There has been a marked increase in debate surrounding the issue of the cognitive foundations of economic behaviour in recent years. This debate seeks to explain the determinants of a variety of activities such as forecasting economic variables, perception and decision under uncertainty and communication in interactive contexts. This volume contains contributions from leading scholars in their respective fields. Themes covered range from behavioural finance to neuroscience. Under the impressive editorship of Dimitri, Basili and Gilboa, this book will be of benefit to all those interested in the 'intersection between cognitive sciences and economics as well as economic theorists.
Nicola Dimitri is Professor of Political Economy at the University of Siena, Italy. Marcello Basili is Associate Professor of Economics at the University of Siena, Italy. Itzhak Gilboa is Professor of Economics at Tel Aviv University, Israel, and Fellow
of the Cowles Foundation for Research in Economics, Yale University, US.
Routledge Siena Studies in Political Economy
The Siena Summer School hosts lectures by distinguished scholars on topics characterized by a lively research activity. The lectures collected in this series offer a clear account of the alternative research paths that characterize a certain field. Different publishers printed former workshops of the School. They include: Macroeconomics A survey of research strategies Edited by Alessandro Vercelli and Nicola Dimitri Oxford University Press, 1992 Intemational Problems of Economic Interdependence Edited by Massimo Di Matteo, Mario Baldassarri and Robert Mundell Macmillan, 1994 Ethics, Rationality and Economic Behaviour Edited by Francesco Farina, Frank Hahn and Stefano Vannucci Clarendon Press, 1996 The Politics and Economics of Power Edited by Samuel Bowles, Maurizio Franzini and Ugo Pagano Routledge, 1998 The Evolution of Economic Diversity Edited by Antonio Nicita and Ugo Pagano Routledge, 2000 Cycles, Growth and Structural Change Theories and empirical evidence Edited by Lionello F. Punzo Routledge, 2001 General Equilibrium Edited by Fabio Petri and Frank Hahn Routledge, 2002 Cognitive Processes and Economic Behaviour Edited by Nicola Dimitri, Marcello Basili and Itzhak Gilboa Routledge, 2003
Cognitive Processes and Economic Behaviour
Edited by Nicola Dimitri, Marcello Basili and Itzhak Gilboa
chapters, the contributors Typeset in by Times New Roman by Newgen Imaging Systems (P) Ltd, Chennai, India Printed and bound by Gutenberg Press Ltd, Malta All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording~ or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Cognitive processes and economic behaviour / [edited by] Nicola Dimitri, Marcello Basili, and Itzhak Gilboa. p. cm. - (Routledge Siena studies in political economy) Includes bibliographical references and index. 1. Economics-Psychological aspects. 2. Economic man. 3. Cognition. 4. Emotions and cognition. 5. Decision making. I. Title: Cognitive processes and economic behaviour. II. Dimitri, Nicola. III. Basili, Marcello, 1959- IV. Gilboa, Itzhak. V. Series. HB74.P8C64 2003 330'.01 '9-dc21
To the memory of Michael Bacharach
List offigures List oftables List of contributors Introduction
ix x xi xii
1 Behavioral finance and markets GUR HUBERMAN
2 A non-expected glance at markets: financial models and
MARCELLO BASILI AND FULVIO FONTINI
3 On the existence of a "complete" possibility structure
4 Correlated communication
5 A survey of Rule Learning in normal-form games
DALE O. STAHL
6 Framing and cognition in economics: the bad news and the good
7 Language and economics
BARTON L. LIPMAN
8 Learning from cases: a unified framework
ITZHAK GILBOA AND DAVID SCHMEIDLER
9 Some elements of the study of language as a cognitive capacity LUIGI RIZZI
10 Rationality, habits and freedom
11 For a "cognitive program": explicit mental representations for Homo Oeconomicus (the case of trust)
12 The structured event complex and the human prefrontal cortex: the economic brain
1.1 1.2 5.1 5.2 5.3 6.4 11.1 11.2 12.1
Stock price of Royal Dutch relative to Shell ENMD closing prices and trading volume Payoff matrices for horse races Evidence of non-best-response behavior Games for testing reciprocity-based cooperation The vase-faces illusion The decision process The different emotional impacts Key components of an SEC mapped to PFC topography
5 7 50 54 57 69 186 187 220
1.1 5.1 5.2 5.3 6.1 6.2 10.1 10.2
Returns, excess returns, trading volume, relative trading volume, and corresponding p-values for BMY In-sample performance measures Out-of-sample performance measures Parameter estimates of enhanced Rule Learning models PD in standard form The game of Hi-Lo Transition matrix for soft-drink choices Transition matrix for automobile purchases
9 52 53 58 67 70 161 162
Michael Bacharach, (formerly) Department of Economics, University of Oxford, Oxford, UK Marcello Basili, Department of Political Economy, University of Siena, Siena, Italy Adam Brandenburger, Stem School of Business, New York University, New York, US Cristiano Castelfranchi, Department of Communication Sciences, University of Siena, Siena, Italy & Institute of Cognitive Sciences and Technologies, CNR, Roma, Italy Nicola Dimitri, Department of Political Economy, University of Siena, Siena, Italy Fulvio Fontini, Department of Economic Sciences, University of Florence, Florence, Italy Itzhak Gilboa, School of Economics, Tel Aviv University, Tel Aviv, Israel & Cowles Foundation, University of Yale, New Haven, US Jordan Grafman, Cognitive Neuroscience Section, National Institute of Neurological Disorders and Stroke National Institutes of Health, Bethesda, US Gur Huberman, Columbia Business School, Columbia University, New York, US Barton L. Lipman, Department of Economics, Boston University, Boston, US Luigi Rizzi, Department of Communication Sciences, University of Siena, Siena, Italy David Schmeidler, School of Mathematical Sciences, Tel Aviv University, Tel Aviv, Israel & Department of Economics, Ohio State University, Columbus, US Dale Stahl, Department of Economics, University of Texas, Austin, US Patrick Suppes, Department of Philosophy, Stanford University, Stanford, US
