Stationarity, Markov switching

[Feruzbek Davletov]

Hi everyone, I was trying to model times series with markov switching regimes. Preliminarily, do I need to have my time series to be stationary? Doesnt the notion of regime switching itself imply that series should be non-stationary? Can I model non-stationary times series with markov switching regimes?

[Justin Niakamal]

Yes, you can. In fact, James Hamilton's seminal work on markov switching did just that. See the reference below.

Hamilton, J. D. 1989. A new approach to the economic analysis of nonstationary time series and the business cycle. Econometrica 57: 357–384.