In early October, Daniel Kahneman and Vernon Smith won the Nobel Prize in Economic Sciences for their research, conducted independently, into how individuals make economic decisions. The two professors discovered that investors are not systematically rational, as traditional economic theory asserts (much of Kahneman’s research, it should be noted, was conducted with his longtime collaborator, the late Amos Tversky). Investors make decisions for emotional reasons, they base their decisions on shaky premises, they are quick to see cause and effect where there may be none, and so on. The professors’ research provided an impetus, respectively, to the burgeoning fields of behavioral finance and experimental economics.
So what does this mean for the markets? Since the stock market is based on the cumulative decisions of individuals, does this suggest that the market isn’t efficient at determining the correct prices of stocks? Do arbitrage opportunities abound?
A few days before Kahneman and Smith got their phone calls from the Nobel Prize committee, Wharton hosted a debate that addressed these questions. Called “Two Views of Market Efficiency: A Discussion of Behavioral Finance and Efficient Market Theory,” the scrimmage took place between Burton Malkiel and Richard Thaler, and was moderated by Wharton finance professor Jeremy Siegel, author of Stocks for the Long Run.
Malkiel, a Princeton University finance professor and author of the famed A Random Walk Down Wall Street, represented the efficient-markets camp. This group argues that the market is a mechanism that utilizes the collective information of stock market participants to produce efficient prices, and that there is no easy money – fresh $100 bills lying on the ground – for investors to make by predicting how stocks will behave. In other words, it’s hard to consistently beat the Standard & Poor’s 500 Index.
Thaler, a finance professor at the University of Chicago and a principal at Fuller & Thaler Asset Management, a money management firm based in San Mateo, Calif., took up the cudgels for the behavioral finance camp. He reflected the view that psychology plays a large role in the movement of stock prices and that patterns, or predictabilities, exist in the market that can be profitably arbitraged.
Both professors staked out common ground – at least initially. The market is not so rational that patterns in stock prices can’t be discerned, observed Malkiel. He admitted that psychological factors affect the stock market and that markets overshoot, creating bubbles such as the Internet technology craze of the late 1990s that lifted stock prices to untenable levels.
However, he said, “none of the patterns that have been discovered have been dependable, many have self-destructed as soon as they have been discovered, and many aren’t economically meaningful.” The patterns that endure are often too small to take advantage of, given the transactions costs associated with the trades. Or the trades are too risky. “I truly believe that our capital markets are remarkably efficient,” he continued. “The stock market is far less predictable than many of my academic colleagues have asserted. While the market is not statistically a perfect random walk, in my judgment investors would be very well-advised to act as if it was essentially unpredictable.”
Thaler also claimed what he described as the “sensible” middle ground. “Securities prices are highly correlated with intrinsic value, but sometimes diverge to a significant degree,” he said. “It’s possible to predict stock prices, but not with great precision – and don’t try this at home.”
That said, the two professors pursued their separate views of the market.
Malkiel gave a few examples of alleged predictabilities that, he said, on closer reflection don’t hold up. One is the popular suggestion that growth stocks perform better over the long haul than value stocks. Even if it were true for a period, he said, there could be two explanations. One is the behavioralist explanation, which is that investors are overconfident in their ability to predict growth stocks and therefore overprice growth stocks. The other is the belief that efficient-market proponents hold – that if there’s a pattern for some time, it could be because the value stocks are actually riskier. In other examples, Malkiel noted that patterns in stock prices that get chalked up to investors being exuberant or pessimistic could simply be the adjustment of stock prices to economic conditions.
For those in the efficient-market camp, however, the most convincing proof of the efficiency of the market is the fact that professional portfolio managers cannot consistently outperform the market. “Surely,” said Malkiel, “if the market were always dominated by irrational investors, if it systematically deviated from what were rational estimates of value, professionals who are richly incentivized to outperform would be able to beat the market.”
But actively managed funds don’t beat the S&P 500 index. “If you look at the median mutual fund, and the S&P index over the last 10, 15 and 20 years, there’s been about a 200 basis points underperformance of the median mutual fund vs. the S&P index,” he said. “If the S&P index were an athlete, they would be testing it for steroids.” Not only that, but there is survivorship bias at work behind the statistics, improving the results of managers. What this refers to is the fact that many funds that didn’t fare well no longer exist and so their performance isn’t incorporated into current statistics.
Some fund managers, of course, do beat the index. But the problem is that investors do not know in advance which managers will rise to the top. The 20 best performing funds in the 1970s, which doubled the returns of the S&P 500 index, underperformed the index in the 1980s, noted Malkiel. The 10 best in the 1980s underperformed in the 1990s. Those who managed the best funds in 1998 and 1999, which did three times as well as the S&P 500 index over the same period, were “written up in Money magazine as the genius portfolio managers,” said Malkiel, “and in 2000 and 2001 it was fly now, pay later, because they did three times worse than the index.”
No Free Lunch
Thaler launched his commentary by pointing out a logical fallacy in the efficient-markets camp. He identified two components of market efficiency. The first, which he calls the No Free Lunch Theorem, simply states that in an efficient market prices aren’t predictable and so it’s not easy to make money. He agreed that this was largely true. The second component, dubbed the Price Is Right Theorem, suggests that asset prices reflect the intrinsic value of the underlying securities and that the prices are therefore rational.
