Expected value and Variance of a stochastic time integral of a deterministic variable (Standard Brownian motion)

497 Views Asked by At

Given a Standard Brownian motion $(B_t)_{t\in\mathbf{R}_{+}}$, define:

$$E(e^{\int_0^tudB_u})=?$$ $$ Var(e^{\int_0^tudB_u})=?$$


I started off assuming (!) that $X_t=$ $\int_0^tudB_u \sim N(\mu,\sigma^2)$; so that I could apply:

$E(e^{X_t})=\int_Re^{X_t}.(pdf) dx = e^{(\mu+\frac[2\sigma^2)}$

And went on to find $\mu$ and $\sigma$ of $X_t$:

1) $E(X_t)=E(\int_0^tudB_u)$

2)Since $X_t$ has no drift, and only a diffusion term, it should be a Martingale assuming (!) that

$E(X_t)^2<\infty$ (is bounded)

Which is since:

$E(X_t)=E(\int_0^tudB_u)^2 = E(\int_0^tu^2dt) = (\frac{t^3}{3}) < \infty$

So since it is a Martingale and $B_0=0$:

3)$E(X_t)=E(\int_0^tudB_u)=0$

4) $Var(X_t)=E(X_t^2) - [E(X_t)]^2$$= E(\int_0^tudB_u)^2-0$ $=(\frac{t^3}{3})$

5)At this point I can state:

$$X_t=\int_0^tudB_u \sim N(0,\frac{t^3}{3})$$

and apply

$E(e^{\int_0^tudB_u})=e^{0+\frac{t^3}{6}}$

6) For the variance:

$Var(e^{X_t})=E(e^{2N(0; \frac{t^3}{3}})-[E(e^{X_t})]^2 = e^{0+ \frac{1}{2} \frac{t^3}{3}} -[e^{0+\frac{t^3}{6}}]^2$


PS: Fixed thanks to comments