Monday, November 9, 2015

Overconfidence: The Ancient Evil

Here's how Ulrike Malmendier and I started our short introduction to a three-paper symposium in the Fall 2015 issue of the Journal of Economic Perspectives on the subject of overconfidence.

Economists have been concerned about issues of overconfidence at least since Adam Smith (1776, Book I, Chapter X), who wrote in The Wealth of Nations: “The over-weening conceit which the greater part of men have of their own abilities, is an ancient evil remarked by the philosophers and moralists of all ages.” Titans of modern economics have had similar reactions to the “ancient evil.” Daniel Kahneman recently told an interviewer that if he had a magic wand that could eliminate one human bias, he would do away with overconfidence. As Shariatmadari (2015) reports: “Not even he [Kahneman] believes that the various flaws that bedevil decision-making can be successfully corrected. The most damaging of these is overconfidence: the kind of optimism that leads governments to believe that wars are quickly winnable and capital projects will come in on budget despite statistics predicting exactly the opposite.” Kahneman argues that overconfidence “is built so deeply into the structure of the mind that you couldn’t change it without changing many other things.” 
Evidence concerning the prevalence of overconfidence is widespread and robust. Some of the results have even become fairly well-known in popular culture, like the findings that most drivers believe they are safer than a typical driver, or that the unskilled tend to overestimate their abilities. The finding about driver overconfidence stems from a Svenson (1981) study, a lab experiment using undergraduate students as subjects, which found that 83 percent of American subjects believed that they were in the top 30 percent in terms of driving safety. The finding about overestimation of ability comes from a Kruger and Dunning (1999) study, which reports: “[P]articipants scoring in the bottom quartile on tests of humor, grammar, and logic grossly overestimated their test performance and ability. Although their test scores put them in the 12th percentile, they estimated themselves to be in the 62nd.” Overconfidence on both sides of a conflict may be linked to the willingness to fight a war (Wrangham 1999; Johnson 2004) or to the conditions that lead a strike to occur (Neale and Bazerman 1985). 
The three papers in the symposium ask about the economics of overconfident consumers, overconfident CEOs, and overconfident investors.

Michael Grubb points out in "Overconfident Consumers in theMarketplace" that if firms know that a substantial share of consumers are overconfident, they will set prices and contract terms accordingly. For example, consumers may be overconfident about future actions. Sure, they will send in that mail-in rebate. Sure, they won't have a car accident, so choosing car insurance with a REALLY high deductible makes sense. Sure, they will subscribe to something and put it on "auto-renew," because they will remember to cancel it when ready. Grubb points out that in some of these situations, competition by firms to take advantage of the overconfident can mean great deals for those who are not overconfidence. Conversely, steps by government to protect the overconfident from themselves can, in some cases, lead to higher costs for others.

lrike Malmendier and Geoffrey Tate discuss "Behavioral CEOs: The Role of Managerial Overconfidence." They suggest a number of ways of measuring the overconfidence of CEOs. One example is based on the insight that CEOs should want to cash in their stock options when they have a chance for a good gain, because it gives them a chance to diversify their wealth by investing it somewhere else. However, an overconfident CEO will holds stock options right up the expiration date before cashing them in, out of a belief that the company is doing better than the market recognizes and for that reason the stock option will keep rising in value. Using measures of when CEOs cash in their stock options, along with estimates of how much they would receive for cashing in their options and various back-of-the-envelope estimates for risk and diversification, they find that about 40% of  CEOs qualify as overconfident. It turns out that those CEOs are more likely to use the company's cash, or borrowed money, to make big investments and acquisitions. They also point out that some companies facing the need for a tough transition might prefer an overconfident CEO to manage the transition. Indeed, you can offer an overconfidence CEO less in stock options, because the overconfident CEO will tend to believe that those options will be worth more.

Kent Daniel and David Hirshleifer study "Overconfident Investors, Predictable Returns, and Excessive Trading." They argue that the volume of trading in financial markets is very high, and that there are a number of ways that have been well-known for decades in which stock market returns are somewhat predictable. They suggest that overconfidence among investors helps to explain the eagerness to trade, and the willingness to believe that you can choose a financial adviser who will make you rich. They also suggest links between overconfidence and predictabilities in stock market pricing related to momentum in stock prices, firms with a low ratio of book-too-market value, small firms outperforming larger firms, and others. As they point out, overconfidence is reinforced by "self-attribution" bias, which is the common mental posture that whatever goes well was due to your own skill, while whatever goes poorly is too to bad luck.

Overconconfidence shouldn't be treated as a giant all-purpose explanation for less-than-rational behavior by consumers, CEOs, and investors. There are lots of ways, in different situations and contexts, that people make less-than-rational decisions. But overconfidence plays a central role in the persistence of lots of less-than-rational behavior. If people were actually rational, most of us (myself very much included) would be pretty humble about our abilities to assess situations, draw inferences, and make decisions. Overconfidence explains why most of us, despite being drenched in the cold rain of reality on a regular basis, still manage believe so strongly in our own points of view and our own ways of making decisions.

And one can indeed argue, with Adam Smith and Daniel Kahneman, that a world in which overconfidence was substantially diminished would be a more productive and pleasant place.