Behavioral Finance PDF

Behavioral Finance Szymon Okoń Assistant Professor of Economics & Finance Collegium of World Economy Warsaw School of Economics 1  Neoclassical versus behavioral approach to financial markets...

Behavioral Finance Szymon Okoń Assistant Professor of Economics & Finance Collegium of World Economy Warsaw School of Economics 1  Neoclassical versus behavioral approach to financial markets 2 Neoclassical cornerstones ▪ rationality of investors (homo oeconomicus) ▪ utility theory - von Neumann and Morgenstern (1944) ▪ efficient market hypothesis ▪ portfolio theory – Markowitz (1952) ▪ neoclassical asset pricing models (CAPM, APT and their variations) 3 Rationality of investors ▪ a rational person correctly interprets the information he receives and knows how to estimate the probability of the future events on its basis ▪ it is assumed that rational market participants are so strong and dominant as a group that they are able to quickly and efficiently eliminate any symptoms of irrationality on the part of other traders ▪ as a result, the market will behave as if all participants acted rationally 4 Rationality of investors ▪ according to neoclassical theory of finance, a rational decision maker follows two general rules: ▪ he displays so-called risk aversion in that he in willing to take risk only when it may lead to further benefits, that is, only when he stands a chance of being rewarded with a risk premium ▪ decision makers always make choices in such a way as to maximaze total expected utility, given that the marginal utility of each additional benefit unit is positive 5 Rationality of investors ▪ The behavioral approach challenges the assumptions on the grounds that risk aversion depends primarily on the context in which decisions are made ▪ Moreover, decision makers attach greater importance to changes in the affluence level when measured against a specific reference point rather then its total value ▪ Many behavioral experiments showed that subjects display risk aversion when they make choices between alternatives leading to lower or higher gains. However, when faced with a decision problem in the domain of losses, they are more prone to take risk 6 Rationality of investors ▪ Von Neuman and Morgernstern (1944) formalized the classical theorem on the existence od the utility function on the basis of a series of assumptions determining preferences of rational decision makers ▪ There are four fundamental axioms that cause the most controversy among representatives of behavioral finance: ▪ axiom of completeness ▪ axiom of transitivity ▪ axiom of continuity ▪ axiom of independence 7 Axiom of completeness ▪ Rational decision maker knows how to compare different options and has well-defined preferences ▪ Results of experimental studies show that people are often not able to correctly interpret the problems they have (Tversky and Kahneman 1981, 1986) ▪ Depending on how the information is framed, decision makers may exhibit different preferences in the same situation 8 Axiom of transitivity ▪ In case a decision maker prefers variant A to B and rates variant B higher than C, then he also prefer variant A to C ▪ Under real-life economic conditions, decision makers may make intransitive choices because they are motivated by several criteria of variant assessment (depending on the situation in which a decision needs to be made, a decision maker may be motivated by the amount of the reward in one case but the probability of success in another) 9 Axiom of transitivity ▪ Tversky (1969) conducted and experiment where a group of subjects were to choose from several pairs of lotteries, but the expected value of each lottery within a pair was equal ▪ In some games the potential payoff was large by with small probability assigned. In others, the payoff was small, but the probability of winning was very high ▪ Tversky noted that when two lotteries are relatively similar in terms of gain profitability (winning is possible, but very unlikely), decision makers will choose the game offering a higher payoff disregarding the risk criterion 10 Axiom of transitivity ▪ However, when the difference in gain probability between each of the lotteries is considerable (winning is probable in both cases, but the probability is much higher in one of them), people will choose the lottery that offers higher chances of winning, paying less attention to the actual amount to be won 11 Axiom of continuity ▪ Choice between two variants should only depend upon differences between them or conditions under which the two variants lead to different results. If both options are changed in the same way, decision maker’s preferences should remain as they were ▪ The investor should not change his mind if the risk level is changed for both variants in the same way. However, psychological experiments have shown that people do modify their behavior in response to the level of risk 12 Axiom of continuity ▪ Certainty effect – decision makers tend to overestimate the value of a lottery where the reward is certain compared to a lottery with a higher expected reward but involving even a marginal level of risk (Allais, 1953; Kahneman and Tversky, 1979, 1984, 1986) 13 Axiom of continuity | example ▪ A decision maker is faced with a choice between the following investment strategies: ▪ strategy A offers certain income in the amount X ▪ strategy B offers a possibility to earn a much higher income Y with the probability determined in such way that the expected value of the strategy exceeds the value of income X, albeit only slightly ▪ Decision makers usually choose strategy A, even though its value of expected reward is slightly lower 14 Axiom of continuity | example ▪ Now, both strategies are changed in the same way, that is, they are burdened with identical additional risk, for instance, by reducing the probability of profit in each of them fourfold ▪ Following a change, none of the strategy offers a 100 percent of income ▪ Face with this choice, the decision maker will change his preferences choosing strategy B as one that offers higher expected value ▪ Naturally, the change in investor preferences contradicts the axiom of continuity 15 Axiom of continuity ▪ Another aspect putting the axiom into question is the issue of sensitivity to the way in which a decision problem is presented (dependence on information framing) ▪ Kahneman and Tversky (1984) present the example of choosing a strategy to fight an epidemic involving the total of 600 people infected. 