Why Weathermen Get It Right More Often Than Analysts
Overconfidence works for some, but it can kill you in the market.
Scott Plous, professor of psychology at Wesleyan University and the author of The Psychology of Judgment and Decision Making, wrote, “No problem in judgment and decision making is more prevalent and more potentially catastrophic than overconfidence.”
John Nofsinger, associate professor of finance at Washington State University, a specialist in behavioral finance, and the author of The Psychology of Investing, had this to say:
People are overconfident. Psychologists have determined that overconfidence causes people to overestimate their knowledge, underestimate risks, and exaggerate their ability to control events.
Does overconfidence occur in investment decision making? Security selection is a difficult task. It is precisely this type of task at which people exhibit the greatest overconfidence.
In fact, studies have shown that the ratio of confidence to accuracy is an inverse one -- that is, a lesser confidence level tends to correlate with a higher degree of accuracy.
A study released 10 days ago by Tadeusz Tyszka and Piotr Zielonka of the Leon Kozminski Centre for Market Psychology at the Academy of Entrepreneurship and Management in Warsaw, titled "Expert Judgments: Financial Analysts vs. Weather Forecasters," asked two groups of experts to “predict corresponding events (the value of the Stock Exchange Index and the average temperature of the next month).”
Tyszka and Zielonka found that, “Although both groups of experts revealed the overconfidence effect, this effect was significantly higher among financial analysts than among the weather forecasters.”
In the group of financial analysts one-third of the participants succeeded in their forecast and in the group of weather forecasters approximately two-thirds of the participants succeeded in their forecast.
Further analysis showed that “Experts who assigned confidence estimates of 80% or higher were correct only in 45% of cases. As quoted by Korn & Laird (1999), a high level of overconfidence seems to be characteristic for finance analysts as well and most probably for many other kinds of experts."
In a slightly larger than usual nutshell, the crux of the issue was as follows:
Weather forecasters deal with the events of a periodic nature: seasons repeat cyclically. This world is partially predictable -- either through data-based climatological models or through theory-based Numerical Weather Prediction models. The weather forecasters are aware that they are working with a gross approximation of the underlying system and that in such an area the uncertainty must be taken into consideration. Presumably, the same awareness of uncertainty causes these experts to manifest a lower overconfidence effect than experts from the other domain. Indeed, as it was observed in the US by Murphy and Winkler (1977), meteorologists were exceptionally well calibrated in the sense that their confidence level was comparable to the actual accuracy of their predictions.
Financial analysts, on the other hand, have to deal with a world which seems to be completely unpredictable, where even weak probabilistic tendencies are rarely observed. Most of the modern financial theories presume that stock prices approximately follow random walk pattern (Cootner, 1964; Samuelson, 1965; Malkiel, 1996). [There] is some important evidence that stock prices' movements deviate from randomness (Lo, 1999; Shleifer, 2000) but this has a limited meaning for the practice of forecasting. In such an area no analytical formula of forecasting can be used.
“Paradoxically,” Tyszka and Zielonka discovered, “financial analysts, having less precise knowledge than the weather forecasters about the underlying system, can be more self-assured.”
...the differences observed between these two groups of experts can also be accounted for by motivational factors. There are perhaps, some features of these two professions that make financial analysts manifest a higher level of self-assurance than the weather forecasters. We found that the financial analysts not only expressed a higher level of overconfidence in relation to the weather forecasters, but also, in contrast to the weather forecasters, they did not decrease their self-evaluation after being motivated to think about reasons why a forecast might have failed. Thus, the financial analysts behaved as if they had to demonstrate the ability of a perfect forecast of the events in question. Professor Raymond Dacey from Idaho University suggested that this mechanism can pertain to the clients. Unless there are severe storms in the area (hurricanes, tornadoes, possible floods), most people who listen to weather forecasts are happy if the forecast is not hopelessly inaccurate. People who listen to financial analysts (i.e., investors) are very unhappy when the forecasts are only slightly inaccurate, i.e., when the reality is slightly below their expectations. Therefore, in order not to lose their clients, financial analysts are very sensitive about their reputation and better skilled than weather forecasters in formulating excuses for their errors.
Copyright 2011 Minyanville Media, Inc. All Rights Reserved.