4
$\begingroup$

Can someone point me into the right direction to calculate this one: $E(B^4_t)=3t^2$

I had tried using the following property with no luck:

$E(B^4_t)=E(B^2_tB^2_t)=E(\int B^2 dt )E(\int B^2 dt )=[E(\int B^2 dt )]^2=[\int E(B^2) dt]^2=[\int t dt]^2$

Any other suggestion will be appreciated. Thanks!

$\endgroup$
2
$\begingroup$

Apply Itô's Lemma to $W_t^4$: $$ \text{d}(W_t^4)=4W_t^3\text{d}W_t+6W_t^2\text{d}t$$

Integrate: $$ W_t^4=4\int_0^tW_s^3\text{d}W_s+6\int_0^tW_s^2\text{d}s$$

The first term is an Itô integral, which is by construction a martingale, with expectation $0$ hence: $$E[W_t^4]=6\int_0^tE[W_s^2]\text{d}s=6\int_0^ts\text{d}s=3t^2$$

$\endgroup$
6
$\begingroup$

At time $t$, Brownian Motion $B_t$ is simply a normal random variable $N(0,t)$.

The Moment Generating Function for a normal $N(\mu,\sigma^2)$ random variable is as follows: $$M(x) = exp(\mu x + \frac{1}{2}\sigma^2 x^2)$$ Furthermore, the fourth moment is given as the fourth derivative of this equation: $$M''''(x) = exp(\mu x + \frac{1}{2}\sigma^2 x^2)\Big( (\mu + \sigma^2x)^4 + 6\sigma^2(\mu + \sigma^2 x)^2 + 3\sigma^4 \Big)$$ So the expectation of $B_t^4$ is just the fourth moment, evaluated at $x=0$ (with parameters $\mu = 0$, $\sigma^2 = t$): $$E(B_t^4) = M''''(0) = 3\sigma^4 = 3t^2 $$

$\endgroup$
  • 1
    $\begingroup$ It is also possible to use Ito lemma with function $f(B_t)=B_t^{4}$, but this is an elegant approach as well. $\endgroup$ – Jan Sila Aug 1 '16 at 9:07
6
$\begingroup$

Other Way

$$\mathbb{E}\left[ \,{{e}^{iuB_t}} \right]=\exp \left( iu\,\mathbb{E}\left[ B_t \right]+\frac{1}{2}{{(\,iu\,)}^{2}}\operatorname{Var}(B_t\,) \right)={\exp \left( -\frac{1}{2}{{u}^{2}}t \right)}$$ We know $$\mathbb{E}\left[{{e}^{iuB_t}} \right]=E\left[1+iuB_t-\frac{1}{2\,!}{{u}^{2}}{{B_t}^{2}}-\frac{1}{3\,!}i{{u}^{3}}{{B_t}^{3}}+\frac{1}{4\,!}{{u}^{4}}{{B_t}^{4}}+\cdots \right]$$ therefore

$${\exp \left( -\frac{1}{2}{{u}^{2}}t \right)}=1+iu\mathbb{E}\left[ B_t \right]-\color{green}{\frac{1}{2!}{{u}^{2}}\mathbb{E}\left[ {{B_t}^{2}} \right]}-\frac{1}{3!}i{{u}^{3}}E\left[ {{B_t}^{3}}\right]+\color{red}{\frac{1}{4!}{{u}^{4}}\mathbb{E}\left[ {{B_t}^{4}} \right]}+\cdots \tag1$$ On the other hand $$\exp \left( -\frac{1}{2}{{u}^{2}}t \right)=1-\color{green}{\frac{1}{2}\,{{u}^{2}}t}+\color{red}{\frac{1}{2!}\left( \frac{1}{4}{{u}^{4}}{{t}^{2}} \right)}-\frac{1}{3 !}\left( \frac{1}{8}{{u}^{6}}{{t}^{3}} \right)+\frac{1}{4 !}\left( \frac{1}{16}{{u}^{8}}{{t}^{4}} \right)-\cdots\tag2 $$ $(1)$ and $(2)$ $$\frac{1}{4!}{{u}^{4}}\mathbb{E}\left[ {{B_t}^{4}} \right]=\frac{1}{2 !}\left( \frac{1}{4}{{u}^{4}}{{t}^{2}} \right)$$ thus $$\mathbb{E}\left[ {{B_t}^{4}} \right]=3t^2$$ Generally, we have

$$\left\{ \begin{align} & E\left[ {{B}^{2n+1}}(t) \right]=0\,\,\,\,\,\,\,\,\,\,\,\,\,\, \\ & \quad E\left[ {{B}^{2n}}(t) \right]=\frac{(2n)!}{{{2}^{n}}n\,!}\,{{t}^{n}} \\ \end{align} \right.$$

$\endgroup$

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.