Components of the strong Markov property

The fact that a process \(X\), adapted to some filtration \(F= (F_t)\), satisfies the strong Markov property can be defined as a combination of the conditional independence \(X_{\tau+ h}\amalg X_\tau F_\tau\) with the homogeneity \(P[X_{\tau+ h}\in\cdot\mid X_\tau]= \mu_h(X_\tau,\cdot)\) for a suitable set of probability kernels \(\mu_h\). Section 2 identifies two sets of conditions, each of which ensures the conditional independence to imply the time-homogeneous Markov property. In this case, the homogeneity and conditional independence are essentially equivalent. In a previous paper [Ann. Probab. 15, 213-240 (1987; Zbl 0614.60063)], the author defined the stronger, ``pairwise version of the homogeneity condition. Section 3 of the paper shows, conversely, that the pairwise homogeneity condition implies conditional independence and hence the time-homogeneous Markov property. Sufficient conditions for these results to be true are simple optimal times \(\tau<\infty\), and \(X\) to be adapted. Under the stronger hypotheses of progressive measurability, the corresponding conditions for more general optimal times \(\tau<\infty\) are considered and appropriate forms of the strong Markov property are entailed. For the entire collection see [Zbl 0895.00024].

Recommendations

Cited in

(3)

This page was built for publication: Components of the strong Markov property

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4218909)