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.

Define

$$X_t := \left( \begin{matrix} \cos W_t \\ \sin W_t \end{matrix} \right).$$

where $W = \left( W_t,\mathcal F_t \right) _{t\ge0}$ is a standard Wiener process. Find the Ito differential of X and show that

$$\mathbb EX_t = \left( \begin{matrix} \exp \left( -\frac t2 \right) \\ 0 \end{matrix} \right).$$

share|improve this question
    
What did you try, what's the difficulty in solving it? Did you apply Ito's lemma to find the differential of $X$? –  Ilya May 30 '13 at 14:59
    
I'm stuck at the beginning... :( I started with the two-dimensional Ito eq. $$df(t,\mathbf{X}_t) = \frac{\partial f}{\partial t} dt + (\boldsymbol{\nabla}_\mathbf{X}^{\mathsf T} f) d\mathbf{X}_t + \tfrac{1}{2} (d\mathbf{X}_t^\mathsf{T}) (\nabla_\mathbf{X}^2 f) d\mathbf{X}_t$$ but I have difficults to derive the integral form –  user79133 May 30 '13 at 15:46
    
@user79133 This approach requires that you know about the Itô differential $dX_t$. But you don't - you want to determine it! Apply the Itô Formula to the Brownian motion instead... –  saz May 30 '13 at 19:05

1 Answer 1

By applying Itô's Formula $$f(W_t)-f(W_0) = \int_0^t f'(W_s) \, dW_s + \frac{1}{2} \int_0^t f''(W_s) \, ds$$

to $f(x) := \cos x$ we obtain

$$\cos(W_t)-1 = -\int_0^t \sin(W_s) \, dW_s - \frac{1}{2} \int_0^t \cos(W_s) \, ds$$

The mapping $$(t,w) \mapsto M(t,w) := \int_0^t \sin(W_s) \, dW_s$$ is a martingale since it's a stochastic integral with respect to a Brownian motion. This implies $\mathbb{E}M_t=0$. Thus,

$$\mathbb{E}(\cos(W_t))-1 = - \frac{1}{2} \int_0^t \mathbb{E}(\cos(W_s)) \, ds$$

Define $\varphi(s) := \mathbb{E}(\cos(W_s))$. The last equation is equivalent to

$$\varphi(t)-1 = - \frac{1}{2} \int_0^t \varphi(s) \, ds$$

Solve this differential equation and you are done with the first component of $X_t$. Similar argumentation works for $\sin(W_t)$.

share|improve this answer
    
one issue: if I do as you suggest, I find $$\phi'+\frac 12\phi=\frac 12, \qquad \phi(0)=1$$ Now, the solution is $C \exp (-t/2) +1$ and $\phi(0)=C+1=1$ that means $C=0$ ?!?! Maybe I'm wrong? –  user79133 Jun 10 '13 at 9:16
    
@user79133 How do you arrive at this differential equation? By differentiating the last equation, I obtain $$\varphi' = - \frac{1}{2} \varphi \qquad \varphi(0)=1.$$ –  saz Jun 10 '13 at 12:51
    
yeah, I was wrong with the integral theorem. yeah I agree with your answer. thanks again –  user79133 Jun 12 '13 at 7:07

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.