1
$\begingroup$

I am currently reading these notes and in particular try to understand the proof of theorem 5.12.

Actually, my only questions are about the last paragraph. He claims that $T_n$ is a uniformly decreasing sequence convering to $T$ (so far so good), but then he writes $\mathcal{F}^+(T_n) \subset \mathcal{F}^+(T)$. This does not make sense to me, if we have a decreasing sequence we should have $\mathcal{F}^+(T) \subset \mathcal{F}^+(T_n)$ and I think this is also what he uses then in the independence argument. So we know that $W_{s+T_n} - W_{T_n}$ is independent from $\mathcal{F}^+(T).$

My second problem is, I don't see how all this shows that $Y_t:=W_{t+T}-W_T$ is a Brownian motion. For a Brownian motion, we need to show that continuity of the path holds (okay I see how this could go) and that $Y_t-Y_s \tilde \ \mathscr{N}(0,t-s)$ is also okay to me, but I just don't see how he tries to argue that $Y_t-Y_s$ is independent from $Y_u-Y_v$ where $u<v<s<t$ and I think he somehow mixed that up, as he is talking about the independence of $Y_t$ from $Y_s$ which is not what Brownian motion is about.

Can anybody clarify this point?- If anything is unclear, please let me know.

$\endgroup$

1 Answer 1

1
$\begingroup$

[Correction: Proof of Theorem 5.13]

It looks like there is a typo in the text, and the claim $\mathcal F^+(T_n)\subset\mathcal F^+(T)$ should read $\mathcal F^+(T)\subset\mathcal F^+(T_n)$. This implies that $\mathcal F^+(T)$ is independent of $\{W_{s+T_n}-W_{T_n}: s\ge 0\}$ for each $n$, hence independent of the limit process $\{W_{s+T}-W_{T}: s\ge 0\}$.

What is $Y_t$? (I don't see it in Leiner's proof of Theorem 5.13.)

$\endgroup$
2
  • $\begingroup$ sorry, I used the short notation $Y(t)=W_{(t+T)}-W_T.$ My problem is therefore in understanding how this exactly implies the independence of the increments. And thus I am talking about the increments $Y(t)-Y(s)$, as we want to show that $Y(t)$ is a Brownian motion. $\endgroup$ Nov 1, 2015 at 18:13
  • $\begingroup$ are you still considering this problem? $\endgroup$ Nov 2, 2015 at 21:05

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .