Let $\{B_t:t\geq0\}$ be a Brownian motion defined on the probability space $(\Omega, \mathcal{F}, \mathbb{P})$. Let $\widehat{B}_0 = B_0$ and set \begin{align} \widehat{B}_t = B_t - \int_0^t \frac{B_s}{s} ds, \qquad \text{for }t>0. \end{align} Now, I want to use the fact that $\widehat{B}$ is a Brownian motion to show that for all $t>0$, one has that $B_t$ is independent of $\widehat{B}_s$ for all $0\leq s < t$.

In order to do so, I first want to show that the function $s \to \frac{B_s}{s}$ is Lebesgue integrable on $[0,t)$. Therefore, I have to show that \begin{align} \int_0^t \mathbb{E}\big[\frac{|B_s|}{s}\big]\ ds < \infty. \end{align} However, I face some difficulties with showing the above inequality and hence to conclude the upper independence statement.


By the scaling property, we have $B_s \stackrel{d}{=} \sqrt{s} B_1$ which implies

$$\mathbb{E}(|B_s|) = \sqrt{s} \mathbb{E}(|B_1|).$$

Note that $\mathbb{E}(|B_1|)$ is finite since $B_1$ is Gaussian. Consequently,

$$\mathbb{E} \left( \int_0^t \frac{|B_s|}{s} \, ds \right) = \int_0^t \frac{\mathbb{E}(|B_s|)}{s} \, ds = \mathbb{E}(|B_1|) \int_0^t \frac{1}{\sqrt{s}} \, ds < \infty.$$


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.