"Miscalibration is a form of overconfidence examined in both psychology and economics. Although it is often analyzed in lab experiments, there is scant evidence about the effects of miscalibration in practice. We test whether top corporate executives are miscalibrated, and study the determinants...

Using a unique 10-year panel that includes more than 13,300 expected stock market return probability distributions, we find that executives are severely miscalibrated, producing distributions that are too narrow: realized market returns are within the executives' 80% confidence intervals only...