Nonzero sum linear–quadratic stochastic differential games and backward–forward equations (Q4238565)

It is considered a nonzero sum linear-quadratic stochastic differential game. The usual methods used in the theory of stochastic games are not functional because the matrices defining the game are random. This paper is concerned with the application of the results on backward-forward equations. It is shown that if the matrices defining the drift are bounded by a constant that depends on the duration of the game \(T\), then there exists a Nash equilibrium point. There are proved two theorems concerning the existence of a solution for backward-forward stochastic differential equations and its implication for the existence for a Nash equilibrium point.