Paper Losses, Real Decisions: The Disposition Effect in US Corporate Officers

Abstract

This paper extends the literature on the disposition effect in US corporate officers through analysis of their open market trading. Using the Layline insider trading dataset with all SEC Form 4 filings from January 2003 to December 2022, this study tests which of the three leading explanations (prospect theory, realization utility, or belief revisions) best characterizes the observed selling patterns, and whether the disposition effect is associated with firm level stock returns. A linear probability model with officer-firm pair and year fixed effects estimates that officers are 2.6 percentage points more likely to sell their firm’s stock when in a month where they hold an unrealized gain than unrealized loss, which is strong evidence of the disposition effect in the sample. The selling schedule and bin indicator regressions find evidence of a strong positive relationship between unrealized gains and sell probability, while sell probability remains roughly flat across the loss domain. A regression discontinuity design finds no statistically significant jump in sell probability at zero unrealized return. Collectively, these patterns observed in these tests align more closely with the predictions of realization utility than competing theories, although the evidence cannot conclusively determine the driving mechanism. In the second part of the analysis, neither the Fama-MacBeth nor the quintile portfolio sorts show evidence that firms level disposition scores tend to be associated with future stock returns. However, this paper cannot rule out the possibility that disposition prone behavior can affect firm outcomes through channels not captured by these tests.