The Disposition Effect and Realization Preferences: a Direct Test

This paper develops a new laboratory test of the hypothesis that individual investors have an irrational preference for selling winning stocks vis-à-vis selling losing stocks. In the experiment, subjects invest in a security that bundles a risky asset, whose price evolves in near-continuous time, with a perpetual put option. Optimal behavior is characterized by an upper and a lower selling thresholds in the asset price space, thus producing a clear rational benchmark and eliminating known confounds. Subjects indeed tend to delay selling losers beyond the optimal point and to sell winners before reaching the optimal liquidation point. The median liquidation points imply the probability of realizing a gain conditional on a sale is 56% larger than optimal. Such behavior is shown to be consistent with a realization utility model and structural estimates reveal that the sensitivity of realization utility to gains and losses decreases faster than what is implied by canonical estimates of prospect-theoretic value functions. A direct estimate of the degree of diminishing sensitivity is an important input for behavioral finance theory as even qualitative results of realization utility models, such as whether investors voluntary realize losses or not, depend on the value of the sensitivity parameter.