BEHAVIORAL FINANCE AND THE SOURCES OF ALPHARUSSELL , CFAPRESIDENT OF FULLER &THALER ASSET MANAGEMENTLUGANO,JUNE 200015The information contained herein is based on data believed to be reliable,but not guaranteed by ,the present publication is to be viewed as being purely informative in nature and does not commit the rights is allowed provided that the source is properly noted.
BSI GAMMA Foundation (Global As-set Management Methods and Ap-plications),is a private Swiss founda-tion established by BSI AG to markthe Bank's 125th mission of BSI Gamma Foundation is topromote the development of theoretical andempirical research in the following domains:-Asset management.-The functioning of capital markets,including asset pricing and the competi-tive structure of the fundmanagementindustry.-The role and form of public regulation infund public announcements,researchersfrom the international academic community areinvited to submit research final se-lection by the Board of the Foundation will bebased on the following criteria:relevance withthe objectives of the Foundation and provenacademic Foundation will facilitate the organization ofpublic conferences to dissiminate the results ofthese research conferences will beaddressed to the general public,while others willbe designed for for making research proposal can be obtained directly from BSI Gamma Foundation c/o BSI AGVia Magatti 2 - CH-6900 LuganoTel.+4191/809 33 38,Fax +4191/809 36 78E-Mail:@ from Dermine,Chairman of BSI Gamma Foundation,c/o INSEAD,Boulevard de Constance,77305 Fontainbleau Cedex (FRANCE),Tel:33 1 60 72 42 27,Fax:33 1 60 72 42 42,E-mail:@
Behavioral Finance and the Sources of AlphaRussell J. Fuller, CFAPresidentRJF Asset Management1300 El Camino Real, Suite 504San Mateo, CA 94402(650) 572-8334February 6, 2000ForthcomingJo,u rnal of Pension Plan Inve, sWtiningter 1998, Vol. 2, No. 3C:\My Documents\Word\Misc\
ABSTRACTBehavioral finance is a new field in economics that has recently become a subject ofsignificant interest to investors. This article provides a general discussion of behavioralfinance and presents some insights from this field that apply to the problems plansponsors face when evaluating and selecting active equity is Behavioral ....?................................................................................1.....Better Expectations: The Mother of All. .A....h..................................................4.....Three Sources of ..h..a...........................................................................................5...Superior (Private) InformationProcess Information BetterBehavioral BiasesTypes of Behavioral Bia..s..e..s..................................................................................8.....Non Wealth-Maximizing BehaviorHeuristic Biases and Sysitce mMeanttal MistakesHeuristics, Heuristic Biases and Optical .....n..s..........................................1..0..Heuristic Biases and Errors in ......s.....................................................1.. Plan Sponsors Should Ask Prospective. .M.....a..g..e...........................................................................................................................................2....s...............................................................................................................2..1.
Behavioral Finance and the Sources of AlphaBehavioral finance is a relatively new field in economics that has become a “hottopic” for investment professionals. For example, a large number of conferences orientedtoward investors have recently featured sessions on behavioral finance. However,because the field is so new, most professionals responsible for large portfolios were notexposed to the principles of behavioral finance in their college curricula and theseprinciples have significant practical implications for investment , this article provides an overview of behavioral finance. The conclusionpresents some insights from behavioral finance that specifically apply to the problem plansponsors face when evaluating and selecting active equity managersWhat is Behavioral Finance?Behavioral finance has recently become a subject of significant interest to it is a relatively new and evolving field in economics and consequently not welldefined, a legitimate question is: “What exactly is behavioral finance?” I personallydescribe behavioral finance in the following ways:• Behavioral finance is the integration of classical economics and finance withpsychology and the decision-making sciences.• Behavioral finance is an attempt to explain what causes some of the anomaliesthat have been observed and reported in the finance literature.• Behavioral finance is the study of how investors systematically make errors injudgement, or “mental mistakes.”
All economic models make simplifying assumptions about both market conditionsand the behavior of market participants. Sometimes the simplifying assumptionsunderlying the model are explicitly stated and sometimes the assumptions are implicit –the latter is often the case regarding the behavioral assumptions underlying the illustrate, consider the efficient market hypothesis (EMH), an economic model ofconsiderable importance to investors. The simplifying assumptmionasrk reetgarding condition tshat underlie the EMH frequently include, among others, assumptions such as:• Transaction costs are zero.• Markets are not segmented.• Easy (even unlimited) entry into the security markets ehavioral assump ttihoants underlie the EMH can be expressed as:• Investors act, in an unbiased fashion, to maximize the value of their portfolios.• Investors always act in their own first behavioral assumption is frequently stated as investors are “rational expectationswealth maximizers” – this means that investors form unbiased expectations of the futureand given these expectations, they buy and sell in the securities markets at prices whichthey believe will maximize the future value of their finance questions whether the behavioral assumptions underlying theEMH are true. For example, consider the assumption that individuals always act in theireconomic self-interest. Suppose you are having dinner at an out-of-town restaurant and itis extremely unlikely that you will ever return to this restaurant. Do you leave a tip?2
Most people do, but in this case leaving a tip decreases, rather than increases one’swealth, and because you won’t be returning to this restaurant there are (presumably) no“costs” associated with not leaving a tip. In this case leaving a tip violates the rationalexpectations and self-interest germane to the EMH, consider “social investing” such as arbitrarily decidingnot to invest in tobacco stocks or deciding to overweight environmentally cleanindustries, etc. Such behavior is not consistent with pure wealth maximization, if for noother reason than opportunities for forming better-diversified portfolios are investors might engage in non-wealth maximizing behavior, and what are theimplications of such behavior for security pricing, are areas of inquiry in aspect of behavioral finance concerns how investors form expectationsregarding the future and how these expectations are transformed into security in cognitive psychology and the decision sciences have documented that,under certain conditions, people systematically make errors in judgement or mentalmistakes. These mental mistakes can cause investors to form biased expectationsregarding the future that, in turn, can cause securities to be considering that investors may not always act in a wealth maximizing manner andthat investors may have biased expectations, behavioral finance may be able to explainsome of the anomalies to the EMH that have been reported in the finance returns such as those associated with “value” stocks, earnings surprises,3
short-term momentum and long-term price reversals are fertile ground for researchers inbehavioral Expectations: The Mother of All AlphasAlpha can be defined as the difference between a portfolio’s risk-adjusted return andthe return for an appropriate benchmark portfolio. Most active investors are trying tomaximize alpha. In contrast, passive investors generally accept the EMH and merely tryto match the benchmark the process of forming portfolios, active investors buy and sell securities based ontheir expectations about the future – typically expectations regarding the futureprofitability and risk characteristics of the firm issuing the securities. For example, acompany’s stock price may be largely determined by the consensus expectation (themarket’s expectation) regarding the company’s future earnings. (From this point forwardI will refer to the consensus of investors, as reflected in current prices, as the “market.”)If today’s stock price is based on the market’s expectation regarding the future, thenin order to predict tomorrow’s stock price change, one must have better expectationsabout the future than the market. In this sense, the mother of all alphas is the ability toform expectations that are better than the market’s expectations. Consequently, for anactive investment manager to claim that he can generate above normal returns (a positivealpha) in the future, he must argue that, in some manner, his expectations regarding thefuture are better than the market’s expectations. 1 These anomalies have been documented by many studies. For example, with respect to the“value” anomaly, see Lakonishok, Shleifer and Vishny [1994]; with respect to earnings surprise,see Bernard and Thomas [1990]; with respect to short-term momentum, see Jegadeesh and4
