Hi Mark:
The problem isn't to estimate any parameters (so, I don't need
use the likelihood function). I use the estimates published in the
article and I'm trying to estimate the state (the root at = 1+ dt ) each
period (the problem is similar to estimate an unobserved component time series
model). I think the error is in the expression of the equations in state space
form, but I'm not sure.
Thank's for your collaboration
Best regards
Román Mínguez