16 Axiom of continuity | example ▪ Kahneman and Tversky (1984) present the example of choosing a strategy to fight an epidemic involving the total of 600 people infected ▪ The choice to be made is between two alternative treatment programs: ▪ program A that gives a 100 percent certainty of saving 200 of the infected ▪ program B that gives a 1/3 probability of saving all 600 infected people and a 2/3 probability of failing to save a single person ▪ Most responders (72%) decided to opt for program A. 17 Axiom of continuity | example ▪ The same respondents had to choose between „other” possible programs: ▪ program C that will cause the death of 400 of the infected ▪ program D that gives a 1/3 probability that no one will die, but a 2/3 probability that all 600 people will perish ▪ Most of the responders (78%) preferred program D. 18 Axiom of continuity | example ▪ Logically speaking, programs A and C as well as B and D are identical, the only difference being that the results of each of them are presented first in the context of the number of people saved, and than in the context of people who will have to die 19 Axiom of independence ▪ If the decision maker treats two options X and Y indifferently, he should also be indifferent about the following two variants for any option Z: ▪ option X with probability p and option Z with probability (1-p) ▪ option Y with probability p and option Z with probability (1-p) 20 Axiom of independence ▪ It is enough to take the example of complementary goods to show that this assumption will not always work in reality ▪ If any option Z turns out to be complementary in relations to X but not to Y, the decision maker may be more inclined to prefer the simultaneous occurrence of X and Z to Y and Z, even though he would be indifferent toward X and Y if he considered them separately 21 Axiom of independence | example ▪ There are two assets X and Y with exactly the same amount of risk and the same expected return ▪ Theoretically, when assessed individually, they will be perceived by the investor as equally good investment opportunities if he adopts the criterion of risk premium only ▪ It may, however happen that the return for asset Y will be strongly correlated with the return of asset Z while fluctuations in the value of asset X will be relatively independent of changes in the value of Z 22 Axiom of independence | example ▪ According to classical portfolio theory, in such case, the investor will prefer to simultaneously invest in X and Z rather than Y and Z because the first option leads to more benefits from diversification ▪ Thus, the axiom of independence falls before it is even necessary to remove the assumption of decision-maker rationality 23 Probability assessment ▪ According to the traditional theory of finance decision makers estimate the probability of different scenarios and they correctly modify their beliefs in the light of new information ▪ Proponents of behavioral finance provide many arguments proving that investors struggle to correctly estimate probability ▪ People overreact to powerful information of a descriptive nature downplaying the importance of underlying statistical data 24 Probability assessment | example ▪ A taxicab was involved in a hit-and-run accident in a certain city. There are two cab companies operating in the city – one has only green cars, the other only blue ▪ We know that 85 percent of the cabs in the city are green and 15 percent are blue ▪ A witness found in the course of the investigation identified the cab was blue ▪ It was tested, however, that, under the same circumstances that existed on the night of the accident, the witness was able to identify correctly each one of two colors eight out of ten times (80% probability) 25 Probability assessment | example ▪ What is the probability that the cab involved in the accident was blue? ▪ To answer this question we can apply Bayes’s rule, i.e. mathematical tool describing how rational investors should modify their opinions in the light of the new facts ▪ The median of answers given by responders in the experiment was as high as 80 percent. ▪ People involved in the survey usually did not attach importance to the low percentage of prior probability (P(A)=15%) and so overestimated their general assessment 26 Probability assessment ▪ Another demonstration of problems with the base of probability is a judgment based on a stereotype ▪ Having noticed a clear personal trait corresponding to a common stereotype, responders are overconfident about the probability that the person will follow the stereotype and underestimate base probability ▪ People frequently disregard the fact that, statistically speaking, it is more probable that the analyzed person belongs to a different group only because this other group is larger that the set of people confirming stereotype 27 Probability assessment | example ▪ „Linda is 31 old, single, outspoken, and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in anti-nuclear demonstrations.” ▪ What was more probable? ▪ Linda is a bank teller ▪ Linda is a bank teller and is active in the feminist movement 28 Probability assessment | example ▪ Over 86 percent of the responders said that statement B was more probable. However, it is obvious that, statistically speaking, option A is more probable as there are more women who are bank tellers that women who are bank tellers and feminists at the same time 29 Limits do Arbitrage ▪ Arbitrage may be defined as the simultaneous purchase and sale of the same, or essentially the same, security in two different markets for advantageously different prices. Its role in capital asset pricing is to sustain asset prices at fundamentals ▪ Active arbitrage triggers the market adjustment mechanism, which quickly brings asset prices back to their real values through changing demand and supply ▪ Higher demand for the cheaper assets pushes the price up while higher supply for the same or essentially the same asset pushes the price down 30 Limits do Arbitrage ▪ Behavioral finance does not negate the principle of arbitrage itself ▪ However, in practice, arbitrageurs face a series limitations that partly or totally constrain their actions contributing to market inefficiency and incorrect asset pricing ▪ The reason is that arbitrage may be not only risky but also costly and so the market is not always able to eliminate the effects of irrational behavior 31 Fundamental Risk ▪ Fundamental risk in arbitrage is due to the fact that financial markets do not always offer an ideal substitute whose price will react to news in exactly the same way as the price of the security to be initially mispriced ▪ The information could be specific for one company only and have considerable impact on the price of one security, but not on the price of the other instrument ▪ Fundamental risk may trigger sever losses for arbitrageurs, which potentially are much higher that initial profits from the difference between prices of both securities at the moment when positions are taken 32 Noise Traders Risk ▪ The risk that arbitrageurs may incur stems from the danger of more intense activities on the part of noise traders who may cause the price of the security to deviate even further from its fundamental value ▪ When funds are withdrawn by capital owners, arbitrageurs are forced to close their positions that are only temporary ▪ Arbitrageurs are often forced to close their positions at the very moment when the deviation of the price from the fundamental value of the security is at its greatest creating the best opportunities for arbitrage 33 Noise Traders Risk ▪ When making transactions arbitrageurs borrow money and securities. As short-term losses accrue, lenders of capital and securities may want their loans back because the no longer perceive the arbitrageurs as credible ▪ In case on short-selling, the obligation to return the borrowed security does not necessarily have to be related to damaged credibility of the arbitrageur. It may simply derive from the fact that the deadline specified in the loan contract has expired 34 Risk of Synchronization ▪ Abreu and Brunnermeier (2002) propose a model based on three main assumptions: ▪ Actions taken by a single arbitrageur cannot influence the market strongly enough