Misbehaving: The Making of Behavioral Economics

by Richard H. Thaler

  The way the spin-off would work is that initially only 5% of the value of Palm would be sold to outside investors. 3Com would retain the rest of the shares. Then, after a period of a few months, each 3Com shareholder would receive 1.5 shares of Palm. Here is where the law of one price comes into play. As soon as the initial shares of Palm were sold and started trading, 3Com shareholders would essentially have two separate investments. A single share of 3Com included 1.5 shares of Palm plus an interest in the remaining parts of 3Com, or what in the finance literature is called the “stub value” of 3Com. In a rational world, the price of a 3Com share would be equal to the value of the stub plus 1.5 times the price of Palm.

  Investment bankers marketing the shares of Palm to be sold in the initial public offering had to determine what price to charge. As excitement about the IPO kept building, they kept raising the price, finally settling on $38 a share, but when Palm shares started trading, the price jumped and ended the day at a bit over $95. Wow! Investors did seem to be wildly enthusiastic about the prospect of an independent Palm company.

  So what should happen to the price of 3Com? Let’s do the math. Each 3Com share now included 1.5 shares of Palm, and if you multiply $95 by 1.5 you get about $143. Moreover, the remaining parts of 3Com were a profitable business, so you have to figure that the price of 3Com shares would jump to at least $143, and probably quite a bit more. But in fact, that day the price of 3Com fell, closing at $82. That means that the market was valuing the stub value of 3Com at minus $61 per share, which adds up to minus $23 billion! You read that correctly. The stock market was saying that the remaining 3Com business, a profitable business, was worth minus $23 billion.

  FIGURE 16

  There is a first principle of finance even more fundamental than the law of one price, which is that a stock price can never be negative. You can throw your shares away if you want to, and shareholders have limited liability, so the absolute lowest a stock price can fall to is zero. No company can be worth minus $100, much less minus $23 billion. But that is what the market was saying.

  Think of it another way. Suppose an Econ is interested in investing in Palm. He could pay $95 and get one share of Palm, or he could pay $82 and get one share of 3Com that includes 1.5 shares of Palm plus an interest in 3Com. That does not seem to be a tough decision! Why buy Palm directly when you can get more shares for less money by buying 3Com, plus get a stake in another company thrown in for free?

  This was a colossal violation of the law of one price. In fact it was so colossal that it was widely publicized in the popular press. Nevertheless, the value of the 3Com stub remained negative for several months.

  How could this happen? Two ingredients are necessary for a violation of the law of one price to emerge and persist. The first is that you need some investors with an inexplicable desire to own a pure, unadulterated version of Palm rather than one watered down with extra money and a share of a profitable company. In other words, you need noise traders, a.k.a. Summers’s IDIOTS. And note that even if some people buy Palm knowing it is overvalued but hoping to sell it to idiots later at an overvalued price—well, you still need some idiots to make it all work.

  The other thing that is necessary for this to happen is that something must be preventing the “smart money” from driving the prices back to where they are supposed to be. The merely “sensible” investor would simply buy 3Com instead of Palm. But a true Econ would go one step further. The smart-money trade in this situation is to buy undervalued 3Com shares and sell short an appropriate number of shares of Palm. Then, when the deal is completed, the investor sells the shares of Palm he receives, uses those shares to repay his loan, and is left with a profit equal to whatever price 3Com is selling for as a stand-alone company. This is a trade that cannot lose. Why wasn’t everyone trying to do it, given that it was so widely known? The problem was that so few shares of Palm were sold in the initial public offering that there were not enough to satisfy everyone who wanted to borrow them: the supply of shares to lend exceeded the demand from people who wanted to borrow them to sell them short. This meant that the smart money was unable to drive the relative prices of Palm and 3Com into a rational equilibrium where the price of 3Com was at least 1.5 times the price of Palm.*

  The Palm/3Com story is not unique.† Back in 1923, a young Benjamin Graham noticed that DuPont owned a large number of shares of General Motors and strangely, the market value of DuPont was about the same as its stake in GM. In spite of the fact that DuPont was a highly profitable firm, its stub value was close to zero. Graham made the smart trade, buying DuPont and selling GM short, and made a bundle when the price of DuPont went up.

  But things don’t always work out so well for the smart investors. For many years, there were two kinds of shares of the merged company Royal Dutch Shell. Royal Dutch shares traded in New York and the Netherlands, and Shell shares traded in London. According to the terms of the merger agreement that created this company in 1907, 60% of the profits would go to Royal Dutch shareholders and 40% would go to Shell shareholders. The law of one price stipulates that the ratio of the prices of the two classes of shares should be 60/40 or 1.5. But did the two share prices always trade at that ratio? No! Sometimes the Royal Dutch shares traded as much as 30% too low, and other times they traded as much as 15% too high. Noise traders appear to have particular difficulty with multiplying by 1.5.

