Even worse are investment strategies that appear to start with a target value, usually a nice round marketable number, and then calculate the rate of growth in stock prices needed to reach the target. Arguments are then retrofitted to the numbers to explain why such a high rate of growth is plausible, or even likely. The stock market bubble of the dot-com era generated a bumper crop of this nonsense. In October 1999, with the Dow Jones Industrial Average at 11,497 after a long run-up, James K. Glassman and Kevin Hassett published Dow 36,000, which forecast that stock prices would more than triple within six years. Their optimism surpassed that of Dow 30,000 but was no match for Dow 40,000, let alone Dow 100,000. (All of these are real books, by different authors, and every one of them was selling for just one cent—plus shipping and handling, of course—on Amazon.com’s used-book marketplace as of April 2009.) The sheer number of these titles testifies to the large market for simple models that investors can easily assimilate and act on because they give a false sense of understanding. By the time the stock market began to recover from the dot-com bust, more titles appeared, including Dow 30,000 by 2008: Why It’s Different This Time.
Illusory Knowledge and a Real Crisis
With hindsight we can see that the implosion of Amaranth in 2006 was a harbinger of the much larger financial crisis that came to a head two years later. Venerable companies like Bear Stearns and Lehman Brothers went out of business, others like AIG were driven into government control, and the economy plunged into a deep recession. The world financial system is perhaps the ultimate complex system: It reflects decisions made by literally billions of people every day, and those decisions are all based on beliefs about how much, or how little, various investors know. Any time you buy an individual stock, you are acting on an implicit belief that the market has undervalued the stock. Your purchase represents a claim that you have better knowledge than most other investors about the future value of that stock.
Consider the biggest investment that most people make: their house.23 Most people view the decision about what house to buy as, at least in part, an investing decision. They wonder whether a house will have good “resale value” or whether it is in an “up-and-coming” or “declining” neighborhood. Some people make a business of buying, improving, and selling the houses they live in, a practice called “flipping” that was promoted heavily by television shows like Property Ladder and Flip That House in the mid-2000s. At that time, the number of people who thought houses were a good investment was rising dramatically.24 Even if you have never been a house flipper, you may still think of your house in part as a savings account, an asset you expect to appreciate in value over the medium-to-long term. Flipping is based on a model of the real estate market in which the prices of houses can also be counted on to increase in the short term, and the demand for them is always strong.
Acting on this model, people with no experience investing in real estate started buying houses on credit with the intent of selling them quickly at a profit. The speculative cycle was exacerbated, of course, by the willingness of banks to make loans that would probably never be repaid. Alberto Ramirez, a strawberry picker who lived in Watsonville, California, and earned about $15,000 a year, was able to buy a house for $720,000 without putting any money down; naturally he soon found that he couldn’t afford the payments. The apotheosis of subprime lending gimmicks was mortgage company HCL Finance’s “ninja” loan—no income, no job, no assets. Harvard economist Ed Glaeser, explaining why he did not foresee the bubble and ensuing crash in the housing market, said, “I underestimated the human capacity to think rosy thoughts about the value of a house.”25
Flawed models of the housing market extended well beyond individual homeowners and speculators, of course. Large banks and government-backed corporations purchased mortgages and resold them in groups to other investors as mortgage-backed securities, which were themselves packaged together into the infamous collateralized debt obligations (CDOs). The bond-rating agencies—Moody’s, Standard & Poor’s, and Fitch—used complex statistical models to evaluate the riskiness of these new securities. But behind these models lay simple assumptions that—when they no longer applied—undermined the entire edifice. As late as 2007, Moody’s was still using a model that had been built using data from the period before 2002—before the era of massive overbuilding, ninja loans, and strawberry pickers buying luxury homes. That is, despite the changes in the market, the model assumed that mortgage borrowers of 2007 would default at about the same rate as the mortgage borrowers of 2002. When the housing bubble burst, a general recession ensued, and the rate of mortgage defaults diverged from historical norms. As a result, many CDOs turned out to be riskier than the models had predicted, and firms that had invested in them lost a lot of money.
It can be difficult to determine how well our simple models correspond to the realities of complex systems, but it is easy to determine three things: (1) how well we understand our simple models; (2) how familiar we are with the surface elements, concepts, and vocabulary of the complex system; and (3) how much information we are aware of, and can easily access, about the complex system. We then take our knowledge of these particular things as signals that we understand the system as a whole—an utterly unwarranted inference that can quickly land us in hot water. Analysts understood their models, they were familiar with the vocabulary of subprime mortgages, CDOs, and the like, and they were swimming in a river of financial data and news, giving them the illusion that they understood the housing market itself—an illusion that persisted until the market collapsed.26 With more and more financial information available at higher speed and lower cost (think CNBC, Yahoo! Finance, and online discount stockbrokers), the conditions for this illusion have spread from professional market participants to ordinary individual investors.
