Journal of Optimization Theory and Applications Management Science and Operations Research
Optimal Stopping Problems for Maxima of Two Ornstein–Uhlenbeck Processes
Abstract
Abstract We present a semi-analytical solution to the optimal stopping problem of the maximisation of the expected reward representing the difference between the maximum $$\max \{ X^1_{\tau }, X^2_{\tau } \}$$ max { X τ 1 , X τ 2 } and the positive costs $$c \tau $$ c τ over all stopping times $$\tau $$ τ with finite expectation of two mean-reverting Ornstein–Uhlenbeck (O–U) processes $$X^i = (X^i_t)_{t \ge 0}$$ X i = ( X t i ) t ≥ 0 , for $$i = 1, 2$$ i = 1 , 2 , driven by constantly correlated standard Brownian motions. The method of proof consists of embedding the initial problem into an optimal stopping problem for the two-dimensional diffusion O–U process $$(X^1 - X^2, X^1 + X^2)$$ ( X 1 - X 2 , X 1 + X 2 ) . The optimal stopping time is shown to be the first time at which the difference process $$X^1 - X^2$$ X 1 - X 2 exits a region formed by continuous stochastic boundaries depending on the current value of the sum process $$X^1 + X^2$$ X 1 + X 2 . These boundaries are characterised as a unique solution to the associated system of (fixed-point) nonlinear Fredholm-type integral equations. The result is applied to determine the best constants in the sharp forms of the corresponding maximal inequalities.