Cognitive and Emotional Aspects
People who exhibit frame dependence do so for both cognitive and emotional reasons. The cognitive aspects concern the way people organize their information, while the emotional aspects deal with the way people feel as they register the information.
The distinction between cognitive and emotional aspects is important. For example, the main cognitive issue in choice problem 3 is whether people ignore having just won $1,500 when deciding whether or not to take an even chance on winning or losing $450. Some do ignore the $1,500, whereas others see themselves as being $1,500 ahead. The cognitive and emotional aspects operate together, in that those who ignore the $1,500 feel a $450 loss as just that, a $450 loss. But those who begin by seeing themselves as $1,500 ahead instead experience a $450 loss as a smaller gain of $1,050. This difference affects behavior: Because of loss aversion, people who ignore having just won $1,500 are much less prone to accepting the gamble than those who see themselves as $1,500 ahead. Thaler and Johnson call this a “house money” effect.8
The term frame dependence means that the way people behave depends on the way that their decision problems are framed. Hedonic editing means they prefer some frames to others. That is the main insight to be gleaned from studying how people chose in the four preceding choices. In a financial context, hedonic editing offers some insight into investors’ preference for cash dividends. When stock prices go up, dividends can be savored separately from capital gains. When stock prices go down, dividends serve as a “silver lining” to buffer a capital loss. Remember Merton Miller’s succinct description of frame independence? Some investors prefer to keep dividends in their right pocket.
The following excerpt, taken from a Forbes magazine interview with closed-end fund manager Martin Zweig, describes how he came to realize the importance of dividends. It began with the fact that his fund was trading at a deep discount relative to net asset value (NAV), the value the shares would trade for if the fund were open-ended instead of being closed.
Then in 1986 we did a closed-end fund. … I always worried about discounts on closed-end funds. … The first nine months out of the gate, we were at a 17 percent discount. I was mortified. I sat down and did a lot of thinking. Bond funds at the time were selling at about parity. Stock funds were all at discounts. It didn’t make sense, because stocks do better than bonds in the long run. And I realized bond funds pay interest. People like the certainty of an income stream. So I said, “Well, we’re going to pay the dividend, whether we earn it or not.” And we went to this 10 percent dividend policy. … The discount narrowed immediately. (Brimelow, 1998)
Self-Control
Self-control means controlling emotions. Some investors value dividends for self-control reasons as well as for reasons that stem from hedonic editing.
Martin Zweig talks about paying a dividend whether earned or not because people “like the certainty of an income stream.” What does a reliable dividend have to do with self-control? Meir Statman and I (Shefrin and Statman 1984) argue that the answer involves the “don’t dip into capital” heuristic. Older investors, especially retirees who finance their living expenditures from their portfolios, worry about spending their wealth too quickly, thereby outliving their assets. They fear a loss of self-control, where the urge for immediate gratification leads them to go on a spending binge. Therefore, they put rules into place to guard against the temptation to overspend.
“Don’t dip into capital” is akin to “don’t kill the goose that lays the golden eggs.” But if you don’t dip into capital, how do you finance consumer expenditures—Social Security and pension checks alone? Not necessarily—this is where dividends come in. Dividends are labeled as income, not capital. And investors tend to frame dividends as income, not capital. Again, this is frame dependence. Investors feel quite comfortable choosing a portfolio of stocks that feature high dividend payouts and spending those dividends.
Regret
Imagine someone who makes a decision that turned out badly and engages in self-recrimination for not having done the right thing. Regret is the emotion experienced for not having made the right decision. Regret is more than the pain of loss. It is the pain associated with feeling responsible for the loss.
For example, imagine someone who has a regular route to work. One day, for the sake of variety, she decides to try a different route. That particular day she winds up in an accident. Now, even if the odds of an accident were no different on the two routes, how will that person feel? Will she chastise herself, thinking “If only I had done what I always do and taken my regular route!” If so, she is experiencing the frustration of regret.
Regret can affect the decisions people make. Someone who feels regret intensely, does not have a strong preference for variety, and thinks ahead, may follow the same route to work every day, in order to minimize possible future regret.
Here is a financial example. Consider the choice of equity-fixed income allocation in a defined contribution retirement plan. In the January 1998 issue of Money magazine, Harry Markowitz explains what motivated his personal choice about allocation. As the Nobel laureate recognized for having developed modern portfolio theory, was he seeking the optimum trade-off of risk and return? Not exactly. He said, “My intention was to minimize my future regret. So I split my contributions fifty-fifty between bonds and equities” (Zweig 1998, 118). In other words, had Harry Markowitz selected a 100 percent equity allocation, and had stocks subsequently done terribly, it would have been to easy, in hindsight, to imagine having selected a more conservative posture—and this would give rise to considerable self-recrimination, meaning regret.
Regret minimization also leads some investors to use dividends, instead of selling stock, to finance consumer expenditures. Those who sell stock to finance a purchase, only to find that shortly thereafter the stock price soars, are liable to feel considerable regret. That is often at the heart of expressions such as “this is my half-million-dollar car.”