to correct incorrect pricing on their own ▪ Some arbitrageurs detect the inefficiency quicker than others; they are not aware when other arbitrageurs will start acting ▪ The longer it takes to eliminate incorrect pricing, the longer the trader is forced to keep arbitrage positions open incurring higher related cost 35 Risk of Synchronization ▪ On the one hand, arbitrageurs are afraid to open their arbitrage positions too soon not to incur excessive costs; on the other, they are motivated by desire to benefit from the opportunity they noticed before other traders manage to do the same ▪ Abreu and Brunnermeier confirmed that when arbitrageurs are faced with this dilemma, they usually eventually decide to postpone their actions ▪ The more price of a security deviates from its fundamental value, the more motivated arbitrageurs are likely to be and the sooner they start acting 36 Implementation Costs ▪ Costs related to borrowing stocks for short-selling, transaction costs as commissions, bid-ask spreads, costs of information services and analysis ▪ The higher the total costs of implementing and maintaining and arbitration strategy are, the less willing rational investors are to benefit from incorrect pricing on the market 37 Regulatory Barriers ▪ Legal requirements and internal regulations that are binding for a large group of international investors 38 Portfolio Theory ▪ Every investment in securities is accompanied by two kinds of risk: systemic (market) and nonsystemic (unique, specific) ▪ Nonsystemic risk is related to specific asset individually and may be eliminated through correct diversification of stock in the portfolio ▪ Systemic risk is related to the market as a whole and cannot be eliminated because returns on stock in the market are correlated to a certain extend ▪ Optimally diversified portfolio should contain about 20-30 different assets 39 Portfolio Theory ▪ Rational investors should diversify their investments in such a way as to eliminate nonsystemic risk and create efficient portfolios displaying minimal variance (risk) for specific expected returns ▪ The investors should focus on covariance of assets with other assets and, consequently, on how much it contributes to the risk of the entire portfolio ▪ If all traders create efficient portfolios, then the whole market portfolio will also be efficient being a function of all investment run by investors at a given moment ▪ It should not be possible to achieve abnormal returns 40 Portfolio Theory ▪ Proponents of behavioral finance challenge the classical portfolio theory ▪ Investors have unstable preferences and are not in position to estimate risk correctly ▪ It cannot be expected that they should always take appropriate steps to maximize expected utility and be motivated in their investment choices only by the relation between the level of systemic risk and expected return 41 Portfolio Theory ▪ Investors remarkably often do not follow the rule of minimizing nonsystemic risk, running portfolios that are not diversified enough (De Bont, 1998; Statman, 2002; Goetzmann and Kumar 2008) ▪ Investors do not attach enough importance to the issue of correlation between different kinds of Assets (Kroll 1998) ▪ Shefrin and Statman (2000) claim that if people optimize the structure of their investments at all, they do it in a gradual fashion (portfolio consisted of many layers); if they pay attention to covariance at all, it is limited to assets contained within one sub-portfolio 42 Portfolio Theory ▪ Low level of diversification and underestimating the importance of mutual correlation between returns undermine Markowitz’s classical portfolio theory ▪ The investors create inefficient portfolios and as a consequence the market portfolio might turn out to be inefficient ▪ As a result it would be possible to make higher return at market risk or the same return at the market portfolio, but at smaller risk 43 Capital Asset Pricing Model ▪ CAPM is based on a series of the assumptions: ▪ Investors are risk averse and try to maximize utility, shaping their preferences in accordance with Von Neumann and Morgenstern’s theory ▪ Investors create efficient portfolios on the basis of Markowitz’s portfolio theory ▪ Market players are homogeneous (adopt the same single-period investment perspective, have the same amount of information, perceive reality in the same way etc.) 44 Capital Asset Pricing Model ▪ There are no obstacles hampering the flow of capital and information ▪ All assets may be bought and sold, including short- selling ▪ There are unlimited possibilities to borrow and lend funds at the same risk-free return ▪ Most of aforementioned theoretical assumptions are disproved in real life 45 Capital Asset Pricing Model ▪ Fama and French (1992, 1993, 1996) suggested a Three Factor pricing model in which the traditional market risk premium is supplemented by two additional elements related to the size of the company and the book-to market equity ratio (elements not taken into account by the previously used beta measure of risk) ▪ They argue that returns for small companies or companies with high book-to-market equity ratio are higher than might be expected from the traditional CAPM because they represent a rational premium for unidentified element of undiversified risk 46 Capital Asset Pricing Model ▪ In light of behavioral observations, asset pricing and market equilibrium models assuming that investors are risk averse and rational and portfolios they hold are efficient ▪ Decision makers cannot always interpret information they receive correctly and do not always know how to estimate probability, it would be difficult to expect them never to make mistakes when assessing risk and expected returns ▪ If they do not create investment portfolios offer the best possible combination between expected returns and variance the market portfolio will not be efficient either 47 Capital Asset Pricing Model ▪ Behavioral approach undermines the entire theoretical foundation on which proponents of the classical theory of finance build their asset pricing models 48 Efficient Market Hypothesis ▪ Efficient market is a market in which prices always fully reflect available information (Fama 1970) ▪ Market prices incorporate both information based on events that have already occurred and on events that, as of now, the market expects to take place in the future 49 Efficient Market Hypothesis ▪ Three basic forms of informational efficiency of capital market depending on the scope of information to be reflected in asset prices: ▪ Weak efficiency: assumes that stock process reflect all important information from historical pricing; investors cannot predict how asset prices will behave in the future solely of information taken into account when pricing assets ▪ Semi-strong efficiency: stock prices do not only reflect the information that can be obtained from historical prices but also all other publicly available information (financial statements, forecasts) 50 Efficient Market Hypothesis ▪ Strong efficiency: no matter whether information is publicly available or private, confidential, and available to a small group of people, it is quickly reflected in stock prices only because it can be induce by observing actions of insiders 51 Efficient Market Hypothesis ▪ According to EMH investors should not hope to consistently beat the market ▪ Achieving abnormal returns