Contrarian Investment Strategies Page 7
Mental Shortcuts
Beginning in the 1970s, researchers began discovering that people adopt an extensive number of such mental shortcuts, or rules of thumb, to make a wide range of day-to-day decisions rather than formally calculating the actual odds of a given outcome. The pioneering work was done by Daniel Kahneman, now of Princeton, and the late Amos Tversky of Stanford. As noted, Kahneman won the Nobel Prize in Economics in 2002 for this work. Tversky would undoubtedly have shared the prize, the recipients of which must be living, had he not died prematurely several years earlier.*8 These judgmental heuristics or cognitive heuristics are simplifying strategies we use for managing large amounts of information. They seem to be hardwired into our minds and are thus rather undetectable under ordinary circumstances. Most of us are entirely unaware of them, especially in the heat of action.
Having evolved out of long human experience, these judgmental shortcuts work exceptionally well most of the time, as does the emotional decision making of Affect, and are enormous time-savers.
Consider, for example, what happens as you drive a car down a highway. You concentrate on dozens of mental shortcuts not only to operate the car but also to monitor highway traffic, road signs, and the visibility ahead, as well as the vehicles around you, all the while staying mostly undistracted by the music playing on your iPod and screening out thousands of other distracting and disruptive bits of information. That you can actually do this “almost without thinking” is a tribute to the efficiency of our mental heuristics.
We use such heuristics in dealing with many of our decisions and judgments and tend to become “intuitive statisticians” in the process. We think, for example, that our odds of survival in a crash are better when we are driving at fifty-five miles an hour than at ninety miles an hour, although few of us have ever bothered to check the actual numbers. We readily assess that a professional team is likely to beat an amateur one, assuming that the “amateurs” are not of Olympic quality. And we might expect to get to a city three hundred miles away faster by air than by ground.*9
We have an immediate sense of the odds in such situations (or we think we do), using our past experiences to make these assessments quickly. Hundreds of times a day, these heuristics work flawlessly for us.
However, being an “intuitive statistician” has limitations as well as advantages. The simplifying processes that are normally efficient time-savers also lead to systematic mistakes in decision making. Despite what many economists and financial theorists assume, people are not really good intuitive statisticians, and as a result, heuristics are consistently shortchanging investors, even professionals. If I had to pinpoint the one primary error that heuristics cause investors to make, it would be this: they do not calculate odds properly when making investment decisions. The distortions produced by our heuristic calculations are often large and systematic, leading even the savviest investors into blunders of considerable magnitude. In addition, these cognitive biases are locked more firmly into place by group pressures.5 Our own biases are reinforced by the powerful influence of experts and peer groups we respect, and the pressure for us to follow becomes more compelling.
The research also indicates that even when people are warned of such biases, they appear not to be able to adjust for the effects. So it will take a good deal of concentration and effort on your part to avoid these pitfalls. Doing so begins with becoming familiar with these heuristics. Once their nature is understood, a set of rules can be developed to help monitor your decisions and provide a shield against serious mishaps—perhaps even profit from them instead. But as you’ll see, it’s much easier said than done, even after people have been made fully aware of them.
The Perils of Availability
Let’s start by taking a closer look at the availability heuristic. According to Tversky and Kahneman, it causes us to “assess the frequency of a class or the probability of an event by the ease with which instances or occurrences can be brought to mind.”6 This is why we think deaths by shark attack are more common than deaths from pieces of airplanes falling on people.
The shortcut answers we get this way are accurate most of the time because our minds normally recall most readily events that have occurred frequently. But they can sometimes be strikingly off the mark. We think something is more likely because the prospect is more available in our minds when an event had an especially important meaning—in other words, it had a big impact on our thinking—and had occurred recently.7 There are two psychological errors involved here. The first of these errors is referred to in the literature as “saliency” and the second as “recency.” Saliency leads people to recall distinctly “good” events (or “bad” events) disproportionally to the actual frequency. It’s so automatic that we barely recognize that we’re doing it.8 For example, returning to the wilds of nature, the actual chance of being mauled by a grizzly bear at a national park is one or two per million visitors; the death rate is even lower. Casualties from shark attacks are an even smaller percentage of all deaths of swimmers in coastal waters. But because the saliency of an actual attack is so powerful, we wind up thinking such attacks happen more often than they really do. Our memories are heuristically biased to bring scary images to mind more quickly.
I was snorkeling alone in the Bahamas’ Exuma National Park several years back, well aware, in the abstract, that the probability of a shark attack was slim. Then a six-hundred-pound bull shark slowly came swimming toward me. As the dorsal fin passed by me inches away, it appeared to have the width of a nuclear submarine. I hoped it knew how unlikely it was to attack as I did. When the shark stopped behind me, my evaluation of attack probabilities skewed sharply higher. Fortunately, the shark didn’t have a taste for snorkeler that day—but I was left with a considerable emotional charge that skewed my own calculations of shark attacks for some time to come.
To see recency in action, consider that after disasters, such as floods and earthquakes, purchases of flood and earthquake insurance rise, even though the underlying likelihood of floods and earthquakes hasn’t changed. People are led to think these disasters are more likely because they’ve happened recently.9 Through such subsconscious mechanisms, recent, salient events often strongly influence decision making in the stock and bond markets that can cause or exacerbate sharp price movements.
