Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top


Let $\left(X_t\right)_{t \in I}$ ($I\subseteq\mathbb R$) be an $E$-valued stochastic process ($E$ being a Polish space with the Borel $\sigma$-algebra $\mathcal{B}\left(E\right)$) equipped with the filtration generated by $X$, $\left(\mathcal F_t\right)_{t\in I}=\left(\sigma\left(X_s\space:\space s\leq t\right)\right)_{t\in I}$. Suppose $E$ is countable.


Why is it the case (as claimed in Klenke, Remark 17.2) that if for all $n\in\mathbb N$, all $s_1<\cdots<s_n<t$ and all $i_1,\dots,i_n,i\in E$ with $\mathbb{P}\left[X_{s_1}=i_1,\dots,X_{s_n}=i_n\right]>0$ we have

$$\mathbb{P}\left[\left.X_t=i\space\right|\space X_{s_1}=i_1,\dots,X_{s_n}=i_n\right]=\mathbb{P}\left[\left.X_t=i\space\right|\space X_{s_n}=i_n\right]$$

then the Markov property applies, namely

$$\forall s\leq t\in I\bullet\mathbb{P}\left[\left.X_t\in A\space\right|\space\mathcal{F}_s\right]=\mathbb{P}\left[\left.X_t\in A\space\right|\space X_s\right]$$

share|cite|improve this question
up vote 2 down vote accepted

Fix $s\leq t$. We can consider that $A=\{i_k\}$, using additivity and the fact that the space is countable. We have to check that $$\forall B\in\mathcal F_s,\quad\int_B\chi_{\{X_t=i_k\}}dP=\int_BE[\chi_{\{X_t=i_k\}}\mid X_s]dP.$$ To see that, note that the equality is true on the finite intersections of sets of the form $\{X_u\in E'\}$, $E'\subset E$ and $u\leq s$, then show that the sets which satisfy the displayed equality is a $\lambda$-system. You will need Dynkin's theorem.

share|cite|improve this answer
Thank you, Davide. That's a terrific idea, though i'd go about it a little differently: Define $f:=\mathbb{P}\left[\left.X_t=i\space\right|\space\mathcal{F}_s\right]$. We need to show: $\mathbb{P}\left[\left.X_t=i\space\right|\space X_s\right]=f\space\space a.s.$ In order to do so, we need to prove two propositions: (a) $f$ is $\sigma\left(X_s\right)$-measurable, (b) $\forall B\in \sigma\left(X_s\right)\bullet\int_B f d\mathbb{P}=\mathbb{P}\left(\{X_t=i\}\cap B\right)$. Since $\sigma\left(X_s\right)\subseteq\mathcal{F}_s$, (b) is immediate. [to be continued...] – Evan Aad Aug 9 '12 at 20:55
[... continued] To show (a) it suffices to consider $B$s that are finite intersections of sets of the form $\{X_r=j\}$, $r\leq s$, $j\in E$, since they constitue a $\pi$-system that generates $\mathcal{F}_s$. – Evan Aad Aug 9 '12 at 20:56

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.