Thursday, 12 April 2007
Discretization of bivariate processes
I have a question for you. Imagine that you have two exogenous and stochastic variables, say TFP and real exchange rate. You estimate a VAR(1) process for them and want to discretize them in order to add them in a model. How would you do that? One obvious idea is to stack the two exogenous variable into one and use the logic of Tauchen-86 to build the transition matrix between states. Do you know of anybody having done that? Other alternatives?
Subscribe to:
Post Comments (Atom)
1 comment:
Tauchen (86) can be applied also to VAR's, by generalizing the logic behind it.
I never did it myself, but Victor used to have a F77 code on his webpage, which I sent you by e-mail.
I read it once and it looked quite easy to understand.
Hope it helps!
Post a Comment