Ask the Experts—Then Do the Opposite?

You’ve no doubt heard the expression, “Hindsight is 20/20.” But, according to extensive research by psychologist and researcher Philip Tetlock, that may not be the case, especially with highly educated, intelligent, and well-informed experts (yes, you read that correctly). Tetlock, who has studied thousands of expert predictions over the past thirty years, has actually found that the more respected and famous the expert, and the more specialized and detailed their knowledge of their subject, the less likely their predictions are to be correct.

David Epstein’s new book Range: Why Generalists Triumph in a Specialized World, details Tetlock’s research as well as other fascinating insights into why some people succeed and others don’t. Tetlock identified two distinct styles of predictors that he dubbed “hedgehogs” and “foxes,” after the analogy formulated by philosopher Isaiah Berlin. Hedgehogs—the experts—know one thing, and they know it in exceeding depth. Thus, their predictions, while based on a large body of data, tend to be drawn solely from their own thoughts and conclusions about the data. In Tetlock’s study, such experts’ predictions were wrong between 75 and 85 percent of the time. Tetlock also made the astounding discovery that, even when hedgehogs’ predictions are proven wrong by events, they typically double down, drawing inferences from actual events to reinforce the beliefs they already hold. In other words, hedgehogs’ judgment about the future may not change much, even with the benefit of hindsight.

Foxes, on the other hand, as generalists, know many things in varying amounts of detail. When they make predictions, they tend to be more cautious about their conclusions and also pay greater heed to varying and even contradictory opinions. They are also more inclined, when one of their predictions proves incorrect, to glean data that may change their assumptions and result in a better subsequent prediction. In other words, foxes tend to listen to more sources, learn from their errors, and improve their accuracy over time.

A related study at Duke University bears out these observations. Each year the business school at Duke surveys the CFOs of the nation’s largest corporations, asking for their predictions of various business indicators, including the following year’s return on the S&P 500. After gathering more than 11,000 such predictions and comparing them to actual events, the researchers found that the correlation of the CFOs’ predictions to the market’s actual performance was less than zero—worse than random chance! Not only that, but the data showed that the CFOs were grossly overconfident in the correctness of their opinions.

Investors can draw several important lessons. First, as we have often stated, most pundits’ opinions, especially in the financial news media, are intended more to attract headlines—and “likes” and “clicks”—than to impart actual insight. When any expert—either someone merely claiming expertise or even a trained, credentialed professional—predicts that the market—or the dollar, or interest rates, or what have you—is about to do so-and-so, investors should be wary of staking their futures on such claims. Second, the overwhelmingly frequent incorrectness of expert opinions should encourage investors to form strategies that do not depend solely on a single trend, outcome, or event.

In other words, the fundamentals of diversification, rebalancing, and disciplined, long-term commitment to a strategy are the best antidotes to fear- or greed-driven responses to the prognostications of experts.