  In this case, the smart trade is to buy whichever is the cheaper version of the stock and sell the expensive version short. Unlike the case of Palm and 3Com, both versions of the stock were widely traded and easy to borrow, so what prevented the smart money from assuring that the shares traded at their appropriate ratio of 1.5? Strangely, nothing! And crucially, unlike the Palm example, which was sure to end in a few months, the Royal Dutch Shell price disparity could and did last for decades.‡ Therein lies the risk. Some smart traders, such as the hedge fund Long Term Capital Management (LTCM), did execute the smart trade, selling the expensive Royal Dutch shares short and buying the cheap Shell shares. But the story does not have a happy ending. In August 1998, because of a financial crisis in Asia and a default on Russian bonds, LTCM and other hedge funds started to lose money and needed to reduce some of their positions, including their Royal Dutch Shell trade. But, not surprisingly, LTCM was not the only hedge fund to have spotted the Royal Dutch Shell pricing anomaly, and the other hedge funds had also lost money in Russia and Asia. So at the same time that LTCM wanted to unwind its position in Royal Dutch Shell, so did other hedge funds, and the spread moved against them, meaning that the expensive version got more expensive. Within weeks, LTCM had collapsed from this and other “arbitrage” opportunities that got worse before they got better.

  The LTCM example illustrates what Andrei Shleifer and his frequent coauthor Robert Vishny call the “limits of arbitrage.” In fact, in a paper they published on this topic in 1997, a year before these events occurred, they quite cannily described a hypothetical situation much like what LTCM experienced. When prices start to move against a money manager and investors start to ask for some of their money back, prices will be driven further against them, which can cause a vicious spiral. The key lesson is that prices can get out of whack, and smart money cannot always set things right.

  Owen and I wrote an academic paper about the Palm–3Com episode boldly titled “Can the Market Add and Subtract?” and presented it at the finance workshop at the University of Chicago. At the end of the workshop Gene Fama questioned the significance of examples such as this one and closed-end funds. He pointed out that these are relatively minor financial assets. So, although the results were in conflict with the EMH, he argued that the stakes were too small to worry about.

  My view was that these special cases are finance’s equivalent of geneticists’ fruit flies. Fruit flies are not a particularly important species in the grand scheme of things, but their ability to quickly reproduce offers scientists the chance to study otherwise difficult questions. So it is with finance’s f
ruit flies. These are the rare situations in which we can say something about intrinsic value. No one can say what the price of 3Com or Palm should be, but we can say with great certainty that after the spin-off, the price of 3Com had to be at least 1.5 times the price of Palm. I suggested that examples like this one were the tip of the iceberg of market mispricing. Gene’s view was that we had beheld the entire iceberg.

  What are the implications of these examples? If the law of one price can be violated in such transparently obvious cases such as these, then it is abundantly clear that even greater disparities can occur at the level of the overall market. Recall the debate about whether there was a bubble going on in Internet stocks in the late 1990s. There was no way to prove at the time, or even now, that the pricing of technology stocks was too high. But if the market could not get something as simple as Palm and 3Com right, it certainly seemed possible that the technology-heavy NASDAQ index could be overpriced as well. It does not seem to be a coincidence that the expensive part of the Palm/3Com trade was the sexy Palm division and the inexpensive part was the sleepier 3Com parent. The same could be said when contrasting the rise in prices of sexy tech stocks versus sleepy industrials.

  So where do I come down on the efficient market hypothesis? It should be stressed that as a normative benchmark of how the world should be, the EMH has been extraordinarily useful. In a world of Econs, I believe that the EMH would be true. And it would not have been possible to do research in behavioral finance without the rational model as a starting point. Without the rational framework, there are no anomalies from which we can detect misbehavior. Furthermore, there is not as yet a benchmark behavioral theory of asset prices that could be used as a theoretical underpinning of empirical research. We need some starting point to organize our thoughts on any topic, and the EMH remains the best one we have.

  When it comes to the EMH as a descriptive model of asset markets, my report card is mixed. Of the two components, using the scale sometimes used to judge the claims made by political candidates, I would judge the no-free-lunch component to be “mostly true.” There are definitely anomalies: sometimes the market overreacts, and sometimes it underreacts. But it remains the case that most active money managers fail to beat the market. And as the story about Royal Dutch Shell and LTCM shows, even when investors can know for sure that prices are wrong, these prices can still stay wrong, or even get more wrong. This should rightly scare investors who think they are smart and want to exploit apparent mispricing. It is possible to make money, but it is not easy.§ Certainly, investors who accept the EMH gospel and invest in low-cost index funds cannot be faulted for that choice.

  I have a much lower opinion about the price-is-right component of the EMH, and for many important questions, this is the more important component. How wrong do I think it is? Notably, in Fischer Black’s essay on noise, he opines that “we might define an efficient market as one in which price is within a factor of 2 of value, i.e., the price is more than half of value and less than twice value. The factor of 2 is arbitrary, of course. Intuitively, though, it seems reasonable to me, in the light of sources of uncertainty about value and the strength of the forces tending to cause price to return to value. By this definition, I think almost all markets are efficient almost all of the time. ‘Almost all’ means at least 90%.”