Money Illusion
Frame dependence also impacts the way that people deal with inflation, both cognitively and emotionally. This is the issue of money illusion. Let us examine the following questions from a study by Eldan Shafir, Peter Diamond, and Amos Tversky (1997).
Consider two individuals, Ann and Barbara, who graduated from the same college a year apart. Upon graduation, both took similar jobs with publishing firms. Ann started with a yearly salary of $30,000. During her first year on the job, there was no inflation, and in her second year, Ann received a 2 percent ($600) raise in salary. Barbara also started with a yearly salary of $30,000. During her first year on the job, there was 4 percent inflation, and in her second year, Barbara received a 5 percent ($1500) raise in salary.
a. As they entered their second year on the job, who was doing better in economic terms, Ann or Barbara?
b. As they entered their second year on the job, who do you think was happier, Ann or Barbara?
c. As they entered their second year on the job, each received a job offer from another firm. Who do you think was more likely to leave her present position for another job, Ann or Barbara?
Most people indicate that Ann is better off, Barbara is happier, and Ann is more likely to look for another job. Now this is somewhat perplexing. If Ann is better off, why is she less happy and more likely to look for another position? Shafir, Diamond, and Tversky suggest that although people can figure out how to adjust for inflation, it is not a natural way for them to think. The natural way is to think in terms of nominal values. Therefore people’s emotional reaction is driven by the nominal values, and those appear more favorable for Barbara than they do for Ann.
Summary
This chapter presents the second theme of behavioral finance, frame dependence, which deals with the distinction between form and substance. Framing is about form. In short, frame dependence holds that differences in form may also be substantive. It reflects a mix of cognitive and emotional elements. The cognitive issues per
tain to the way that information is mentally organized, especially the coding of outcomes into gains and losses. There are several emotional issues, the most fundamental of which is that people tend to feel losses much more acutely than they feel gains of comparable magnitude. This phenomenon has come to be known as loss aversion. Therefore, people prefer frames that obscure losses, if possible—and engage in hedonic editing. People tend to experience losses even more acutely when they feel responsible for the decision that led to the loss; this sense of responsibility leads to regret. Regret is an emotion. People who have difficulty controlling their emotions are said to lack self-control. Some people use framing effects constructively to help themselves deal with self-control difficulties.
Chapter 4 Inefficient Markets: The Third Theme
This chapter discusses the following:
• representativeness, and the market’s treatment of past winners and losers
• anchoring-and-adjustment, and the market’s reaction to earnings announcements
• loss aversion, and the risk premium on stocks
• sentiment, and market volatility
• overconfidence, and the attempt to exploit mispricing
Cause and Effect
One of the most fiercely debated questions in finance is whether the market is efficient or inefficient. Remember the hedge fund Long-Term Capital Management (LTCM)? How did it advertise itself to investors? LTCM members promoted their firm as an exploiter of pricing anomalies in global markets. In this regard, consider the following heated exchange between Myron Scholes, LTCM partner and Nobel laureate, and Andrew Chow, vice president in charge of derivatives for potential investor Conseco Capital. Chow is quoted as saying to Scholes, “I don’t think there are that many pure anomalies that can occur”; to which Scholes responded: “As long as there continue to be people like you, we’ll make money.”1
That last remark might not be the best way to win friends and influence people. But Scholes is correct about cause and effect—investors’ errors are the cause of mispricing. Is the market efficient?
The fact is that from 1994 through 1997, LTCM claims to have successfully made leveraged bets—bets that exploited mispricing identified by the option pricing theory for which Scholes and Merton jointly received the Nobel prize. In this regard Merton Miller, another Nobel laureate, is quoted as having said, “Myron once told me they are sucking up nickels from all over the world. But because they are so leveraged, that amounts to a lot of money.”2 “Sucking up nickels” is indicative of inefficiency. Of course, then came LTCM’s 1998 fiasco, but more on that later.
Effects Stemming from Representativeness
Let’s begin with the De Bondt-Thaler winner-loser effect. De Bondt and Thaler (1985) argue that investors who rely on the representativeness heuristic become overly pessimistic about past losers and overly optimistic about past winners, and that this instance of heuristic-driven bias causes prices to deviate from fundamental value. Specifically, past losers come to be undervalued and past winners come to be overvalued. But mispricing is not permanent; over time the mispricing corrects itself. Then losers will outperform the general market, while winners will underperform.
De Bondt and Thaler (1989) present evidence in support of their claim. Figure 4-1 displays the returns to two portfolios, one consisting of extreme losers and the other of extreme winners. In both cases, the criterion used to judge performance is past-three-year returns. Extreme losers are the stocks that lie in the bottom tenth percentile, while the stocks that lie in the top tenth percentile are the extreme winners.
Figure 4-1 shows cumulative returns to the two portfolios for the sixty months after formation, relative to the overall market. Notice that the cumulative returns are indeed positive for losers, about 30 percent, and negative for winners, about—10 percent. De Bondt and Thaler suggest that this pattern signifies a correction to mispricing.
