I am given the following exercise, where I don't see how to do this

enter image description here

Here $B_t$ is a Brownian motion. The expectation value $E^x$ means that we take the expectation value w.r.t. a Brownian motion starting in $x$ instead of zero. $\theta_s$ maps a Brownian motion $B_t$ to $B_{t+s}.$

What I don't understand here is now how to use this Markov property.

I mean sure we have $P(\tau_{a+b}< \infty; \tau_a < \infty) = E(1_{\tau_{a+b}< \infty} ; \tau_a < \infty), $ but how can I apply then the Markov property to actually show this relation for the conditional probability?

If anything is unclear, please let me know.


Assume $B(0)=0$. On $\{\tau_a<\infty\}$,

$$ \{\omega:\tau_{a+b}<\infty\}=\{\omega: \tau_{a+b}-\tau_a<\infty\}=\{\omega: \tau_{a+b}\circ \theta_{\tau_a}<\infty\}. $$


$$ \mathbb{E}_0\left[1\{\tau_{a+b}<\infty \}\times 1\{\tau_a<\infty\}\right] $$ $$=\mathbb{E_0}\left[\mathbb{E_0}[1\{\tau_{a+b}<\infty\}\circ \theta_{\tau_a}\mid \mathcal{F}_{\tau_a}]\times1\{\tau_a<\infty\}\right] $$ $$=\mathbb{E_0}\left[\mathbb{E_{B(\tau_a)}}[1\{\tau_{a+b}<\infty\}]\times1\{\tau_a<\infty\}\right] $$ $$ =\mathbb{E_0}\left[\mathbb{E_0}[1\{\tau_{b}<\infty\}]\times1\{\tau_a<\infty\}\right]=P_0\{\tau_b<\infty\}P_0\{\tau_a<\infty\}, $$


$$ P_0\{\tau_{a+b}<\infty\mid \tau_a<\infty\} $$ $$=\mathbb{E}_0\left[1\{\tau_{a+b}<\infty \}\times 1\{\tau_a<\infty\}\right] / P_0\{\tau_a<\infty\}=P_0\{\tau_b<\infty\}. $$


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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