Repeated Moral Hazard in Multi-Stage R&D

This paper studies a continuous-time principal-agent problem in which a principal hires an agent to do a multi-stage R&D project. The agent’s effort is unobservable and the transition from one stage to the next is modeled by a jump process with a constant Poisson arrival rate. We characterize the optimal dynamic contract that solves the repeated moral-hazard problem. In it, the agent’s payment decreases over time in case of failure and jumps up to a higher level after each success. We also provide an implementation of the optimal contract, in which a primary component of the agent’s compensation is a risky security and the principal lets the agent choose the consumption and effort levels. This implementation gives a theoretical justification for the wide-spread use of stock-based compensation by firms that rely on R&D.