Throughout most of the twentieth century, the official position of mainstream economics ignored cognition. The revealed preferences paradigm held that economics could and should be based on observable choice behavior alone, and that any other type ofdata would be at best irrelevant, if not meaningless and detrimental to the progress of science. This position is perhaps best epitomized by Samuelson's (1937) canonical contribution. But it had its roots in the preceding seven decades or so. To understand the popularity of the revealed preference paradigm, as well as its decline in the late twentieth century, one might wish to distinguish among three types of explanations, relating to science, to the philosophy of science, and to the sociology of science. On the scientific, substantial level, the 1870s were crucial times. At the beginning of this decade, Marshall (1890), Menger (1871), and Walras (1873) suggested that marginal utility is key to understanding consumer demand. Some two hundred years after the invention of the calculus, economists highlighted the role of the derivative of the utility function, rather than the level of the function itself, as a determinant of demand. This insight resolved Aristotle's diamond-water paradox and paved the road to Marshall's theory of determination of price by both supply and demand. Marginal utility is not a behavioral concept. Indeed, the intuition behind the assumption of decreasing marginal utility has to do with cognition and affect. 1 However, when coupled with Walras's general equilibrium setup, the so-called marginalism led to the realization that only ratios among derivatives, namely marginal rates of substitution, mattered for the determination of demand. It followed that the utility function was only ordinal, and that direct measurement of utility or of marginal utility was not related to behavior. It was at the beginning of the twentieth century that Pareto (1916) suggested a concept of efficiency that was to change the course of normative economics. Pareto pointed out that this concept did not resort to a cardinal utility function or to interpersonal comparisons of utility. This sufficed for economics to have a nontrivial normative question that, in principle, did not require data beyond revealed preference. It so happened that Pareto efficiency was also the most demanding normative criterion about which the profession was in agreement. The result was
that economics chose to focus on this criterion, and many economists view it as the normative forefront of scientific research, arguing that the discipline should not attempt to go beyond it. Thus, by the time that neoclassical economics was taking center stage, it appeared that cognition does not matter. Economists observed that, for descriptive and normative purposes alike, they only had any use for behavior data. The philosophy of science also had an impact on the course of economics. The rise of logical positivism at the end of the nineteenth century and the beginning of the twentieth, culminating in Carnap's (1923) formulation of the Received View, provided a template for scientific disciplines. In particular, all theoretical concepts were to be based on observable and measurable data. Applying the Received View to economics, one would have to ask how utility is measured before one could use the concept. It was probably rather natural to identify the notion of observability with behavior, and to conclude that any economic theory that cannot be firmly based on observed choice behavior was meaningless and therefore more likely to hinder scientific progress than to facilitate it. The sociology of science cannot be ignored as well. Economics was traditionally considered the branch of social science most closely related to the natural sciences. At the beginning of the twentieth century it found itself at a crossroad. It could choose to associate with the "hard" sciences, relying on measurable data and on mathematical theories, or to be a "soft" science, closer to psychology and sociology. As Loewenstein suggests, it is possible that the rise of psychoanalysis, fiercely attacked for its unscientific nature (see Popper (1934)), made psychology a dubious discipline to associate with, and pushed economics to the arms of the "hard" sciences. It is also possible that the dual role of Marxism, as a socio-economic theory and as a political agenda, made sociology an uneasy partner for western economists. Finally, economists were probably also attracted by the sheer beauty and parsimony of the revealed preference paradigm, and by the modernist promise of harnessing mathematical tools for the understanding of human behavior. In conclusion, it was a conjunction of purely scientific reasons with philosophical and sociological ones that made mainstream economics focus on behavior as the sole legitimate and meaningful source of data, and use cognition as anything beyond a vague source of inspiration. By the end of the twentieth century, however, the validity of the revealed preference approach was questioned on several fronts. In fact, none of the- reasons for its rise to a status of a dominant paradigm could sustain it any longer in this role. On the scientific level, various cracks began to appear in the revealed preference view of the world. As a descriptive theory, economics has not proven capable of providing the type of precise predictions of individual agents' behavior, or of market equilibrium. Moreover, the most basic assumptions of rationality came under forceful attack by Kahneman and Tversky (see Kahneman and Tversky (1979, 1984), and Tversky and Kahneman (1981)). It became clear that economics couldn't continue to ignore cognition based on the argument that it is successful enough without it. The dissatisfaction with economics as a descriptive science led economists to ask what went wrong and where more help could be found.
