STOCK OPTION EXERCISE AND GIFT EXCHANGE RELATIONSHIPS: EVIDENCE FOR A LARGE US COMPANY


We investigate gift exchange relationships in real jobs, making use of a field quasi-experiment associated
with the exercise of stock options for roughly 4500 managers in a large public company. In this company,
option grants are set equally for all employees within occupational categories, and financial markets
set the price at which the options are ultimately exercised. We assert that the considerable variation
that we observe across employees and over time in profits from those sales is beyond the control of
the individual employee and can be thought of as effectively randomized. We also assert that employees
perceive the profit they receive from exercising these options at least in part as the equivalent of a
gift: Higher profits in turn cause them to reciprocate with better job performance in the subsequent
period. We find significant and economically meaningful positive relationships between the variation
in profit per share of the options sold and standard measures of subsequent job performance for individual
employees. These effects exist in real jobs and persist over long periods, extending previous studies.
Non-parametric and parametric fixed effects models, other controls for sample heterogeneity, and
alternative specifications address possible concerns about the randomization assumption and associated
statistical issues.