Page 1 of 3

Robert Shiller, a professor of economics at Yale University, made a prediction in 2005 that a massive bubble was developing in the housing market, and was proved right just two years later, it seemed a mortal blow for classical finance. Shiller is one of the founders of behavioral finance, a school of economics that believes that the psychological behavior of investors can have a big impact on markets. As he had done with his earlier prescient forecast of “irrational exuberance” in the stock market bubble of the late ’90s, Shiller seemed to be staging a direct attack on the Efficient Market Hypothesis (EMH), which University of Chicago economist Eugene Fama had developed three decades earlier. According to Fama, investors are always rational, and markets accurately reflect all publicly known information. In this utopian world, securities will always be appropriately priced, and no amount of analysis can result in outperformance. Shiller, for his part, vehemently disagrees.

“The Efficient Market Hypothesis is one of the most egregious errors in the history of economic thought,” he says. “It’s a half-truth.”

As Shiller suggests, the financial crisis of 2008–2009 seems to have given a major boost to behavioral finance theory, and its advocates are not shy in declaring victory. “If the argument is that people are perfectly rational, then we have won the argument,” says Dan Ariely, a professor of behavioral economics at Duke University’s Fuqua School of Business.

Yet, when the bubble burst, very few investors actually made any money from the subsequent market crash. Even funds that employ behavioral techniques to influence their investing fell sharply in 2008 along with the rest of the market. It’s true that some hedge funds made huge profits betting against subprime-mortgage-backed securities, but Richard Thaler, a professor at the Booth School of Business at the University of Chicago and a founding theorist of behavioral finance, says it’s almost impossible to earn a living making such investments.

“The world isn’t structured in a way that somebody could create a fund that will bet against bubbles when they appear, because you’d be on the sidelines a lot of the time and you’d go through really hard times,” says Thaler, who works as a principal at Fuller & Thaler Asset Management in San Mateo, California, when he isn’t teaching finance or doing academic research.

Like Thaler, the fund managers employing behavioral finance are not betting against bubbles. Instead, they believe that investors make mistakes because of cognitive and emotional biases — such as a presumption that a stock that has performed well in the past will continue to do so far into the future — that cause equity prices to either overreact or underreact to market news. It is these mispricings that behavioral finance strategies attempt to exploit.

Over the past 15 years, there has been a steady increase in the number of fund managers that are using behavioral finance concepts to select stocks and construct portfolios. One estimate is that half of the 200 listed small-cap value funds use some form of behavioral finance in selecting their portfolios. Such firms as Fuller & Thaler, Chicago-based LSV Asset Management and even fund behemoth J.P. Morgan Asset Management have deployed strategies that use behavioral concepts to select equities for their portfolios. And all of them are beating their market benchmarks over the long term.

In addition, a growing number of investment companies, ranging from Des Moines, Iowa–based Principal Global Investors to Catalpa Capital Advisors, a New York hedge fund firm headed by Joseph McAlinden, a former chief investment officer for Morgan Stanley, are using behavioral finance tools to help shape their portfolios, even if they are not the mainstay of their investment strategy.

Yet critics are dubious of behavioral claims. Ray Ball, a professor of accounting at the University of Chicago, complained recently that behavioral finance is not a theory, but merely a “collection of ideas and results” that depend for their existence on the EMH. Other critics cite the fact that behavioral finance has not produced much data to support its arguments. Fama tells Institutional Investor he still believes behavioral investing is really just another name for choosing value stocks that have a higher cost of capital, which gives them higher expected returns.

“Active managers as a whole can’t beat the market,” he says. “That’s impossible.”

Even Meir Statman, a renowned behavioral finance expert at the Leavey School of Business at Santa Clara University in California, says he doubts you can make any money investing in behavioral finance funds. “Maybe you can make money at the margins, but I doubt it very much,” Statman explains. Even if behavioral investing can earn a few basis points of alpha, he adds, that would be consumed in managers’ fees.

When Russell Fuller and Richard Thaler set up their asset management firm in 1998, Fuller, who had been chairman of the finance department at Washington State University, had already been in business five years. As an academic, Thaler had written a classic paper on why the stock market overreacts to new information, but he says he was still dubious about using behavioral finance actually to invest. He thought Fuller’s results might have been the product of luck as much as skill.