Cognition is one field, which has been waiting patiently for almost a century, and may finally offer its help in predicting economic phenomena. On the normative level there has always been some degree of dissatisfaction with the revealed preference paradigm. Throughout the twentieth century authors within the economics profession pointed out the fundamental inadequacy of choice data in determining people's well-being, and in accordingly choosing economic policies. Thus, Duesenberry (1949) pointed out that well-being is determined by an individual's relative standing in society, Foley (1967) suggested the concept of envy-free allocations, and so forth. The philosophy of science has greatly changed since the 1930s. The Received View came under numerous attacks even when applied as a guideline for the progress of the natural sciences (see Quine (1953), Hanson (1958)). It became abundantly clear that it cannot serve as a template for economics. In fact, in the last quarter of the twentieth century economist theorists were increasingly promoting the view that economics models were merely metaphors or illustrations that were designed to make a certain point rather than predict data precisely (see Gibbard and Varian (1978)). Finally, the sociology of science has also significantly evolved over the twentieth century. Psychological treatment is no longer dominated by psychoanalysis. Importantly, psychological research has become a very respectable and responsible branch of science, drawing very careful distinctions between theoretical concepts and observable data and elevating issues of measurability to the level of a scientific discipline on its own. Specifically, cognition is a realm of carefully documented phenomena that offers relevant insights into the nature of economic activities. Past decades have also witnessed major developments in brain research. Neuroscience has established itself as an interdisciplinary scientific field, holding a promise to provide better understanding of mental phenomena. Admittedly, neuroscience has not yet produced any specific economic predictions or insights. Yet, the very existence of the field and its potential applications convince economists that there are relevant observable data beyond choice behavior. Moreover, the hallmarks of neuroscience, involving electrodes and tMRI images, project an image of a "hard" science. As such, brain research constitutes a desirable scientific ally, which partly legitimizes cognitive data as well. As a result of the processes described above, the revealed preference paradigm has become the target of numerous attacks. In particular, a growing number of economists voiced their dissatisfaction with a definition of rationality that is based solely on revealed choice. The concept of procedural rationality, originally proposed by Simon (1986), became more popular in recent years. (See Rubinstein 1998) At present, economists begin to be interested in the cognitive and mental processes that lead to behavior, and not just in behavior per see Relatedly, economic theorists expand the scope of cognitive phenomena that they find relevant, or potentially relevant, to the understanding of economic activity. Thus, topics such as emotions (Frank (1988), Rabin (1998), Elster (1998)) and language (Rubinstein (2000)) have become legitimate objects of study for economics.
The present volume includes several recent contributions to the study of cognitive processes and rationality in economic theory. To a large extent, these contributions reflect the causes for and the highlights of the renewed interest in cognition within economics. The volume begins with descriptive economics, discussing new cognitive-based approaches to classical economic problems. It proceeds with normative issues. Several chapters are devoted to less traditional economic issues, and a final chapter surveys neurological evidence that might prove relevant to economics in the long run. The first two chapters are devoted to financial markets. While this is a classical topic of descriptive economics, there is a widespread sensation that classical economic theory fails to provide a perfectly accurate account of these markets. The chapter by Gur Huberman describes recent development in behavioral finance,2 attempting to further our understanding of these markets. It is followed by a chapter by Marcello Basili and Fulvio Fontini, discussing financial markets with Knightian uncertainty. The latter refers to situations in which uncertainty is not quantifiable by a single probability measure. Both chapters thus deal with models of behavior in financial markets, inspired by intuition regarding cognitive processes that underlie trade choices. Chapters 3-5 deal with game theory. Adam Brandenburger's contribution deals with what people know, what they can know, and what they can conceive of. It presents an impossibility result regarding the scope of beliefs people might possibly entertain. This chapter belongs to a tradition of epistemology in game theory, starting with Aumann (1976). While this literature retains a formal linkage to Savage's (1954) behavioral foundations of beliefs, most of