Convergence of values in optimal stopping and convergence of optimal stopping times (Q2461967)

In this paper the problem of stability of values of optimal stopping, and of optimal stopping times, under approximations of the process \(X\) are investigated [see \textit{D. Lamberton} and \textit{G. Pagès}, Sur l'approximation des réduites. Ann. Inst. Henri Poincaré, Probab. Stat. 26, No. 2, 331--355 (1990; Zbl 0704.60042)]. Let us consider a sequence \((X^n)_n\) of càdlàg processes which converges in probability to a càdlàg process \(X\). For all \(n\), let us denote by \({\mathcal F}^n\) the natural filtration of \(X^n\) and by \({\mathcal T}^n_T\) the set of \({\mathcal F}^n\) stopping times bounded by \(T\). Denote the values in the optimal stopping problem \(\Gamma^n(T)= \sup_{\tau\in {\mathcal T}^n_T}{\mathbf E}[\gamma(\tau,X_\tau^n)]\). In this paper first are given conditions under which \(\Gamma^n(T)\) converges to \(\Gamma(T)\), and second, when it is possible to find a sequence \((\tau_n)\) of optimal stopping times w.r.t. the \(X^n\), to give further conditions under which the sequence \((\tau_n)\) converges to an optimal stopping time w.r.t \(X\). The results are obtained under hypothesis of inclusion of filtrations or convergence of filtrations.