is possible, but only as a simple result of luck, and not due to whether trading strategy used or resources spent on analysis ▪ The best strategy is „buy & hold” approach (well- diversified portfolio) ▪ Changes to the portfolio are not recommended, as active trading only generates transaction costs and cannot help at all to achieve long-term abnormal returns 52 Efficient Market Hypothesis ▪ According to behavioral finance because the market is not always efficient, investors who make better than average use of available information are capable of making abnormal returns ▪ It might be worth to seek good investment opportunities and to spent resources on the investigation of the mispricing that occur from time to time in the market ▪ Active trading strategies might be better than passive „buy and & hold approach” ▪ Achieving higher returns is possible, thanks to better analysis and strategies, but requires better self-control 53 Cornerstones of Corporate Finance ▪ Behavioral corporate finance can be divided into two main areas based on distinctively different assumptions ▪ The first approach emphasizes the effect of market inefficiency on corporate policies, assuming that executives act as rational professionals; it focuses on how smart manager modifies corporate policy in order to exploit investor irrationality and market inefficiency ▪ The second approach replaces the assumption of managers’ rationality with evidence-driven psychological foundations; it shows how managerial biases may impact managerial practice 54 Psychological Aspects of Decision Making 55 Perception of Information ▪ Psychologists believe that the human mind is limited in its liability to focus and process all the incoming information ▪ They also stress that the final form our decisions take is often strongly influence by the signal received by our subconscious 56 Perception of Information ▪ One of the widely discussed psychological phenomena is the so-called framing effect ▪ Analyzing problems without paying attention to their wider context or even in and extremely isolated way ▪ Decision makers may display changing preferences and make radically different choices depending on the way in which the same logically identical problem was presented to them ▪ Examples of framing effect: money illusion, mental accounting, anchoring bias, illusion of truth 57 Perception of Information ▪ One version of the framing bias is the so-called money illusion ▪ Even though people generally know the difference between the real and nominal value of money, there is a lot of evidence suggesting that their perception is remarkably often dominated by nominal values ▪ For example, a pay rise of 2% with inflation running at 4% will be much better than a reduction of 2 percent when there is no inflation (Shafir 1997) 58 Perception of Information ▪ Another instance of framing is mental accounting in which individual aspects of financial decisions are perceived in isolation (Thaler 1985, 1990, 1999) ▪ People create different „accounts” in their mind for various expenses or revenues ▪ People are easier about spending money won at a lottery but more careful with their hard-earned savings (even though economic value od 1 dollar is the same) 59 Perception of Information ▪ Anchoring bias is most often illustrated by the experiment conducted by Khaneman and Tversky (1974) in which subjects where asked to estimate the percentage of African countries in the United Nations ▪ First responders were requested to spin a „wheel of fortune” to generate a number between 0 and 100 ▪ Next they had to say whether they believe the percentage of African countries in the UN is lower or higher than the randomly selected number 60 Perception of Information ▪ Next they were asked to estimate the actual percentage of Africa’s representatives among all UN members as precisely as possible ▪ Finally, it turned out that the initial, totally arbitrary anchor value significantly influenced participants’ responses ▪ For example, the median estimates were 25 for the group that received 10 as a starting point and 45 for the group that started the study with 65 61 Perception of Information ▪ Northcraft and Neale (1987) proved that, when estimating the value of a house, professional real estates appraisers are very much influenced by the price suggested by the owner ▪ In this case most probably, the anchoring effect works on the subconscious level if over 90 percent of the experts in the study insisted that the owner’s expected price had no impact on the value they estimated 62 Perception of Information ▪ Stephan and Kiell (2000) asked to groups of professional financial analysts to forecast the value of the German Stock Index (DAX) for next 12 months ▪ First, responders were only asked if they thought that next year the index was going to be below or above threshold of 4,500 points (group one) or 6,500 (group two) and next, they were asked to provide an accurate forecast ▪ The value suggested in the first question proved to be an important „anchor” for the forecasts as a result of which the average forecast value in group one was significantly lower than the same value in group two 63 Perception of Information ▪ Reception and assessment of incoming information is also frequently influenced by its clarity to the recipient and the ease with which he can take in and process new messages ▪ Our perception is often marred by the so-called illusion of truth, whereby we are more willing to accept as true those statements that seem clear to us even though they might be in fact false ▪ At the same time, we reject as untrue the information that we perceive as complicated and difficult to decipher (Reber and Schwarz 1999) 64 Overconfidence, optimism, narcissism ▪ A lot of evidence suggests that decision makers are generally overconfident. This can be manifested in three basic ways: above average effect, calibration effect and illusion of control ▪ Overconfidence is usually closely linked with excessive optimism and unrealistic wishful thinking and it is also related to narcissism 65 Above-average effect ▪ When making assessments and constructing beliefs about reality, people consider their knowledge and skills to be above the average ▪ For example, in different surveys, 60 percent to 90 percent of the responders claimed that they had above- avaerage driving skills or better-than-average sense of humor ▪ They also said that they stood less-than-average change of developing a specific disease and were less-than- average likely to be victims od mugging (Weinstein, 1980; Svenson, 1981; Barberis and Thaler, 2003) 66 Calibration bias ▪ Overconfidence is also evident in the so-called calibration bias (Lichtenstein 1982, Yates 1990; Keren 1991) ▪ When asked to provide information or make a forecast without being precise but estimating within a certain confidence range, people usually give answers which indicate that they are overconfident as to the precision of their knowledge ▪ Alpert and Raiffa (1982) demonstrated that responses given by respondents with an alleged 98 percent certainty actually turn out to be correct only in about 60 percent of cases 67 Illusion of control ▪ Overconfidence also manifests itself in the illusion of control ▪ People are frequently convinced that their actions may positively affect totally random outcomes ▪ Langer (1975) documents that lottery players place