the epistemological discussion is motivated by purely cognitive notions of knowledge and beliefs. The contribution by Nicola Dimitri, in Chapter 4, also belongs to the epistemological tradition in game theory, with a stronger emphasis on economic applications. In particular, it studies the electronic mail game and explores the possibility of risky coordination with noisy and correlated communication. Finally, in Chapter 5, Dale Stahl surveys rule learning in normal-form games. Dealing with learning in games, this chapter belongs to a long tradition in economic theory. However, it is distinguished from the bulk of the literature in that it deals with rule learning, and compares it to other types of learning methodologies. All of these are motivated by hypothesized cognitive processes, rather than by axiomatic derivation based on behavioral data. The next five chapters deal with cognitive phenomena that are beyond the scope of classical economics. In Chapter 6, Michael Bacharach offers a model of framing effects. This phenomenon has been ignored by economic theory. In fact, formal modeling in economics has, almost with no exception, implicitly assumed that representation does not matter, and thus that framing effects do not exist. Bacharach's chapter paves the way for an extension of the scope of formal modeling that would include framing effects in a way that may alter the directions of economic research. Barton Lipman, in Chapter 7, discusses language and economics. This chapter surveys the extension of economic formal modeling to the use of language. Language is another realm of cognitive activity that has been largely ignored by
economists. Yet, its importance in everyday economic decisions can hardly be disputed. This chapter also surveys the study of debates as strategic games, which is a new topic of study to economists. Chapter 8, by Itzhak Gilboa and David· Schmeidler, discusses prediction. Whereas belief formation has always been implicit in economic behavior, this chapter takes the approach that prediction is an independent cognitive activity, which can be axiomatized based on cognitive data. The goal of the axiomatization is to char~cterize and generalize standard statistical techniques to situations that are not readily formulated numerically. An analysis of language as a cognitive capacity is offered in Chapter 9 by Luigi Rizzi. Taking as a starting point the Chomskian approach to the issue, the author stresses the view of linguistic knowledge as a computational capacity, also discussing questions related to optimality and efficiency of such ability. Patrick Suppes has contributed Chapter 10, in which he discusses rationality and freedom. Freedom is an essential philosophical concept that is intuitively clear to human beings. Yet, it serves no role in economic modeling. Suppes highlights the importance of this concept and relates it to the notion of uncertainty. Another recent addition to the topics that economics finds relevant is exemplified in Chapter 11. In it, Cristiano Castelfranchi uses the case of trust to argue for a cognitive program for economics. This chapter discusses the notion of trust and argues that it cannot be reduced to other concepts, more familiar to economists. The volume concludes with a chapter by Jordan Grafman, describing the activities recorded in the human prefrontal cortex. Initiating economists into neuroscience, this chapter gives an inkling of the new directions in which economics may proceed. The present collection reflects a scientific discipline at the point of transition. After a century of domination of a behavioral, non-cognitive paradigm, economics opens up to other fields and to other ways of looking at the phenomena of interest. While it is too early to tell which directions will prove useful, one can hardly fail to be excited by the intellectual activity we are currently witnessing. It is the editors' hope that the reader would share this excitement while reading the following chapters. The Editors
Notes Affect is a psychological term encompassing phenomena such as emotion, mood, feeling, and so forth. While these are distinct from cognition in the psychological literature, we will henceforth not be too meticulous, and will use "cognition" to refer to the various mental processes that are, in principle, open to introspection, but that are not directly reduced to behavior. 2 The terms "behavioral finance" and "behavioral economics" probably hark back to "behavioral decision theory," which refers to the experimental study of decisions, following, for the most part, the works of Kahneman and Tversky. The epithet "behavioral" in these titles means "how people actually behave" as opposed to "how economic theory assumes they behave." However, all these fields show much greater interest in cognitive phenomena than does classical economics. The latter, by adhering to the revealed preference paradigm, attempted to be "behavioral" in the sense that it restricted
its attention to allegedly observable behavior. Hence, somewhat ironically, "behavioral" finance/economics should perhaps be called "cognitive" or "cognitive-behavioral" finance/economics.
