accounting theory revision

10 Association for Investment Management and Research

Behavioral Finance and Its Implications for Stock-Price Volatility

Robert A. Olsen

An increasing number of academic and professional articles are being published about research on and potential applications of behavioral finance. This article offers a more complete picture of the origin, content, and rationale behind this emerging area of study than previously presented. In the process, the traditional dominance in finance of the economic concepts of subjective expected utility and rationality are discussed. In addition, the article argues that the newer theories of chaos and adaptive decision making, which have a place in behavioral finance, can help explain the puzzle of stock-price volatility.

recent controversy in the field of finance involves the evolution of what has become known as behavioral finance. An increasing number of academic and pro-

fessional articles are being published devoted to research about and potential applications of new behavioral finance findings. Few authors, however, have attempted to present a complete picture of the origin, content, or rationale behind this emerging area of study. The primary purpose of this article is to present just such a general perspective. A second- ary objective is to provide an example of how behavioral finance can be of value for understand- ing financial markets by showing how it helps explain what some theorists see as “excessive” stock-price volatility.

The Origins of Behavioral Finance One of the earliest (if not the first) call for a scientific melding of psychological and financial research came from University of Oregon Business Professor O.K. Burrell in a 1951 article titled “Possibility of an Experimental Approach to Investment Studies.” The article explored the need for the construction of laboratory experiments to test theories, and it was followed in 1967 by “Scientific Investment Analysis: Science or Fiction?” by Oregon Finance Professor W. Scott Bauman. Burrell and Bauman called for a new area of research focusing on the benefits that might be obtained from combining the “new”emphasis on

quantitative investment models with information from the more traditional behavioral approaches to finance.

In 1969, Oregon Psychology Professor Paul Slovic published a detailed study of the investment process from a behavioral perspective. In 1972, Bauman and Slovic continued this line of inquiry and Slovic published the first seminal paper in the area, “Psychological Study of Human Judgment.”

In spite of an articulate appeal by Gail Farrelly in a 1980 article, interest in behavioral finance did not gain momentum until well into the late 1980s. At that time, Professors Richard Thaler at the Uni- versity of Chicago, Robert Shiller at Yale Univer- sity, Howard Kunreuther at the University of Pennsylvania, Werner De Bondt at the University of Wisconsin, Josef Lakonishok at the University of Illinois, and Meir Statman and Hersh Shefrin at Santa Clara University, among others, began to publish research relevant to behavioral finance. The renewed interest in the topic appears to have been triggered by two developments. The first was mounting empirical evidence that existing finan- cial theories appeared to be deficient in fundamen- tal ways. The second was the development of prospect theory by Professors Daniel Kahneman of Princeton University and Amos Tversky of Stan- ford University. Prospect theory presented a model of decision making that was an alternative to sub- jective expected utility theory with more-realistic behavioral assumptions.

Between 1972 and the revival of interest in behaviorally oriented finance in the late 1980s, academic accounting researchers conducted the most extensive research into the impact of psycho- logical processes on financial decision making.

Robert A. Olsen is a professor of finance at California State University at Chico and a research scholar at Decision Research in Eugene, Oregon.

A

Financial Analysts Journal • March/April 1998 11

Accounting’s traditional focus on the individual and firm, as opposed to the market, appears responsible, in part, for this foresight (see Ashton 1995).

By and large, the academic financial commu- nity has remained cautious about embracing devel- opments in this field, although practicing financial professionals believe that it is about time academi- cians became realistic about investor behavior. Dreman Value Management of New York was a pioneer in psychologically based “contrarian” investment strategies. And some newer investment management firms—notably, RJF Asset Manage- ment of San Mateo, California, and LSV Asset Management of Chicago—have recently started offering behaviorally oriented investment advice. Research in the area is now being supported by a number of university institutes, by the Association for Investment Management and Research, and by a number of investment firms through the Institute of Psychology and Markets.

Interestingly, what is currently known as mod- ern or standard finance also had its birth in about 1951. It prospered, however, whereas behavioral finance languished, in part at least, because of some unique social and academic conditions that I will explain.

Defining Behavioral Finance The new paradigm of behavioral finance seeks to replace the behaviorally incomplete theory of finance now often referred to as standard or modern finance. Even as it seeks to be a replacement for the existing financial paradigm, however, behavioral finance recognizes that the existing paradigm can be true within specific boundaries.

Behavioral finance is part of science, in that it starts from fundamental axioms and asks whether a theory built on the axioms can explain behavior in the financial marketplace. Contrary to some assertions, behavioral finance does not try to define “rational” behavior or label decision making as biased or faulty; it seeks to understand and predict systematic financial market implications of psycho- logical decision processes. In addition, behavioral finance is focused on the application of psycholog- ical and economic principles for the improvement of financial decision making. Behavioral finance does not reject economic concepts and principles that are sound.