higher value on tickets for which they themselves picked the numbers that on the tickets filled in by the quick-pick lottery machine 68 Excessive optimism ▪ Overconfidence is also related to excessive optimism and unrealistic wishful thinking, resulting in various kinds of planning errors due to the underestimation of negative factors and overestimation of positive outcomes ▪ Montgomery (1997) collected macroeconomic forecasts about inflation, gross domestic product (GDP) growth, unemployment, and so on, estimated by various experts over may years. ▪ He then performed an ex post comparison of the forecasts with the values actually observed. It turned out that the forecasts of unfavorable effects (inflation, unemployment) were systematically underestimated 69 Narcissism ▪ Narcissism in a personal trait that is related to but distinct from overconfidence. ▪ The central idea of narcissism is that individuals are not only convinced about their superiority but also have a need to maintain a positive sense of self by engaging in ego-defensive high-stake activities ▪ Campbell, Goodie and Foster (2004) document in experimental studies that narcissist display greater overconfidence and more willingness to bet ▪ Liu (2009) and Aktas, de Bodt, Bolleart adn Roll (2012) investigate narcissism among corporate executives in the real setting of mergers and acquisitions 70 Investor Behavior 71 Forecasting the Future on the Basis of Past Events 72 Extrapolation Bias ▪ Extrapolation bias consists in overemphasizing trends in the past, especially over a relatively short period ▪ One example of extrapolation bias is when investors make long-term financial forecasts assuming that the company they analyze will have same sales or profit growth rate as the one observed in the last reporting periods ▪ If the assumed growth rate is too high, financial forecasts will be overestimated resulting in overly optimistic pricing of stock 73 Following Trends ▪ Some market players are convinced that returns on stocks develop in a predictable way following a specific trend. There is even a popular market saying: Trend is your friend. ▪ Investors often observe historical stock prices very closely to identify patterns in totally random time series. They wrongly interpret consecutive series of rises and falls in stock prices as a trend forgetting that observations might be purely coincidental 74 Following Trends ▪ If sufficiently large group of investors is strongly convinced that the observed series of changes mark a beginning of a new trend, and if this conviction is buttressed by signals coming from popular strategies based on technical analyses, a self-fulfilling prophecy may set in and indeed trigger a further wave of rises or falls ▪ This will convince more investors that the trend has really turned and it may be expected to continue. Actions takes by subsequent „trend observers” boarding the „leaving train” will only confirm the direction of changes 75 Following Trends ▪ The investors are more willing to wait for the continuation of a trend when the market is bullish rather than bearish. ▪ This seems understandable in the context of the psychologically documented phenomenon of excessive optimism and wishful thinking 76 Minimum and Maximum Prices ▪ Statistical data most often include the minimum and maximum price listed in the last 52 weeks ▪ One possible reason why the statistics are published is that the readers want to read them. Investors might be interested to compare the current price to its historical peaks and the media simply try to satisfy their curiosity ▪ The investors do attach a lot of importance to whether current prices are close to their historical minimum or maximum values 77 Minimum and Maximum Prices ▪ Such observations are difficult to explain from the perspective of the classical theory of finance. Behavioral finance offers such clues as the anchoring effect combined with excessive optimism ▪ Investors become attached to minimum and maximum price levels and treat them as natural reference points when making investment decisions ▪ If the price exceeds its last maximum level, it is interpreter as a signal that the company will continue the positive trend justifying further increase in value. Reaching the minimum value will in many cases be understood as an incentive to buy as the price seems very attractive. 78  Perception of Value and Investment Selection 79 Good Company vs Good Investment ▪ In might be argued that if the company has prospered as a result of good management, organization, know-how, and other unique factors, it is to be expected that it will continue to perform well and keep growing ▪ It might be a good idea to check whether some of the companies that let investors down in the past, lost much value as a result, and are not very popular have not undergone a process of restructuring, for example, which has improved their prospectus for the future ▪ A good investment is one that generates returns which are as high as possible given the level of the attached risk 80 Beauty Contest ▪ Deciding to invest investors try to predict the price for which it will be possible to resell stocks in the future; the future price will depend on the value that other market players will be ready to assign to the stocks at the time ▪ Keynes (1936) compares investor behavior to a newspaper beauty contest. ▪ One of the London newspapers published photographs of beautiful women asking its readers to select the prettiest. It also founded a prize awarded to the most beautiful women selected and prizes for readers who would select the women most often selected by the contestants 81 Beauty Contest ▪ In order to win the contest one had to choose not by following his subjective judgement, but rather by anticipating what preferences other participants will have ▪ The situation on the capital market is similar. It does not matter what we like and what the fundamental values of the company are. What is important is that others like it in the future and be prepared to pay even more for it 82 Familiarity and Home Bias ▪ People prefer to invest in companies they know ▪ The mere fact that the company is recognized, that is people have all information on its line of business, owned brands, products, locations and so on, does not have to mean that it enjoys information advantage over other, average market players ▪ Focusing only on well-known companies without looking for new investment opportunities may lead to nonoptimal diversification and inefficient portfolios 83 Familiarity and Home Bias ▪ Active search for and assessment of new signals is one of the conditions for informational efficiency of the market ▪ If investors concentrate only on what they are familiar with, staying clear of some sophisticated and less recognizable securities, efficiency and the market pricing may be distorted at least in some segments of the market 84 Incorrect Perception of Information ▪ Market participants sometimes find it difficult to assess particular bits of information and their reactions often depend on the way facts about the company are presented ▪ There have been documented cases when the publication of unimportant data or media attention around facts that were already publicly known triggered investor reaction influencing stock prices 85 Incorrect Perception of Information ▪ Copper, Dimitrov and Rau (2001) point out