Zero restriction in one regime for switching regression
Posted: Mon Feb 22, 2016 7:07 am
Dear Moderator,
I'm using Markov-switching models and would like to know whether is it possible to impose a correlation coefficient to be equal to zero in one particular regime and to be freely estimate in the other. The aim is to esimate probabilities of the regime which captures the correlation between two variables and use it for different purpose.
Thank you
I'm using Markov-switching models and would like to know whether is it possible to impose a correlation coefficient to be equal to zero in one particular regime and to be freely estimate in the other. The aim is to esimate probabilities of the regime which captures the correlation between two variables and use it for different purpose.
Thank you