1
$\begingroup$

I know that if we have the following SDE:

$$dX_t = \alpha X_t dt + \sigma X_t dW_t$$ where $\alpha$ and $\sigma$ are constants and $X_0 = x_0$, then the solution is the geometric brownian motion: $$X(t) = x_0e^{(\alpha-\frac{1}{2}\sigma^2)t + \sigma W_t} $$

Now, what if we have more than one $dW$ term? Specifically, say we have the following SDE:

$$dX_t = \alpha X_t dt + \sigma_1 X_t dW^1_t+\sigma_2X_tdW^2_t$$ where $\alpha, \sigma_1, \sigma_2$ are all constants and $W^1_t$ and $W^2_t$ are independent standard Brownian motions, can we find a solution to $X(t)$? Can we say anything about the distribution of $X(t)$?

$\endgroup$
1
$\begingroup$

The respective solution would become $$ X_t=X_0\exp\left[\left(\alpha-\frac{1}{2}\sigma_1^2-\frac{1}{2}\sigma_2^2\right)t+\sigma_1W_t^1+\sigma_2W_t^2\right]. $$

Note that the SDE reads $$ {\rm d}X_t=X_t\left(\alpha{\rm d}t+\sigma_1{\rm d}W_t^1+\sigma_2{\rm d}W_t^2\right). $$ This gives the quadratic variation of $X_t$ by, intuitively, \begin{align} {\rm d}\left<X\right>_t&={\rm d}X_t{\rm d}X_t\\ &=X_t^2\left(\alpha{\rm d}t+\sigma_1{\rm d}W_t^1+\sigma_2{\rm d}W_t^2\right)^2\\ &=X_t^2\left(\alpha^2{\rm d}t^2+\sigma_1^2{\rm d}W_t^1{\rm d}W_t^1+\sigma_2^2{\rm d}W_t^2{\rm d}W_t^2+\right.\\ &\quad\quad\left.2\alpha\sigma_1{\rm d}t{\rm d}W_t^1+2\alpha\sigma_2{\rm d}t{\rm d}W_t^2+2\sigma_1\sigma_2{\rm d}W_t^1{\rm d}W_t^2\right)\\ &=X_t^2\left(\sigma_1^2{\rm d}t+\sigma_2^2{\rm d}t\right), \end{align} where the ${\rm d}W_t^1{\rm d}W_t^2$ term drops due to the independence of $W_t^1$ and $W_t^2$.

Thus thanks to Ito's lemma, \begin{align} {\rm d}\log X_t&=\frac{{\rm d}X_t}{X_t}-\frac{1}{2}\frac{{\rm d}\left<X\right>_t}{X_t^2}\\ &=\left(\alpha{\rm d}t+\sigma_1{\rm d}W_t^1+\sigma_2{\rm d}W_t^2\right)-\frac{1}{2}\left(\sigma_1^2+\sigma_2^2\right){\rm d}t\\ &=\left(\alpha-\frac{1}{2}\sigma_1^2-\frac{1}{2}\sigma_2^2\right){\rm d}t+\sigma_1{\rm d}W_t^1+\sigma_2{\rm d}W_t^2\\ &={\rm d}\left[\left(\alpha-\frac{1}{2}\sigma_1^2-\frac{1}{2}\sigma_2^2\right)t+\sigma_1W_t^1+\sigma_2W_t^2\right]. \end{align} This, together with the initial condition, eventually yields $$ X_t=X_0\exp\left[\left(\alpha-\frac{1}{2}\sigma_1^2-\frac{1}{2}\sigma_2^2\right)t+\sigma_1W_t^1+\sigma_2W_t^2\right]. $$

$\endgroup$

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.