Currently, no unified theory of behavioral finance exists. Shefrin and Statman (1994) began work in this direction, but so far, most emphasis in the literature has been on identifying behavioral decision-making attributes that are likely to have

systematic effects on financial market behavior. Indeed, even though the paradigm is new and evolving, behavioral finance theorists have already identified a number of potential psychological decision attributes with overarching potential axi- omatic status. These include the following: • Decision makers’ preferences tend to be multi-

faceted, open to change, and often formed only during the decision process itself.

• Decision makers appear to be adaptive, in the sense that the nature of the decision and the environment in which the decision is made contribute to their selection of a decision pro- cess or technique.

• Decision makers seek satisfactory, rather than optimal, solutions.

A more inclusive list of narrowly defined decision behaviors is given in Exhibit 1.

Although the list of decision attributes is ten- tative and the nature of their interactions and impli- cations for the market are incomplete, empirical evidence suggests that they contribute to the fol- lowing investment-related characteristics: • formally chaotic stock prices (prices imper-

[Advertisement]

1 2

 A

sso cia

tio n

fo r In

ve stm

e n

t M a

n a

g e

m e

n t a

n d

R e

se a

rch

Exhibit 1. Decision Behaviors Narrowly Defined 1. Decision makers tend to rely more heavily on stereotypical, analogic, heuristic (rule-of-thumb), or

other intuitive or experiential decision processes as decisions become more complex, time grows short, or emotions run high.

21. Decision makers tend to overestimate the probability of conjunctive events.

2. Decision makers’ choices are influenced by affect, or feeling, because an objective of a good decision is positive affect.

22. Decision makers treat small samples as overly representative of populations.

3. Decision makers tend to discount future outcomes at rates that vary inversely with the size of the outcome.

23. Decision makers are very imperfectly aware of the actual weight they give to separate pieces of information in a decision.

4. Decision makers tend to overweight confirming evidence and underweight disconfirming evidence. 24. Decision makers weight information that seems personal more heavily than impersonal information.

5. Decision makers tend to overweight probabilities of favorable outcomes and underweight probabilities of unfavorable outcomes.

25. Decision makers tend to overestimate their ability to have correctly forecasted past events (hindsight bias).

6. Decision makers tend to overweight the more salient or more memorable facts and evidence. 26. Decision makers tend to be overconfident about their ability to forecast and make correct decisions.

7. Decision makers tend to overweight low-probability events and underweight high-probability events. (Events with very low probability tend to be treated as certain to not occur; events with very high probability are treated as certain to occur.)

27. Decision makers tend to use information in the form given.

8. Decision makers tend to be improperly regressive and thus fail to understand that outcomes tend to revert to the mean.

28. Decision makers tend to overestimate the variability associated with a random series.

9. Decision makers are social animals. Thus, decisions are influenced by needs for self-control and group acceptance as well as the fear of experiencing regret.

29. Decision makers tend to confuse precision with reliability and view quantitative information as more reliable than nonquantitative information.

10. Decision makers appear to discount distant future losses at higher rates than distant future gains. 30. Decision makers often forecast by establishing an “anchor” value, and the adjustment they make to the anchor based on specific information is usually insufficient.

11. Decision makers’ choices appear to be influenced by the existence of separate mental financial accounts based on social and economic criteria.

31. Decision makers give greater weight to information that has been made to seem more complete by the addition of nondiagnostic facts.

12. Decision makers tend to be loss averse rather than risk averse. 32. Decision makers give greater weight to nondiagnostic information as diagnostic information becomes ambiguous.

13. Decision makers tend to exhibit a strong preference, all else being equal, for an ascending pattern of returns.

33. Decision makers weight information based on the order in which it is obtained. Primacy (earliest data gets most weight) appears to dominate for simple decisions, whereas recency (latest data gets greatest weight) appears to dominate for complex decisions.

14. Decision makers tend to experience diminishing returns for gains and losses. 34. Decision makers exhibit greater inconsistency in choice when under stress.

15. Decision makers appear to focus on changes in, as opposed to absolute levels of, decision attributes. 35. Decision makers tend to overestimate the accuracy of consensus judgments.

16. Decision makers engage in outcome gain and loss segregation or aggregation among decisions to enhance perceived net benefit.

36. Decision makers give greater weight to nonnumeric information as numeric information becomes ambiguous.

17. Decision makers’ perceptions of risk vary inversely with perceptions of control. 37. Decision makers’ accuracy, speed, and decision processes can be influenced by the physical format of the information supplied.

