Example of non measurable stochastic process

measurable-functionsstochastic-analysisstochastic-calculusstochastic-processes

I know that a Stochastic process $X=(\Omega,\mathcal{F},\{\mathcal{F}_t\}_{t\in T},\{X_t\}_{t\in T}),P)$ is said to be measurable if the map:
\begin{equation}
G:(T\times\Omega,\mathcal{B}(T)\times\mathcal{F})\to(E,\mathcal{B}(E))
\end{equation}

such that $(t,\omega)\mapsto X_t(w)$ is a measurable map. ($\mathcal{B}(E)$ denotes borelian of $E$ )

I'm looking for an exampke of a non measurable stochastic process but I still struggle on finding it. Can someone give me an hint?

Best Answer

Take $\mathcal{F}=\{\emptyset,\Omega\}$. Then any non-deterministic process is non measurable. For a more interesting example, consider a a family of i.i.d. random variables $(X_t:t\in[0,1])$ s.t. $\mathsf{E}X_0=0$ and $\mathsf{E}X_0^2=1$. It is easy to see that if $\{X_t\}$ is measurable, then for any $0\le a<b\le 1$, $$ \mathsf{E}\left[\int_{[a,b]}X_t\, dt\right]^2=\int_{[a,b]^2}\mathsf{E}[X_sX_t]\, dsdt=0 $$ (by Fubini), i.e., $X_t=0$ for almost all $\omega\in\Omega$ and $t\in[0,1]$. Thus, $$ \mathsf{E}\left[\int_0^1X_t^2\,dt\right]=0\ne 1=\int_0^1 \mathsf{E}X_t^2\, dt. $$