Contrarian Investment Strategies
Page 27
Neuroeconomics is an important new discipline within economics. Major contributors to this rapidly growing field have written about a significant amount of new research work that appears to strongly support the hypothesis that at least a part of the earnings surprise findings, particularly event triggers and reinforcing events, can be attributed to relatively new research in this field: specifically, the four different types of surprise we saw in Figure 9-7.
We do not need to be experts in biology, chemistry, and neuropsychology to understand the powerful answers that this new discipline gives for why all four kinds of surprises play out the way they do. Fortunately, understanding the interactions themselves is much simpler.
Let’s start with dopamine, a chemical naturally released by the body that is commonly associated with the pleasure systems of brain chemistry. It provides feelings of enjoyment and reinforcement, which motivate a person to undertake certain activities. Dopamine, a neural transmitter, transmits nerve impulses, which are released by naturally rewarding experiences such as food, sex, certain drugs, and the neural stimuli associated with them. Of the brain’s approximately 100 billion neurons, only about one-thousandth of 1 percent produce dopamine. But this minuscule group has a major effect on how your brain makes certain choices, including investment decisions. It also plays a major role in both alcohol and drug abuse. As Jason Zweig, a respected columnist and author, wrote in Your Money and Your Brain, alcohol, marijuana, cocaine, morphine, and amphetamines all seem to be related to the release of dopamine. “All hook their users by affecting in a variety of ways the trigger zones for dopamine in the brain.”10 A hit of cocaine, for example, tells the brain to release dopamine about fifteen times as fast as normal, indicating that it may somehow help transmit the euphoric kick of the drug. “Dopamine spreads its fingers all over the brain,” says a neuroscientist, Antoine Bechara of the University of Southern California, of this process.11 When the dopamine neurons are lit up, they send gushers of energy throughout the brain that make and put decisions into actions. Ironically, the brain patterns look almost identical when dopamine is quickly activated after the inhalation of cocaine and when an investor is excited by an investment decision he has just made.12
Dopamine is more than simply a rush on its own. An investor must know more: that there’s a sizable reward in taking a given action. He must also take the required actions that he thinks will capture the reward.
Next, let’s look at the role of dopamine in the various forms of earnings surprises examined. Dopamine neurons are released by rewarding events that are better than predicted, remain uninfluenced by events that are as good as predicted, and are depressed by events that are worse than predicted. This, it appears, is the key reason why effects of the four forms of earnings surprises we just examined have been so consistent over time. Remember: the possibility that the event trigger and the reinforcing events are not random is about 1,000 to 1. If the link between neuroeconomics and earnings surprises stands—and the evidence certainly points that way—this might forge a strong bond between predictable and repetitive economic events and neuroeconomic and Affect findings. These correlations, given the high odds that event triggers and reinforcing events are not pure chance, are highly probable. If this is the case, the similarity of findings in neuroeconomic experiments to those on earnings surprise presents a fascinating topic of research. If the current experiments do show a tie between the two, neuroeconomics as a major economic research tool will certainly become far more important.
Let’s briefly look at neuroeconomics research that appears to explain the four categories of classes of earnings surprises, which, as we know, divide into two distinct categories, event triggers and reinforcing events. The researchers Wolfram Schultz of the University of Cambridge (Department of Physiology, Development and Neuroscience), Read Montague at Baylor College of Medicine, and Peter Dayan at University College London have made a number of important research findings about dopamine and reward. They discovered that getting what you anticipated produces no dopamine rush, sending out electromechanical pulses at their resting rate of about three bursts per second. Even though a reward is expected to make investors enthusiastic, it does not. This is almost precisely what we see with the high-P/E phenomenon with positive surprises and negative surprises on out-of-favor stocks with reinforcing events. The reward—in this case a positive surprise on a favored stock—is actually more than investors expected, but the reaction is almost nebulous and sometimes negative. This may also explain why drug addicts require larger hits to get the same high and why investors require larger earnings hits on “best stocks to see their prices move higher.”
