Shift Operators and Stopping Times

158 Views Asked by At

I have some questions to the proof of the recurrence of the Brownian motion where we consider the stopping time $\tau_a=\text{inf}\{t\geq 0:B_t=a\}$.

In the proof we consider a shift-operator $\theta_t$. (I give the definition at the bottom.) Together with the stopping time we get these equalities:

$\theta_{\tau_{-a}}^{-1}\{\tau_a<\infty\}= \{\omega \in \Omega_{\tau_{-a}}:\tau_a(\theta_{\tau_{-a}(\omega)}(\omega))<\infty\}=\{\tau_{-a}<\infty\}\cap\{\tau_a(\tau_{-a})<\infty\}$

where $\Omega_{\tau_{-a}}= \{{ \tau_{-a}}<\infty\}$.

By the definition of $\Omega_{\tau_{-a}}$ I think that the second equality uses that $\tau_a(\theta_{\tau_{-a}(\omega)}(\omega))=\tau_a(\tau_{-a})$, right? If yes, why does this hold and if not, why is the second equality satisfied?

And why does the first equality hold?

Definition of the shift-operators of a Markov process $(X_t,\mathcal{F}_t,P_x)$:

It is a family of maps $\theta_t:\Omega \rightarrow \Omega$ such that

  1. $\theta_0=Id_{\Omega}$
  2. $\theta_t \theta_s = \theta_{t+s}$
  3. $ X_s(\theta_t) = X_{s+t}$