References Bazerman, M. (1986), Judgment in Managerial Decision Making. New York: John Wiley and Sons. Brams, S. (1997), "Game Theory and Emotions," Rationality and Society, 9: 91-124. Camerer, C. and M. Weber (1992), "Recent Developments in Modeling Preferences: Uncertainty and Ambiguity," Journal ofRisk and Uncertainty, 5: 325-370. Carnap, R. (1923), "Uber die Aufgabe der Physik und die Andwednung des Grundsatze der Einfachstheit," Kant-Studien, 28: 90-107. Damasio, A. (1994), Descartes' Error. New York: Putnam. Duesenberry, J. S. (1949), Income, Saving, and the Theory of Consumer Behavior. Cambridge, MA: Harvard University Press. Elster 1. (1998), "Emotions and Economic Theory," Journal of Economic Literature, 36: 47-74. Foley, D. (1967), "Resource Allocation and the Public Sector," Yale Economic Essays, 7: 45-98. Forgas, J. P. (1991), "Mood Effects on Decision Making Strategies," Australian Journal of Psychology, 41: 197-214. Forgas, J. P. and G. H. Bower (1987), "Mood Effects on Person Perception Judgments," Journal of Personality and Social Psychology, 53: 53-60. Forgas, J. P. and G. H. Bower (1988), "Affect in Social Judgment," Australian Journal of Psychology, 40: 125-145. Frijda, N. H. (1986), The Emotions. Cambridge; England: Cambridge University Press. Gibbard, A. and H. Varian (1978), "Economic Models," Journal of Philosophy, 75: 664677. Gilboa, I. and D. Schmeidler (1997), "Cumulative Utility Consumer Theory," International Economic Review, 38: 737-761. Gilboa, I. and E. Gilboa-Schechtman (2003), "Mental accounting and the absent-minded driver" In Brocas, I. and Carrillo, J. D. (eds), The Psychology of Economic Decisions, vol. 1 Rationality and Well-being, pp 127-136, Oxford: Oxford University press. Hanson, N. R. (1958), Patterns of Discovery. Cambridge, England: Cambridge University Press. Harless, D. and C. Camerer (1994), "The Utility of Generalized Expected Utility Theories," Econometrica, 62: 1251-1289. Isen, A. M. and B. Means (1983), "The Influence of Positive Affect on Decision Making • Strategy," Social Cognition, 2: 18-31. Isen, A. M. and N. Geva (1987), "The Influence of Positive Affect on Acceptable Level of Risk and Thoughts about Losing: The Person with the Larger Canoe has a Large Worry," Organizational Behavior and Human Decision Processes, 39: 145-154. Isen, A. M., K. A. Daubman, and G. P. Nowicki (1987), "Positive Affect Facilitates Creative Problem Solving," Journal of Personality and Social Psychology, 52: 1122-1131. Kahneman, D. and A. Tversky (1979), "Prospect Theory: An Analysis of Decision Under Risk," Econometrica, 47: 263-291. Kahneman D., P. Slovic, and A. Tversky (1982) (eds), Judgment under Uncertainty: Heuristics and Biases. Cambridge and New York: Cambridge University Press.
LeDoux, J. E. (1996), The Emotional Brain. New York: Simon and Schuster. Lewin, S. (1996), "Economics and Psycholog~: Lessons For Our Own Day From the Early Twentieth Century," Journal of Economic Literature, 34: 1293-1323. Lewinsohn, S. and H. Mano (1993), "Multi-Attribute Choice and Affect: The Influence of Naturally Occurring and Manipulated Moods on Choice Processes," Journal of Behavioral Decision Making, 6: 33-51. Loewenstein, G. (1992), "The Fall and Rise of Psychological Explanation in the Economics of Intertemporal Choice." In G. Loewenstein and J. Elster (eds), Choice over Time (pp. 3-34). New York: Russell Sage. Loewenstein, G. (1996), "Out of Control: Visceral Influence on Behavior," Organizational Behavior and Human Decision Processes, 65: 272-292. Loomes G. and R. Sugden (1982), "Regret Theory: An Alternative Theory of Rational Choice under Uncertainty," Economic Journal, 92: 805-824. Mann, L. (1992), "Stress, Affect, and Risk Taking." In Yates, J. F. (eds), Risk Taking Behavior (pp. 202-230). Wiley Series in Human Performance and Cognition. Chichester: John Wiley and Sons. Mano, H. (1990), "Emotional States and Decision Making." In Goldberg, M., G. Gorn, and R. Pollay (eds), Advances in Consumer Research, 17: 577-584. Mano, H. (1992), "Judgment under Distress: Assessing the Role of Unpleasantness and Arousal in Judgment Formation," Organizational Behavior and Human Decision Processes, 52: 216-245. Marshall, A. (1890), Principles of Economics. 9th edn (1961), London: Macmillan. Menger, C. (1871), Principles of Economics. Reprinted (1951), Glencoe, Ill.: Free Press. Ortony, A., J. L. Clore, and Collins, A. (1988), The Cognitive Structure of Emotions. New York: Cambridge University Press. Pareto, Vilfredo (1916), Trattato di Sociologia Generale. 4 vols, Florence: Barbera. Translated into English and edited by Arthur Livingston as The Mind and Society, New York: Harcourt Brace & Co., 1935. Piattelli-Palmarini (1994), Inevitable Illusions: How Mistakes of Reason Rule Our Minds. New York: John Wiley and Sons. Popper, K. R. (1934), Logik der Forschung; English edition (1958), The Logic ofScientific Discovery. London: Hutchinson and Co. Reprinted (1961), New York: Science Editions. Prelee, D. and G. Loewenstein (1998), "The Red and the Black: Mental Accounting of Savings and Debt," Marketing Science, 17: 4-28. Quine, W. V. (1953), "Two Dogmas of Empiricism," in From a Logical Point of V~ew. Cambridge, MA: Harvard University Press. Rabin, M. (1993), "Incorporating Fairness into Game Theory and Economics," American Economic Review, 83: 1281-1302. Rabin, M. (1998), "Psychology and Economics," Journal of Economic Literature, 36: 11-46. Rubinstein, A. (1998), Modeling Bounded Rationality, Cambridge: MIT Press. Rubinstein, A. (2000), Economics and Language, Cambridge: Cambridge University Press. Samuelson, P. A. (1937) "A Note on Measurement of Utility," Review ofEconomic Studies, 4: 155-161. Simon, H. A. (1986), "Rationality in Psychology and Economics," Journal ofBusiness, 59: 5209-S224. Tversky, A. and D. Kahneman (1981), "The Framing of Decisions and the Psychology of Choice," Science, 211: 453-458. Walras, L. (1873), Elements ofPure Economics. Reprinted (1954), Homewood, Ill.: Richard D, Erwin.
