Overconfidence impacts important corporate policies
Whether you are on Wall Street or Main Street, it is tough to forecast the stock market. Most of us know it is tough and would offer a wide range of possible outcomes if pressed for a prediction.
Not true for financial executives. Chief financial officers (CFOs), who are constantly asked to explain movements in their own company's stock prices, are astonishingly overconfident in their ability to get the forecast correct.
A new study finds the range of outcomes that CFOs provide for one-year stock market forecasts are unrealistically narrow, according to research by professors from Ohio State University's Fisher College of Business and Duke University's Fuqua School of Business.
Using 10 years of quarterly CFO surveys conducted at Duke, Itzhak Ben-David, associate professor of finance at Ohio State, and professors John R. Graham and Campbell R. Harvey of Duke looked at more than 13,300 forecasts of the S&P 500 made by finance professionals. They found the range of possible outcomes that CFOs provide for where the stock market will be in one year to be unrealistically narrow.
"These executives are too sure of their ability to predict the future," Ben-David said. "And we found this overconfidence was linked to decision-making at their firms, so there is a real-world impact."
The study is published in the November issue of the Quarterly Journal of Economics and can be accessed at the following link.
"We asked them to tell us the range of outcomes that would lead them to be right 80 percent of the time," said Harvey. "However, the ranges were way too narrow. Instead of working 80 percent of the time, they barely got more than 1 in 3 correct."
"For example, they might say next year's return should fall in the range of -20 to +20 percent. If the market return is 10 percent, they are correct. But that is not what we found. They would give really tight ranges, like 11 to 18 percent. If the market return is 10 percent, they are wrong."
"This is a classic trait of severe overconfidence," Harvey added. "The magnitude of the miscalibration is astounding and led us to investigate whether this overconfidence impacted the way they ran their businesses."
Graham said executives were also asked to predict the range of possible outcomes for their firm's own investment projects, "and again they proved to be extremely overconfident. It is well known that companies' investment outcomes are very volatile. Surprisingly, CFOs seem oblivious to this risk."
As a result, Graham said, "We found their overconfidence was linked to decision-making at their firms. CFOs who are more overconfident about the stock market were more aggressive in their investing and borrowing, which means their overconfidence appears to filter into corporate policy."
A number of psychology laboratory experiments show that participants in these studies, often college students, also tend to be prone to this type of overconfidence.
"These studies are often critiqued because the subjects are college students and are not representative of experienced senior executives," Graham said. "We show that these seasoned financial executives also suffer from significant miscalibration."