Dear Tom,
For the procedure of Markov regime switching multivariate regression @MSSysRegression, I see that the independent variables on the right hand side are set to be the same. Is there possibility that the right hand side variables can be different across equations? For example, I want to run the following regression under regime switching:
yt = constant + residual
xt = constant + yt + residual
where yt and xt follow independent Markov chain.
A reference for this model is "Volatility and shocks spillover before and after EMU in European stock markets" by Monica Billio and Loriana Pelizzon (2003). I would appreciate your commments and help!
Regime switching multivariate regression
Re: Regime switching multivariate regression
@MSSysRegression first of all does require the same RHS variables in each equation; that greatly simplifies EM and Gibbs Sampling processes. And second, it only does a single Markov chain.