In a brilliant article for Condé Nast Portfolio, journalist Michael Lewis tells the story of a hedge fund manager named Steve Eisman who was one of the few to see through the smoke and mirrors of the housing boom and the CDO markets. Eisman looked into some complicated mortgage securities and had trouble understanding their terms, despite his many years of experience as a trader. Dan Gertner, a writer for Grant’s Interest Rate Observer, had a similar experience; he actually read through the several hundred pages that constituted the complete documentation for a CDO—something none of its investors probably ever did—and after days of study still couldn’t figure out how it really worked.
The central issue for any complex investment is how to properly determine its value. In this case, the value was obscured by layer upon layer of untestable assumptions, but buyers and sellers deceived themselves into thinking they understood both the value and the risk. Eisman would go to meetings and ask CDO salespeople to explain their products to him, and when they spouted some gobbledygook, he would ask them to explain what exactly they meant. Essentially, he played Leon Rozenblit’s “why boy,” gradually exposing whether the CDO vendors really knew their own products. “You figure out if they even know what they’re talking about,” said one of Eisman’s partners. “And a lot of times, they don’t!” He might just as well have asked them to explain how their toilets worked.
You don’t have to be a seller of newfangled securities to let the surface familiarity of financial terms and concepts blind you into thinking you know more about the markets than you really do. For a few years, Chris made a specialty of investing in small biotechnology and pharmaceutical companies that focused on developing treatments for brain diseases. A couple of his stocks did well for a time, increasing by over 500 percent in one case. He started to believe that he actually had some talent for picking stocks in this sector, and easily came up with reasons why: He knew a lot of neuroscience and some genetics, and he was competent at designing experiments and analyzing data, which is the core discipline behind the clinical trials that are used to decide whether drugs can jump over all the regulatory hurdles to reach patients. But the sample of his stock-picking experience was orders of magnitude too small
to demonstrate any real skill—luck was the most likely explanation for his success. That interpretation seems to have been confirmed: Most of his picks lost three-quarters or more of their value in the end.
If you can’t escape the illusion entirely and still think of yourself as a knowledgeable stock picker, you might try to limit how much the illusion can affect you by allocating just a small proportion of your assets to active investment decisions, and thinking of those investments at least partly as a hobby. The rest of your money could be dedicated to strategies that are less subject to the illusion of knowledge, such as passively investing in index funds that just track the movements of the overall market. That’s also a reasonable plan for a gambler who wants to keep his or her hobby under control: Set aside a small bankroll and focus on the entertainment that comes from the practice rather than counting on it to generate significant income. Chris has abandoned stock picking entirely, and he keeps his poker money in a separate bank account.
Sometimes More Is Less
Imagine that you are a subject in the following experiment, conducted by pioneering behavioral economist Richard Thaler and his colleagues.27 You are told that you are in charge of managing the endowment portfolio of a small college and investing it in a simulated financial market. The market consists entirely of just two mutual funds, A and B, and you start with a hundred shares that you must allocate between the two. You can put all of your shares into A, all of them into B, or some into A and the rest into B. You will be running the portfolio for twenty-five simulated years. Every so often, you will be informed of how each fund has performed, and thus whether your shares have gone up or down in value, and you will then have the opportunity to change how your shares are allocated. At the end of the simulation, you will be paid an amount that is proportional to how well your shares have performed, so you have an incentive to do as well as you can. Before the game begins, however, you have to choose how often you would like to receive the feedback and have the chance to change your allocations: every month, every year, or every five years (of simulated time).
The correct answer seems obvious: Give us information, and let us use that information, as often as possible! Thaler’s group tested whether this intuitive answer is right—not by giving people the choice, but by randomly assigning them to receive feedback monthly, yearly, or every five years. Most people initially tried a 50/50 allocation between the two funds since they knew nothing about which might be better. As they got information about the performance of the funds, they shifted their allocations. Since the simulated length of the experiment was twenty-five years, the subjects in the five-year condition got feedback and could change their allocations only a few times, compared with hundreds of times for the subjects in the monthly condition. By the end of the experiment, subjects who only got performance information once every five years earned more than twice as much as those who got monthly feedback.
How could having sixty times as many pieces of information and opportunities to adjust their portfolios have caused the monthly-feedback investors to do worse than the five-year ones? The answer lies partly in the nature of the two funds the investors had to choose from. The first had a low average rate of return but was fairly safe—it didn’t vary much from month to month and rarely lost money. It was designed to simulate a mutual fund consisting of bonds. The second was like a stock mutual fund: It had a much higher rate of return, but also a much higher variance, so that it lost money in about 40 percent of the months.