18. Decision makers place more weight on decision information that is presented in a format consistent with the choice format. For example, if an answer is sought in numerical terms, verbal information will be underweighted.

38. Decision makers’ task participation is enhanced by providing immediate and dynamic feedback.

19. Decision makers tend to weight negative evidence more heavily when they are under stress than when they are not.

39. Decision makers are “locked into the present”; they overweight current beliefs and feelings and are very inaccurate when forecasting future hedonic (pleasurable) states.

20. Decision makers tend to reduce the portion of information considered when they are under stress. 40. Decision makers give greater weight to anchor values as forecasting becomes more difficult.

Note: For further information on these attributes, see Baron (1994), Plous (1993), Robertson (1991), and Yates (1992, 1993).

Financial Analysts Journal • March/April 1998 13

fectly predictable), • excessive stock-price volatility and bubbles in

prices, • follow-the-leader or herding behavior among

investors, • misestimation of the risk of loss, • selling winning investments too early and sell-

ing losing investments too late, • differing preferences among investors for cash

dividends, • belief in the value of time diversification (that

risk diminishes with time), • popular investments earning poorer-than-

desired returns, • investors mistaking “good” companies for

“good” investments, • asset prices appearing to over- or underreact to

new market information, • individual investors holding poorly diversi-

fied portfolios, and • superior short-run and inferior long-run per-

formance of initial public offerings. Investigation into the source of the decision

attributes is just beginning, but many of the attributes appear to be robust and cross-cultural. In addition, Nisbett and Ross (1980) noted that these attributes can be responsible simultaneously for correct solutions to complex problems and for errors in simple daily judgments. They contended that human inferential strategies are designed to deal with a wide range of common general prob- lems and that when applied beyond this range, the strategies result in errors in judgment. Humans are seen as applied scientists, guided not only by the goal of epistemic correctness (that is, correct knowl- edge) but also by the need to share their fellow humans’ beliefs about reality, fairness, justice, and so forth. The methods of the pure scientist, who is focused on truth at any cost, must often be set aside in practical situations, especially when action is demanded. Nisbett and Ross noted that humans’ decision making can be improved by formal statis- tical training and by education about the correct application of intuitive decision procedures but the improvement is never complete.

Recent research in evolutionary psychology suggests that the tenacity and generality of many of the listed decision attributes indicate that their roots lie in human evolution. Barrow (1992) noted that a universal human nature exists at, primarily, the level of evolved psychological mechanisms and these mechanisms are adaptive and conditioned by natural selection. The current human mind, given the slow rate of biological evolution, appears to be most adapted to life in a Pleistocene hunter/ gatherer society. From this vantage point, decision

attributes that focus on negative events during stress, loss aversion, and a preference for concrete over abstract information, among others, make sense because they have survival value.

Similarly, Simpson and Kenrick (1997) have pointed out that genes can influence complex behavior, culture is not independent of evolving psychological mechanisms, and evolutionary pro- cesses are relevant to a wide range of social phenom- ena. Thus, the way is open for natural selection to play a role in the development and application of decision strategies when the biological objective is not optimization, or even survival, at the individual level but satisficing (finding the minimally accept- able solution) that leads to survival of the species. For example, evidence for a partial biological influ- ence on decision behavior is reported in “Under- standing and Assessing Financial Risk Tolerance: A Biological Perspective” (Harlow and Brown 1990). Researchers found that individuals with low levels of the enzyme monoamine oxidase were more will- ing than those with high levels to accept economic risks. A similar finding has been reported in the case of physical risk taking (Zuckerman 1994).

The Need for Change Because finance is considered a social science, it comes as a surprise to most social scientists, as well as to lay people, that the academic study of finance before behavioral finance involved little or no examination of individual decision making. In this respect, finance followed its mother discipline, eco- nomics: Deduction was dominant, and the decision- making process was treated as a black box.

Finance and economics assume the existence of mental models of choice. Psychology makes the same assumption, but finance and economics are primarily concerned with prediction, as opposed to description or explanation; thus, finance and eco- nomic theorists feel free to construct abstractions of the decision process. Some finance and economic researchers seem to have avoided the content of the decision box for fear of losing sight of the forest for the trees, but others seem to have simply assumed that psychology has little to offer finance and eco- nomics. One finance professor was overheard to say when asked by a student about the role of psychology in finance, “Finance types don’t carry excess baggage.”