Three Sources of AlphaIf having better expectations than the market is the mother of all alphas, the issue ofhow expectations are formed is important in order to understand the sources, or causes, ofalpha. In general, when forming expectations people use:• a set of information• procedures (models) for processing the informationThis suggests two potential sources of alpha:1. Superior (Private) Information: Most traditional investment managerstry to generate a better information set. For example, they may try togenerate a superior earnings forecast, or they may try to better understandthe economics underlying a particular industry’s profitability. These typesof managers are frequently referredt ratod itiaosn al mana georsfundamental . Process Information Better: Some investment managers assume thatmost information is commonly available to all investors and focus theirenergy on trying to develop better procedures for processing thisinformation. Managers that try to do this in a formal way are frequentlycalledq uantitative is important to note that if a particular investor has superior information, this doesnot necessarily imply that the market’s expectations are biased – the market cannotincorporate information it does not have into its expectations in either a biased orunbiased fashion. Rather, in this case, the market’s expectations are simply not as good Titman [1993]; with respect to long-term price reversals, see DeBondt and Thaler [1985, 1987,and 1990].5
as they would be if the market had the private information. The key question to consider,however, is: What are the probabilities of any individual investor or investment managerconsistent lgyathering superior, private information when so many other investors aretrying to do the same thing?A similar point can be made with respect to processing information. If the marketdoes not know the best way to process information, it may still form unbiasedexpectations that are simply not as good as they might be if the market knew the bettermethod for processing information. To illustrate, suppose a three-factor model generatesreturns, but the market uses a two-factor model in processing information. In this casethe market’s expectations might be unbiased, but not as good as the expectations of anindividual investor using the correct three-factor model. Again, the key question to askis: What are the probabilities of any single investor or investment manager discoveringthe true factor model when so many other investors are trying to do the same thing?There is a third possible source of alpha:3. Behavioral Biases: Scholars in psychology and the decision makingsciences have documented that in some circumstances investors do not tryto maximize wealth and in other circumstances investors make systematicmental mistakes. Both of these cases can result in mispriced securities andboth are the result of behavioral conditions under which behavioral biases occur and how they might affect securitypricing are discussed in the next section. For now, simply note that there are threepossible sources of alpha: Superior (private) information, better methods for processinginformation and behavioral biases. These are illustrated in Exhibit
EXHIBIT 1THREE SOURCES OF ALPHAExploitInformationBetter“Model”ExploitBehaviorQuantitativeTraditionalManagersBehavioralManagersManagersAs mentioned earlier I generally think of investors who try to generate superiorinformation as fundamental or traditional, manadn atgher smajority of active investorsare in this group. For example, almost all of Wall Street’s research represents an attemptto generate superior information. I also tend to think of investors who try to developbetter procedures (better models) for processing information as quantitative managers try to exploit situations where securities are mispriced by themarket because of behavioral course there is overlap between all three types of managers. For example, manytraditional managers who primarily focus on generating, say, better earnings estimates7
may also try to develop better ways for processing such information. Similarly, manytraditional and quantitative managers may also try to exploit behavioral of Behavioral BiasesIt would be convenient if a taxonomy existed for classifying behavioral biases thatwas generally accepted by scholars in the field. However, because behavioral finance is arelatively new field, no such classification scheme currently exists, at least to myknowledge. Thus, I will present my personal method for classifying behavioral biasesthat affect investor behavior, and the reader should be aware that scholars maysubsequently develop better classification , it is important to understand that we are not dealing with truly irrationalbehavior. For example, suppose someone prefers a gift, with no strings attached, of$1,000 to a gift of $2,000. In the economic sense this would be truly irrational behaviorand I, for one, do not expect to observe such behavior in the securities types of behavioral biases we do observe in securities markets fall into two broadcategories:1. Non Wealth-Maximizing Behavior: The economist's view of rationalbehavior assumes that investors act only to maximize the expected valueof their portfolios. In fact, investors may maximize other things that aremore important to them than their . Heuristic Biases and Systematic Mental Mistakes: Heuristic biasescause investors to make systematic mental mistakes and as a resultincorrectly process available information. Before the fact investorsbelieve they are correctly processing information and acting in a manner8