To neuroeconomists, the two event triggers would be the real surprises. They are major unexpected events. According to a research study by Pammi Chandrasekhar, C. Monica Capra, Sara Moore, Charles Noussair, and Gregory Berns,13 higher-than-expected rewards are unanticipated and as a result release dopamine. Looking at event triggers shows that a positive earnings surprise for an unfavored stock is also unanticipated by investors holding or interested in the stock. They are likely to experience “rejoice” or elation when their stock shows a positive earnings surprise, and their brains release dopamine.
Research findings on monkeys also show strong results for unanticipated positive surprises. (I apologize to readers who might be offended by the comparison of monkeys to humans. In my defense I can only point out that monkeys, chimpanzees, and pigeons scored higher on some of these neuropsychology tests than did people.)
Schultz studied the brains of monkeys and found that when dopamine comes as the result of a surprise the dopamine neurons fire more strongly and for a longer time than for a reward that was anticipated beforehand.14 This research, although only now being tested on event triggers, appears to support the unanticipated positive surprise on low-P/E stocks in the event trigger. When investors receive unexpectedly higher earnings on out-of-favor stocks, their dopamine is also likely to fire up almost instantaneously and strongly—Schultz’s studies show from three to forty times a second. In Schultz’s words, “This kind of positive reinforcement creates a special kind of attention dedicated to rewards. Rewards are what keep you coming back for more.”15 Schultz and Anthony Dickinson, in the 2000 Annual Review of Neuroscience, wrote, “In summary, the reward responses depend on the difference between the occurrence and the prediction of reward (dopamine response = reward occurred – reward predicted).”16
Chandrasekhar and colleagues’ results suggest that activation of a neural network consisting of the rostral anterior cingulate, left hippocampus, left ventral striatum, and brainstem/midbrain is correlated with rejoicing.17 Don’t worry, this is not a test; it’s merely here to show how complex these neural interactions can be. Similarly, the second event trigger, a negative surprise on a favorite stock, might cause regret and disappointment. In this case another neural network, the cortical network, is activated. Neuroeconomists can measure the two classes of surprises by using functional magnetic resonance imaging (fMRI). This network activates the degree of regret.18
Schultz, Montague, and Dayan also found that if rewards we expect don’t pan out, dopamine dries up.19 Dopamine neurons activate when people spot a signal that the reward is coming, but if it doesn’t come, they will instantly stop firing. The brain is thus deprived of its expected shot of dopamine, and disappointment sets in. This is similar to the reaction we see in the earnings surprise on a favored stock that has a negative surprise (an event trigger). Schultz and Dickinson also observed that omitted rewards induce opposite changes in dopamine neurons compared with unpredicted rewards. If a predicted reward fails to occur, dopamine neurons are depressed at the time the reward would have occurred. This suggests a form of error coding that is compatible with the idea that the error directly controls learning about the prediction.20 The fact that there is no positive reward, but a negative consequence, for each security in the best-stock negative-surprise portfolio quite possibly is responsible for the significan
t, immediate, and then continuing drop in price. As we have seen, both the positive and the negative effects last for at least four quarters.
Reinforcing events, the two other classes of earnings surprises we noted—a positive surprise on a favored company and a negative surprise on an out-of-favor stock—do not seem to have much impact on our neuroprocessing or, for that matter, in the marketplace, from what the neuropsychologists report in related studies. For example, winning a bet when it is highly probable evokes less rejoicing than winning a bet when the outlook was unlikely. Similarly, losing a bet when the outcome is probable evokes less regret than losing a bet when the outlook is improbable. Chandrasekhar and colleagues indicate that the degree of regret or rejoicing correlates with the perceived probabilities of winning or losing. The higher the expectation of winning, the lower the amount of rejoicing; and the lower the expectation of winning, the higher the amount of rejoicing. Also noted was that different brain regions exhibited activation that increased with the levels both of regret and rejoicing. The authors conclude, “Our results suggest that distinct but overlapping networks are involved in the experiences of regret and joy.”21 What seems apparent from this analysis is that expected positive earnings surprises for favorite stocks result in only a small amount of rejoicing. The same is likely of regret for negative surprises on out-of-favor stocks.