So Fuller produced what his colleague now calls “Thaler’s favorite chart.” The chart shows that, using a small-cap growth strategy, an investor can identify companies that produce consistently positive earnings revisions over a long period of time. The idea is that the market consensus incorrectly interprets positive information about a company’s profitability. That’s because analysts unconsciously make cognitive mistakes in their estimates. Fuller says his firm is able to identify these companies in four out of five cases. “What this says is, it’s working for the reason we think it does,” Thaler explains.

So how does it work? The small-cap growth strategy begins with an earnings surprise. According to Thaler’s research, if you bought every stock that had an earnings surprise, you’d actually make a little money. But Fuller & Thaler takes the strategy one step further. The firm has a portfolio manager go through earnings surprises that are likely to generate an underreaction, which means that the market’s estimates of earnings are biased on the downside and the stocks tend to be underpriced. The trick, according to Thaler, is to determine whether the source of the surprise is permanent. So, for example, if an oil company had an earnings surprise because the price of oil went up, Fuller & Thaler wouldn’t be interested, but if the company had found a more efficient way of refining oil the money manager would be.

Fuller gives this example: Analysts are expecting $1 in earnings, but a company reports $1.50. Many analysts will be slow to revise their estimate to $1.50 because of a behavioral bias, or what academics refer to as a heuristic called anchoring — the attachment to previous information. So, although analysts will eventually catch on, they will catch on slowly, maybe increasing their estimate to $1.30, and there will be a time lag before the estimate goes all the way up to $1.50. It’s that time lag that Fuller & Thaler tries to exploit.

The opposite effect is overreaction. Here, Fuller & Thaler is seeking out companies whose share prices have been beaten down by bad news about their profitability, but where the news is likely to be temporary. If their book value (total assets minus liabilities, preferred stock and intangible assets) remains constant, these equities become low price-to-book stocks, which investors normally term value stocks. Many money managers that have embraced behavioral finance use this value strategy.

Another bias commonly associated with overreaction is representativeness — using past experience to interpret new information. This is based on psychological studies that show when people suffer a long string of losses, their relative risk aversion changes. When the market thinks a stock is a dog, has always been a dog and always will be a dog, that is representativeness. But Thaler and DePaul University finance professor Werner De Bondt showed in an academic paper in 1985 that long-term losers often become future winners. Fuller & Thaler watch for information that indicates that the future earnings of some of these “dog” stocks will make positive changes. That’s when the firm buys them.

Fuller emphasizes that the investing technique is considered “judgmental” because the firm uses three portfolio managers and three analysts to look at events to see if they are correctly reflected in share prices. But they don’t perform fundamental research on the companies in question, because to do so would be to introduce their own behavioral biases. “We don’t try to forecast earnings; we don’t try to be industry experts,” says Fuller, who as president and CIO oversees the 20-person firm’s investment and research activities. “We understand in what circumstances people are likely to make mistakes given a certain type of information event.”

The decision to exclude fundamental analysis is controversial even within the behavioral finance community. Charles Lee, a professor of accounting at Stanford Graduate School of Business and a former adviser to Barclays Global Investors, says that even with a behavioral approach you need to look at fundamentals to determine the difference between price and value. “If you don’t look at fundamentals, you’re looking at market indicators and sentiment indicators,” Lee says. “Behavioral finance is trying to look for the departures and why they are happening, but you have to measure them first before you know they are happening.”

Thaler says that the strategy works best with small-cap stocks because that sector is less efficient. “Not surprisingly, the alpha that seems to exist goes up as the market cap of the stock goes down and even more so as the liquidity goes down,” says Thaler, who in his role as a principal at the firm talks on the phone almost every day with Fuller.

With large-cap stocks, Thaler says, his firm has to go both long and short and use both underreaction and overreaction to make any money, because of the increased competition in that sector. In fact, the firm is about to roll out a new large-cap strategy.

According to Fuller, the firm has about $1.1 billion under management. The small/midcap growth strategy, which started in 1992, has average annualized returns of 13.18 percent, while the Russell 2500 growth index for the same period returned 7.17 percent and the Standard & Poor’s 500 index earned 7.95 percent. The small-cap value strategy, started in 1996, has earned annualized returns of 13.62 percent, compared with 9.38 percent for the Russell 2000 value index and 6.20 percent for the S&P 500.