which maximizes their expected wealth. After thme afya dcits cthoevye rthe mental mistake, but frequently they are not even aware of the are many examples of non wealth-maximizing behavior. Agency problemsrepresent one broad class of this type of behavior. "Window dressing" at the end of aquarter or year is an example. Selling stocks just before the end of a quarter which havebeen big losers and buying stocks that have been big winners will not raise the portfolio'sreturn and the associated transaction costs of trading may actually lower the , the portfolio manager may have an easier time at the quarterly client meeting ifhighly visible "losers" are not in the portfolio at the end of the concept of regret can cause another class of problems that result in non wealth-maximizing behavior. Kahnamen and Tversky's (1979) prospect theory formallyaddresses the fact that for most investors the pain associated with losses exceeds thepleasure of gains. One manifestation of this is the fact that investors tend to hold ontotheir loser too long and tend to sell their winne2rs too also suffer from a lack of self-control, which can lead to non-wealthmaximizing behavior. Statman (1995) shows that dollar cost averaging is sub-optimalwith respect to maximizing wealth. Nevertheless, dollar cost averaging is used by manyinvestors who apparently lack the discipline (and fortitude) to invest all of their wealth inrisky assets at one point in , for non-wealth maximizing behavior to result in mispriced securities, themarket as a whole must engage in this type of behavior, as opposed to isolated , for an investment strategy to be able to exploit such mispricings, the market mustengage in non-wealth maximizing behavior in a systematic fashion. These two conditions9
are not likely to be met. On the other hand, heuristic biases can cause the majority ofinvestors to make systematic mental errors. These mental mistakes, in turn, cause themarket to have biased expectations and, as a result, misprice securities. Thus, heuristicbiases are potentially , Heuristic Biases and Optical IllusionsHeuristics are rules of thumb, or mental shortcuts, the human brain uses to quicklysolve complex problems. For example, a billiard player does not solve the trigonometricand differential equations needed to determine at what angle and speed to hit the cue ballin order to put another ball in the correct pocket. Rather, a billiard player uses rules ofthumb and mental shortcuts that allow him to play the game, even though he may notunderstand the mathematics. Thus, heuristics are very useful, powerful problem , when used in the wrong situation, heuristics can cause people to makesystematic mental mistakes. Optical illusions are a simple way of illustrating heuristicbiases. Vision is a very complex problem for the brain to solve. The eye generates atremendous amount of information which must be quickly analyzed and interpreted bythe brain to form the images which we "see." Over many thousands of years the humanbrain has developed mental shortcuts for interpreting vision data and these visionheuristics typically work quite well. However, when these mental shortcuts for solvingvision problems are used in the wrong context, a heuristic bias results in an opticalillusion. To illustrate, look at Exhibit 2 on the next page and answer the question: Whichof the two vertical straight-line segments appears to be longer? 2 See Shefrin and Statman (1985) and O'Dean (1996).10
EXHIBIT 2OPTICAL ILLUSIONS AND HEURISTIC BIASESThe answer to this question (assuming the respondent is human and not cheating) isthe vertical straight line on the left "looks" longer. In fact, the two lines are drawn to beexactly the same length. The optical illusion results from the brain using heuristics forsolving three-dimensional vision problems when the lines are drawn on a two-dimensional surface. What "tricks" the brain into thinking it is receiving three-dimensional data from the eye are the arrowheads drawn on the end of each vertical linesegment. The reverse arrowheads drawn on the left line segment give the illusion ofdepth and the line appears to be farther away from the eye than it really is. The regulararrowheads on the right line also give the illusion of depth, but in this case the brain istricked into thinking the line is closer to the eye that it actually is. The brain uses thefollowing heuristic for converting three-dimensional vision information: Objects that are11
farther away are bigger than they appear and objects that are close are smaller than theyappear. (For example, a house observed from a long distance creates a very small imageon your eye's retina, but you still know the house is a large object.) Thus, while bothvertical line segments create the same image on your eye's retina, your brain converts thisinformation so that the vertical line on the left, which appears to be farther away, islonger than the vertical line on the right, which appears to be interesting, and important, aspect of heuristic biases is that they are very difficultto overcome. For example, now that you understand what causes the optical illusion,look again at the two vertical line segments in Exhibit 2. Which vertical line appearslonger? If you are typical, the line on the left still appears longer, even though you knowthe two line segments are the same length and you understand what causes the opticalillusion. Heuristics, which