Figure 9-8 demonstrates that this is exactly what happens with the stocks we have labeled as reinforcing events.*54 Also of interest is the difference in the size of event triggers and reinforcing events, as the neuroeconomists’ work would suggest. The event trigger’s impact on stock prices is about four times as much as that of reinforcing events in the quarter of the surprise, and almost twenty-four times as much after one year (removing signs on both). The chart is statistically significant at the 0.1 percent level; this means that there is only a 1-in-1,000 possibility that it could be sheer chance. For the investor it clearly provides some robust neuroeconomics findings that appear to strongly support the purchase of out-of-favor stocks.
The Effects of Surprise over Time
We have seen the results of surprises on “best” and “worst” stocks for up to one year after the surprise is announced. Are there lingering effects beyond that? Figure 9-9, which measures the performance of the “best” and “worst” groups of stocks by P/E ratios for five-year periods following an earnings surprise, using a buy-and-hold strategy, provides the answer.*55 The figure indicates that the lowest-P/E group showing positive earnings surprises (low-P/E positive) outperformed the market in all twenty quarters after the surprises and recorded an above-market return of 30.3 percent for the five-year period. Conversely, the highest-P/E group receiving negative surprises (high-P/E negative) underperformed in every quarter for the following five years, lagging behind the market by 30.4 percent for the full period. As we can see, the differential between the two groups continued to increase significantly through the five years measured.
Is the entire difference in performance between the two types of event triggers caused by earnings surprises? Did the original surprises change investors’ perceptions permanently? These questions are impossible to answer statistically at this time. We do know that investors were far too confident of their prognostications for both “best” and “worst” stocks; consequently, “best” stocks were significantly overvalued and “worst” stocks were undervalued. When the dark or rose-colored glasses were removed, perhaps they were swapped for each other. As was also noted earlier, not one but a series of surprises may occur, some in later quarters, including surprises other than analysts’ forecast errors, that continue to reinforce the price reevaluations.
What we can say, however, is that the enormous market outperformance by the low-P/E and other low-value groups*56 and the underperformance by the highest quintile in each value group indicate that there certainly had to be an event or a series of events that changed investors’ perceptions of what were “best” and “worst” stocks.
We also see the effects of reinforcing events. “Worst” stocks with negative surprises (low-P/E negative) consistently outperformed after the surprise quarter and for the next nineteen quarters, while “best” stocks with positive surprises (high-P/E positive) just as consistently underperformed. Although the differences are not as large as for “best” and “worst” stocks that experienced an event trigger in the surprise quarter, they are still major. “Best” stocks underperformed the market by 21.3 percent in the full five-year periods, while “worst” stocks outperformed by 20.4 percent. (The results for the two other value measures, price to cash flow and price to book value, are again similar.)
The middle group is not shown. However, the long-term findings differ little from those at the end of the first year. Surprise has a major effect only in the quarter in which the news is announced. After that the stocks perform in line with the market. Overall, positive and negative surprises almost cancel each other out, and this is pretty much what we should expect, since being in the middle group shows that the stocks are not overvalued or undervalued by much.
A Surprising Opportunity
Psychological Guideline 18 positioned us in out-of-favor stocks to take advantage of analysts’ forecast errors and other surprises. We can now go further in delineating the effect of surprises, which will prove an essential tool for the strategies to be outlined shortly. Psychological Guideline 20 summarizes our findings on surprise.
PSYCHOLOGICAL GUIDELINE 20(a): Surprises, as a group, improve the performance of out-of-favor stocks, while impairing the performance of favorites.