interesting phenomenon from days of the dot-com boom. The mere information about company changing its name in such a way that the new name was associated with the Internet caused a sharp increase in the company’s stock price 86 Financial Forecasts ▪ Analysts tend to be overly optimistic when preparing forecasts and drafting recommendations. There is a rich literature documenting how analysts’ predictions are overestimated compared to later, actual results ▪ Some publications suggests that analysts might show signs of optimism on purpose as the feeling relates to the conflict of interests they have being hand in glove with financial institutions that are active on the market ▪ It is also sometimes argued that analysts consciously overestimate the companies they assess in order to have good relations with their management, a strategy allowing them to have access to nonpublic information 87 Portfolio Management 88 Insufficient and Naive Diversification ▪ Campbell, Lettau, Malkiel and Xu (2001) document a clear tendency on the American market to narrow the correlation between rates of return on individual stocks ▪ Empirical studies demonstrate that, both in the past and currently, individual investors keep portfolios that are drastically under-diversified ▪ Groetzman and Kumar (2001) who studied 40,000 individual active brokerage accounts, showed that investors keep an average of four companies in their portfolios with the median standing at 3 89 Insufficient and Naive Diversification ▪ The least diversified portfolios are kept by young people with relatively low income and professional position. The number of companies in the portfolio increases strongly in proportion to the level of income, education, age. ▪ Stocks are often treated as lottery tickets. People buy a few stocks in the hope that if they bet correctly the „wins” will let them jump to a higher level of consumption and realize their aspirations 90 Insufficient and Naive Diversification ▪ Goetzmann and Kumar (2008) find that investors underestimate the meaning of cross-correlation among securities, do not apply correct weights to particular components of the portfolio, and overlook the impact of a single security on the overall variance portfolio ▪ A different version of naive diversification is observed when an investor decides to spread his investment over funds of the same type (similar strategy) ▪ Benartzi (2001) employees display clear preferences to invest directly in the shares of the company they work for 91 Insufficient and Naive Diversification ▪ De Bondt (1998) finds the problem with insufficient or naive diversification among experienced individual traders. ▪ Investors who have systematically traded on the stock market have strong preferences to focus only on a few selected picks. ▪ Most of the investors erroneously believed that they could manage portfolio risk better by understanding a handful of companies thoroughly rather than by wide diversification (overconfidence) 92 Insufficient and Naive Diversification ▪ Poor diversification and underestimating the mutual correlation between returns on individual assets undermine Markowitz’s portfolio theory and the assumptions behind traditional asset-pricing models 93 Excessive trading ▪ There is a lot of evidence suggesting that investors make transactions too often and too aggressively leading to an unreasonably high level of trading ▪ The first behavioral explanation of excessive trading volumes suggested refers in general to the presence of noise traders making frequent transactions on the basis of unconfirmed rumors and irrelevant information ▪ Other points at overconfidence as the answer to why investors attach too much importance to the doubtful information they have 94 Excessive trading ▪ Overconfident investor decides to make the transaction believing that his analytical skills are better than those of other traders (above-average bias) and overestimates the precision of the signals he received (calibration bias) ▪ Odean (1999) shows that investors who speculate most actively have the worst average performance. This is caused mainly by transaction costs but also he points out systematically wrong decisions of active investors who decide to buy and sell securities at the wrong moment 95 Myopia in Asset Allocation ▪ Benartzi and Thaler (1995) suggested that the relative reluctance to invest in stocks in the result of two simultaneous psychological phenomena ▪ First, investors get less satisfaction from making a profit that they feel pain for suffering a loss of the same magnitude ▪ Second, most investors, even those with a very long investment perspective, evaluate their portfolios relatively often assessing the results of the evaluation separately for each short period 96 Myopia in Asset Allocation ▪ It is obvious that stock prices may fluctuate radically in short term periods of time. ▪ Due to the habit of mental accounting investors register achieved results separately for each of the short periods that they evaluate. It is harder to them to perceive returns in long term perspective ▪ At the same time they are very much affected to losses even if these are short term in nature. Trying to avoid negative feeling related to potential losses they limit the level of investment in stocks 97 Myopia in Asset Allocation ▪ Myopic allocation of funds in investment portfolios in aggravated by institutional factors. The basic reporting period for most financial institutions in one year ▪ Managers of all types of funds and asset managers are also typically evaluated and rewarded annually ▪ Another factor exacerbating investors’ myopia in the capital market in the tax system forcing individual and institutional investors to account for profits and losses annually 98 Disposition Effect ▪ Investors are more ready to keep in their portfolio Assets whose prices fell from the time of purchase rather than the stocks that might be sold with profit ▪ From the perspective of behavioral finance, the phenomenon can be explained it two different ways ▪ First, investors may fall victim to the anchoring bias becoming too much attached to the price at which they bought stocks ▪ This attitude may be related to overconfidence and unrealistic optimism. Investors are reluctant to admit that they have made a mistake 99 Disposition Effect ▪ The other behavioral explanation of the disposition effect refers to the prospect theory ▪ When the investors have shares whose current price is higher that the purchase price, that is, they may be sure of making profit, they will display risk aversion ▪ But if they have shares whose price has recently dropped so that they are faced with suffering a loss, they are prone to take more risk 100  Asset-Pricing Anomalies and Investment Strategies 101 Violation of the Law of One Price 102 Violation of the Law of One Price ▪ The law of the one price is a principle saying that, in a perfect market, the same assets should have the same price. Identical pricing is possible by arbitrage ▪ When we look at capital market practice, we can distinguish three major categories of events where the law of one price is violated: pricing shares in close-end funds, listing of the so-called twin stocks and situation when shares of mother and daughter companies are simultaneously listed on the market 103 Closed-End Funds Puzzle ▪ The price of the close-end fund share on the