have evolved over thousands of years, can be thought of asbeing part of the brain's hardware. Unlike software, these mental shortcuts are such usefuland powerful problem solving tools they simply cannot be reprogrammed. However,when a heuristic is used in the wrong situation (when the wrong rule-of-thumb is used tosolve a problem), a heuristic bias causes the person to make a mental Biases and Errors in ExpectationsToday's stock prices are based on the market's expectations about the future. If themarket has biased expectations, then stocks may be mispriced. If an investor has unbiasedexpectations, and knows the direction of the market’s bias, she can predict future pricechanges and consequently form portfolios that have a future alpha. As the optical illusionillustrates, heuristics, when used in the wrong situation, can cause mental mistakes -- in12
the case of the investment process, heuristics can cause investors to form use many different heuristics in solving problems. This section willconcentrate on only four heuristics that are closely related to the most common mentalmistakes, or errors in judgement, made by3 iTnhveseto are:1. Representativeness: This heuristic is the source of the adage, "if itlooks like a duck and quacks like a duck, it probably is a duck." Withrespect to forming expectations, people evaluate the probability of anuncertain future event by the degree to which it is similar to recentlyobserved events. Representativeness can causeo ivnevresatcotrs to to new information, ., investors give new information too muchweight in forming their expectations about the . Saliency: For events which occur infrequently, people tend tooverestimate the probability of such an event occurring in the future ifthey have recently observed such an event. For example, commercialairplane crashes occur infrequently. However, if an airplane crash hasrecently been prominently reported in the media, people will greatlyoverestimate the probability of a crash occurring in the can cause investovres rteoa ct to new . Overconfidence: People are grossly overconfident regarding theirability and their knowledge. For example, when people say that theyare 90 percent sure that an event will happen or that a statement is true,they typically are correct less than 70 percent of time. Overconfidencecan cause investorusn tdoe rreact to new information. 3For a more complete treatment of heuristics and heuristic biases, see Shiller (1997) or DeBondtand Thaler (1994).13
4. Anchoring: Psychologists have documented that when people makequantitative estimates, their estimates may be heavily influenced byprevious values of the item. For example, it is not an accident that aused car salesman always starts negotiating with a high price and thenworks down. The salesman is trying to get the consumer anchored onthe high price so that when he offers a lower price, the consumer willestimate that the lower price represents a good value. Anchoring cancause investorsu ntod erreact to new are based on a set of information and procedures for processing theinformation. A simple way of thinking about how the market can form biasedexpectations is that the market can either overreact or underreact to new means investors place too much weight on recent, new information informing their expectations regarding future events. Underreaction means investors givetoo little weight to recent, new noted above, the representativeness and saliency heuristics can cause investors tooverreact to new information. These heuristic biases are probably the source of alpha formost value and contrarian strategies. For example, suppose earnings changes follow arandom walk for most comp4 aIfn itehsis. is true, then tomorrow's earnings change is justas likely to be positive, as it is negative for any particular company. However, if themajority of investors are vulnerable to representativeness bias, they might naivelyextrapolate a recent negative earnings change for a company far into the future,particularly if the recent earnings announcement is very vivid, or salient. In this case, themarket's expectations regarding the company's future profitability are biased downward, 4 Many researchers have shown that a random walk is a reasonable model for earnings a recent article, see Fuller, Huberts and Levinson (1993). However, these authors also argue14
the price falls and the stock sells at a P/E that is “too low” relative to trailing mispricing results in the stock subsequently generating a positive is important to note that the use of heuristics results in moennlyta ul nmdiesrtakes certain nTshese conditions must occur infrequently, or the heuristic would nothave successfully evolved over time. Thus, most stocks with low P/E ratios are probablycorrectly priced. To find the mispriced stock from among the universe of low P/E stocks,one should try to determine which stocks are subject to the conditions that causeinvestors, as a group, to be vulnerable to the heuristic biases that cause flip side of overreaction is underreaction. It is important to note that differentheuristics (anchoring and overconfidence) cause underreaction and the conditions underwhich investors are vulnerable to these heuristics are different from the conditions thatcause