3
$\begingroup$

I am trying to compute this quantity:

$\frac{d}{dt}\int_{0}^{t} W_s ds $

Where $W_t$ is a Wiener process. Is there a theorem which tells how this can be computed?

I have tried https://en.wikipedia.org/wiki/Leibniz_integral_rule , however here the functional form is $\frac{d}{dx}\int_{a(x)}^{b(x)} (f(x,t) dt) $, that is the function $f$ takes 2 inputs - $x$ and $t$.

Using the above rule,

$\frac{d}{dt}\int_{0}^{t} W_s ds = W_t\frac{d}{dt}t - W_0\frac{d}{dt}0 + \int_{0}^{t}\frac{\partial{}}{\partial{t}}W_sds$
I am not sure what the term $\int_{0}^{t}\frac{\partial{}}{\partial{t}}W_sds$ means.

Any help is appreciated.

Improve this question
$\endgroup$
5
  • 3
    $\begingroup$ You are confusing the inputs for the Wiener process. The process is a function of the underlying state of the world and time, which you indicate as $s$, so the Wiener process is $W(\omega,s)$. So basically $t$ is a constant and you don't have that integral issue. So the answer should be $W_t$. So basically think of $t$ as a user input time that we want to see how the Wiener process will act up until. But the time variable for the process itself is different, which you wrote as $s$. Hope that makes sense. $\endgroup$
    – Slade
    Feb 28, 2019 at 14:36
  • $\begingroup$ Please consider adding the tag "Homework" or "Self-study" $\endgroup$
    – Sanjay
    Feb 28, 2019 at 15:11
  • 1
    $\begingroup$ Thanks for your reply, however your answer is not entirely clear to me. So $W(w,s)$ is the weiner process, where $w$ denotes the paths and $s$ denotes the time. If I fix $w$, then the path is fixed. Then, for that particular path, the integral from $0$ to $t$ should exist, i.e. $s$ will vary from $0$ to $t$. $\endgroup$
    – SPaul
    Feb 28, 2019 at 15:13
  • 1
    $\begingroup$ Yes the integral exists, but once you take its derivative with respect to $t$, it follows the integral rule as you said. I was pointing out that the variable $t$ is different than the variable $s$, and therefore the Wiener process does not have any dependence on $t$. So for example, a student wants to simulate a Wiener Process from time 0 to time $450$. So here $t$ is $450$, but this does not affect the Wiener process's ($W(\omega,s)$) values for any $s$. So we'd have values for $W$ from 0 to 450, but the fact that $t = 450$ does not affect the Weiner process. $\endgroup$
    – Slade
    Feb 28, 2019 at 15:30
  • $\begingroup$ I am not sure why you think you need a theorem to show the result. Since you are integrating with respect to ${ds}$, you are integrating (pathwise) continuous functions which enables you to use regular calculus rules. $\endgroup$
    – phantagarow
    Feb 28, 2019 at 19:21

1 Answer 1

Reset to default
3
$\begingroup$

You can actually use Leibniz here. And your computation above is correct. Just be aware that $\frac{d}{dt} W_s=0$. $W_s$ is a function of $s$ differentiating wrt $t$ is zero.

To sum up: $$\frac{d}{dt}\int_{0}^{t} W_s ds =W_t $$

Improve this answer
$\endgroup$
0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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