exchange market should be as close as possible to NAV per one share. In practice, however, share prices deviate significantly from NAV ▪ Attempts to provide rational explanations point out at deferred tax liabilities or limited liquidity. Agency costs, understood here as the uncertainty as to future actions of fund managers, are also considered ▪ Nonetheless, the aforementioned explanations, even though partially justified, cannot fully account for all observations, related to the pricing of closed-end fund shares 104 Closed-End Funds Puzzle ▪ Lee, Shleifer and Thaler (1990, 1991) offer a behavioral explanation of the closed-end fund puzzle referring to the concept of noise-trader risk ▪ Initial listings above the NAV stem from the advantage of rational professionals who benefit from the ignorance of noise traders to organize new funds in the periods of increased market optimism and take actions with the aim to sustain the high initial price ▪ Lee, Shleifer and Thaler (1991) document that public offerings of closed-end funds are particularly abundant on the market when the discount on the closed-end funds is especially low or there is even a premium to NAV 105 Closed-End Funds Puzzle ▪ Weiss, Lee and Seguin (1996) also point out that over the first 30 sessions after new shares are placed on the stock market, the average value of one sell order in much higher than the average value of the buy order. This suggests that bigger players sell to small individual investors just after fund gets listed ▪ After a time, the ownership structure of the new fund shares is dominated by small investors. Because of the high percentage of small players, irrational factors are more likely to have influence on prices 106 Twin Stocks ▪ A twin company’s charter fixes the division of current and future equity cash flows to each twin ▪ Examples of these kinds of twin companies are Royal Dutch Petroleum – a company registered in the Netherlands whose shares are primarily listed on the Dutch and American markets – and Shell Transportation and Trading PLC, which is registered in the UK with assets traded mostly on the London Stock Exchange ▪ Years ago, shareholders of these companies decided to joint up forces on a global scale and divide the net profit they generated in the following proportion: 60% RDP and 40% STT 107 Twin Stocks ▪ Shareholders of those two entities are to be paid dividends in equal amounts on the basis of totaled net results ▪ Bearing in mind the agreement, we should assume that share prices of RDP and STT will hover around the 60:40 proportion. ▪ In practice, relative prices of twin shares were markedly different from the theoretical assumptions ▪ Froot and Dabora (1999) have demonstrated that deviations from the theoretical parity of listings cannot be explained 108 Twin Stocks ▪ Froot and Dabora (1999) have demonstrated that deviations from the theoretical parity of listings cannot be explained neither by differences in voting rights and control over dual listed companies nor by fluctuations of exchange rates or other international transaction cost ▪ The only factor able to explain, but only just, such deviations are different tax regimes for investors located in various countries ▪ Particulary, it is difficult to justify why in some periods there is relative overvaluation while others experience relative underpricing between twin companies 109 Pricing of Mother and Daughter Shares ▪ It is sometimes the case that shares of mother and daughter companies are listed on the market ▪ An especially interesting situation takes place when the mother company decides to carve out a part of its business and float a daughter company on the market offering a part of its shares for cash in IPO while allocating the remaining part to shareholders of the mother company ▪ The investors interesting in purchasing daughter shares can do it in two ways (buy directly or buy the shares of the mother and wait until allocation of shares) 110 Calendar Anomialies 111 Month-of-the-Year Effect ▪ The month-of-the-year effect, also called January effect, is one of the best described anomalies od the seasonal distribution of returns ▪ Many publications demonstrated that average returns are much higher in the first month of the year compared to other months ▪ Gu (2003) and Schwert (2003) point out that the phenomenon was much weaker over the last two decades of the twentieth century, which might have suggested that the fact it had been discovered and publicized made the market more efficient, however the effect seems to be back in a few recent years 112 Month-of-the-Year Effect ▪ The most widely known explanation of the January effect is the tax-loss selling hypothesis ▪ At the year end, investors sell stocks whose prices fell over the last 12 months in order to deduct the realized loss from their tax base ▪ Early January, investors start buying stocks that are now underpriced due to the December sellout ▪ Keim (1983) and Roll (1983) find that the market is most active during the first five sessions of the new year 113 Month-of-the-Year Effect ▪ Reinganum (1983) documents that the great majority of the total number of stocks sold in order to settle losses are small company shares ▪ Supply pressure has more impact on prices of small company stocks that are not usually traded intensively ▪ Gultekin (1983) documented that the month-of-the-year effect was present in 16 other countries. These markets did not, however, display a relationship between the seasonality of returns and company size that would be as strong as the one observed in the US 114 Month-of-the-Year Effect ▪ While tax-loss selling hypothesis seems to play the most important role, it is definitely not the only explanation behind the month-in-the-year effect, nor can be considered to be fully rational ▪ It may be understandable that investors want to sell shares in order to offset their taxes, but it is harder to explain why they would wait to realize losses right up to the last moment in December 115 Month-of-the-Year Effect ▪ From behavioral perspective it may be assumed that keeping open positions for assets whose prices was systematically falling over the year is related to strong loss aversion and the disposition effect ▪ Investors wait until the last moment hoping that the trend will turn allowing them to avoid losses. This attitude may be related to overconfidence and unrealistic optimism ▪ As psychologically driven biases are more typical for individual investors who usually dominate small companies, supply pressure at year end will mostly affect small-cap stocks 116 Month-of-the-Year Effect ▪ Postponing investment decisions until beginning of January is in turn caused by the psychological phenomenon of mental accounting ▪ In December, people do not think of new investments by examine last year’s profits and losses. It is the new calendar year that is traditionally associated with a new investment period bringing fresh expectations that are usually built up ignoring last year’s results 117 Month-of-the-Year Effect ▪ The behavior of individual investors is additionally enhanced by actions of institutional investors who also usually reallocate their portfolios at the turn of the year ▪ Wanting to include a portfolio where most shares are blue-chip and have recently performed well, many managers sell blocks of less-well-known shares, that is, small-company assets, especially it they have generated losses lately (window dressing) ▪ In the new year, managers open their positions again, purchasing shares that they consider to be most underpriced 118 Weekend Effect ▪ A number of studies have demonstrated that average returns between session close on Friday and Monday were much lower than the average rates on other weekdays, their typical values being even negative ▪ The discovery of the weekend effect raised the question of when exactly observed negative returns are realized. Does it happen between session close on Friday and the Monday opening or perhaps during the Monday session? 