Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

For a standard one-dimensional Brownian motion $W(t)$, calculate:

$$E\bigg[\Big(\frac{1}{T}\int\limits_0^TW_t\, dt\Big)^2\bigg]$$

Note: I am not able to figure out how to approach this problem. All i can think of is that the term $\frac{1}{T}\int\limits_0^TW_t\,dt$ is like 'average'. But not sure how to proceed ahead. I'm relatively new to Brownian motion. I tried searching the forum for some hints..but could not find one. I will really appreciate if you could please guide me in the right direction. Thanks!

share|improve this question

3 Answers 3

up vote 5 down vote accepted

If you recall that $\mathrm d(t W_t) = W_t\mathrm dt + t\mathrm dW_t$ you can write your integral in the other form $$ \int_0^T W_t\mathrm dt = TW_T - \int_0^Tt\mathrm dW_t. $$ If we forget about the factor $\frac1T$ as it does not affect the derivation much, we obtain $$ \mathsf E\left(\int_0^T W_t\mathrm dt\right)^2 = \mathsf E\left[T^2W_T^2\right] - 2T\mathsf E\left[W_T\int_0^T t\mathrm dW_t\right]+\mathsf E\left(\int_0^T t\mathrm dW_t\right)^2 $$ $$ = T^3- 2T\int_0^Tt\mathrm dt+\int_0^Tt^2\mathrm dt $$ where we applied the Ito isometry a couple of times. Hopefully, the most hard part now is done an you can finish the derivations.

share|improve this answer
    
Thanks a lot!!. Got the answer as T/3. –  Prakhar Mehrotra Dec 4 '12 at 7:34
    
@PrakharMehrotra: you are welcome –  S.D. Dec 4 '12 at 7:35

Another approach would be to show that the random variable

$$\omega \mapsto \int_0^T W_t(\omega) \, dt$$

is a centered normal random variable with variance $\int_0^T (s-T)^2 \, ds=\frac{T^3}{3}$. You can find a proof here. (The proof is a bit lenghthy, but you don't need Itô-Calculus to prove it.)

share|improve this answer
    
The variance is not T. –  Did Dec 4 '12 at 10:34
    
Sorry, I corrected it. –  saz Dec 4 '12 at 10:54

Expand the square as $$ \left(\int_0^TW_t\, \mathrm dt\right)^2=2\int_0^T\int_0^t2W_tW_s\,\mathrm ds\mathrm dt, $$ and use Fubini theorem and the identity $\mathbb E(W_tW_s)=s$ for every $s\leqslant t$ to conclude that the expectation you want to compute is $$ \frac1{T^2}\int_0^T\int_0^t2s\,\mathrm ds\mathrm dt=\frac{T}3. $$

share|improve this answer
    
thanks for your reply. I read about [Fubini theorem] (en.wikipedia.org/wiki/Fubini's_theorem) and also follow your explanation about $E(W_tW_s)$. Could you please elaborate on expanding the square of the integral? –  Prakhar Mehrotra Dec 4 '12 at 7:54
    
No: this is straightforward and if you think one second about it you will see this is true in full generality. –  Did Dec 4 '12 at 10:35

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.