12
$\begingroup$

Girsanov theorem seems to have many different forms. I've got a problem matching the form in wiki to the one in Shreve's book, due to the difficulty of quadratic variation calculation.

Below is the Girsanov Theorem from wiki:

Let $\{W_t\}$ be a Wiener process on the Wiener probability space $\{\Omega,\mathcal{F},P\}$. Let $X_t$ be a measurable process adapted to the natural filtration of the Wiener process $\{\mathcal{F}^W_t\}$.

Given an adapted process $X_t$ with $X_0 = 0$, define $Z_t=\mathcal{E}(X)_t,\,$ where $\mathcal{E}(X)$ is the stochastic exponential (or Doléans exponential) of X with respect to W, i.e. $\mathcal{E}(X)_t=\exp \left ( X_t - \frac{1}{2} [X]_t \right )$, where $[X]_t$ is a quadratic variation for $X_t$. Thus $Z_t$ is a strictly positive local martingale, and a probability measure $Q$ can be defined on $\{\Omega,\mathcal{F}\}$ such that we have Radon–Nikodym derivative $\frac{d Q}{d P} |_{\mathcal{F}_t} = Z_t = \mathcal{E} (X)_t$. Then for each $t$ the measure $Q$ restricted to the unaugmented sigma fields $\mathcal{F}^W_t$ is equivalent to $P$ restricted to $\mathcal{F}^W_t.\,$

Furthermore if $Y$ is a local martingale under $P$ then the process $\tilde Y_t = Y_t - \left[ Y,X \right]_t$ is a $Q$ local martingale on the filtered probability space $\{\Omega,F,Q,\{F^W_t\}\}$.

Below is the Girsanov Theorem from Shreve's book "Stochastic calculus for finance II"

Theorem 5.2.3 (Girsanov, one dimension). Let $W(t)$, $0 \leq t \leq T$, be a Brownian motion on a probability space $(\Omega, \mathscr F, \mathbb P)$, and let $\mathscr F(t)$, $0 \leq t \leq T$, be a filtration for this Brownian motion. Let $\Theta(t)$, $0 \leq t \leq T$, be an adapted process. Define $$Z(t) = \text{exp} \left\{ -\int_0^t \Theta(u)dW(u) - \frac{1}{2} \int_0^t \Theta^2(u) du \right \}, \tag{5.2.11}$$ $$\widetilde W(t) = W(t) + \int_0^t \Theta(u) du, \tag{5.2.12}$$ and assume that $$\mathbb E \int_0^T \Theta^2(u) Z^2(u) du < \infty \tag{5.2.13}$$

Set $Z = Z(T)$. Then $\mathbb E Z = 1$ and under the probability measure $\widetilde P$ given by (5.2.1), the process $\widetilde W(t)$, $0 \leq t \leq T$, is a Brownian motion.

Seems the Girsanov theorem form wiki is more general than the one on Shreve's book.

Now my questions is: How to derive the latter from the former?

It seems only need to take $Y(t) = W(t)$ and $X(t) = \int_0^t \Theta(u) du$ in the wiki definition. This left to prove

$$[W(t), \int_0^t \Theta(u) du]_t = - \int_0^t \Theta(u) du$$

, but how to calculate the quadratic variation?

Quadratic variation definition is $$[X,Y](T) := \lim_{\|\Pi\|\to 0} \sum_{j=0}^{n-1} \left[ X(t_{j+1}) - X(t_j) \right] \left[ Y(t_{j+1}) - Y(t_j) \right]$$ , where $\Pi := \{ t_0, t_1, \cdots, t_n \}$ . But I'm a bit stuck here.

Could you please kindly give me some hint how to proceed?

$\endgroup$
9
$\begingroup$

Shreve's theorem also called "Girsanov II" indeed represents a special case of the general "Girsanov I" from Wiki above, with $$Y_t:=W_t,$$$$X_t:=-\int_0^t\Theta_udW_u$$

We can show: $$[Y,X]=-\int_0^t\Theta_udu$$ by using general Stochastic Calculus rules (e.g. p.37, 6.6 here):

$$[Y,X]=[W_t,-\int_0^t\Theta_udW_u]=-\int_0^t\Theta_ud[W_u,W_u]=-\int_0^t\Theta_udu$$

as $[W,W]=[W]=t$.

$\endgroup$

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.