The beginning of standard or modern finance is generally considered to be the publication of “Portfolio Selection” by Harry Markowitz in 1952. This article was more analytical, normative, and optimization oriented than anything written on finance to that date. Markowitz’s approach to

14 Association for Investment Management and Research

finance was fresh and rigorous, and it quickly gained adherents. The next 25 years witnessed an avalanche of theorizing that resulted in the capital asset pricing model (CAPM), arbitrage pricing the- ory, and option-pricing models. These approaches are highly analytical and normative and, from a behavioral perspective, assume a world dominated by homo economicus, a virtually omniscient decision maker who is completely “rational” and focused on utility (wealth) maximization.

In hindsight, the appeal and direction of this theorizing is understandable. First, the time was an era when “hard” science emphasizing quantitative methods and optimization was in favor. It was the era of the arms race and space exploration. In addi- tion, on the psychological front, cognitive psychol- ogy was in its formative years and the field of behavioral decision making had yet to be devel- oped. Second, finance academics were generally unhappy with the rigor of finance as an applied area of “business administration” and wished to beef it up by the adoption of more concepts and techniques of the mother science, economics. In the early 1950s, positivism (in which the primary concern is predic- tion as opposed to explanation) was a dominant theme in economic theorizing, so again, financial economists saw little to be gained from examining financial decision making at the individual level.

In their drive for rigor, not only did most financial economists ignore psychology, but some also misapplied and confused the meaning of the economic concepts of “rationality” and “subjective expected utility” (SEU). The result was that some later financial models were unfalsifiable, implied that economic truth could reliably be known and reflected in asset prices, and implied that a unique normative decision process (namely, SEU) existed that all rational decision makers should follow. For this reason, economic theorist Nicolaas Vriend of the Santa Fe Institute noted in 1996 that the concept of “instrumental economic rationality” is essentially contentless because it means only that decision makers’ actions are determined by their perceived preferences and opportunities. This said, Vriend demonstrated that two advocated approaches to identifying rationality—“internal consistency” and “reasonableness of decision-making procedures”— have no fundamental relevance for economic ratio- nality. In support of this position, economic theorist B. Lipman noted in 1991 that the assumption of “reasonableness in the decision procedure” leads to an infinite mental regress or, at least, an unidentifi- able decision procedure. L.J. Savage, a founder of SEU theory, noted, however, that normatively, SEU theory has no special ground to stand on in economic analysis. Economic theorist Paul Anand

noted in 1993 that SEU theory, with its principal virtue of choice consistency, is an incomplete standard by which to judge decision quality. Frisch and Clamen (1994) noted that for improving deci- sions, the emphasis should be placed on experience utility as opposed to decision utility; that is, research- ers must be concerned with how decision makers evaluate potential outcomes, the extent to which they identify relevant consequences, and the way they make final choices.

Although some financial economists will accept these criticisms of rationality, they might advocate a third approach. This approach suggests that rationality might be empirically verified by finding that decision makers do not make system- atic errors in financial judgments over time. Heap (1989) noted that this so-called Muth–Lucas version of rationality implicitly assumes that subjective and objective beliefs can coincide. Muth himself (1961) questioned the reasonableness of this assumption. In addition, many researchers, notably Hogarth (1988), have indicated that in most real-world social and economic situations, the learning conditions necessary for objective and subjective beliefs to coincide do not exist and that learning would itself create “nonstationarity” in belief distributions.

In essence, this third definition of rationality depends on the standard neoclassical economic assumptions that the chain-of-event causality is recoverable and that the relevant universe is fun- damentally deterministic. Financial asset prices, however, as will be discussed later, appear to exhibit chaotic behavior, especially in the long run. Chaotic behavior results from a complex nonlinear process in which causality is incompletely recover- able. Moreover, even if prices were not formally chaotic, Rosser (1996) has noted that the introduc- tion of stochasticity in the presence of nonlinear behavior would ruin any hope of improving fore- casting ability and render meaningless the idea of rational expectations.

It is ironic, yet a sign of potential for renewal, that the first cracks in the standard finance edifice were discovered through the application of sophis- ticated econometric techniques by standard finance pioneers. As early as 1977, Roll had noted that a cornerstone of standard finance, the CAPM, was probably unverifiable. Throughout the 1980s and 1990s, statistical anomalies, as they were called, continued to appear, which suggested that the existing models, if not wrong, were probably incomplete. Beta risk, a key concept in standard finance, was shown to be only weakly associated with stock returns. In addition, studies seemed to show that stock returns were excessively volatile and prone to bubbles. More disturbingly, investors