investors to be vulnerable to is probably the primary source of the alpha associated with earningssurprise and short-term momentum strategies. This is easy to see with respect to earningssurprises. Suppose a company had reported EPS of approximately $ for manyquarters and the analysts' current estimates are also around $1. Suppose the company thenreports quarterly EPS of $, which represents a large change in earnings and a largeearnings surprise. If the analysts are overconfident and also anchored to their most recentestimate, they may be reluctant to give as much weight as they should to the informationin the current earnings announcement and not raise their estimate. However, if the changein earnings that caused the surprise is permanent, over time the analysts will figure thisout, slowly raise their estimates and the stock price will drift upward after the earnings that the market forecasts earnings better than a simple random walk model and as a result, lowP/E stocks tend to have lower future growth than high P/E
announcement, generating the well-documented post-announcement alpha associatedwith earnings , it is important to note that analysts are unlikely to be vulnerable to theheuristics biases associated with anchoring and overconfidence, except under unusualconditions. And, of course, if the earnings change that caused the surprise is notpermanent, the analysts will be correct if they do not raise their estimates. Thus, the keysto exploiting the true source of the earnings surprise alpha are determining under whatconditions analysts are likely to be overconfident and anchored, and whether the earningschange associated with the surprise is permanent in 3 summarizes the manner inin wvehsictohr s incorporate new informationinto their expectations regarding future 3BEHAVIORAL BIASES AND ERRORS IN EXPECTATIONSA Distribution of Expectations Based on New Information100%80%60%40%20%0%Over ReactionCorrectly ProcesseUdnder Reaction16
Most of the time the information is correctly processed and the market's expectationsare unbiased. Under some conditions the market may overreact to the information,resulting in biased expectations. This overreaction is likely due to biases associated withthe representativeness and saliency heuristics and may be the source of the alpha for mostvalue and contrarian strategies. Under other conditions the market may underreact to thenew information, which again results in the market's expectations being biased and stocksbeing mispriced. Underreaction is likely due to biases associated with theoverconfidence and anchoring heuristics and may be the source of the alpha for mostmomentum and earnings surprise Plan Sponsors Should Ask Prospective ManagersBehavioral finance provides a number of insights that should be useful to plansponsors when evaluating and selecting active investment managers. The primary, if notthe sole reason for selecting and hiring an active manager is the belief that the managerwill provide a positive alpha in the future. If one agrees with the proposition that currentprices are based on the market's expectations regarding the future, then the genesis of allalphas is having better expectations than the expectations are formed by processing an information set, two potentialsources of alpha are: 1) Generating a better information set by obtaining information thatthe market does not have, ., private information; 2) Processing information better thanthe market. tAh ird potential source of alpha is: 3) Exploiting behavioral factors thatcause the market's expectations to be
One of the first questions to ask a prospective manager is: "What is the source of youralpha?" If the manager cannot describe the source of his alpha, you may want to politelyend the interview. Most managers, of course, will have an answer to this question, butthe answer should fall into one or more of the three sources of alpha noted second question might be: "Can you specifically relate the source of your alpha tothe past returns of portfolios you have managed?" If a manager answered the firstquestion by indicating he generates private information that the market does not have,then ask him to give specific examples of private information he obtained for specificstocks. Just as importantly, ask him to explain how he obtained this information. Forexample, he may argue that his firm employs the best analyst in, say, the semiconductorindustry, and this analyst formed an earnings estimate for Intel that was superior to theconsensus estimate. Follow up questions might include: "Is the analyst still employed bythe manager? Why does the manager believe this particular analyst is superior to themany smart, highly trained analysts following Intel? Does the manager have othersuperior analysts and, if so, how is the manager able to hire so many superior analystswhen so many other managers are also trying to hire these superior analysts?” Mypersonal view is that, given the amount of competition among analysts and thecompetition among managers to hire superior analysts, it is not very likely that anyinvestment manager will have enough superior, private information to generate asignificant alpha in the the other hand, if the manager claims the source of his alpha is that his firmprocesses information better than the market, ask for specific details as to how his firm18