PSYCHOLOGICAL GUIDELINE 20(b): Positive surprises result in major appreciation for out-of-favor stocks, while having minimal impact on favorites.
PSYCHOLOGICAL GUIDELINE 20(c): Negative surprises result in major drops in the price of favorites, while having virtually no impact on out-of-favor stocks.
PSYCHOLOGICAL GUIDELINE 20(d): The effect of an earnings surprise continues for an extended period of time.
In this chapter, we have examined the role of surprises and have found that they consistently favor stocks that investors believe have poor outlooks and just as consistently work against those believed to be la crème. Because of the frequency of earnings surprises demonstrated in the preceding chapter, we know they are a powerful force acting to reverse previous over- or undervaluations of stocks.
Just how important surprises and the resulting changes in investors expectations are in developing powerful investment strategies will be shown in the next chapter. With our probabilities of winning moving increasingly higher, it’s almost time to roll up our sleeves and sit down at a table in the investment casino. Now, what sort of odds would you care to have?
Part IV
Market Overreaction: The New Investment Paradigm
Chapter 10
A Powerful Contrarian Approach to Profits
ONLY DIE-HARD New Yorkers or sports masochists—I guess I fall into the latter category—rooted for the New York Giants in Super Bowl XLII. The bookies put the odds at an enormous 17 points favoring the New England Patriots. This meant that if you wanted to bet on the Patriots and collect, they would have to win over the Giants by more than two touchdowns and a field goal. So confident were the public and the experts that this would not happen that The Boston Globe had already begun preselling a book on Amazon.com. It was entitled 19–0: The Historic Champion Season of New England’s Unbeatable Patriots. To be released immediately after the inevitable win.
The fans and sportscasters believed it was a mismatch of almost historic proportions. The Giants, underdogs by far entering the playoffs, thought otherwise and made it through to the championship round. In February 2008, a TV audience estimated at more than 50 million saw a hard-fought game. When it was over and the artificial dust had cleared on Super Bowl XLII, the Giants had beaten the mighty Patriots. It took a Giants touchdown in the last thirty-five seconds of play to seal the Patriots’ fate, but that’s what happened, and it sent the stun
ned fans into a state of shock.
How could the most powerful football team in decades be reduced to such an ignoble outcome? Many money managers and other professional investors, immaculately dressed in their Paul Stuart suits and Ferragamo ties, sitting in their well-appointed offices, flanked by the latest and most powerful market technology of the day, must wonder the same thing: how can they perform so dismally when they buy the top research and investment advice?
It must be a real puzzle to them.
But . . . are the stocks the experts like really the ones to buy? We have seen that the answer is a strong no. The favorite stocks of analysts and money managers are consistently punished by earnings surprises, while the stocks nobody loves or wants just as consistently benefit from surprises. Investors’ enthusiasm often causes popular stocks to become overpriced, while the lack of it dumps them into the bargain basement. Earnings and other surprises result in a reevaluation of both groups and more realistic pricing.
This is certainly a large, critical piece of the puzzle, but how do we fit it into a practical investment strategy? After all, we’re not just trying to analyze market dynamics; we’re seeking to learn patterns of how to profit from them. For a change, rather than trying to build a deductive case, I’m going to put the answer right out on the table for your inspection. While I’m at it, let’s identify the core of this proven investment approach and make the next Psychological Guideline.
PSYCHOLOGICAL GUIDELINE 21: Favored stocks underperform the market, while out-of-favor companies outperform the market, but the reappraisal often happens slowly, even glacially.
This reevaluation process, heavily influenced by the new psychological findings of Affect and neuroeconomics, is the key to large and consistent profits in the marketplace. From the dawn of markets, people’s reaction to “best” and “worst” investments has been consistent and predictable. And here we come to the bottom line for the practical reader: investors’ behavior is so predictable in this respect that the average reader can take advantage of it. There are proven yet uncomplicated contrarian strategies that should allow you to outperform the market handily, with relatively little risk.