Classic financial agency theory recommends compensation through stock options rather than shares to induce risk neutrality in otherwise risk averse agents. In an experiment, we find that subjects acting as executives do also take risks that are excessive from the perspective of shareholders if compensated through options. Compensation through restricted company stock reduces the uptake of excessive risks. Even under stock-ownership, however, experimental executives continue to take excessive risks—a result that cannot be accounted for by classic incentive theory. We develop a basic model in which such risk-taking behavior is explained based on a richer array of risk attitudes derived from Prospect Theory. We use the model to derive hypotheses on what may be driving excessive risk taking in the experiment. Testing those hypotheses, we find that most of them are indeed borne out by the data. We thus conclude that a prospect-theory-based model is more apt at explaining risk attitudes under different compensation regimes than traditional principal-agent models grounded in expected utility theory.
prospect theory expected utility theory risk attitude executive compensation reference dependence experimental finance