We estimate Cumulative Prospect Theory parameters for fund managers using monthly data from nearly 200,000 funds across all asset classes, styles, and regions (1990–2022) via hierarchical Bayesian methods. Managers exhibit behavioral traits documented in experimental psychology, but with significantly different parameter values than laboratory studies. We find statistically significant differences across asset classes, fund categories, and fund sizes, while manager tenure has limited impact. Prospect-theory values computed under our market-estimated parameters explain 60% more cross-sectional variation in ex-post Sharpe ratio than conventional laboratory parameters and deliver superior fit for Morningstar ratings, demonstrating that parameter choice is economically material for valuing fund return distributions.
Harris et al. (2026) studied this question.