# Martingale conditional expectation [closed]

Let $(\Omega, \mathcal{F}, P)$ be a probability space and let $X$ be such that; $$P(X=(\frac{3}{2})^n) = 2^{-n}.$$

For $n \geq 1$, define $A_n = {\{X = (\frac{3}{2})^n)}\}$, and let $\mathcal{F_n}$ be the sigma field generated by $A_1, ..., A_n$.

Set $Z_n = E(X|\mathcal{F_n})$. It is easy to show this is a martingale with respect to the sigma field above.

Now, for $n \geq 1$ define $B_n = {\{X > (\frac{3}{2})^n)}\}$, then ${\{A_1, A_2, ..., A_n, B_n}\}$ forms a finite partition of $\Omega$ that generates $\mathcal{F_n}$.

I have been asked to show that $$Z_n = \sum_{k=1}^{n}(\frac{3}{2})^k\cdot I_{A_k} + 2(\frac{3}{2})^{n+1}\cdot I_{B_n}$$

Where $I_{A_k}, I_{B_n}$ are indicator functions.

I don't know where to start here. I don't really understand the variable $Z_n$, which is conditioned on the sigma field generated by ${\{A_1, A_2, ..., A_n, B_n}\}$.

Any help would be appreciated, thank you.

## closed as off-topic by BCLC, Namaste, Leucippus, JonMark Perry, ShaileshJun 1 '18 at 1:42

This question appears to be off-topic. The users who voted to close gave this specific reason:

• "This question is missing context or other details: Please improve the question by providing additional context, which ideally includes your thoughts on the problem and any attempts you have made to solve it. This information helps others identify where you have difficulties and helps them write answers appropriate to your experience level." – Namaste, Leucippus, JonMark Perry, Shailesh
If this question can be reworded to fit the rules in the help center, please edit the question.

• The sigma field generated by ${\{A_1, A_2, ..., A_n, B_n}\}$ is $\mathcal{F}_n$ since $B_n = (\bigcup_{i=1}^nA_i)^{\text{c}}$ – clark Jan 15 '18 at 18:15
• @clark Ok thank you. I don't see how this helps solve the problem, though? – the man Jan 15 '18 at 18:22
• Since $\cup A_i \cup B_n$ forms partition of the sigma field $\mathcal{F}_n$ is suffices to verify the desired equality for each of the events. To show that for each event you can use $E(X|A)=E(X1_A)/\mathbb{P}(A)$ where $A$ is an event. – clark Jan 15 '18 at 18:23
• @clark So $Z_n = E(X|A_1) + ... + E(X|A_n) + E(X|B_n)$? – the man Jan 15 '18 at 18:31
• Not quite, notice that your RHS, as is, is a number not a random variable, we should write it as $E(X|A_1)1_{A_1}+\ldots$ . More specifically, on the event $A_i$ $Z_n= E(X|F_n)= E(X|A_i)$, which you can compute as mentioned in the above comments. – clark Jan 15 '18 at 18:38

To find $Z_n$, it suffices to calculate $E(X|A_i)$ and $E(X|B_n)$. We can calculate $E(X|B_n)$, and, the others can be computed similarly. We set $B_{n,i}=\{X=\left (\frac{3}{2}\right)^{n+i}\}$ for $i\geq 1$.
• If $T =$ min${\{n \geq 1 : Z_n = X}\}$ is a stopping time, could you possibly give me any hints on how to show that $T = \sum_{n=1}^{\infty} n \cdot I_{A_n}?$ My thoughts are that $X = E[X|\mathcal{F_n}]$, so $X$ is independent of the sigma field, but this isn't getting me anywhere! – the man Jan 15 '18 at 20:43