Optimal Dividend Distribution Under Drawdown and Ratcheting Constraints on Dividend Rates

TL;DR

This paper studies an optimal dividend distribution model with drawdown constraints, extending Dybvig's ratcheting consumption problem, and derives a semi-explicit solution for the optimal dividend policy considering ruin risk and habit formation.

Contribution

It introduces a stochastic control framework for dividend policies with drawdown constraints and provides a semi-explicit solution to the associated nonlinear free-boundary problem.

Findings

Optimal dividend rate depends on surplus and historical maximum dividend rate.

Dividend policy involves multiple regimes based on surplus levels.

Maximum dividend rate is proportional to the historical peak of surplus.

Abstract

We consider the optimal dividend problem under a habit formation constraint that prevents the dividend rate to fall below a certain proportion of its historical maximum, the so-called drawdown constraint. This is an extension of the optimal Duesenberry's ratcheting consumption problem, studied by Dybvig (1995) [Review of Economic Studies 62(2), 287-313], in which consumption is assumed to be nondecreasing. Our problem differs from Dybvig's also in that the time of ruin could be finite in our setting, whereas ruin was impossible in Dybvig's work. We formulate our problem as a stochastic control problem with the objective of maximizing the expected discounted utility of the dividend stream until bankruptcy, in which risk preferences are embodied by power utility. We semi-explicitly solve the corresponding Hamilton-Jacobi-Bellman variational inequality, which is a nonlinear free-boundary problem. The optimal (excess) dividend rate $c^*_t$ - as a function of the company's current surplus $X_t$ and its historical running maximum of the (excess) dividend rate $z_t$ - is as follows: There are constants $0 < w_{\alpha} < w_0 < w^*$ such that (1) for $0 < X_t \le w_{\alpha} z_t$, it is optimal to pay dividends at the lowest rate $\alpha z_t$, (2) for $w_{\alpha} z_t < X_t < w_0 z_t$, it is optimal to distribute dividends at an intermediate rate $c^*_t \in (\alpha z_t, z_t)$, (3) for $w_0 z_t < X_t < w^* z_t$, it is optimal to distribute dividends at the historical peak rate $z_t$, (4) for $X_t > w^* z_t$, it is optimal to increase the dividend rate above $z_t$, and (5) it is optimal to increase $z_t$ via singular control as needed to keep $X_t \le w^* z_t$. Because, the maximum (excess) dividend rate will eventually be proportional to the running maximum of the surplus, "mountains will have to move" before we increase the dividend rate beyond its historical maximum.