According to Thaler, Malkiel was making the mistake that many economists make – arguing that the first finding leads logically to the second. Even if prices are unpredictable, said Thaler, “they may be very wrong.” He noted that Bob Schiller, whose book Irrational Exuberance argued forcefully that prices were not rational during the Internet bubble, called the confusion about this “one of the most remarkable errors in the history of economic thought.”
Although Malkiel acknowledged that bubbles can exist, even in an efficient market, he pointed out that markets always come back down to earth. But in Thaler’s view, this position misses a significant point. In addition to bubbles, there may be stock market “funks” – times when stock prices are irrationally low. “If bubbles break, what makes us think they break to the right level? Do we know that Japan’s stock prices are rational now?” asked Thaler. “If we’re going to concede that they were irrational in 1989, like [Malkiel] says Internet prices were irrational in 1999, then we also have to grant the possibility that they’re irrationally low now. Stock prices have steadily fallen for over a decade in Japan. If there was a bubble then, maybe they’re in a funk now.”
The idea that prices are rational at any point is up for grabs. To prove his case, Thaler offered examples of companies whose stock prices were, by most lights, incorrect at some point. In 1907 Royal Dutch and Shell Group formed a single company and merged their interests 60/40, but continued to trade as separate stocks. However, the shares of the respective companies didn’t always trade at that ratio, and over the last two decades in particular there were large deviations from their theoretical relationship.
Thaler gave another example. In 1999, 3Com decided to spin off its Palm division, which made handheld computers. 3Com held onto 95% of the shares and announced that each 3Com shareholder would probably get about 1.5 shares of Palm for each share of 3Com. The stocks of the two companies should have moved in tandem, but on the day of the IPO, Palm shot up in value while 3Com lost ground. In other words, what happened wasn’t rational, given the numbers.
Malkiel agreed that the 3Com/Palm example was “a $1,000 bill,” adding that he had tried to sell Palm short at the time but was unable to borrow shares. The Royal Dutch/Shell trade, he said, wasn’t as simple an example since the two stocks were under different national regulatory authorities. More broadly, these and other examples of market inefficiencies are “very small examples,” he said, adding that the market “is not a perpetual tulip bulb craze.”
Both professors agreed that stock market bubbles eventually deflate but that it’s difficult to predict when that will occur. Similarly, they noted that it’s hard to take advantage of mispricings because it might take too long for prices to return to a more sensible level. Thaler pointed out that Long-Term Capital Management, the Greenwich, Conn.-based hedge fund that capsized in 1998, had put on the Royal Dutch/Shell trade by going long Shell and shorting Royal Dutch. “But if you’re doing it with other people’s money, like LTCM was, you’ve lost your clients before you’re right,” he said. LTCM was right about Royal Dutch/Shell, but prices didn’t fall back into line until 2001.
The point is not simply that there is often no free lunch. “What we learn is that prices are wrong,” argued Thaler. These examples are the tip of the iceberg of market inefficiency, not just minor blips. “We can argue till the end of our days about whether value firms outperform growth firms, and if so whether that’s rational,” he said. “I don’t think we can argue for long about whether the Royal Dutch and Shell stocks were priced right. If there’s anything the market should be able to get right, it’s these things.”
The Internet Mistake
What troubles Malkiel most as a proponent of efficient markets, he conceded, “is that when the market got it wrong, the market was not giving the right signals to businesses about what the true cost of capital was – and we had an enormous overinvestment in not only Internet companies but the telecommunications structure to make the Internet run.” About 95% of the long-distance fiber is currently unused. “We had a tremendous misallocation of resources,” he said.
For Thaler this misallocation of capital is a big iceberg. “The fact that it’s hard to predict prices and that most money managers don’t earn their fees do not tell us anything about whether the capital markets are doing a good or bad job of allocating capital,” he said. “We know there was a $7 trillion mistake in the late 1990s.” He doesn’t think the government or any other entity could do a better job of allocating resources, but that doesn’t mean the market is efficient.
Thaler also pointed out that a consequence of the Internet bubble could be a reduction in the availability of venture capital for a number of years. Siegel weighed in at this point to note that, unlike, say, the railroad bubble of the late 1800s, the Internet bubble involved a lot of people switching jobs but very little destruction of physical capital. “The railroads spent millions of man-years producing iron and cutting through mountains, but with the Internet you didn’t have a lot of [physical] resources, so there wasn’t that [kind of] misallocation,” he said.
Thaler took another swipe at the idea of efficient markets, floating the idea that the runaway boom in stock options was a consequence of the belief in efficient markets. “This was pushed by economists who felt that markets are sending the right signal through stock prices and that the right way to reward CEOs was to load them up with stock options,” he said. Siegel agreed that there was overissuance of stock options, but noted that tax issues had fueled the trend. In 1993, he said, the government decided that a CEO’s salary in excess of $1 million would not be deductible from the corporate tax bill unless it was deemed incentive-based, and the IRS subsequently ruled that options were incentive-based.
So given the difficulty of predicting stock prices reliably, how does Thaler make investment decisions? Thaler noted that one strategy his money management company follows is to try to predict analyst revisions. “We know people make mistakes,” he said. “We think we can predict their mistakes.” He compared it to baseball. When a sinker-ball pitcher throws a pitch, batters typically swing too high and hit ground balls. “We try to predict the ground balls of analysts,” he explained. “We try to find stocks where we think they’re going to revise up next quarter.” He joked that he didn’t know if the prices were becoming more rational. But, he added, quoting Keynes, “in the long run we’re all dead.”