Behavioral finance and markets* Gur Huberman
Economics is interested primarily in prices and aggregate quantities. The study of individual behavior is a building block to derive implications about social outcomes. Until the behavioral approach became fashionable, individuals were usually assumed to make choices so as to optimize a well-defined objective subject to well-defined constraints. This very simple idea is also very powerful, in that it lends the analysis to aggregation, and thereby affords the study of markets and equilibrium. The main contribution of the behavioral approach has so far been to question the validity of modeling the individual decision maker as optimizing a simple objective. The earlier pioneers are Allais (1953) and Ellsberg (1961). More recently, the profuse work of Kahneman and Tversky (1979) (with various coauthors) has had the strongest impact. Their joint paper on Prospect Theory in Econometrica (Kahneman and Tversky 1979) is reputed to be the most cited paper in that highly esteemed journal. Once scholars acknowledged that the optimizing foundations were not as solid as had been assumed, they ventured to modify them, and felt freer to discover anomalies that would not have existed had economic agents (or at the least, the important agents, the marginal ones) been neoclassical optimizers. "Is the asset price right?" is the question at the heart of financial economics. To answer it directly, one has to agree on what "right" means in this context. An early commentator was Adam Smith. The value of a share in a joint stock is always the price which it will bring in the market; and this may be either greater or less, in any proportion, than the sum which its owner stands credited for in the stock of the company. Adam Smith, The Wealth ofNations, 1776 The efficient market hypothesis that "the price is right" is difficult to study directly. A circuitous, but profitable route, calls for the study of implications of
* This chapter is based on a lecture given at a workshop on Cognitive Processes and Rationality in Economics at the International School of Economic Research, University of Siena in July 2001.
the "price is right" statement. One of them is that price changes are unpredictable. This implication has stood up to empirical scrutiny very well. There are no obvious and reliable ways to predict which way the prices of securities will go. But price changes can seem unpredictable even if the price is not right, especially when it comes to securities with open-ended payoffs such as common stocks. One problem for the "the price is right" school is not that price changes are unpredictable; it is that ex post they are poorly explained. A single dramatic day best illustrates how poorly stock price changes are understood. On October 19, 1987, world stock markets crashed; in the United States, the S&P500 index lost 20.47 percent of its value. The New York Times' "explanation" was "worry over dollar decline and trade deficit, fear of US not supporting the dollar." Motivated by the 1987 crash, Cutler et ale (1989) list the top major world news in 1941-87 and the stock market reaction to them, as well as the top fifty market moves, and the New York Times, "explanations" to them. Remarkably, although the major news produce some big price movements, they do not produce any of the top five and only seven of the top fifty price movements. Thus, it seems that fundamentals move prices, but major price movements cannot be explained as reaction to changes in major fundamentals. The Law of One Price states that two securities that represent identical claims to cash flows should trade for the same price. In financial economics the most interesting anomalies are violations of the Law of One Price. They are important because they constitute a direct assault on the efficient market hypothesis that the market price is right, or at least approximately right. Examples ofviolations ofthe Law ofOne Price include closed-end mutual funds, Siamese twin stocks, and the case of EntreMed. Together they allow the outlines of a coherent story to emerge. The story is about the influence of the demand side of financial markets on asset prices. The demand side may be affected by investor sentiment, whose fluctuations may be independent of fundamentals. Shleifer and Summers (1990) summarize this approach. Prices are the main focus of financial economics. Trading volume receives much less attention. In fact, the motives of security trading are poorly understood. But it is those who trade who also determine prices. Therefore an acceptable model of trading may herald a better understanding of security prices. The neoclassical approach has not adequately explained the huge trading volume, but the behavioral approach may offer some hope of doing just that. The balance of this chapter has two main sections. The next section describes various violations of the Law of One Price. The section that follows it considers a related, but very different a~d fundamental issue: Why do people trade?
1.2 Violations of the Law of One Price 1.2.1
Closed-end fu nds
Closed-end funds are investment companies that raise equity when they are formed and use it to acquire tradable securities. After the inception period, the fund sells
Behavioral finance and markets
and buys tradable securities and its shareholders are free to trade its shares. The fund does not redeem outstanding shares unless it liquidates or changes its status to an open-end fund. The Law ofOne Price suggests that shares ofclosed-end funds should trade close to net asset value (NAV). This is not the case, as a quick look at the appropriate table on Monday's Wall Street Journal (Or Saturday's New York Times or Barron's) will attest. Lee et ale (1991) summarize the main empirical regularities associated with closed-end funds as follows: • • • • •
Most of the time they trade at a discount relative to NAV. The discounts fluctuate. The discounts as well as changes in them across funds are positively correlated. They are issued at a premium relative to NAV. When liquidation or open-ending of a fund is announced, its price quickly converges to the NAV.
