Given two independently constructed Brownian bridges (from their marginal means and variances), is there a way to correlate the sample paths?
Asked
Active
Viewed 1,041 times
5
-
hmm can we consider bridges that are pinned to non-zero locations? Also are the bridges pinned at the same time $t$? If it is as you have linked, where they are constrained to (1,0), then simply: $$cBB^1_t = BB^1_t$$ and $$cBB^2_t = \rho BB^1_t + \sqrt{ 1- \rho^2 }BB^2_t $$ should work. – Cam.Davidson.Pilon Aug 15 '12 at 14:52
-
Thanks. So in general to construct the multivariate Brownian bridge from the one-dimensional (based on each marginal), one just need to multiply them by the square root of the variance-covariance matrix? – Tim Aug 15 '12 at 15:23
-
Yup, that should do it. – Cam.Davidson.Pilon Aug 15 '12 at 18:23