Markov Switching Model of the Business Cycle

[Bairam Kara]

Hello everyone,

Can anyone guide me about Markov Switching Model of the Business Cycle, I have the following model;
deficit=b1deficit+b2debt+b3outputgap


I want to check the regime impacts of outputgap on deficit

Thanks and Regards

[Justin Niakamal]

Here's an example that you can hopefully build off of.

webuse "rgnp.dta", clear
mswitch ar rgnp, ar(1/4) nolog

predict pr_state1 pr_state2, pr  smethod(smooth)

label var pr_state1 "probability of recession"
label var recession "recession"

twoway  (area recession date,   color(gs14))     ///        
              (tsline pr_state1,       lcolor(blue)),       ///
              ylab(, angle(h)) ytitle("")  xtitle("") legend(region(lcolor(white)))

[Bairam Kara]

Thanks Justin,
But the model you mention has the dependent variable GDP already hence the figure shows the recessions.
Is it possible to do that with a different dependent variable?
Thanks again.
Best

[Justin Niakamal]

Yes. The model here isn't explicitly predicting a recession, but it is predicting the transition into a state. The default, if not specified is two states. Hope this helps.

[Bairam Kara]

Thank you Justin