Contrarian Investment Strategies Read online

  The finding is important in gauging the overreaction to either a positive or a negative earnings surprise as well as a host of other market positive or negative events. To take one example, oil exploration and development companies plummeted sharply in the spring of 2010, after the BP rig blew up and a major oil spill ensued in the Gulf of Mexico. The prevailing feeling at the time was that no deep drilling would be allowed again in the gulf or other continental waters around the United States for years. Too, the cleanup costs were at the time thought to be prohibitive for both BP and its partners. Within a year, deepwater drilling was allowed again off U.S. shores, and the costs, though high, were manageable by all the firms involved. Some of the stocks originally believed to be the most exposed more than doubled in less than a year.

  This finding, as we shall see in Parts III and IV, would seem to be helpful in explaining both the superior performance of “worst” over “best” stocks and the consistent but opposite reaction to surprise events by these two categories.

  4. Temporal Construal

  Affect leads us to make misjudgments related to time. Events nearer in time are likely to be represented in terms of more concrete and specific detail. Short-term events will include sales and earnings that are higher or lower than expected, and many other expectations that analysts report.16 Meanwhile, the farther an event occurs in the future, the more probable returns are likely to be represented in terms of a few abstract or general features that contain the perceived essence of the concept of the stock under consideration. This phenomenon is called “Temporal Construal.” This Affect characteristic causes investors to extend their views of the prospects of stocks both in and out of favor far into the future. If the future outlook is negative, the results will be reversed.

  Affect results in a smoothing effect on the judgment of longer-term prospects. Rather than focusing on dozens of shorter-term inputs—most positive, but some negative—focusing on the more general thoughts about exciting longer-term prospects tends to result in more favorable forecasts than those made for the short term.

  As a result of a strong positive or negative Affect, a stock, an industry, or the market itself can be priced too high or too low. Professors Trope and Liberman reported that great optimism toward long-term returns might be explained by this Affect characteristic.

  Temporal Construal helps explain the consistent outperformance of contrarian investment strategies over time, as well as the major overpricing of technological stocks in bubbles. In the first case (as we shall see in detail in Part IV), investors expect the worst from out-of-favor contrarian stocks and discount their prospects far into the future. In the second case, investors generally extend expectations for concept companies’ positive results as they expand their markets rapidly too far into the future. We’ve now seen a number of instances of just how overvalued such concept stocks can get as investors overestimated the length of time a positive Affect would last for favored stocks such as the dot-com issues in the tech bubbles. Professor Trope and his colleagues have produced a good deal of experimental evidence that documents the difference in the way long-term and short-term estimates are constructed.

  A glaring case in point is the expectations before the dot-com bust about Yahoo!’s growth. Yahoo! went public in April 1996 and was the hottest of the hot IPOs through the dot-com bubble.

  The stock appreciated 18,000 percent from late July 1996 to its high of $119 (adjusted for stock splits) in January 2000. Investors considered it the Web site extraordinaire, with the potential for almost unlimited growth of advertising, strong revenue streams, and enormous user growth. Yahoo!’s potential growth was considered unequaled.

  The company’s sales growth in the 1996–2000 period was enormous, rising from $23.8 million to over $1.1 billion (116 percent compounded annual growth over the five-year period). However, earnings did not do quite as well. The company lost money in 1996–1998 before making pennies in earnings per share in 1999 and 2000. It made only a couple of pennies a share for the entire 1996–2000 period on a stock that at its high was, as noted, close to $120. But concept ruled supreme.

  In early 2000, Yahoo! had a P/E of 5,938, and most investors expected huge earnings growth in the next ten years—well over 50 percent a year. As good a company as it was, these long-term expectations were wildly overestimated. In the dot-com crash from 2000 to 2002, both a severe reappraisal of its future growth rate and the fact that Yahoo!’s earnings had grown at a far lower rate than the market expected resulted in the stock’s dropping 97 percent to its low in early fall of 2001, as the dot-com crash approached its nadir.

  Another interesting instance of Temporal Construal can be observed in studies that asked investors to estimate the future performance of their stock portfolios during the 1996–2000 bubble. In 1998, near the height of the bubble, U.S. investors were extremely optimistic about the future returns of stocks. Late that year Paul Slovic, Stephen Johnson, Donald MacGregor, and I were coauthors of a study to pinpoint the expectations of a large number of mutual fund investors for the next ten years. According to our study,17 when subjects were asked to forecast their average return over the next ten years, investors were highly optimistic, estimating 14 percent annually on average. Stocks have in fact returned about 10 percent over time since the 1920s, so those enthusiastic investors appeared to be anticipating market returns in the decade ahead some 40 percent above those of the past seventy years.

