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

This is a problem from the book "Markov chains". Let $(X_n)_{n\ge 0}$ be a Markov chain on $I$ and let $A$ be an absorbing set in $I$. Set $$T=\inf\{n\ge 0 : X_n \in A\}$$ and $$h_i = \mathbb{P}_i(X_n \in A \text{ for some } n \ge 0) = \mathbb{P}_i(T<\infty)$$ Show that $M_n = h(X_n)$ is a martingale.

This is my solution: \begin{align*} E[M_{n+1} \mid \mathcal{F}_n] = E[h(X_{n+1}) \mid \mathcal{F}_n] &= \sum_{j \in I} p_{X_n,j}h(j) \\ &= h(X_n) \text{ (First step analysis)}\\ &= M_n \end{align*} Am I correct or not? I'm afraid I missed something. Thanks!

share|cite|improve this question
up vote 1 down vote accepted

Some notions first. To show that $h(X_n)$ is a martingale it is just sufficient to check whether $$ \mathsf E[h(X_{n+1})|X_n] = h(X_n) \iff \mathsf E_x[h(X_1)] = h(x) \iff h(x) = Ph(x) \tag{1} $$ since $(X_n)$ is a Markov process. Here $P$ is a transition operator $$ Pf(x):=\mathsf E_x[f(X_1)] = \sum_{y\in I}p_{x,y}f(y). $$ and all functions satisfying $Pf=f$ are called harmonic. As a result, $(1)$ shows that $f$ is harmonic if and only if $f(X_n)$ is a martingale.

So, your step is right where you get $$ \mathsf E[M_{n+1}|\mathscr F_n] = \sum_{y\in I}p_{X_n,y}h(y) $$ is correct. To show that the latter is $h(X_n)$, just use the formula $$ h(x) = 1_A(x)+1_{A^c}(x)Ph(x). \tag{2} $$ In general, $(2)$ describes the hitting probability for any set $A$ and then $h(X_n)$ is a supermartingale in general. It is a martingale provided $A$ is absorbing.

share|cite|improve this answer

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.