I'm having trouble computing Itô's calculus. Take $\int_0^tB_sdB_s$ for example, can I solve it base solely on Itô-Doeblin Formula, instead of assuming $f(x)=\frac{x^2}{2}$? Please help!
Asked
Active
Viewed 65 times
1
-
What exactly do you mean by "base it solely on the Itô-Doeblin formula"...? Is your question how to come up with the idea to consider $f(x)=x^2/2$...? – saz Mar 01 '19 at 15:59
-
@saz Yes, sorry about my English. – Bubblethan Mar 01 '19 at 16:20
-
Don't worry; I just wasn't sure whether I understood your question correctly. – saz Mar 01 '19 at 17:48
1 Answers
2
Itô's formula states that
$$f(B_t) -f(B_0) = \int_0^t f'(B_s) \, dB_s + \frac{1}{2} \int_0^t f''(B_s) \, ds $$
for any twice differentiable function $f$. Equivalently,
$$\int_0^t f'(B_s) \, dB_s = f(B_t)-f(B_0) - \frac{1}{2} \int_0^t f''(B_s) \, ds \tag{1}.$$
Now if we are interested in the stochastic integral
$$\int_0^t B_s \, dB_s$$
then we can write
$$\int_0^t B_s \, dB_s = \int_0^t f'(B_s) \, dB_s$$
for any function $f$ which satisfies
$$f'(x)=x. \tag{2}$$
Clearly, the function $f(x)=x^2/2$ satisfies $(2)$. Using $(1)$ for $f(x)=x^2/2$ we get
$$\begin{align*}\int_0^t B_s \, dB_s &= \int_0^t f'(B_s) \, dB_s \\ &= f(B_t)-f(B_0)- \frac{1}{2} \int_0^t f''(B_s) \, ds \\ &= \frac{1}{2} B_t^2 - \frac{1} {2} t.\end{align*}$$
saz
- 120,083