  The Implications of Affect on Security Analysis

  In this chapter, we have seen how powerful various forms of Affect were in both fueling the bubbles and manias we observed in chapter 1 and then, as circumstances changed, sharply escalating the terror and panic when each bubble imploded. The Affect heuristic is both wondrous and frightening: astounding in its speed, subtlety, and sophistication; frightening in its power to lead us astray. According to Paul Slovic, “It is sobering to contemplate how elusive meaning is, due to its dependence upon Affect.”18

  We should pause for a second to consider the forms of meaning which we take for granted and in which we justify spending immense time and expense to gather information. Will the use of “meaningful information,” such as thorough security or market analysis, be in many cases illusionary, since as we’ll see, these cases are all too often laced with Affect?

  Obviously, it is not going to be easy to protect ourselves from such dangers. That is why the book will introduce disciplines you can follow that will, like a stun gun, stop Affect when it’s working against your interests. The real question is whether you can pull the trigger when the time comes. It’s harder than it seems. Still, being aware of Affect is a good first step to a better investment strategy.

  The Disconnect Between Fundamentals and Price

  One of the most notable characteristics of an Affect-induced bubble, as we have seen, is the enormous disconnect between fundamentals and price for analysts, money managers, and other highly trained financial professionals. The standard guidelines used to evaluate a company’s outlook are derived from the rational-analytic system.

  The CFA Institute and virtually all of the academic training materials teach analysts and money managers to behave in a rational manner and thus to look at all the important fundamentals of a company to determine its price. A bible for this training is Benjamin Graham and David Dodd’s Security Analysis.19 A myriad of other books adhere to very similar methods. As we saw earlier, Graham and Dodd use a large number of financial ratios as well as detailed company information to decide on the value of a stock. The standard guidelines used to evaluate a company’s outlook sketched out earlier are most often bypassed or short-circuited.

  Unfortunately, most professionals—despite their training, the state-of-the-art information available to them, and the best research available—often do no better and sometimes do worse than the average investor. This gives us a good idea of just how strong the forces of Affect actually are.

  Insensitivity to probability appears to be the most important Aff
ect characteristic that leads financial professionals to make exactly the same mistakes as the average investor by focusing on a grossly exaggerated view of the values for companies that appear at the time to have almost unlimited growth and profitability. As we’ve also seen, the finding that judgments of risk and benefit are negatively correlated, the Durability Bias, and Temporal Construal all can knock our portfolios down mercilessly. Affect can turn the playing field upside down. Unfortunately, Affect can take down the professional investor despite his extensive knowledge and training as quickly as it does the individual investor.

  Almost no price seemed too high for companies that have already had explosive price appreciation as well as an exciting, if not almost irresistible, story to tell. Strong positive Affect attaches to the image that the dot-com or IPOs issues create instant wealth, and quite likely overwhelms the rational-analytic system, resulting in a major swing away from optimum decision making.

  The Consequences for Security Analysis

  We should ask if analysts or money managers use any rational evaluation models at all to justify extreme overpricing. While almost all claim they do, the enormous disparity between the maximum price a fundamental model would allow and the substantially higher market prices makes it clear that, as Table 2-1 demonstrated, the models could not in fact be based on accepted valuation techniques.

  Looking at the sky-high prices so many analysts paid in recent years, Graham would have shaken his head, although I’ve heard from some of his still-living contemporaries that the response would have been a little more lively. Graham was ultraconservative on price-earnings and other valuation ratios, normally arguing not to buy a stock with a P/E much above 20, regardless of its outlook, as he believed “Mr. Market,” the name he coined for the market, could go to major extremes quickly.

  Stocks with average prospects will normally have P/Es of 12 to 15; those with well above average earnings growth prospects will be in a 25-to-35 P/E range; and those with extraordinary growth will be at possibly forty to fifty times earnings. How, then, can we justify the fact that many stocks, recommended by scores of analysts, trade at 100 to 1,000 or more times earnings?

  Can Affect Fatally Flaw Security Analysis?

  Unfortunately, what comes out clearly is that investment professionals cannot accurately gauge prices in the range of approximately plus or minus 25 to 35 above or below fundamental value that financial theory holds a good analyst should, thereby allowing her or him to profit through buying undervalued or selling overvalued securities respectively. As we just saw, analysts often missed the proper valuations of Internet and high-tech stocks by fifty- to 100-fold as evidenced repeatedly in the dot-com bubble and other tech bubbles since the early 1960s. This created, at least temporarily, valuations sharply higher than any considered acceptable that were based on standard security analysis.*6

  What we see, then, is that positive Affect, in extreme circumstances such as bubbles overpowers analysts following contemporary security analysis and raises prices significantly above what long-term valuation standards will sanction. In a panic, negative Affect results in the opposite case, reserved for investors who want to sell at any price. Unfortunately, the very people we depend on most to guide us to safety when a market storm appears are the first to be swept away when the terrible weather hits.

  The Great Escape

  The standard guideline used to evaluate a company’s outlook is based on the rational-analytic system. According to the CFA Institute and virtually all sophisticated academic training in the field, analysts and money managers behave in a rational manner and so look at all the important fundamentals of a company to determine its price, as we discussed earlier. The bible for this training, as we saw, is normally Graham and Dodd’s Security Analysis, or a similar text, as well as hundreds of pages of CFA instruction, written by highly experienced investment professionals.