One key way in which this occurs is that the recent trend is thought to be the new permanent trend. Take the market when the base rate, meaning the long-term rate of returns of stocks, is trampled under by the case rate, the more recent short-term but sharply higher rate of return stocks are currently generating.*10 Even though we know that the case rate in most instances does not last and reverts back to the base rate, whether for a stock, a group of stocks, or an industry, our minds cling to a false hope that it won’t be so this time. Recency frequently convinces us that the case rate is now the new base rate.
Consider a recent example. Bill Gross, the outstanding head of PIMCO, the largest bond house in the world, and Mohamed El-Erian, its CEO, coined the term “the New Normal” as the economy was put through the shredder in 2007 and 2008. The New Normal anticipated that low earnings and stock prices would be the norm for years into the future because the growth rate of the global economy was altered downward permanently. This chant was widely picked up by both money managers and the media and became the prevailing wisdom about stocks for more than two years.
But the stock market, in its contrarian way, completely ignored the doddering role it was assigned by the experts and started to rally. It had risen more than 100 percent by mid-2011 from its New Normal lows of March 2009. After the market doubled, the New Normal lost much of its luster. This has happened in similar circumstances in the past. Recall the New Era of the 1920s, when stocks roared, or the New Economy during the Internet bubble. I expect we’ve not heard the last of such clever coinages.
Now consider an even more glaring example. Conventional financial thinking holds that IPOs should be offered at a small discount to its ac
tual value (in the neighborhood of around 10 percent), to ensure that the issue is fully subscribed for.
Figure 3-1 demonstrates the large overpricing of IPOs during the 1996–March 2000 Internet bubble. As the figure indicates, the average IPO was priced at a small premium of 10 percent or less at the first day’s closing price for the eight years from 1987 through 1994. Over the next four years, premiums expanded significantly, averaging somewhat under 20 percent. Returns of 20 percent in a day are very rare. This major appreciation was by itself enough to attract many investors away from more conventional investments into the far more speculative arena of IPOs. Speculation took over in 1999 and the first three months of 2000, with the average IPO premium rising to a staggering 90 percent, respectively, at the close of the first day’s trading. (For those unfamiliar with initial public offerings, this meant that if you were lucky enough to get an IPO at its issue price, you made 90 to 95 percent on average on your investment by the close of the first day of trading.) Moreover, for the first three months of 2000, the peak of the bubble, the average closing price on IPOs of dot-com stocks (not shown in the figure) increased to 135 percent from the original offering price to the close of the first day’s trading. IPO valuation premiums thus multiplied ten times or more above the normal premiums over the course of the Internet bubble. The recency and saliency of the enormous price movements resulted in investors vividly recalling the sharp gains these stocks provided while downplaying their considerable risks.
A retrospective analysis indicated that the quality of this group of IPOs was certainly no better and, judging by the high rate of company failures, probably a good deal worse than IPOs of earlier time periods.
Since the 1960s, as chapter 1 documented, four major technology bubbles have occurred, with the 1996–2000 mania surpassing any in the past in terms of its size, the enormous appreciation, and the magnitude of the eventual crash.
In the bubble of the early 1980s, for example, the Value Line New Issue Survey analyzed a group of proposed IPOs and found that many were start-ups, perhaps 95 percent dream and 5 percent product. The survey also found that quite a few had only one or two full-time employees and some had none. The majority attempted to go public with absolutely no earnings at 20 to 100 times their book value prior to the offering.10
In 1994, Professors Jay Ritter of the University of Illinois at Urbana-Champaign and Tim Loughran of the University of Iowa, two of the pioneers in the field of behavioral finance, completed the most comprehensive study of new issues made to date. The study followed the returns of 4,753 IPOs traded on the New York Stock Exchange, the AMEX, and NASDAQ between 1970 and 1990.11 The average return for IPOs was 5 percent annually, compared with 10.8 percent for the S&P 500. Put another way, investing in the S&P 500 over those twenty-one years would have returned 762 percent versus 179 percent for the IPOs. The far safer stocks of the S&P did more than four times as well!
But perhaps even more telling was that the median five-year return for these almost five thousand initial public offerings was a decline of 39 percent from the original offering price. If an investor couldn’t buy the handful of red-hot IPOs that doubled or even tripled on the first trade—and nobody but the largest money managers, hedge funds, mutual funds, and other major investors could—he would lose a good chunk of his original investment.
As in the case of the Value Line study, Loughran and Ritter’s results indicate that many IPOs were start-ups, usually high on expectations and low or nonexistent on actual revenues and earnings. They also found that most new issues go public near the top of an IPO market, when the demand is the greatest and the value of the merchandise is at its lowest. Not coincidentally, it is precisely at this time that investors are most excited about stocks with supposedly excellent visibility—further evidence of the strength of recency and saliency.
A 1991 study by Ritter12 showed that fully 61 percent of IPOs went public in 1983, the peak of the 1977–1983 mania. How many went public in the first five years, when the quality was at its best? Only 6 percent.