Financial Analysts Journal • March/April 1998 15

were shown not to react “logically” to new infor- mation but to be overconfident and to alter their choices when given superficial changes in the pre- sentation of investment information. Defenders of standard finance began to argue that the test data were noisy and could not be trusted. They sug- gested the presence of measurement error, selection bias, and survivability bias. Even though econo- mists Vernon Smith at the University of Arizona and Colin Camerer at Chicago had by now followed Burrell’s call and begun developing the field of “experimental” markets, some financial econo- mists referred to experimental outcomes that hinted at behavioral deficiencies in existing models as “parlor tricks” and unrepresentative of actual market behavior. Finally, in 1992, Eugene Fama, a key figure in the development of the CAPM, with- drew his support from the model, and in 1995, economist Werner De Bondt wrote,

The sad truth is that modern finance theory offers only a set of asset pricing models for which little support exists and a set of empirical facts for which no theory exists. (p. 8)

Even economist, Nobel Laureate, and standard finance pioneer Merton Miller admitted, in an April 23, 1994, interview with the Economist, that conven- tional economics had failed to explain how asset prices are set. He added, however, that he believed the new mix of psychology and finance would lead nowhere. Miller might be correct, but as pointed out by Langer (1997), strong commitment to existing paradigms can obscure one’s vision of promising alternatives.

Behavioral Finance Explanations of Stock-Price Volatility The potential explanatory value of behavioral finance may be seen by focusing on the volatile nature of stock prices. The relationship between many other market behaviors and the decision attributes previously listed have been discussed elsewhere (see Thaler 1991 and 1993 and Wood 1995). To my knowledge, the following analysis is unique.

Behavioral finance offers an explanation for the empirical evidence suggesting that the poor descriptive power of some current financial models stems from the fact that financial asset prices arise from a statistically complex and nonlinear process. Peters (1994) noted that stock prices and returns are cyclical, imperfectly predictable in the short run, and unpredictable in the long run and that they exhibit nonlinear, and possibly chaotic, behavior related to time-varying positive feedback. The con- nection between Peters’ observations and behav- ioral finance is that, as noted in Eve, Horsfall, and

Lee (1997), most complex social phenomena exhibit this nonlinear behavior. In addition, studies of indi- vidual economic decision making, such as those by Baumol (1989), Richards (1990), and Sterman (1988), show that the nature of the decision process provides many opportunities for positive feedback and the appearance of nonlinear behavior in asset prices. For example, studies show that when faced with a complex purchase, decision makers tend to anchor on the price and changes in the price as indicators of value. Also, decision makers over- weight more recent evidence, tend to be improperly regressive (that is, to forecast the continuance of trends), overweight the value of consensus beliefs, seek confirming evidence, and wish to be part of the group, all of which encourage positive feedback. Baumol noted that, given the complexity of stock evaluation, decision makers will end up focusing on similar pieces of information, called focal points. This behavior, coupled with the fact that new stock supply tends to be relatively inelastic in the short run, tends to exacerbate positive-feedback effects.

The observation that stock-price breaks are more severe than upward adjustments is also con- sistent with investors being loss averse, tending to focus on negative information when under stress, overweighting the probability of negative events, and becoming more loss averse as downward movements in the value of their portfolios remind them of their incomplete personal control. Abolafia (1996) demonstrated how professional investment managers might encourage positive feedback as a result of acceptance of and commitment to a com- mon institutional culture, set of beliefs, and prac- tices. Neoclassical economic models assume that negative feedback always dominates, however, and that prices tend toward stability. These assumptions explain why current financial models, based on stable equilibriums, tend to do poorly, especially in times of financial market upheaval when uncertainty is great and positive feedback is likely to become prominent.

The discovery of nonlinearity in security prices and the fact that outcomes can be predicted only within wide limits also have normative implica- tions for financial decision making. In particular, Hammond (1996) noted that optimization strate- gies tend to be brittle, in that when they fail, they create greater variance in outcomes than more- robust satisficing strategies. Therefore, given imperfect market predictability, to avoid catastro- phe, investors might prefer to use decision pro- cesses that preserve appropriate future financial flexibility. Empirical evidence indicates that inves- tors do attempt this approach, but the ways they do it can be improved with education and training.

16 Association for Investment Management and Research

Behavioral finance also appears to offer an explanation for what some observers label exces- sive stock-price volatility. In the lively debate going on among academics (see White 1990) about whether stock prices exhibit excessive volatility, the bogey is a price based on underlying concepts of present value, rational economic man, and mar- ket efficiency. Reliable evidence of excessive vola- tility is scarce because empirical tests using market data imply that researchers have some knowledge of the (unknowable) valuation model used by investors and because experimental market inves- tigations assume similar investor behaviors in syn- thetic and in real markets. But financial economists appear to agree that security-price volatility and trading volume should vary directly with the divergence of investor opinion (see Schwartz 1988). At this point, standard finance is unable to explain a wide divergence of opinion except to invoke the concept of asymmetrical information. In public markets for widely traded securities, however, where asymmetries are likely to be small, it seems unlikely that differential information among inves- tors could create the kind of divergence of opinion necessary to account for many instances of high stock-price volatility.