  I am not sure whether “90% of the time” is a satisfactory definition of “almost all” of the time, but more importantly, a factor of 2 strikes me as a very wide margin to call a market efficient. Just think about all the housing units built during the real estate bubble that are still worth only half of what their value was at the peak. The people who bought those homes would probably disagree with an assessment that the housing market was acting efficiently during the boom. Furthermore, Black died in 1996, before the technology and real estate bubbles. I think if he were still around he could be convinced to revise that assessment to “right within a factor of 3.” The NASDAQ index lost more than two-thirds of its value from its peak in 2000 to its trough in 2002, and the decline is almost certainly due to initial exuberance. (It can certainly not be blamed on the Internet proving to be a disappointment.)

  My conclusion: the price is often wrong, and sometimes very wrong. Furthermore, when prices diverge from fundamental value by such wide margins, the misallocation of resources can be quite big. For example, in the United States, where home prices were rising at a national level, some regions experienced especially rapid price increases and historically high price-to-rental ratios. Had both homeowners and lenders been Econs, they would have noticed these warning signals and realized that a fall in home prices was becoming increasingly likely. Instead, surveys by Shiller showed that these were the regions in which expectations about the future appreciation of home prices were the most optimistic. Instead of expecting mean reversion, people were acting as if what goes up must go up even more.

  Moreover, rational lenders would have made the requirements for getting a mortgage stricter under such circumstances, but just the opposite happened. Mortgages were offered with little or no down payment required, and scant attention was paid to the creditworthiness of the borrowers. These “liar loans” fueled the booms, and policy-makers took no action to intervene.

  This lesson is one of the most important to take away from the research about market efficiency. If policy-makers simply take it as a matter of faith that prices are always right, they will never see any need to take preventive action. But once we grant that bubbles are possible, and the private sector appears to be feeding the frenzy, it can make sense for policy-makers to lean against the wind in some way.

  Central banks around the world have had to take extraordinary measures to help economies recover from the financial crisis. The same people who complain most about these extraordinary recovery measures are also those who would object to relatively minor steps to reduce the likelihood of another catastrophe. That is simply irrational.

  ________________

  * It was possible to find some shares to borrow if you had time on your hands. In fact, at the time, there was a finance PhD student at the University of Chicago who was determined to make money on 3Com/Palm. He opened accounts at every discount brokerage house and spent all of his time trying to borrow shares of Palm to sell short. Whenever he got Palm shares he would sell them short and use the proceeds to buy the number of shares of 3Com required to hedge his bet. When the deal was finalized a few months later, he made a tidy profit and bought a sports car that he nicknamed the Palm-mobile. The moral of this story is that it was possible to make tens of thousands of dollars from this anomaly, but not tens of millions.

  † A similar situation arose in mid-2014 when Yahoo’s holdings of Alibaba were calculated to be worth more than the whole of Yahoo (Jackson, 2014; Carlson, 2014).

  ‡ When I described this anomaly to the CEO of a large pension fund sometime in the 1990s, he said I must be wrong because surely the smart money would just buy whichever shares were cheaper. I said, “Really? I believe your fund owns millions of dollars of the more expensive version,” and offered to bet a fancy dinner that I was right. He wisely didn’t bet. His fund was partly indexed to the S&P 500, which then included the Dutch version that was selling at a premium.

  § Full disclosure: Since 1998 I have been a partner in a money management firm called Fuller and Thaler Asset Management that invests in U.S. equities by finding situations where investors’ behavioral biases are likely to cause mispricing. The fact that we are still in business suggests that we have either been successful at using behavioral finance to beat the market, or have been lucky, or both.

  VII.

  WELCOME TO CHICAGO:

  1995–PRESENT

  During what amounted to a job interview at the University of Chicago for a position at what is now called the Booth School of Business, I had a lunch meeting with several of the finance faculty members. As we left the business school to walk over to the faculty club where we would have lunch, I spotted a twenty-dol
lar bill lying on the sidewalk, right outside the building. Naturally I picked it up, and then everyone started laughing. We were laughing because we all realized the irony of this situation. There is an old joke that says a Chicago economist would not bother to pick up a twenty-dollar bill on the sidewalk because if it were real, someone would already have snagged it. There is no such thing as a free lunch or a free twenty-dollar bill. But to a heretic like me, that twenty looked real enough to be worth bending over.

  My appointment was not without some controversy in the school. Predictably, Merton Miller was not too happy about it, even though my primary appointment would not be in finance. Instead, I would join the behavioral science group that was made up primarily of psychologists, which I viewed as a plus. I would have the opportunity to build the kind of group of behavioral scientists with strong disciplinary training that I had long thought should exist in a top business school, and while so doing, I would have a chance to learn more about psychology, a field in which my knowledge was quite narrow.

  I am not privy to the internal conversations the faculty had at the time my appointment was considered, but a magazine reporter interviewed Gene Fama and Merton Miller after I arrived, wondering why they were letting a renegade like me join them. Gene, with whom I have always had a good relationship, replied that they wanted me nearby so that they could keep a close eye on me, his tongue firmly in cheek. The reporter pressed Miller a bit harder, specifically asking him why he had not blocked my appointment. This was obviously an impertinent question, to which Miller might well have replied, “None of your business.” Instead, he said that he had not blocked the appointment “because each generation has got to make its own mistakes.” Welcome to Chicago!