A 2011 working paper by Vladimira Ilieva, of the Dreman Foundation, and myself calculated the price of 1,547 IPOs ($2.00 or more in price) that went public between January 1, 1997, near the beginning of the dot-com bubble, and March 10, 2000, its high point. We then measured the drop from the market high to the low price of each issue through December 31, 2002. As Figure 3-2 shows, the average decline from the high to the low price in the dot-com bubble was a startling 97 percent.13 Finally, the median decline from the IPO offering price was 73 percent higher than the median decline that Ritter and Loughran calculated in this far more severe crash.
Also of interest was that the quality of the IPOs was consistently bad. Of the 1,547 companies in our original sample, only 524 were still listed in mid-2011. The other 66 percent had merged or just plain gone out of business.
Including the work of Ritter and Loughran, we now have an over-forty-year record of how dismal investing in IPOs has actually been. Recency and saliency appear to have played a not-insignificant hand in these results. Other studies support the findings. H. Nejat Seyhun reported that the market beat a sample of 2,298 IPOs for six years,14 and Mario Levis showed that a group of British IPOs underperformed the U.K. averages for three years.15 Further research found that IPO fundamentals fell after the offerings, indicating a deteriorating business picture precisely when investors were most excited by the stock.16
Loughran and Ritter (1995) concluded, “Our evidence is consistent with a market where firms take advantage of transitory windows of opportunity by issuing equity when, on average, they are substantially overvalued.”17 That’s a gentlemanly way of stating that the real mountains of gold were made not by the investors but by the concept weavers, aka the investment bankers, as they were in bubbles hundreds of years back. It seems the investment bankers have known about heuristical errors for centuries and made very good money exploiting them—too good to want to give their discovery to investors or science.
The strength of the subconscious effect of recent and salient events cannot be exaggerated, as the research above indicates. Whether the pain of the 1996–2000 dot-com bubble takes another few years to forget or not, there is little doubt that a new and powerful IPO bubble is out there with results as predictable as any witnessed in this section. Another Psychological Guideline should prove helpful here.
PSYCHOLOGICAL GUIDELINE 2(a): Don’t rely solely on the “case rate.” Take into account the “base rate,” the prior probabilities of profit or loss. Long-term returns of stocks (the “base rate”) are far more likely to be reestablished.
PSYCHOLOGICAL GUIDELINE 2(b): Don’t be seduced by recent rates of return (the “case rate”) for individual stocks or the market when they deviate sharply from past norms. If returns are particularly high or low, they are likely to be abnormal.
Who’s on First?
In the preceding chapter I spent a fair amount of time detailing the power of Affect in manias and crashes. In the current chapter I have so far detailed the errors that can result from indiscriminately following the availability heuristic lures of recency and saliency. Now, someone may ask, how can all three be present in the same bubble or any other investor error? This question brings us close to the state-of-the-art research cognitive psychologists are currently undertaking.
Affect, although not discovered until much later than cognitive heuristics, is considered by many senior researchers to be the driving force behind price movements. But to try to give a more precise answer about the contributions of recency, saliency, and Affect at this time would be like getting us into the famous (and endless) Abbott and Costello skit “Who’s on First?” But whatever the relative contributions, we receive strong warnings of imminent danger ahead that we can easily pick up.
Although the final research on the contribution of each of these heuristics may take many years to resolve, by now having a knowledge of each, we can build defenses so we can recognize bubbles and get out while there is
still time. Naturally, a similar approach could be adapted if stocks dropped to levels that were far too cheap.
The best defense against recency and saliency, in particular, is to keep your eye on the longer term. Though there is certainly no assured way to put recent or memorable experiences into absolute perspective, it might be helpful during periods of extreme pessimism or optimism to wander back to your library. If the market is tanking, reread the financial periodicals from the last major break. If you can, look up The Wall Street Journal of late February 2009, turn to the market section, and read the wailing and sighing by expert after expert, just before the market began one of its sharpest recoveries in history. Similarly, when we have another speculative market, it would not be a bad idea to check the Journal again and read the comments made during the 1996–2000 or 2005–2006 bubble. Though rereading the daily press is not a magic elixir, I think it can help.
I will also point out some defensive tips that should help. By themselves, these tips are not going to put your investing strategy on a firm footing. But you may want to add them to the strategies we are going to discuss later. Think of them as useful personal sidearms until we can bring in the heavy artillery, which will defend you from the heuristics we have discussed.
A Picture Is Not Worth a Thousand Words
The second important cognitive bias that Professors Kahneman and Tversky identified was the one they labeled “representativeness.” What they showed in rigorous experimental studies is that it’s a natural human tendency to draw analogies and see identical situations—where none exist.
In the market, representativeness might take the form of labeling two companies or two market environments as the same when the actual resemblance is superficial.18 Give people a little information, and click! they pull out a mental picture they’re familiar with, though it may only remotely represent the truth. The two key ways the representativeness bias leads to miscalculations are that it causes us to give too much emphasis to the similarities between events and does not take into account the actual probability that an event will occur, and that it reduces the importance we give to variables that are actually critical in determining an event’s probability.