- #1
mckammer
- 2
- 0
Hello everybody! I'm working on a Stochastic Processes course based on Ross' Introduction to Probability Models.
I hope you can help me work through this problem.
{B(t), t >= 0} is a standard Brownian motion process
Compute E[B(t1)B(t2)B(t3)] for t1 < t2 < t3.
See my attempt at a solution?
E[B(t1)B(t2)B(t3)] =E[E[B(t1)B(t2)B(t3)|B(t1)]]
=E[B(t1)E[B(t2)B(t3)|B(t1)]
Then, I name another Brownian Process C(t), such that C(t) = B(t) - B(t1)
=E[B(t1)E[(C(t2)-B(t1))(C(t3)-B(t1))|B(t1)]
Expanding, I get
E[B(t1)^3 - B(t1)^2(E[C(t2)+C(t3)]) + B(t1)E[C(t2)C(t3)]]
The second term evaluates to 0, since we have the sum of two Brownian motion processes with mean 0
The third term evaluates to
E[C(t2)[C(t3) + C(t2) - C(t2)]] = E[C(t2)(C(t3)-C(t2))] + E[C(t2)^2]
= 0 + t2? (Because the first term here is multiplied by E[C(t2)], and the second term is the second moment, which is equal to the variance due to the mean being 0?)
My final answer is
B(t1)^3 + (t2)B(t1)
Does this look reasonable?
I hope you can help me work through this problem.
Homework Statement
{B(t), t >= 0} is a standard Brownian motion process
Compute E[B(t1)B(t2)B(t3)] for t1 < t2 < t3.
Homework Equations
See my attempt at a solution?
The Attempt at a Solution
E[B(t1)B(t2)B(t3)] =E[E[B(t1)B(t2)B(t3)|B(t1)]]
=E[B(t1)E[B(t2)B(t3)|B(t1)]
Then, I name another Brownian Process C(t), such that C(t) = B(t) - B(t1)
=E[B(t1)E[(C(t2)-B(t1))(C(t3)-B(t1))|B(t1)]
Expanding, I get
E[B(t1)^3 - B(t1)^2(E[C(t2)+C(t3)]) + B(t1)E[C(t2)C(t3)]]
The second term evaluates to 0, since we have the sum of two Brownian motion processes with mean 0
The third term evaluates to
E[C(t2)[C(t3) + C(t2) - C(t2)]] = E[C(t2)(C(t3)-C(t2))] + E[C(t2)^2]
= 0 + t2? (Because the first term here is multiplied by E[C(t2)], and the second term is the second moment, which is equal to the variance due to the mean being 0?)
My final answer is
B(t1)^3 + (t2)B(t1)
Does this look reasonable?