7
$\begingroup$

$Let \space X(t) =\begin{cases} 2, \qquad\text{if} \space 0\le t \le 1 \\ 3, \qquad\text{if} \space 1 < t \le 3 \\ -5, \qquad\text{if}\space 3 < t \le 4 \end{cases} $

or in one forumala $X(t) = 2I_{[0,1]}(t)+3_{(1,3]}(t)-5_{(3,4]}(t)$. Give the $It\hat{o}$ integral $\int_0^4 X(t)dB(t)$ as a sum of random variables, give its distribution, specify the mean and the variance.

Here is what I tried:

$\int_0^4X(t)dB(t) = \int_0^12dB(t)+\int_1^33dB(t)+\int_3^4-5dB(t)$

$=2(B(1)−B(0)) + 3(B(3)−B(1))−5(B(4)−B(3))$

We can write this as the sum of Normal Random variables:

$\int_0^4X(t)dB(t) = 2N(0,1) + 3N(0,2)−5N(0,1)$

$= N(0,2^2) + N(0,2(3^2)) + N(0,(−5)^2)$

= $N(0,47)$

Thus, the $It\hat{o}$ integral $\int_0^4X(t)dB(t)$ is Normally distributed with mean $0$ and variance $47$.

Am I on the right track here?

$\endgroup$
  • 3
    $\begingroup$ Yes you are. You can also use the property that for any $L^2$ function $f$, $\int_0^T f(t) dB(t)$ is normally distributed with mean zero and variance $\int_0^T f(t)^2 dt$ $\endgroup$ – Antoine Conze May 22 '18 at 11:15
4
$\begingroup$

To verify @AntoineConze's suggestion, the variance should be: $$\int_0^4 (2_{[0,1]}(t)+3_{(1,3]}(t)-5_{(3,4]}(t))^2\,dt.$$ Since the supporting domains are disjoint, the product of any two of the terms $2_{[0,1]}(t), 3_{(1,3]}(t), 5_{(3,4]}(t)$ is identically 0, so the integral is just $$\int_0^4 2^2_{[0,1]}(t)+3^2_{(1,3]}(t)+5^2_{(3,4]}(t)\,dt=4(1-0)+9(3-1)+25(4-3)=47.$$

$\endgroup$

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.