Business Report - New breed of investor exploits human weakness

New breed of investor exploits human weakness

By Bloomberg

On the morning of January 28, Frederick Stanske sat in his San Mateo, California, office staring at his computer screen as AirTran Holdings stock surged. The Orlando-based airline had just reported fourth-quarter earnings a share of $0.10, and its stock was en route to a 9 percent gain that day.

Stanske remembers thinking the rise wasn't enough - not after AirTran's profit had blown away Wall Street's average forecast of 3.6c a share by 178 percent.

Stanske, senior vice-president at Fuller & Thaler Asset Management, pounced, buying blocks of AirTran stock. As of June 30, Fuller & Thaler held 1.26 million shares, about 2 percent of the firm.

And so far, Stanske's bet has paid off big: AirTran shares rose more than threefold to $16.88 on September 19 since January 28.

Stanske's trading that day was based on the theory that most investors react too slowly to information that will boost stocks over the long haul.

It's one of many errors in judgment that investors make, according to Fuller & Thaler, and it's among the missteps that are being exploited by a growing number of adherents to an investment discipline called behavioural finance.

Fusing classical economic theory with studies about human psychology and the decision-making process, behavioural finance explores how investors systematically make judgment errors or mental mistakes.

They follow the herd, trade excessively, become "anchored" to ideas and estimates, hold on to losing investments too long, sell winners too quickly and react too slowly to unexpected news.

When they commit any of those sins, financial pros like Stanske are poised to strike.

Confined to academia for more than a decade, behavioural finance is gaining prestige.

Daniel Kahneman, an Israeli-born psychology professor at Princeton University, won the 2002 Nobel Prize in economics for work showing that, while markets are generally efficient, investors still behave irrationally.

Behavioural finance theorists have become regular speakers on the investment conference circuit these days, and the How Humans Behave lecture was the headline event at the Federal Reserve Bank of Boston's gathering in June.

Behavioural finance has attracted critics as well.

Richard Michaud, the president and chief investment officer at New Frontier Advisors in Boston, dismisses the discipline as a marketing tool.

"They've done well, taking a very old way of solving every valuation problem and calling it behavioural finance," says Michaud, who has a PhD in mathematics from Boston University and received a patent in 2000 for an asset allocation strategy he designed.

His lectures boast sobering titles such as The Behavioural Finance Hoax, and his views often place him at the opposite end of the dais from behavioural finance adherents at panel discussions on the subject.

"I'm extremely sceptical, bordering on total cynicism, that any of this stuff actually works," he says.

That's not what some big investors are saying. Margaret Stumpp, senior managing director and co-head of a 40-person team at Prudential Investment Management in Newark, New Jersey, invests $35 billion in assets.

More than $6 billion of the total is earmarked for a computer-driven method that identifies stocks trading out of sync with their valuation because of what she calls "behavioural biases" of investors - in particular their slow reaction to companies' earnings improvements.

"The trick has been to identify the mistakes that people make and exploit them to make money," says Stumpp, who has a doctorate in economics from Brown University, one of seven PhDs held by members of Prudential's quantitative management department.

"No one disputes the notion that people screw up - especially investors."

Kahneman drilled home that point last May in a speech to 1 000 money managers in Washington. When he asked how many considered themselves to be among the room's above-average drivers, hands shot up all over.

"No," Kahneman told investors from Capital Research & Management, Fidelity Investments, Pacific Investment Management and other firms attending the I

nvestment Company Institute's annual unit trust industry conference.

"About 95 percent of any group of people think they have above-average driving skills. That means 45 percent of you are wrong."

Kahneman's point: Overconfidence is an inherent human trait that can sabotage many decisions, and stock picking is no exception.

"The first industry that comes to mind with the most people who believe they have above-average skills is the financial services industry," he said.

Behavioural finance zealots don't hold that markets are completely irrational.

"We're not talking crazy here," says Russell Fuller, president and chief investment officer of Fuller & Thaler, which he calls the first pure-play behavioural finance investment firm.

Fuller is a former stock analyst with a PhD in economics who once headed the finance department at Washington State University. He spent years analysing data showing that the longer analysts and investors study a company, the harder it is for them to evaluate new data and adjust their views.

That's because, he says, their brains rely on mental shortcuts to make decisions. One such shortcut is "categorisation".

For example, mail-order catalogue retailer Fingerhut in the mid-1990s consistently fell short of analysts' profit and sales estimates. Its stock, which traded as high as $31.63 in March 1994, fell more than 60 percent to $11.75 in November 1996. The company, according to Fuller, was categorised as a "bad stock".