119 Weekend Effect ▪ Damodaran (1989) put forward a hypothesis that low Monday returns may be rationally explained by information-related factors, the usual time when companies announce negative news being Friday afternoon ▪ DellaVigna and Pollet (2005) have partially confirmed Damodaran’s intuitions documenting that the Friday announcements of financial results are 25 percent more likely to contain a negative surprise ▪ But they also argue that investors attach less importance to the information announced on Fridays even if it is made public when trading is not finished yet 120 Weekend Effect ▪ It could therefore be argued that investors are much less concerned by news they receive on Fridays as they are anxious to begin their rest at the weekend ▪ Corporate managers exploit this habit selecting Fridays as the days of announcing bad news ▪ Monday activity of institutional investors is less intense. So it might be possible that low Monday returns are the product of small investors reacting too nervously and slightly too late to negative information published much more often on Fridays than other weekdays 121 Incorrect Market Reaction to Information 122 Contrarian Investing ▪ In the strategy of contrarian investing, it is generally recommended to purchase assets that have recently lost a lot of value and sell those whose price has increased (winner-loser effect) ▪ Proponents of behavioral finance explain the winner-loser effect by citing market overreaction. The investors fall victim to the extrapolation bias ▪ Noise traders attach to much importance to historical performance of companies. They are too optimistic extrapolating prospects of those companies that have so far been successful on the market underestimating the potential of weaker companies 123 Contrarian Investing ▪ Market overreaction my be related to overconfidence. The investors overestimate the value of the information they have which makes them react to strongly ▪ Having made the position the investors will be focused on signals confirming the position they took, downplaying the importance of information to the contrary 124 Momentum strategy ▪ Momentum strategy recommends to buy shares of companies whose value has recently increased significantly and to sell those that have lost the most ▪ The principle seems to be the exact opposite of contrarian investing. The key difference is the time frame in which companies are assessed before portfolios are constructed as well as the length of the holding period ▪ While contrarian strategies were concerned with very short (weeks) or very long periods (three to five years), the momentum strategy falls somewhere in between 125 Momentum strategy ▪ Trying to explain the momentum effect, we have so far tried to determine if it is cause by under- or overreaction to information ▪ Another question is whether the effect is the result of a wrong reaction to company-specific or systemic information ▪ Moskowitz and Grinblatt (1999) argued that industry momentum effect could almost entirely explain the profitability of momentum strategies using individual assets. Studies by Grundy and Martin (2001) showed that even though industry factors may play a role, it is not generated by incorrect reaction to company information 126 Momentum strategy ▪ The investors have problems to determine correctly to what degree changes in a company’s financial standing are cause by specific factors and what should be attributed to events affecting the entire industry ▪ This is why the market react slowly not only to company- specific signals, but also to the gradual diffusion and delayed influence of the information on the pricing of other companies from the same sector 127 Momentum strategy ▪ The momentum effect may be also potentially triggered or at the very least strengthened by activities of professional market players ▪ Womack (1996) documents that analysts give average better recommendation for companies whose prices have previously had a series of rises. Such information may make investors believe that the growth is likely to continue ▪ Chen, Jegedeesh and Wermers (2002) demonstrate that investment funds are more likely to buy previously rising shares, selling those that lost value 128 Momentum strategy ▪ Burch and Waminathan (2001) showed that managers decisions to buy or sell assets depend more on the returns these assets have recently generated than on financial results reported by the companies 129 Forecasting Returns on the Basis of a Firm’s Characteristics 130 Small-Size Effect ▪ Banz (1981) analyzed returns on shares listed on NYSE in 1936-1977 dividing the population of companies by size into five portfolios. The difference between average annual returns on portfolios containing the smallest and biggest stocks amounted to 20 percent ▪ Fama and French (1992) observed the difference between portfolios covering 5 percent of the smallest and biggest companies was much lower and amounted 9 percent annually 131 Small-Size Effect ▪ Supporters of traditional theory of finance tried to explain the small-size effect by identifying additional risks and microstructural factors typical of this market segment (liquidity, high bid-ask spread) ▪ Financial behaviorists interpret the phenomenon differently. They relate higher small-cap returns to the presence of noise traders whose actions are motivated more by speculation and emotions that rational information analysis 132 Book to Market Equity ▪ Book value of equity is based on historical data and does not factor in the current earnings not the company’s prospects ▪ Market equity represents the total pricing of all shares on the basis of demand and supply the assets elicit on the market at the moment ▪ Fama and French (1992) documented that the difference between returns on portfolios with the highest and lowest book-to-market ratios amounted to almost 20 percent on average annually 133 Book to Market Equity ▪ Fama and French claim that higher returns on companies with high book-to-market ratios constitute a rational premium for extra risk ▪ They argue that value firms are often the „fallen stars” that is, former growth companies that got into some trouble ▪ As the economic condition is likely to be poor, it may be expected that they will be exposed to additional risk common for this segment of the market 134 Book to Market Equity ▪ Proponents of behavioral finance contest this explanation. They are on the opinion that higher returns on value stocks as compared with growth stocks are the result of market overreaction ▪ Investor overestimate companies that have recently had good financial results, underestimating those whose performance was relatively weak. This is due to the extrapolation bias and a naive belief in trends 135