does this. If the manager responds that they simply understand an industry better, or theysomehow have a better sense of value, I would be very the manager is a pure quantitative manager and argues that his firm has the bestmodel, ask for specific details about their "black box." Although not as ubiquitous astraditional fundamental managers, there are quite a few quantitative managers, all of whoare trying to develop the best model. Ask this manager: "Why do you believe your firm'smodel processes information better than the rest of the market?"Finally, if the manager responds that the source his alpha is the result of exploitingbehavior, ask him to describe the specific behavioral bias, or biases, his firm exploits. Ifhe answers in generalities, such as his firm exploits overreaction, ask: "What are thespecific heuristic biases causing the overreaction?" Just as importantly, ask: "Under whatconditions is the market vulnerable to these heuristic biases?” If the manager cannotanswer these questions, he probably does not know how to exploit behavioral summarize, ask the manager to identify the source of his firm's alpha and toclearly document the relationship between past portfolio returns and this source of the manager passes these tests, ask why he expects his firm's source of alpha to persistin the future. Most things change over time. Truly superior analysts change firms, orretire. Better models for processing information may be replicated by of the appealing things about managers who exploit behavioral biases (assumingthey really do) is that human behavior changes very slowly. Thus, strategies that exploitbehavior and have generated a positive alpha in the past are likely to continue to besuccessful in the
ConclusionBehavioral finance offers many useful insights for investment professionals. Fromthe plan sponsor’s viewpoint, behavioral finance provides a framework for evaluatingactive investment managers, as illustrated by the “questions to ask” in the insight that was not discussed, but is potentially useful to plan sponsors,is the issue of diversification across active investment managers. If the source of alphafor most managers is derived either by generating private (or superior) information or byprocessing information better, then managers whose alpha is derived by capitalizing onbehavioral biases will be better “diversTihfiies rsfo.” llows from the idea that if thesource of alpha for behavioral managers is different from the source of alpha for the vastmajority of managers, then the returns generated by behavioral managers will tend tohave a low correlation with the returns generated by more traditional are many other potentially useful applications of behavioral finance thatwere not discussed in this article due to space constraints. For example, principles ofbehavioral finance, and psychology in general, might be quite useful in dealings betweenthe plan sponsor staff and the plan’s board of directors or
REFERENCESBasu, Sanjoy, 1977, “Investment Performance of Common Stocks in Relation to TheirPrice-Earnings RatioJos,u” rnal of Finan 3ce2, , Victor L. and Jacob K. Thomas, 1990, “Evidence that Stock Prices Do NotFully Reflect the Implications of Current Earnings for FuturJeo uEranranlings,” of Accounting and Econo m13ic, , Werner F. M. and Richard H. Thaler, 1985, “Does the Stock MarketOverreact?Jo” urnal of Finan 4ce0, , Werner F. M. and Richard H. Thaler, 1987, “Further Evidence on InvestorOverreaction and Stock Market SeasoJonuarlnitayl, ”o f Finan 4ce2, , Werner F. M. and Richard H. Thaler, 1990, “Do Security Analysts Overreact?”American Economic Rev 8ie0w, , Werner F. M. and Richard H. Thaler, 1994, “Financial Decision-Making inMarkets and Firms: A Behavioral Perspective," in Jarrow, R., V. Maksimovic . Ziemba ()n,a nc,e Series of Handbooks in Operations Research andManagement Science, Elsevier-North , Russell J., Lex Huberts and Michael Levinson, 1993, "Returns to E/P Strategies,Higgledy-Piggledy Growth, Analysts' Forecast Errors and Omitted Risk Factors,"Journal of Portfolio Manage m19e n(Wtinter), , Narasimhan and Sheridan Titman, 1993, “Returns to Buying Winners andSelling Losers: Implications for Stock Market EffJiocuierncayl ,o” f Finan 4ce8,, Daniel and Amos Tversky, 1979, "Prospect Theory: An Analysis of DecisionMaking Under RiskE,c"o nometri c4a7, , Josef, Andrei Shleifer and Robert W. Vishny, 1994, “Contrarian Investment,Extrapolation, and RiJsoku,”r nal of Finan 4ce9, 'Dean, Terrance, 1996,"Are Investors Reluctant to Realize Their Losses," unpublishedpaper, University of California at , Hirsh and Meir Statman, 1985, "The Disposition to Sell Winners Too Early andRide Losers Too LonJgo,u" rnal of Finan 4ce0, , Robert J., 1981, “Do Stock Prices Move Too Much to be Justified by SubsequentChanges in DividendAsm?"e rican Economic Rev 7ie1w,
Shiller, Robert J., 1997, "Human Behavior and the Efficiency of the Financial System,"paper on the Internet, Meir, 1995, "A Behavioral Framework for Dollar CostJ oAuvrenralg oinfg," Portfolio Managem 2e2n t(Fall), , Amos and Daniel Kahneman, 1992, "Advances in Prospect Theory: CumulativeRepresentation of UncertaJionutyrn,"a l of Risk and Uncert a5i,n