Behavioral finance offers a second source of opinion difference. This source originates in the decision processes investors use to manipulate data and arrive at estimates of securities’ values. As noted in Payne (1993), research supports the notion that an investor’s decision process is adap- tive to the perceived nature of the problem and the environment in which the problem is framed. In particular, studies have documented that, despite appearing to be minor, differences in the perceived complexity of a decision, in the decision maker’s emotional state, in the time available to make the decision, in the reversibility of the decision, and even in the format of the presented information have important influences on the process used to arrive at a decision. As a result of the adaptive nature of the decision process itself, even if all investors were presented with the same set of data, differences in data perception, selection, weight- ing, and manipulation would prevail. If the deci- sion process is adaptive, market volatility will not only vary directly with information asymmetry but will also be greater in situations characterized by greater difference in decision processes.

Because of the complexity of stock selection, the decision process is likely to vary widely among investors. In particular, stock valuation appears to be the quintessential complex, ill-structured task. “Complex, ill-structured tasks,” as they are for- mally called, dominate in the social sciences. Such

tasks are characterized ex ante by lack of a unique set of characteristics that clearly define the method and information needed to arrive at a single, well- defined goal. Such tasks are also usually distin- guished by statistical overdetermination (in which too many variables are present relative to the num- ber of pieces of data), low base rates of occurrence, and ambiguous antecedent behavior states (that is, a lack of clarity about the conditions before the phenomenon was described). Ex post, the perfor- mance of few “experts” on complex, ill-structured tasks surpasses naive strategies, and when the experts do outperform, the margin of superior per- formance is small and inconsistent. The majority of empirical studies of the performance of profes- sional investment managers report just such results. Moreover, Charles Ellis (1993), the “Dean of Portfolio Management,” implicitly defines invest- ment management as a complex, ill-structured task when he says,

Portfolio Management is neither art nor science. It is instead a special problem of determining the most reliable and efficient way to reach a goal, given a set of policy constraints, and working within a remarkably uncertain, probabilistic, always changing world of partial information and misinformation all filtered through the inexact prism of human interpretation. (p. 51)

Finally, a survey of the current literature on security analysis and portfolio management clearly identi- fies the large number of techniques and wide range of opinions that exist among investment profession- als concerning approaches to stock valuation.

Complex, ill-structured tasks or decisions give rise to great variability in decision outcomes because they tend to lie more toward the experien- tial or intuitive end of the decision spectrum than the objective end and make greater use of idiosyn- cratic information and procedures that are per- sonal, concrete, holistic, affective (emotional), and based on such associative conventions as the use of analogies and stereotypes (Forgas 1991, Epstein 1994, Hammond 1996, and Busemeyer 1995). In addition, because information is both qualitative and quantitative, “stories” are often constructed to achieve integration and to evaluate the coherence and completeness of a proposed decision or fore- cast. Payne suggested that decision makers tend to trade off decision accuracy against effort. Thus, when decisions become complex, decision makers shift toward using rules of thumb and noncompen- satory procedures (that is, decision rules in which trade-offs do not involve weighting of attributes) because their use is less costly than more complete quantitative approaches. The impetus toward less effortful strategies will strengthen when accurate

Financial Analysts Journal • March/April 1998 17

predictions are hard to achieve and responsibility is difficult to apportion. Both of these conditions apply to stock evaluation, where return causality is difficult to estimate.

Empirical studies have demonstrated that decision procedures become more intuitive or experiential the more reversible the decision, the more emotional the decision, and the greater the time stress. Therefore, the more reversible, emo- tional, or subject to time stress the trading in a stock is, the greater price volatility might be expected. In addition, as Miller (1977) noted, unless arbitrage opportunities are complete, larger divergence of opinion will lead not only to greater price volatility but also to higher equilibrium market prices.

Where To for Financial Research? Ellen Langer of Harvard University, in her recent review of research on learning and creativity, made three points that might serve as touchstones for appraising future initiatives in financial research. First, data should not be viewed as stable commod- ities but as sources of ambiguity and opportunities for creative thought. Second, the way in which any person tends to construct a world vision of reality is usually only one construction among many. Third, those who have considerable familiarity with the available data but have not let themselves become locked into a particular perspective are the most likely to make conceptual contributions that advance understanding in any area of research.

