Behavioral Finance, Osa 1

(Kuva: Chris Lott)

Käyttäytymisperusteinen rahoitus eli behavioral finance on aihealue, josta minua on pyydetty kirjoittamaan useampaan eri otteeseen. Sitä opetetaan myös yliopistossa. Tämä kirjoitussarja koostuu kolmen tunnin rahoituksen luennosta, joka johdattaa behavioral financeen. Sen on kirjoittanut Oulun yliopiston rahoituksen professori Hannu Kahra, joka ystävällisesti antoi luvan materiaalin käyttöön blogissa. Kahran koostama materiaali perustuu pitkälle John Nofsingerin kirjaan Psychology of Investing. Luento tiivistää minusta hyvin olennaisimmat asiat behavioral financesta. Olen palastellut sen useampaan pienempään osaan lukemisen ja omaksumisen helpottamiseksi. Teksti on englanniksi, mutta sen verran helppolukuista, että uskon suurimman osan selviytyvän sen lukemisesta helposti.

"Advice is the only commodity on the market where the supply always exeeds the demand."

Anonymous

1. Introduction

The behavioral finance challenge to the efficient market hypothesis. In recent years there has been a rapid growth in a branch of economics known as behavioral finance. The field of behavioral finance seeks to explain why and how financial markets can be inefficient by bringing in insights from human psychology to explain the behavior of financial market participants. As long ago as 1912 C.C. Selden wrote a book entitled Psychology of the Stock Market which was based 'upon the belief that the movements of prices on the exchanges are dependent to a very considerable degree on the mental attitude of the investing and trading public'. But it was not until the publication of a paper entitled 'Does the Stock Market Overreact?' by De Bondt and Taylor in 1986 that the study of behavioral finance began. However, cognitive psychologists Daniel Kahneman and Amos Tversky are considered to be the fathers of behavioral economics/finance, having published over 200 papers. Kahneman was awarded the Nobel Prize in Economics in 2000 for his contribution.

In their papers, Kahneman and Tversky reported on a series of experimental studies in which subjects answered questions that entailed making judgments between two monetary decisions. These decisions involved prospective losses and gains, and an example of two questions is given below:

Question 1 

You have $1,000 and you must pick one of the following choices:
• Choice A: You have a 50 per cent chance of gaining $1,000, and a 50 per cent chance of gaining $0.
• Choice B: You have a 100 per cent chance of gaining $500.

Question 2 

You have $2,000 and you must pick one of the following choices:
• Choice A: You have a 50 per cent chance of losing $1,000, and a 50 per cent of losing $0.
• Choice B: You have a 100 per cent chance of losing $500.

There is no correct answer to the above questions. If the subjects answer the questions logically then they should give the same answer to both with people answering 'B' being more risk-averse than those answering 'A'. However, in their study a large majority of people chose 'B' for question 1 and 'A' for question 2. The implication is that people are willing to settle for a reasonable level of gain even if they have a reasonable chance of earning more, but are willing to engage in risk-seeking behavior where they can limit their losses. The conclusion is that losses are weighted more heavily by economic agents than an equivalent amount of gains. Kahneman and Tversky created prospect theory to explain this 'disposition effect'. The disposition effect is also observed in real-life situations; for example, people tend to cash in winning stocks for too little and yet hold losing stocks for too long, when a more logical approach would be to hold onto winning stocks for further gains whilst selling losing stocks to prevent mounting losses.

Another concept is that of anchoring, whereby economic agents tend to focus, or anchor, on a piece of irrelevant information, even though that information is of no particular relevance to the share price. In their 1974 paper 'Judgment Under Uncertainty: Heuristics and Biases', Kahneman and Tversky spun a wheel showing showing the numbers 1 to 100. Once the wheel was spun, subjects were asked whether the percentage of African countries making up United Nations membership was greater or lower than the number on the wheel. They were then asked to estimate the actual percentage. Kahneman and Tversky found that the value of the subjects' estimate was heavily influenced by the random value from the wheel spin. For example, if the wheel spin read 60 then the estimates tended to be high, like 50 per cent, whereas if the wheel spin read 20 the estimates might be only 16 per cent. The value given by the wheel spin was acting as an anchor to the response given, even though it was irrelevant to the question in hand. Anchoring can occur in financial markets. For instance, people tend to look at the year high of a share, and if the current price is well below that high they think they are getting a bargain. This perception endures even if the share has been falling in value so that the recent high is of no relevance to whether it is undervalued.