Lee et ale (1991) also report that the discounts are negatively correlated with the returns on small-company stocks. Presumably, it is individual investors who tend to hold and trade both closed-end funds and small stocks; correlation between the returns on these very different sets of assets suggests that a common sentiment moves their prices. These observations lead them to argue that noise traders affect the prices of closed-end funds, and, by extension, of securities in general. Closed-end country funds (often referred to as country funds) are an interesting subset of closed-end funds because their assets trade in a foreign market. A reason for the formation of country funds is the segmentation of international financial markets. Country funds afford the study of the segmentation of investor sentiment internationally and a novel approach to the speed-of-adjustment question: how quickly do prices react to news, and how dependent is the speed on the salience of the news? Hardouvelis et ale (1994) have done an exhaustive study of the sources of temporal variation in country fund discounts. The article's main finding is in its table 8.8 where it estimates a linear regression of the relation between weekly changes in the premiums and the discount itself (positive), the return on the foreign market (negative), the dollar return on the exchange rate (negative), the dollar return on the world stock market index (positive), the return on large US stocks (positive), and the difference in return on small and large US stocks (positive). The direction of all these relations is consistent with the investor sentiment hypothesis: sentiment in the United States is mean-reverting (hence the negative relation between changes in the discount and the discount itself), not sensitive to pricing of foreign stocks (hence a negative relation with the foreign market), related to US (or world) sentiment about the foreign market (hence the negative relation with the changes in the exchange rate), related to world and US stock returns (hence the positive relations with these two variables) and is primarily correlated with
small stock returns (hence the positive relation with the small minus large stock returns). Country funds (and closed-end funds in general) are important not because they manage a lot of assets, but because they present fairly clean setups in which the examination of standard predictions is clearer than in other contexts. Country funds allow the researcher to entertain a difference in sentiment between the country where the assets are and the country where the funds' shares are traded, and study the extent to which the difference affects temporal variations in the discount of the country fund. Hardouvelis et ale (1994) study a cross-country potential difference in investor sentiment. Klibanoff et al. (1998) study cross-country difference in the impact of news on asset prices. Klibanoff et ale (1998) examine how fast share prices of country funds adjust to news about the relevant foreign markets. They show that in normal weeks, typically, a country fund's return lags significantly by a few weeks behind the return on its underlying assets, which are traded on the foreign market. Then they consider weeks with salient news about the foreign country, which are weeks in which news about the foreign country appear on the front page of the New York Times. In these weeks the prices of country-funds shares (which trade on the New York Stock Exchange) react more robustly to changes in the prices of the funds' underlying assets (which trade on the foreign markets).
1.2.2 Siamese twin stocks Siamese twin stocks afford a similar trading and sentiment structure. These are two classes of shares of the same firm. Their relative property rights are well specified, and the bulk of the trading of each class of shares takes place in different stock markets. The contractual specification of the relative property rights implies that the shares should trade at the same relative prices. On the other hand, if they trade on different markets which are subject to different sentiments, relative prices may diverge, and the divergence should be correlated with the relative movements in the respective markets. Following the early work of Rosenthal and Young (1990), Froot and Dabora (1999) revisit the Siamese twin stocks. These companies are: Royal Dutch and Shell, Unilever NV and Unilever pIc, and SmithKline Beecham class A and class E shares. All three are large international publicly held firms whose stocks trade at various markets. But in each case, the two stock classes trade primarily on different markets. Calculation of the theoretical relative values of the two types of equity are straightforward, and derived directly from the original agreement which gave rise to the two stock classes in each case. Nonetheless, hardly ever do the two stock classes trade at the theoretically correct relative prices. Figure 1.1 demonstrates the disparity for Royal Dutch and Shell. Froot and Dabora (1999) go further, and estimate the relation between relative prices in the stock markets in which the two stocks trade and the relative prices of the stocks themselves. It turns out that indeed, when the London Stock Exchange (where Shell trades) rallies relative to the Amsterdam or New York Stock Exchange
Figure J. J Stock price of Royal Dutch relative to Shell (deviation from 60/40 value). Source: Froot and Dabora (1999).
(where Royal Dutch trades), so does Shell relative to Royal Dutch stock. The other two Siamese twin stocks display similar patterns. The upshot of the Siamese twin study is that not only is the Law of One Price violated but also investor sentiment models suggest that the violations are correlated with local stock market behavior, and this indeed seems to be the case.