Some have suggested that finance as an aca- demic discipline is in danger of ossification or trivi- alization because it has become locked into a general conceptual scheme (mathematical modeling) and into a specific conceptual scheme (market efficiency based on economic rationality). Without a doubt, the collaboration between economics and finance that brought these schemes into prominence in finance has been productive. For example, our understand-

ing of diversification, market equilibrium, and asset pricing have been much improved through the development of the CAPM, arbitrage pricing theory, and option-pricing models. Nevertheless, sooner for some and later for others, the implications of the principle of diminishing returns becomes obvious, and those who are interested in sustaining the rate of progress in a discipline begin to look outside the discipline for new sources of hybrid vigor. Just as the label “financial economist” was attached to those who sought hybrid vigor from economics, “financial behaviorist” may be the label attached to those who search for hybrid vigor in the mature discipline of psychology, especially behavioral deci- sion making.

To those finance traditionalists who are anxious about the entrance of the behaviorist into the finance tent, I offer the following perspective. First, the fact that financial researchers have almost always found it necessary to offer ex post behavioral explanations for numerical results indicates that behavior has always been important and worthy of study in the field of finance. Second, the recent expansion of traditional financial research into agency theory and corporate governance suggests that individual and group behavior is already of interest to some financial researchers. Third, traditional financial research has recently empha- sized the role of information, especially asymmet- rical information, in asset pricing and market behavior. Such an emphasis is not inconsistent with the thrust of behavioral finance, which seeks to expand the information set from a focus on the decision facts themselves to include information about the context in which decisions are made. Finally, in the past, finance and economics have gained by introducing behavioral concepts into explanations of the decision process. For example, consider the benefit of the substitution of the concept of utility for money by Bernoulli in his solution to the St. Petersburg Paradox.

References

Abolafia, Mitchel. 1996. Making Markets: Opportunism and Restraint on Wall Street. Cambridge, MA: Harvard Press.

Anand, Paul. 1993. Foundations of Rational Choice under Risk. New York: Oxford Press.

Ashton, Robert H. 1995. Judgment and Decision Making Research in Accounting and Auditing. Cambridge, United Kingdom: Cambridge Press.

Baron, Jonathan. 1994. Thinking and Deciding. Cambridge, United Kingdom: Cambridge Press.

Barrow, Jerome H. 1992. The Adapted Mind. New York: Oxford Press.

Bauman, W. Scott. 1967. “Scientific Investment Analysis: Science or Fiction?” Financial Analysts Journal, vol. 23, no. 1 (January/ February):93–97.

Baumol, William. 1965. The Stock Market and Economic Efficiency. New York: Fordham University Press.

———. 1989. “Chaos: Significance, Mechanism and Economic Applications.” Journal of Economic Perspective, vol. 3, no. 2 (Winter):77–105.

Burrell, O.K. 1951. “Possibility of an Experimental Approach to Investment Studies.” Journal of Finance, vol. 6, no. 2 (June):211–19.

18 Association for Investment Management and Research

Busemeyer, Jerome. 1995. Decision Making from a Cognitive Perspective. San Diego, CA: Academic Press. De Bondt, Werner F.M. 1995. “Investor Psychology and the Dynamics of Security Prices” In Behavioral Finance and Decision Theory in Investment Management. Edited by Arnold S. Wood. Charlottesville, VA: Association for Investment Management and Research:7–13.

Ellis, Charles. 1993. Investment Policy, 2nd ed. Homewood, IL: Business One Irwin.

Epstein, Seymour. 1994. “Integration of the Cognitive and the Psychodynamic Unconscious.” American Psychologist, vol. 149, no. 8 (August):709–24.

Eve, Raymond, Sara Horsfall, and Mary E. Lee. 1997. Chaos, Complexity and Sociology. Thousand Oaks, CA: Sage Publications.

Fama, Eugene. 1992. “The Cross Section of Expected Stock Returns.” Journal of Finance, vol. 47, no. 2 (June):427–67.

Farrelly, Gail. 1980. “A Behavioral Science Approach to Financial Research.” Financial Management, vol. 9, no. 3 (Autumn):15–28.

Forgas, Joseph. 1991. Emotion and Social Judgment. New York: Pergamon Press.

Frisch, Deborah, and Robert Clamen. 1994. “Beyond Expected Utility.” Psychological Bulletin, vol. 116, no. 2 (July):46–54.

Hammond, Kenneth. 1990. The Integration of Research in Judgment and Decision Theory. Boulder, CO: University of Colorado.

———. 1996. Human Judgment and Social Policy. New York: Oxford Press.

Harlow, W.V., and Keith Brown. 1990. “Understanding and Assessing Financial Risk Tolerance: A Biological Perspective.” Financial Analysts Journal, vol. 46, no. 6 (November/ December):50–62.