Another concept is that of herd behavior whereby people tend to mimic the actions of a large group, although they might act differently if they acted individually. Herd behavior can be explained by people not wanting to appear as outcasts or by a belief that a large group must be 'in the know'. During the dot com bubble many people bought Internet stocks at extremely inflated prices because they observed others doing so, even though in normal times they would have avoided such stocks. In 1985 Werner De Bondt and Richard Thaler published a paper in the journal of Finance entitled 'Does the Market Overreact?'. The authors examined stock returns on the New York Stock Exchange over a three-year period. Stocks were separated into two portfolios, the 35 best-performing stocks were placed into a 'winners portfolio' and the 35 worst-performing stocks were placed into a 'losers portfolio'. The authors then tracked both portfolios against a representative market index for three years. They found that the losers portfolio consistently outperformed the market index while the winners portfolio consistently performed worse than the market index. The cumulative difference between the losers portfolio and the winners portfolio was a significant 25 per cent. As such the original losers became the future winners while the original winners became the future losers. De Bondt and Thaler's study suggested that the price of winning stocks rose too much in the winning years while the price of losing stocks fell too much, suggesting that market participants pushed the prices too high and down too low to begin with. One possible reason for over-reaction could be availability bias whereby people tend to attach too much importance to recent news and fail to take a longer term perspective. For instance a company may report a very good quarter and so its share price increases rapidly in the short term, but had a longer-term perspective been taken the quarter might be viewed as a one-off and share price would not rise as much or as quickly.

Another problem is that of overconfidence. In a 2006 study entitled 'Behaving Badly' James Montier found that 74 per cent of professional fund managers believed they had delivered above-average performance, with the vast majority of the remaining 26 per cent classifying themselves as average performers. In other words, nearly 100 per cent believed that their performance was average or better, suggesting that a degree of overconfidence prevails in the industry. In addition, there is confirmation bias and hindsight bias to consider. Most of us have preconceived ideas about things and we tend to seek out information that confirms our preconceived ideas while ignoring, or not fully taking on board, evidence to the contrary. The confirmation bias can result in poor decision making because the one-sided view of the investment decision means that the downside and risks have not been fully considered. For instance, you might think that biotech stocks are the next big thing and worth investing in. You might then look for all the positive information about how they have the potential to produce new medical cures, generate huge potential revenues and profits, but ignore negative information that suggests only 1 in 10 biotech companies actually succeeds in generating a profit for its shareholders.

Obviously there is a lot more to behavioral finance than we can cover here and it is a rapidly growing area of finance. There now exist funds that seek to make money out of known behavioral patterns. While behavioral finance has provided many useful insights it has not yet been fully accepted as a new paradigm by the finance profession. Indeed, the father of the efficient market hypothesis, Eugene Fama mounted a robust defense of the EMH in a 1998 article entitled 'Market Efficiency, Long Run Returns and Behavioral Finance'. He argued that behavioral finance is riddled with contradictions and that many so called anomalies can disappear if a proper methodology is applied. For instance, the winner-loser anomaly could be due to differences between the risk premium for losing shares going up, resulting in higher future returns, and the risk premium for winning stocks going down, resulting in lower future returns. Fama argued that some anomalies are really the result of chance and that markets are as likely to under-react as they are to over-react to new information.

2. Psychology and finance

Traditionally, a formal education in finance has dismissed the idea that one's personal psychology can be a detriment in making good investment decisions. For the past three decades, the field of finance has evolved based on the following two assumptions.

• People make rational decisions.
• People are unbiased in their predictions about the future.

By assuming that people act in their own best interests, the finance field has been able to create some powerful tools for investors. For example, investors can use modern portfolio theory to obtain the highest expected return possible for any given level of risk they can bear. Pricing models (such as the Capital Asset Pricing Model, the Arbitrage Pricing Theory, and option pricing) can help value securities and provide insights into expected risks and returns. Investment texts are full of these useful theories.