While Thaler was honing his theories about cognitive biases, three other academics with a clear behavioral bias also decided to get into money management. LSV Asset Management was founded in Chicago in 1994 by Josef Lakonishok, a former professor of finance at the University of Illinois; Andrei Shleifer, an economics professor at Harvard University; and Robert Vishny, a professor of finance at the Booth School of Business.

Although Vishny and Shleifer have since retired from LSV, Lakonishok is CEO and chief investment officer. In 1992 he wrote an academic paper that laid out his investment idea: Value investing works because it is contrary to naïve strategies of other investors, which include extrapolating past earnings growth too far into the future and assuming that a well-run company will always be a good investment, regardless of price.

Today, Lakonishok plays down the behavioral aspects of his firm’s investment strategies and says what he practices is classic value investing. “People were doing this type of investment — buying value stocks — before academics got excited about doing research in this area, and academics should not be taking credit for inventing a new field,” Lakonishok says. “Perhaps it became a little more systematic and we pointed some things out, but a lot of people had these ideas going back many years.”

Lakonishok is reluctant to talk about exactly how his strategy works, but his firm’s Web site makes clear that he is looking for behavioral mistakes, such as investors’ habits of ignoring relevant statistical evidence. But unlike Fuller & Thaler, which invests both long and short, LSV follows a strictly value-oriented long-only approach that attempts to choose out-of-favor stocks with the potential for near-term appreciation and that uses momentum as part of the strategy.

In contrast with Fuller & Thaler, LSV uses a purely quantitative approach, employing a complex computer model rather than portfolio managers to screen for stocks that meet its requirements. For example, the firm has identified mathematical proxies for extrapolation by processing data about analyst recommendations and analyst forecasts. Behavioral biases are not so easy to overcome — that’s one of the advantages of a quantitative approach.

Like Fuller & Thaler, LSV does not do fundamental analysis of companies in the sense of talking to management and visiting firms. But, Lakonishok says, “you cannot start building strategies without looking at companies’ financial statements.”

LSV has about $60 billion under management. Its large-cap fund, which began in December 1993, has generated annualized returns of 11.2 percent, compared with 8.2 percent for a large-cap value benchmark and 7.5 percent for the S&P 500. Its international large-cap fund, which was started in 1998, has produced annualized returns of 9.1 percent, versus 4.1 percent for its benchmark.

Despite their philosophical differences, Dick Thaler and Gene Fama remain close friends at the Booth School of Business. According to Fuller, Fama was instrumental in getting Thaler his job there in 1995. The two men are not only friendly academic rivals — they have a fanatical rivalry on the tennis court. Thaler has invented something he calls “equilibrium tennis,” in which a player can move only two steps before hitting a ball, a clever ruse designed to give him a fighting chance against Fama, who is the more physically fit.

Although their friendship thrives on the court, the philosophic divide between Fama and Thaler is as great as ever. Fama maintains that though there are some behavioral biases on an individual level, the greater claim that they influence the market just doesn’t make sense. One of his biggest criticisms is that the behavioralists haven’t replaced the EMH with a single, comprehensive explanation of their own.

“They develop a new explanation for every thing that comes along, and there’s nothing that ever contradicts their story,” he says. “That’s not a real theory.”

Richard Bernstein, chief executive of New York–based Richard Bernstein Capital Management and a former chief investment strategist at Merrill Lynch & Co., agrees that the lack of an overall, coherent theory is one of behavioral finance’s weaknesses. “The lack of a unifying principle is a big problem,” Bernstein says. “It’s not like value investing, where you say value investing is low price-to-­earnings, and it’s very easy to describe to people. With behavioral finance, there is no Graham and Dodd valuation formula,” he adds, referring to the authors of the investing classic Security Analysis.

But Thaler shrugs off the lack of a unifying principle. “There’s never going to be a theory that’s as simple as the one that says everybody is rational, they have rational expectations, and they make decisions according to expected utility theory,” he says. “If you want a neat and tidy theory, that’s the one. It just happens to be wrong.”

For his part, Fama argues that behavioral finance strategies are just rechristened value approaches that succeed because they involve taking on more risk. If cognitive biases were the explanation, he insists, they would be a short-term phenomenon because arbitrageurs would learn about them and they would soon cease to work. (Of course, behavioral finance adherents argue that investors are hardwired to make the same mistakes over and over again.)