In traditional finance, the pattern depicted in figure 4-1 would reflect compensation for risk. That is, losers would be associated with higher returns because they are riskier than the average stock; the opposite holds for winners. But De Bondt and Thaler contend that an investor who bought losers and sold winners short would have beaten the market by about 8 percent on a risk-adjusted basis. I discuss this issue further in chapter 7.
Figure 4-1 Cumulative Average Residuals For Winner and Loser Portfolios Of 35 Stocks (1–60 Months Into the Test Period)
Cumulative abnormal returns for two portfolios, one consisting of past losers and the other consisting of past winners. Past losers subsequently outperform, while past winners subsequently underperform.
Effects Stemming from Conservatism
Analysts who suffer from conservatism due to anchoring-and-adjustment do not adjust their earnings predictions sufficiently in response to the new information contained in earnings announcements. Therefore, they find themselves surprised by subsequent earnings announcements. Unanticipated surprise is the hallmark of overconfidence. However, there is more at work here than plain overconfidence. Conservatism in earnings predictions means that positive surprises tend to be followed by positive surprises and negative surprises tend to be followed by negative surprises.
Does conservatism in analyst earnings predictions cause mispricing? If it does, then we should find that stocks associated with recent positive earnings surprises should experience higher returns than the overall market, while stocks associated with recent negative earnings surprises should earn lower returns than the overall market. Figure 4-2, taken from an article by the late Victor Bernard and Jacob Thomas (1989), summarizes the evidence.3 The figure shows the behavior of cumulative returns to portfolios formed based on the size of the most recent earnings surprise. In the sixty days following an earnings announcement, the stocks with the highest earnings surprises outperformed the overall market by about 2 percent, while the stocks with the most negative earnings surprises underperformed the overall market by about 2 percent.
Figure 4-2 Cumulative Abnormal Returns (CAR) To Portfolios Based Upon Standardized Unexpected Earnings (SUE)
What happens to stock prices after earnings surprises: price momentum is greater for bigger surprises. Cumulative abnormal return pattern is steeper with the magnitude of the surprise (SUE).
Behavioral finance suggests that heuristic-driven errors cause mispricing. As an example, look at the pricing pattern depicted in figure 4-2. Traditional finance holds that this pricing pattern occurs because stocks associated with positive earnings surprises are riskier than the stocks associated with negative earnings surprises. In chapter 8, I discuss why this pricing pattern cannot be explained in terms of risk.
Effects Stemming from Frame Dependence
Does frame dependence have an impact on price efficiency? Shlomo Benartzi and Richard Thaler (1995) suggest that the answer is a strong yes. They argue that in the past, loss aversion caused investors to shy away from stocks; therefore, stocks earned very large returns relative to risk-free government securities.
Economist Jeremy Siegal documents that over the last two centuries the real return to stocks has been about 7 percent more than riskfree securities. From a theoretical perspective, a premium of 7 percent is enormous, and this differential has come to be called the equity premium puzzle (Mehra and Prescott 1985). To understand the character of the puzzle, consider the following question on risk tolerance.
Suppose that you are the only income earner in your family, and you have a good job guaranteed to give you your current (family) income every year for life. You are given the opportunity to take a new and equally good job, with an even chance it will double your (lifetime family) income and an even chance that it will cut your (lifetime family) income. Indicate exactly what the percentage cut x would be that would leave you indifferent between keeping your current job or taking the new job and facing a 50-50 chance of doubling your income or cutting it by x percent (Barsky et al. 1997).
When I administer this question to general audiences, th
e average response comes out at about 23 percent. But the kind of response necessary to justify the historical equity premium is somewhere around 4 percent. The difference between 23 percent and 4 percent is not small. In fact, being willing to tolerate no more than a 4 percent decline seems very extreme, relative to the way people normally respond to the preceding question.
Shmuel Kandel and Robert Stambaugh (1991) suggest that people might be less tolerant of risks whose magnitudes are smaller than those described in the preceding question. However, I find that when the stakes are smaller, people actually become more tolerant of risk, not less tolerant.4
Benartzi and Thaler (1995) suggest that individual investors’ historical reluctance to hold stocks may have stemmed from their evaluation horizons being too short. They call this reluctance myopic loss aversion. Benartzi and Thaler suggest that investors who are prone to myopic loss aversion can increase their comfort with equities by monitoring the performance of their portfolios less frequently, no more than once a year. It appears that investors who hold individual stocks monitor those stocks much more frequently than that. John Pound and Robert Shiller (1989) found that individual investors spent over a half-hour per day following the most recent stock they bought.5 Nicholas Barberis, Ming Huang, and Tano Santos (1999) use the Thaler-Johnson “house money effect” discussed in chapter 3, to take the argument one step further. They suggest that after a market runup, the house money effect kicks in, raising investors’ tolerance for risk, and lowering the equity premium. In a downturn the reverse occurs.
Beyond Greed and Fear Page 8