Abstract:This paper is concerned with two-player zero-sum linear-quadratic stochastic differential games in a regime switching model. The controlled inhomogeneous system coefficients depend on the underlying noises, so it is a non-Markovian regime switching model. Based on a new kind of multidimensional indefinite stochastic Riccati equation (SRE) and a multidimensional linear backward stochastic differential equation (BSDE) with unbounded coefficients, we can provide optimal feedback control-strategy pairs for the two players in a closed-loop form. The main contribution of this paper, which is important in its own right from the BSDE theory point of view, is to prove the existence and uniqueness of the new kind of multidimensional indefinite SRE. Interestingly, the components of the solution can take positive, zero and negative values simultaneously. We also obtain the corresponding optimal feedback control-strategy pairs for homogeneous systems under closed convex cone control constraints. Finally, these results are applied to portfolio selection problems with different short-selling prohibition constraints in a regime switching market with random coefficients.
Optimization and Control,Probability,Mathematical Finance,Portfolio Management,Risk Management
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to find the optimal feedback control strategy pair for a two - player zero - sum linear - quadratic stochastic differential game in non - Markovian regime - switching models. Specifically, the paper focuses on how, in a system affected by regime - switching, when the coefficients of the system depend not only on regime - switching but also on the underlying noise, to provide the optimal feedback control strategy pair for the two players in a closed - loop form by solving a new class of multi - dimensional indefinite stochastic Riccati equations (SRE) and a multi - dimensional linear backward stochastic differential equation (BSDE) with unbounded coefficients. In addition, the paper also explores the application of this type of problem in portfolio selection, especially in markets under different short - selling restriction conditions.
The main contribution of the paper lies in proving the existence and uniqueness of this new class of multi - dimensional indefinite SREs, and showing that the components of its solution can take positive, zero and negative values simultaneously. This characteristic is particularly important for solving problems in actual financial markets, because market parameters (such as interest rates, stock appreciation rates and volatilities) are affected by noise caused by various factors (such as global or local politics, economic growth rates, laws, military, corporate governance, etc.). Therefore, allowing market parameters to depend not only on the market state reflecting the macro - economic level but also on other stochastic resources is crucial for constructing more realistic financial models.