Behavioural investing is quite different from the common practices of fundamental analysis and quantitative strategies.

Behavioural investing consists of three key ingredients, of which the first is simply that investors are human. Investors do not act like supercomputers. As the seminal work of researchers Daniel Kahneman and Amos Tversky documented, people often use mental short-cuts or "heuristics" to solve complicated problems, and while these heuristics have an important role in our life, they often bias our judgment and predictions.

One of the many heuristics that affect our investment decisions is representativeness, or simply stereotyping. Our tendency to chase performance is an example of the representativeness heuristic, as we misjudge past performance to be representative of future performance. Computers tell us past performance has little to do with future performance, yet most plan sponsors and fund of funds managers will only hire investment advisers whose past performance has been superb.

Another example of the representativeness heuristic has to do with the recent crisis in the US banking and insurance industries. The collapse of Bear Stearns, Lehman Brothers and AIG caused virtually every bank to drop in value significantly, even if they had no exposure to the subprime crisis.

Note that applying representativeness and other heuristics inappropriately could bias virtually every stock.

The second ingredient of behavioural investing is biases persist as it is not easy to stop stereotyping. I bet most consultants and plan sponsors reading this article will keep focusing on managers with top past performance, despite the lack of correlation between past performance and future performance.

The third ingredient of behavioural investing is that arbitragers will eliminate some biases and mispricing, but they will not eliminate all of them. Using the terminology of Professor Andrei Shleifer of Harvard University and Professor Robert Vishny of the University of Chicago, arbitrage is limited.

Professor Richard H Thaler of the University of Chicago, who is also co-founder of Fuller and Thaler Asset Management, has an interesting anecdote illustrating the limits of arbitrage. In the mid to late 1990s, he asked analysts and portfolio managers for their views of the valuation of dotcoms. The typical response was that dotcoms were vastly overpriced.

However, when the same portfolio managers were asked to predict the six-month change in price, they indicated irrational exuberance would keep pushing the valuation of dotcoms even higher.

Researchers Markus Brunnermeier of Princeton University and Stefan Nagel of Stanford University investigated the trading patterns of hedge funds during the dotcom bubble and documented that the average hedge fund profited from riding the bubble.

Putting the three ingredients together, the idea is to identify biases in investor behaviour that will persist even in competitive stock markets. Investors should trade on biases that arbitragers cannot easily eliminate. Opportunities to profit from the psychology of investors do exist, especially in smaller and "neglected" stocks.

So how is behavioural investing different from fundamental analysis? Fundamental managers capitalise on their ability to see things in the data that others often miss. And, a few legendary investors, such as Warren Buffett, have convinced most of us they have unique insights. However, the difficulty with selecting fundamental managers is that it takes decades to conclude that a particular fund manager has skill rather than luck.

Behavioural investing also differs from quantitative strategies. If you invest on the basis of a statistical relationship in past data, you never know whether you identified a true opportunity or just statistical randomness. However, understanding the psychology of investors and markets increases the likelihood of identifying true opportunities.

Behavioural investing is not only a tool for fund managers. It is equally relevant to plan sponsors, endowment managers and fund of funds managers. Instead of selecting investment managers based on past performance, plan sponsors should ask why a fund manager outperformed the benchmark. And, is the source of outperformance truly capitalising on sustainable biases and mispricing?

Shlomo Benartzi is professor and co-chair of the behavioural decision making group at UCLA, and principal at Fuller and Thaler Asset Management

sbenartzi@fullerthaler.com