In the long run, the best returns resulted from investing all of the money in the stock fund, since the higher return made up for the losses. Over a one-year or five-year period, the occasional monthly losses in the stock fund were canceled out by gains, so the stock fund rarely had a losing year and never had a losing five-year stretch. In the monthly condition, when subjects saw losses in the stock fund, they tended to shift their money to the safer bond fund, thereby hurting their long-term performance. Subjects who received feedback every year or every five years saw that the stock fund outperformed the bond fund, but they did not see the difference in variability. At the end of the experiment, the subjects in the five-year condition had 66 percent of their money in the stock fund, compared with only 40 percent for the subjects in the monthly condition.
What went wrong for the subjects who got monthly feedback? They got a lot of information, but it was short-term information that was not representative of the true, long-term pattern of performance for the two funds. The short-term information created an illusion of knowledge—knowledge that the stock fund was too risky, in this case. The monthly-feedback subjects had all the information they needed to generate actual knowledge—that the stock fund was the better long-term investment—but they didn’t manage to do so.
The same thing happens in the real world of investing decisions. Brad Barber and Terrance Odean managed to obtain six years of trading records for sixty thousand accounts from a brokerage firm and compared investment returns between people who bought and sold stocks frequently and those who traded rarely. Presumably investors who make lots of trades believe that they have lots of knowledge and good ideas about stocks—that each of their trades will make money because it is anticipating a market move. But once their returns were adjusted for the costs and tax payments generated by all the trades they made, the most active traders earned one-third less per year than the least active ones.28
Professional and amateur investors alike should seek the best rates of return they can get, balanced against the level of risk they are taking. Individual investors in particular may be better off paying more attention to the riskiness of their portfolios than they currently do. Earning an extra few percentage points on your money may not be worth the anxiety, lost sleep, and bad temper that can accompany the volatility of large price swings. To make truly informed financial decisions, you must have an accurate picture of the long-term returns and short-term volatility you should expect from each of your investment options, and you must evaluate these factors in light of your own ability to tolerate risk.
We are generally taught that it is better to have more information than to have less. Who wouldn’t want to consult Consumer Reports before purchasing a car or a dishwasher? Who wouldn’t want to know the price of a flat-screen TV at three different stores rather than just one? And in these cases, more information does make for better decisions (at least up to a point). The studies we just presented, and others like them, suggest that investors who have more information also believe that they have better knowledge. But when that information is in fact uninformative, it only feeds the illusion of knowledge. In reality, most short-term fluctuations in value are unrelated to longer-term rates of return and should not determine your investment decisions (unless you are investing money that you might need in the near future, of course). When it comes to assessing the long-term characteristics of an investment, sometimes having more information can result in less real understanding. What the Thaler group’s experiment showed was that paradoxically, people who got the most feedback about the short-term risks were least likely to acquire knowledge of the long-term returns.
The illusion of knowledge can’t predict the timing and magnitude of each financial bubble—in fact, knowing about the illusion should make us just as wary of attempts to predict price drops as price increases. The illusion of knowledge does appear to be a necessary ingredient for the formation of bubbles, though. Each historical bubble has been associated with a piece of new “knowledge” that was disseminated so widely that it eventually reached people who knew nothing else about finance except for that one piece of information (tulip bulbs are a can’t-lose investment, the Internet will fundamentally change what companies are worth, the Dow is going to 36,000, real estate never loses value, and so on). The proliferation of information about finance, from cable news networks to websites to business magazines, is a recipe for the illusory feeling that we know how the markets work, when all we really have is a lot of information about what they a
re doing at the moment, what they have done in the past, and how people think they work, none of which necessarily predicts what they will do in the future. Familiarity with the language of finance and the immediacy of market changes often masks a lack of deep knowledge, and the increasingly rapid flow of information may even shorten the cycle of booms and busts in the future.
The Power of Familiarity
Much as we cannot focus attention on more than a limited subset of our world and we cannot remember everything around us, the illusion of knowledge is a by-product of an otherwise effective and useful mental process. We rarely need to explain why something works. Rather, we just need to understand how to work it. We need to understand how to unclog a toilet, but we don’t need to know how flushing the toilet causes water to empty from the bowl and then to fill it back up. Our ability to operate a toilet when we need to—and to do so without even thinking about the process—gives us a sense that we understand it. And for most practical purposes, that is all the understanding we really need.
In Chapter 2, we discussed the error of “change blindness blindness”—the idea that people think they will notice changes that, in reality, they rarely do. People easily confuse what they actually remember with what they potentially could remember if given the chance to study things further. Stop reading now and draw a picture of the face of a penny, or form an image of one in your mind. Odds are that your image has at least a couple of errors—you might have Lincoln facing the wrong direction, or you might have put the date in the wrong place, or you might have forgotten to include the date altogether. You have seen pennies every day for years, and before now you probably thought you knew what a penny looked like. You do know enough to tell a penny apart from other coins, which is the only knowledge you really need.29
The Invisible Gorilla: And Other Ways Our Intuitions Deceive Us Page 16