Lakonishok says his firm’s 15-year track record with value investing has taught him differently. “I am absolutely convinced, without any doubt, that it is not risk that is driving the higher returns of value companies,” he says. Instead, it’s investing anomalies, like extrapolating past results into the future, that account for these returns. According to Lakonishok, value stocks don’t seem to be riskier in terms of higher volatility and are not perceived to be more risky by investors: “I am a very big skeptic of the idea that the secondary risk measures that were invented to accommodate the outperformance of value companies are actually risk measures.”

Meir Statman of the University of Santa Clara agrees. “There are enough papers now that show risk is not what underlies outperformance,” he says. “It is emotion; it is sentiment.”

Stanford’s Charles Lee says that if you want to understand why there is sometimes a difference between price and value, it is necessary to study investor sentiment and have a theory of how people make mistakes in investing. In the efficient market world, sentiment doesn’t exist.

It is such disputes that have led some academics to question the accuracy of both efficient market theory and behavioral finance. Andrew Lo, a finance professor at the Massachusetts Institute of Technology Sloan School of Management, says the biggest contradiction of the EMH is the claim that there are no excess profits to be made, while investors such as hedge fund manager John Paulson and a number of others have been successful in beating the market. The inconsistency of the behavioralists, Lo maintains, is the idea that investors are all irrational and the EMH doesn’t hold, yet it is still very hard to beat the market.

Lo has proposed another theory, which he calls the Adaptive Markets Hypothesis, in which markets are neither efficient nor irrational, but some combination of both. Using the principles of evolution to describe stock market behavior, he notes that markets wax and wane like the populations of animals in the jungle — it’s either feast or famine. “There are periods when the market is highly efficient, and there are those that aren’t,” Lo says. “The better question is: What is the relative efficiency at a given point in time? That is something that can be measured.”

In the models Lo has developed, markets are adaptive — they take into account the level of competition, the level of sophistication, even the amount of capital deployed. The more capital deployed in a given market, the more efficient it will be and the smaller the alpha each manager can earn. “Humans may be more sophisticated than animals, but we are nevertheless animals, subject to the same kind of forces that nature has built into all of creation,” he says.

In the early 1990s, J.P. Morgan Asset Management began experimenting with behavioral finance in the U.K. The strategies proved successful and were adopted in the U.S. At first, J.P. Morgan began offering two funds managed by Fuller & Thaler — the Undiscovered Managers Behavior Growth Fund and the Undiscovered Managers Value Fund. But in late 2000, Christopher Blum was chosen to begin managing a small-cap behavioral fund for J.P. Morgan. Blum had been working as an analyst valuing private equity for alternative asset manager Pomona Capital in New York. With a staff of just two, the behavioral unit spent a significant amount of time conducting research into value- and momentum-based anomalies, such as how investors overreact to bad news. A decade later the team has grown to nine, and J.P. Morgan now offers ten different behavioral strategies, seven within the Intrepid family of funds, and has $10 billion under management.

Blum, who is CIO for the U.S. behavioral strategies, says his group employs a hybrid approach using both quantitative and qualitative analyses. First, J.P. Morgan’s quantitative research group screens the universe of all U.S. stocks, looking for behavioral finance anomalies using its valuation and momentum models. The valuation model tries to identify stocks that the market has overreacted to, such as drugmaker Merck & Co. at the peak of the Vioxx lawsuits. Its momentum model tries to pick out stocks driven up by the herd chasing recent winners. The models rank each stock according to the anomaly they are targeting.

Next, J.P. Morgan’s qualitative research team, composed of industry experts, performs due diligence on the names the computers have generated. They make sure that the anomalies they are looking to exploit are reflected in the rankings. They try to identify false positives — say, when a stock ranks highly because it looks cheap but really isn’t. For example, the computer may identify a company as being undervalued on a free cash flow basis, but the qualitative team finds it benefited from a one-time legal settlement that artificially inflated cash. The team also tries to avoid unnecessary risks (say, a good momentum stock that has asbestos liability or uncertain regulatory drug trials).

J.P. Morgan’s analysts hunt through financial statements “to confirm what our quantitative metrics are telling us is actually true,” Blum explains. His group wants to make sure there aren’t one-time events that may be distorting earnings or cash flow. On the momentum side, J.P. Morgan’s analysts are looking for price-related signals and clues about operating profitability. Like Fuller & Thaler, J.P. Morgan is hoping to exploit the fact that Wall Street research analysts are often slow to react to a change in performance.

