The Dangers of Quick Thinking

For example, (LMVTX Quote)Legg Mason Value Trust manager Bill Miller, whose long-term record is among the very best, has said he had trouble getting institutional investors on board until he developed a concise explanation of his style -- it's all about behavioral finance theory and the market's misgauging of some companies, he now says. That style led Miller to buy beaten-down stocks like Tyco (TYC Quote) alongside fast-growers like eBay (EBAY Quote) and Amazon.com (AMZN Quote).

To help explain his unorthodox strategy, Miller even went so far as to hire Michael Mauboussin from CS First Boston as "chief investment strategist."

But Mauboussin doesn't tell Miller how to invest. Asked at a press conference at the time of the hiring what would happen if his view on a stock conflicted with Mauboussin's, Miller quipped: "I didn't hire that kind of strategist."

Instead, Mauboussin writes lengthy think pieces that seek to illuminate the way in which Miller and his team make decisions with plenty of statistical and academic evidence. The fact that the essays come out long after Miller has already taken positions might make a cynic wonder if Mauboussin's role as "strategist" is mainly to provide intellectual hand-holding to the fund's more traditional investors and prospective ones.

The second cautionary theme for investors comes from chapters in Gladwell's book about subconscious biases that can deeply mar intuitive decision making. Warren G. Harding, generally reckoned to be one of the worst U.S. presidents, was elected because he looked and sounded "presidential," not because he possessed the actual qualities needed to run the country, Gladwell notes. Helping Harding was the American prejudice in favor of tall men. Before Bush beat Kerry last year, the taller candidate had won the popular vote in all but two presidential elections since 1888.

This is exactly the kind of subconscious bias that investors sometimes suffer. People sometimes avoid selling losing stocks because they don't want to concede an error or feel regret, as University of California professor Terrance Odean has written. People also do a terrible job estimating probabilities.