Heap, Shaun. 1989. Rationality in Economics. New York: Blackwell.

Hogarth, Robin. 1988. Judgment and Choice. New York: John Wiley & Sons.

Kahneman, D., and A. Tversky. 1979. “Prospect Theory: An Analysis of Decision under Risk.” Econometrica, vol. 47, no. 2 (March):263–91.

———. 1992. “Advances in Prospect Theory: Cumulative Representation of Uncertainty.” Journal of Risk and Uncertainty, vol. 5, no. 4 (October):297–323.

Langer, Ellen. 1997. The Power of Mindful Learning. New York: Addison–Wesley.

Lipman, B. 1991. “Modeling Limited Rationality.” Econometrica, vol. 59, no. 4 (July):1105–25.

Markowitz, H. 1952. “Portfolio Selection.” Journal of Finance, vol. 7, no. 1 (March):77–91.

Miller, Edward. 1977. “Risk, Uncertainty and Divergence of Opinion.” Journal of Finance, vol. 32, no. 4 (September):1151–68.

Muth, M.F. 1961. “Rational Expectations and the Theory of Price Movements.” Econometrica, vol. 29, no. 3 (July):315–35.

Nisbett, Richard, and Lee Ross. 1980. Human Inference: Strategies and Shortcomings of Social Judgment. Englewood Cliffs, NJ: Prentice-Hall.

Payne, John W. 1993. The Adaptive Decision Maker. Cambridge, United Kingdom: Cambridge Press.

Peters, Edgar. 1994. Fractal Market Analysis: Applying Chaos Theory to Investment and Economics. New York: John Wiley & Sons.

Plous, Scott. 1993. The Psychology of Judgment and Decision Making. New York: McGraw-Hill.

Richards, Diana. 1990. “Is Strategic Decision Making Chaotic?” Behavioral Science, vol. 35, no. 3 (July):219–32.

Robertson, Thomas. 1991. Handbook of Consumer Behavior. Englewood Cliffs, NJ: Prentice-Hall.

Roll, Richard. 1977. “A Critique of the Capital Asset Pricing Theory’s Tests.” Journal of Financial Economics, vol. 4, no. 2 (June):129–76.

Rosser, J. Barkely. 1996. “Chaos Theory and Rationality in Economics.” In Chaos Theory in the Social Sciences, edited by L.D. Kiel. Ann Arbor, MI: University of Michigan Press:199–213.

Schwartz, Robert A. 1988. Equity Markets: Structure, Trading and Performance. New York: Harper Business.

Shefrin, Hersh, and Meir Statman. 1994. “Behavioral Capital Asset Pricing.” Journal of Financial and Quantitative Analysis, vol. 29, no. 3 (September):323–49.

Shiller, Robert. 1989. Market Volatility. Cambridge, MA: MIT Press.

Simpson, Jeffrey A. and Douglas Kenrick. 1997. Evolutionary Social Psychology. Mahwah, NJ: Lawrence Erlbaum.

Slovic, Paul. 1969. “Analyzing the Expert Judge: A Study of a Stockbroker’s Decision Process.” Journal of Applied Psychology, vol. 53, no. 4 (August):255–63.

———. 1972. “Psychological Study of Human Judgment: Implications for Investment Decision Making.” Journal of Finance, vol. 27, no. 4 (September):779–801.

Slovic, Paul and W. Scott Bauman. 1972. “Analyzing the Use of Information in Investment Decision Making: A Methodological Proposal.” Journal of Business, vol. 45, no. 2 (April):283–301.

Sterman, John. 1988. “Deterministic Chaos in a Model of Human Behavior.” Systems Dynamics Review, vol. 4, no. 1 (January):148–78.

Thaler, Richard. 1991. Quasi Rational Economics. New York: Russell Sage Press.

———. 1993. Advances in Behavioral Finance. New York: Russell Sage Press.

Vriend, Nicolaas. 1996. “Rational Behavior and Economic Theory.” Journal of Economic Behavior and Organization, vol. 29, no. 2 (March):263–85.

White, Eugene. 1990. Crashes and Panics: The Lessons from History. Homewood, IL: Irwin One.

Wood, Arnold. 1995. Behavioral Finance and Decision Theory in Investment Management. Charlottesville, VA: Association for Investment Management and Research.

Yates, Frank. 1992. Risk Taking Behavior. New York: John Wiley & Sons.

———. 1993. Judgment and Decision Making. Englewood Cliffs, NJ: Prentice-Hall.

Zuckerman, Marvin. 1994. Behavioral Expressions and Biosocial Bases of Sensation Seeking. Cambridge, United Kingdom: Cambridge Press.