scientific article; zbMATH DE number 2069455
The authors consider stochastic optimal control problems involving a singular control parameter \(\xi\) and discretionary stopping times \(\tau\) at reflecting or absorbing states. The cost function is of the form \[ J_x (\xi,\tau)={\mathbf E} \left\{\int^\tau_0e^{-\alpha t}[\lambda x^2_t\,dt+ d\widehat \xi_t]+e^{-\alpha \tau} h(x_t)] \right\}, \] where the state process is \(x_t=x+W_t+\xi_t\), \(t\geq 0\), with a constant initial position \(x_0=x\), \(W_t\), \(t\geq 0\), is a standard Brownian motion, \(\xi_t\), \(t\geq 0\), is left continuous, adapted of bounded variation on compact intervals, and \(\tau \in\Gamma\), the set of stopping times. Finally \(h\) is a given \(C^2\)-function. The goal is to find the optimal control \((\xi^*, \tau^*)\), that is, a pair consisting of a process \(\xi^*\) and a stopping time \(\tau^*\) such that \[ J_x(\xi^*,\tau^*) \inf_{\xi,\tau}J_x(\xi,\tau). \] This is an interesting problem which was treated for the first time by \textit{M. H. A. Davis} and \textit{M. Zervos} [Ann. Appl. Probab. 4, No. 1, 226--240 (1994; Zbl 0796.93111)], where the existence and the optimal control is established in the case \(h(x)=x^2\). Inspired by the study of Davis and Zervos, the authors of the paper under review present a model which can be taken as an extention. By using variational inequalities techniques they prove the existence of optimal control and give explicit descriptions of the value function and the strategies. The involvement of a reflecting and an absorbing barrier allows them to tell what action is optimal. The final comment is about the text, which clearly has not be edited appropriately. It is a pity that even names of well-established scientists are given incorrectly (e.g., we meet several times Doléans-Made and Zerros, instead of the correct Deléans-Dade and Zervos).
- scientific article; zbMATH DE number 2064421
- Singular stochastic control and optimal stopping
- Connections Between Optimal Stopping and Singular Stochastic Control II. Reflected Follower Problems
- A class of singular stochastic controls with stopping time
- Problems of singular stochastic control with stopping, drift and diffusion
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4463387)