In 1998, Fingerhut announced it was developing an online trading department. Company insiders began to buy stock, and sales started to improve. For months, the stock barely budged as most investors remained convinced Fingerhut would continue to perform poorly.

Then, in February 1999, Federated Department Stores agreed to buy Fingerhut for $25 a share in cash, a 33 percent premium to the company's stock price.

"There are powerful behavioural trends ingrained in people," says Fuller. "Overconfidence is one - the 'I can't be wrong' school of thought - and so is 'anchoring', where analysts are glued to prior assumptions."

Fuller & Thaler's managers use a proprietary computer program to spot investor missteps.

That program flagged AirTran for Stanske last January, noting the earnings surprise and then screening for dozens of signs - wider gross and operating margins, cost reductions and stock buy-backs among them - that might determine whether the earnings gain was sustainable or a one-time event.

"We care about three things: is the earnings surprise real? Is it permanent? And are we early in recognising it?" says Stanske.

"Everyone else either dismissed the new data or was opting to wait to see if it was a fluke."

Fuller says his company's investment track record has convinced him that exploiting errors others make is a valid approach.

Its small- and mid-cap growth fund, with $575 million in assets, gained 16 percent after fees from January 1992 through June 30, 2003, compared with a 6.6 percent gain for the Russell 2500 Total Return Growth index.

Behavioural finance traces its roots back to the mid-1970s, when Kahneman and fellow psychology professor Amos Tversky, who died in 1996, published a paper arguing that people make decisions based on mental shortcuts that lead to "systematic and predictable errors".

In the past two decades, Kahneman and other researchers at the University of Chicago and the University of California began studying the behaviour of investors. They concluded that the main obstacle to sound investing is overconfidence.

"During the stock market bubble, investors stayed heavily invested even when they knew it was a bubble," says Kahneman.

"They thought they could get out in time. This is clearly an example of both exaggeration of skill and the illusion of control."

Behavioural finance researchers contend that people take risks because they're deluded - not brave.

"Most people lose money by selling winners and hanging on to losers and then buying stocks they're not certain of," says Kahneman.

"Why do people frantically trade assets? Because people are overconfident. They are overoptimistic."

Terrance Odean, a management professor at the Graduate School of Management at the University of California, Davis, and researcher Brad Barber explored investor optimism in 1999 by looking at the trading records of 88 000 households.

The data spans 10 years and 2 million stock trades, and it showed that overconfident investors underdiversified, that men were more overconfident and traded more often than women, and that the more active investors were, the worse they performed.

Odean's study also showed that having "ideas", such as buying or selling a stock, cost investors money.

The investor often wasn't selling to realise a deliberate tax loss or to raise money to pay off a debt, Odean's study found. Rather, the investor believed the newly purchased stock would outperform the one that was sold.

Looking forward one year from having the idea, the study showed that the stocks sold were worth 2 percent more than the ones bought.

Psychologists and economists also have learned that quantitative estimates are influenced by previous benchmarks.

A car salesman, for example, starts negotiations with a high price. The goal is to "anchor" the customer to the high price so he would think he negotiated a good deal by lowering the amount.

Behavioural finance adherents say this applies to the stock market, offering as an example the $400 presplit price target that Oppenheimer analyst Henry Blodget placed on Amazon.com in 1999. Investors held on to Amazon.com too long because they were anchored to that price.

All of these traits and behavioural characteristics can't be overlooked, some investors say.

"Many of them are embedded into human nature," says Arjen Pasma at ABN Amro Asset Management in Amsterdam, who has built a model that he says accounts for analysts' reaction to earnings reports and herding behaviour.

"People will always have the tendency to be overconfident."

Michaud doesn't deny that researchers can find patterns in market behaviour. Rather, he questions whether the patterns can be predicted - and whether they really are superior to old-fashioned stock-picking tools.

"It's really just repackaging and marketing old strategies," he says. "Many of these look like value managers."

That doesn't explain why a shrewd investor might buy a stock like Genentech, the world's second-biggest biotechnology company. It's the kind of stock a value manager would avoid, because its price:earnings ratio is close to 70 and it trades at about seven times its book value.

Even so, Stumpp bought Genentech shares based on behavioural finance principles during the first week of May. On April 9, Genentech had announced first-quarter earnings of 35c a share, beating the average estimate of about 28c.

After the earnings release, 28 analysts raised their estimates, but probably didn't go far enough, Prudential concluded.

Genentech's stock price rose 144 percent to $87.72 on September 19 from $35.90 during the first week of May. And every share Stumpp bought came from an investor who probably made a mistake in selling it.

Published on the web by Business Report on September 26, 2003.