However, psychologists have known for a long time that these are bad assumptions. People often act in a seemingly irrational manner and make predictable errors in their forecasts. For example, traditional finance assumes that people are risk averse. They prefer not to take risk but will do so if the expected rewards are sufficient. People should also be consistent in their level of risk aversion. But in the real world, people's behaviors routinely violate these assumptions. First, people exhibit risk aversion when buying insurance and simultaneously exhibit a risk-seeking behavior by buying lottery tickets.

The finance field has been slow to accept the possibility that economic decisions could be predictably biased. Early proponents of behavioral finance often were considered heretics. Over the past decade though, the evidence that psychology and emotions influence financial decisions became more convincing. Today, the early proponents of behavioral finance are no longer heretics but visionaries. Although the controversies of when, how, and why psychology affects investing continue, many believe that the 2002 Nobel Prize in Economics awards to psychologist Daniel Kahneman and experimental economist Vernon Smith have vindicated the field.

Financial economists are now realizing that investors can be irrational. Indeed, predictable decision errors by investors can affect the function of the markets. The contributions of behavioral finance include (1) documenting actual investor behavior, (2) documenting price patterns that seem inconsistent with traditional models with rational investors, and (3) providing new theories to explain these behaviors and patterns.

Perhaps most important, people's reasoning errors affect their investing and ultimately their wealth. Investors who understand the tools of modern investing still can fail as investors if they let psychological biases control their decisions.

2.1 Prediction

The brain does not work like a computer. Instead, it frequently processes information through shortcuts and emotional filters to shorten analysis time. The decision arrived at through this process is often not the same decision you would make without these filters. These filters and shortcuts can be referred to as psychological biases. Knowing about these psychological biases is the first step toward avoiding them. One common problem is overestimating the precision and importance of information. The following demonstration illustrates this problem.

Let's face it, investing is difficult. You must make decisions based on information that might be inadequate or inaccurate. Additionally, you must understand and analyze the information effectively. Unfortunately, people make predictable errors in their forecasts. Consider the 10 questions in following table.

Although you probably do not know the answers to these questions, enter the most probable range based on your best estimate. Specifically, give your best low guess and your best high guess so that you are 90 percent sure the answer lies somewhere between the two. Don't make the range so wide that the answer is guaranteed to lay within the range, but also don't make the range too narrow. If you consistently choose a range following these instructions, you should expect to get 9 of the 10 questions correct. Go ahead, give it your best shot.

If you have no idea of the answer to a question, then your range should be large to be 90 percent confident. On the other hand, if you think you can give a good educated guess, then you can choose a smaller range to be 90 percent confident. Now let's check the answers (they are (1) 113 500 kgs, (2) 1513, (3) 192 countries, (4) 16 974 kms, (5) 206 bones, (6) 8.3 million, (7) 19 million, (8) 6 440 kms, (9) 1 681 kms per hour, and (10) 9.5 million). Count your response correct if the answer lies between your low and high guesses. How many did you get right?

Most people miss five or more questions. However, if you are 90 percent sure of your range, then you should have missed only one. The fact is that you are too certain about your answers, even when you have no information or knowledge about the topic. Even being educated in probability is no help. Most finance professors miss at least five of the questions, too.

This demonstration illustrates that people have difficulty evaluating the precision of their knowledge and information. Now that you see the difficulty, you can have a chance to redeem yourself. Consider the following question.

In 1896, the Dow Jones Industrial Average (DJIA) was at 40. At the end of 2006, the DJIA was at 12 463. The DJIA is a price- weighted average. Dividends are omitted from the index. What would the DJIA average have been at the end of 2006 if the dividends were reinvested each year?

Notice that table has room for your DJIA minimum and maximum guesses. Again, you should be 90 percent sure that the correct value lies within the range you choose. Because you are 90 percent sure that the correct value lies within the range you chose, you should get this one correct. Are you ready for the answer? (If dividends were reinvested in the DJIA, the average would have been 1 037 090 at the end of 2006. Does this surprise you? It surprises most people.) Even after learning that most people set their range too narrowly in their prediction and experiencing the problem firsthand, most people continue to do it.

This example also illustrates another aspect of investor psychology called anchoring. When you read the question, you focused on the DJIA price level of 12 463; that is, you anchored your thinking to 12 463. You probably made your guess by starting at this anchor and then trying to add an appropriate amount to compensate for the dividends. Investors anchor on their stock purchase price and the recent highest stock price.