Optimal dividend payout with path-dependent drawdown constraint

Abstract

This paper studies an optimal dividend payout problem with drawdown constraint in a Brownian motion model, where the dividend payout rate must be no less than a fixed proportion of its historical running maximum. It is a stochastic control problem, where the admissible control depends on its past values, thus is path-dependent. The related Hamilton-Jacobi-Bellman equation turns out to be a new type of two-dimensional variational inequality with gradient constraint, which has only been studied by viscosity solution technique in the literature. In this paper, we use delicate PDE methods to obtain a strong solution. Different from the viscosity solution, based on our solution, we succeed in deriving an optimal feedback payout strategy, which is expressed in terms of two free boundaries and the running maximum surplus process. Furthermore, we have obtained many properties of the value function and the free boundaries such as the boundedness, continuity etc. Numerical examples are presented as well to verify our theoretical results and give some new but not proved financial insights.