1.2.3 EntreMed Can stories that appear in the New York Times cause stock price movements even when they don't report any new information? Huberman and Regev (2001) narrowly focus on implications of the Law of One Price for a biotech firm, EntreMed (ENMD), and related firms. Their work is prompted by a front page story in the Sunday, May 3, 1998, edition of the New York Times which reported on a recent breakthrough in cancer research, and mentioned ENMD, a company with licensing rights to the breakthrough. The story's impact on the stock prices was immediate, huge, and to a large extent permanent. The new-news content of the Times story was nil, though: the substance of the story had been published as a scientific piece in Nature and in the popular press (including the Times itself) more than five months earlier, in November 1997. The cover of the November 27, 1997, issue of Nature prominently features the lead headline, "Resistance-free cancer therapy" as well as a related image. In that issue, Boehm et ale (1997) report on a breakthrough in cancer research achieved by a team led by Dr Judah Folkman, a well-known Harvard scientist.
In a "News and Views" piece in the same issue, Kerbel (1997) explains and comments on the findings, suggesting that, "[T]he results of Boehm et ale are unprecedented and could herald a new era of cancer treatment. But that era could be years away." Reports on the discovery of Dr Folkman's team appeared also in the popular press, such as the Times and Newsday on November 27, 1997 as well as in the electronic media, such as CNN's MoneyLine and CNBC's Street Signs. It seems that an effort was made to bring the news to the attention of circles wider than the scientific community. The November 27 Times article appeared on page A28. It, as well as CNN and CNBC, mentioned ENMD. On November 28, ENMD itself issued a press release that covered the news and the company's licensing rights to the proteins developed by the team of Dr Folkman. The closing price of ENMD was 11.875 on November 26, and on November 28 it was 15.25; thus, the news caused a price appreciation of 28.4 percent, an observation made in the Business Section of the November 29 edition of the Times. The unusually high trading volume on November 28 and December 1 indicates that the market paid attention to the news. On the whole, an adherent of the efficient market hypothesis would argue that the market digested the news in a timely and robust fashion. In the months between November 27, 1997 and May 3, 1998, ENMD's stock traded between 9.875 and 15.25. Kolata's Times article of Sunday, May 3, 1998, presented virtually the same information that the newspaper had reported in November, but much more prominently; namely, the article appeared in the upper left comer of the front page, accompanied by the label "A special report." The article featured comments from various experts, some very hopeful and others quite restrained (of the "this is interesting, but let's wait and see" variety). The article's most enthusiastic paragraph was " ... 'Judah is going to cure cancer in two years,' said Dr James D. Watson, a Nobel Laureate... Dr Watson said Dr Folkman would be remembered along with scientists like Charles Darwin as someone who permanently altered civilization." (Watson, of The Double Helix fame, was later reported to have denied the quotes.) ENMD's stock, which had closed at 12.063 on the Friday before the article appeared, opened at 85 and closed at 51.81 on Monday, May 4. The Friday-closeto-Monday-close return of 330 percent was truly exceptional: bigger than all but two of the over 28 million daily returns of stocks priced at $3 or more between January 1, 1963 and December 31, 1997. Not surprisingly, the Times story, and ENMD, received tremendous attention in the national media (print and electronic) in subsequent weeks. In the May 10 issue of the Times, Abelson (1998) essentially acknowledged that its May 3 article contained no new-news, noting that "[p]rofessional investors have long been familiar with [ENMD's] cancer-therapy research and had reflected it in the pre-runup price of about $12 a share." (The Times did not question its own editorial choice of essentially re-reporting the November 27 article, by a different reporter, with the label, "A special report," on the upper left comer of the front page. Gawande (1998) did that in the New Yorker's May 18 issue, which hit the newsstands on May 11.)
Figure 1.2 ENMD closing prices and trading volume 10/1/97-12/30/98. Source: Huberman and Regev (2001).
Figure 1.2 gives the distinct impression that while some of the May 4 price run-up was temporary, a substantial portion of it was permanent. ENMD's stock price fell in the days following May 4, to close the week at 33.25 - still, almost three times higher than its price a week earlier. Moreover, ENMD's closing price did not fall below 20 until late August 1998, and by late fall it had not closed below 16.94, which was 40 percent higher than its May 1 price. (Between mid-July and late August 1998, the S&P500 lost almost 20 percent and the NASDAQ Combined Biotechnology Index lost almost 24 percent of its value.) By early November 1998 ENMD was trading at the upper 20s and lower 30s. On November 12, 1998, another piece of new-news came to light: on· its front page, the Wall Street Journal reported that other laboratories failed to replicate the results described earlier in the Times. ENMD stock price fell from 32.625 on November 11 to close at 24.875 on November 12 - still more than twice ENMD price on May I! Contagion: Can old news reported in the New York Times cause prices of related stocks to increase? A look at the stock prices of other biotechnology stocks magnifies the puzzle. On average, the number of members of the NASDAQ Biotechnology Combined Index, excluding ENMD, went up by 7.5 percent on Monday, May 4, 1998. The returns of seven of the stocks in the index (other than ENMD) exceeded 25 percent on a trading volume that was fifty times the average daily volume. That news about a breakthrough in cancer research affects not only the stock of a firm that has direct commercialization rights to the development is not