Shefrin and Statman offer a behavioral explanation using the conoepts of mental accounting and self-control. In their theory, firms pay dividends because stockholders like theml Andrei Shleifer and Robert Vishnv take up the important topic of short-term thinking in corporations. Do financial markets, and especially arbitrageurs, help or hurt? Shleifer and Vishny show that rational behavior of arbitrageurs may contribute to the problem, rather than solve it, because arbitrage is more costly for long-term assets than for short-term assets. Richard Roll's chapter is concerned with explaining the puzzling phenomenon of corporate takeovers. The puzzle is that while the target firm stockholders do very well when their firm is purchased, stockholders in the acquiring firm do not appear to make any money. Why do firms make takeovers if they cannot expect to make a profit? Roll offers the "hubris" hypothesis as an answer. Put simply, managers of acquiring firms, flush with cash from recent successes (perhaps due to luck), are convinced that they are "good managers" and will be able to run the target firm better than the current managers. This is consistent with a massive literature in psychology showing that individuals tend to be overconfident.P The
3. See, for example, Sarah Lichtenstein, Baruch Fischhoff, and Lawrence Phillips, "Calibration of Probabilities: The Stale of the Art to 1980," in Daniel Kahnernan, Paul Slovic, and Amos Tversl .. eds .. Judgment Under Uncertainty: Heuristics and Biases (Cambridge ry, University Press, 1982).
final chapter in this section, by Jay Ritter, concerns the curious case of the pricing of initial public offerings, also called new issues. IPOs are interesting because they seem to come in waves and are initially underpriced. Ritter finds that over longer periods IPOs are actually overpriced, and offers an explanation in which investors are periodically overoptimistic about the future of young companies, a bias that firms exploit. The final section of the book is about individual behavior, and begins with Robert Shiller's chapter on popular models. Although the behavior of people has been an important theme in the papers in this book, people remain in the background most of the time. In this chapter, Shiller takes the direct approach of asking real investors what they think. While economists may prefer to study actual behavior rather than responses to survey questions, surely we can learn something from asking actual market participants some simple questions. Shiller reports here three sets of surveys on three interesting topics: the crash of 1987, the real estate boom in California in the late 1980s, and the hot IPQ market mentioned above. Shiller tries in these surveys to discover what models of price dynamics actual investors seem to have. One of his conclusions is that participants think that "investor psychology" is important, particularly in the stock market. Shefrin and Starman's chapter is concerned with the relation between price and volume. In many markets, including stock markets and especially real estate markets, when prices fall volume seems to go down. In stock markets this is particularly surprising, since there are tax advantages to selling losers. Shefrin and Statman offer a mental accounting explanation of the phenomenon that is based on the premise that investors do not like to close a mental account with a loss. The last chapter of the book, by Lawrence Ausubel, describes the fascinating market for credit cards. Although there are over 4,000 banks that issue credit cards in the U. S., the market appears to be strangely noncompetitive. In particular, the interest rate that a bank charges does not seem sensitive to the bank's cost of funds. Ausubel offers an explanation based, in part, on individuals selecting a credit card as if they would never fail to repay their balance on time, though their objective chance of incurring interest charges is actually quite high. What are the main conclusions to be drawn from this collection? I would suggest two. First, it is possible to do good economic research even if the assumption of universal rationality is relaxed. Second, we can understand much more about the behavior of markets, even financial markets, if we learn more about the behavior of the people who operate in these markets.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.