“We have this fundamental belief that, over the long run, cheap stocks should outperform expensive stocks, stocks with superior momentum attributes should outperform stocks with inferior attributes, and higher-­quality stocks should outperform lower-­quality stocks,” says Theodore Dimig, a behavioral finance specialist at J.P. Morgan Asset Management. “We think this not only is a great stand-alone discipline but also complements other strategies out there because we’re buying stocks generally at different points of time than our fundamental counterparts are.”

So far, the behavioral finance strategy appears to be paying off. Since inception in 2003, the JPMorgan Intrepid Value Fund has delivered annualized returns of 8.09 percent, compared with 6.74 percent for the Russell 1000 value index. JPMorgan Intrepid Growth has returned 6.79 percent a year since its 2003 inception, beating the Russell 1000 growth index return of 6.20 percent.

Although relatively few money managers actually use behavioral finance to choose stocks and construct portfolios, many have adopted behavioral concepts as tools to help them figure out where they should be investing. Principal Global Investors, with $60.4 billion in equities under management, uses quantitative analysis to identify stocks with rising fundamentals and attractive valuations. The firm’s 74 equity investment professionals then take a deeper look at the stocks on the computer-generated lists and refine the lists into portfolios.

Even with all the fundamental analysis taking place, Jeff Schwarte, a portfolio manager at Principal, says the firm decided a year and a half ago to hire Cabot Research, a Boston-based consulting firm, to look at all of his fund’s transactions on a daily basis to try to identify behavioral mistakes its managers were making. “They then try to identify opportunities that we can improve upon,” Schwarte adds.

Michael Ervolini, CEO of Cabot, says his company uses analytical software developed with the help of Terrance Odean, a finance professor at the Haas School of Business at the University of California, Berkeley, to look for patterns of behavioral errors like holding losers too long or selling winners too soon. “We do this up-front analysis to give managers more self-awareness, and then we do ongoing analysis to give them feedback about individual decisions they are making,” he says.

To determine if a fund is holding losers too long, Cabot will do an analysis of what would happen if those stocks had been sold earlier on a monthly basis. When the adjusted history is compared with the actual history, managers can tell if they would have done better by selling sooner. “Usually a fix is pretty obvious,” Ervolini says. “But overcoming those emotional triggers is hard.”

Another behavioral bias Cabot looks for is regret aversion. Some managers like to test a theory of a rising stock by buying only a small portion and then investing the full position at a later time. They could, for example, buy a stock at $10 hoping it will go to $40. When it reaches $15, that’s the trigger to start putting more money in. But Ervolini says a lot of managers hesitate at that point. “They unconsciously kick themselves for not buying more at $10,” he says. “We see regret aversion commonly.”

Cabot avoids the problem of hindsight by doing analysis over five to six years. In the first two years, it will look at trading history to see if the manager is selling losers too slowly. Then it will formulate a rule — sell all losers after 18 months, for example — and apply that rule to the next two years. The process is repeated over and over again. “If the results have statistical significance, then we know that the behavior is highly persistent and that altering it is likely to have predictive benefits in the future,” Ervolini says.

Cabot is currently working with nine investment firms.

Another manager who uses behavioral finance as an investing tool is Joseph McAlinden, the former Morgan Stanley executive who set up a discretionary global macro hedge fund, Catalpa Capital, in March 2007. While many of his competitors make trades based on mean reversion — the idea that a value will eventually return to an average over time — McAlinden looks for what he calls “discontinuous change,” a turning point that’s not widely recognized and will result in a new mean being established. “All of us who run money are human beings and are susceptible to falling into these behavioral patterns,” he says. “The more you know about them, the better money manager you are going to be.”

One of the factors McAlinden looks for is status quo bias — the idea that markets are stuck on old information and slow to recognize a change, creating a near-term trading opportunity. His firm also spends a lot of time analyzing confirmation bias, in which investors latch onto any news that they think confirms their own views about a stock. When all the earnings estimates are going up and all analysts are bullish, that is when McAlinden wants to bail out. “We try to fight it, but we know we’re human like everybody else,” McAlinden says. “When we get into a theme at the very beginning, we try to outline a set of conditions that would drive our exit from that theme.”

From his vantage point in New Haven, Connecticut, Yale professor Robert Shiller says he can’t imagine making investments without incorporating the wisdom of behavioral finance. “We’re talking about playing a game against other people,” he explains. “How do you ever play a game without thinking about their psychology?”

Single Page    1 | 2 | 3