  Worth repeating is that security analysis can be effective in a range of approximately plus or minus 25 to 35 above or below fundamental value. So if a company is undervalued at, say, 10 times earnings, an astute analyst can buy it and perhaps double his money. However, fundamental analysis cannot work when prices move into outer space. In a bubble, a price-earnings ratio or price-to-cash-flow ratio can go from 10 to 100, or even sometimes to 1,000 or more. A price-to-book-value ratio of 4 can go to 40 or even 400 and so forth. At these levels fundamental security analysis breaks down entirely.

  In a bubble or another period of extreme overvaluation, the focus is on grossly exaggerated prospects for a company or an industry that analysts or money managers forecast. Importantly in this situation, standard fundamental analytic tools, such as earnings discount models, which project highly optimistic earnings estimates several dozen years into the future, repeatedly show that the current price of the bubble stock is far too high, as we saw in Table 2-1.

  What analysts, along with most people, don’t know is that most of the investment rational-analytic standards are often bypassed or short-circuited by Affect as well as by other psychology we will view. Analysts and money managers have enormous difficulty in attempting to work within the rational-analytic system when the experiential system is under full sail. Most professionals do try to get it right, but the force of powerful positive or negative Affect is difficult to cope with, as we now know. I’ve spoken at half a dozen of the CFA Institute’s National Conferences on investor psychology over more than three decades, as well as to many large societies of security analysts and to leading business schools, and I know professionals make a genuine effort to follow their fundamental methods.

  Unfortunately, psychology, although widely acknowledged by these groups, is impossible to square with current investment theory, because the latter sits on a foundation solidly built on the rational-analytic approach. When this happens, analysts and money managers abandon this approach all too frequently and shift over to the experiential system, where Affect, in extreme circumstances, as we saw, easily overwhelms our rational-analytic processes.

  The bottom line is that no analyst could recommend and no money manager could buy bubble stocks using rational-analytic methods. Analysts and money managers didn’t need a financial GPS to calculate how far off course the stocks’ pricing was. With this disparity in price, a sextant would have worked just as fine.

  The analysts and money managers were hijacked by their expectations, which were heavily weighted by their strong positive Affect for these issues. We should see clearly now that professionals cannot accurately gauge prices in the relatively limited range*7 that financial theory holds a good analyst should, thereby allowing him to profit through buying undervalued or selling overvalued securities respectively.

  Analysts missed the proper valuations, as we saw, by 50 to 100 times on many occasions in bubbles. They created valuation paradigms that justified price-earnings almost exponentially higher than any considered acceptable, employing standard analytical techniques. Does security analysis work using today’s rational-analytic methods? Possibly for brief periods of time when the water is tranquil. But don’t go out too far: remember, you’re paddling a twelve-foot kayak; you’re not in a 700-foot ocean liner.

  I will be introducing a number of Psychological Guidelines that, if followed, should limit your losses in the psychological traps we will be examining through the course of the book. If they are adhered to, they will protect you from making some of the mistakes most investors are entrapped by. We should also find ways to use these predictable errors to our advantage. Here is the first one.

  PSYCHOLOGICAL GUIDELINE 1: Do not abandon the prices projected by careful security analysis, even if they are temporarily far removed from current market prices. Over time the market prices will regress to levels similar to those originally projected.

  Next, let’s extend our look into other major heuristics or mental shortcuts that we use daily to simplify our lives. Each, although extremely helpful, can also open the door to bias in our investment decisions, resulting in
other serious errors that can also be very damaging. Knowing what they are and how they work can help you develop some strong defenses against their excesses.

  Chapter 3

  Treacherous Shortcuts in Decision Making

  WHICH IS THE more likely cause of death, being attacked and killed by a shark or getting struck by falling airplane parts?1 Most people don’t have any experience of either, fortunately, but when asked this question they more often than not decide that the shark is the more likely culprit. That’s wrong, however. In the United States, death from falling airplane parts is actually thirty times as probable as dying from a shark attack.2

  What’s going on here? Shark attacks, no matter how rare, receive considerable media attention and are horrifically easy to imagine, especially if you’ve seen the classic movie Jaws.3 But can you recall a media report about falling airplane parts? If any graphic images come to mind, it’s likely to be a plane in a cartoon humorously coming apart à la Bugs Bunny. We just don’t think about small pieces of airplanes coming off. The images we generally have for airplane failure are of a catastrophic crash that kills passengers rather than anyone on the ground. Lacking firsthand experience of seeing airplane parts falling out of the sky and never having read about the problem, we mentally choose the more available image of shark mayhem as the more likely event.

  This is an example of the availability heuristic, which leads us to subconsciously give more weight to images that come readily to mind than those that are “fuzzier” to recall.4 Next, let’s look more closely at cognitive heuristics.