Page 1 of 1

Markov Switching Regression

Posted: Tue Sep 17, 2019 3:17 am
Hi, all.

Please, can you assist me with running a Markov Switching Regression such that the output displayed provides the standard error of residuals and adjusted R-squared measures for each regime? Currently, the output displayed provides the standard error of residuals and adj R-squared measures for the complete sample and not for each regime - which is what I require. I am using the Lab License Eviews 10

Damien.

Re: Markov Switching Regression

Posted: Tue Sep 17, 2019 5:03 pm
It's not a built-in feature.

Note that the residuals that we compute aren't regime specific but rather an expected value using the regime probabilities. This accords with the notion that you never know what regime you are in.

For the computation that you described, you'd have to use the coefficients to construct your own set of regime specific residuals and then to compute an R2 measure yourself. Note that since you don't know the actual regime, you'd presumably have to compute the measure using the entire sample -- you'd have regime specific full sample measures of R2 under the assumption that you are always in the specified regime.

To be honest, I'm not quite certain what this computation would illustrate.

Good luck.

Re: Markov Switching Regression

Posted: Wed Oct 16, 2019 10:06 am
¿how can i get confidence interval in MS Model in Eviewss with 2 regimes?

Check box does not on.

Re: Markov Switching Regression

Posted: Thu Oct 17, 2019 8:45 am
I'm don't understand the request.

Re: Markov Switching Regression

Posted: Thu Oct 17, 2019 11:38 am
EViews Glenn wrote:I'm don't understand the request.

In forecast, after estimate MS Model, does not appear check box to standard. error, to make confidence interval.

How can i make confidence interval for this models?

Re: Markov Switching Regression

Posted: Wed Oct 30, 2019 11:28 am
Unfortunately, computing forecast standard errors for MS switching is extremely difficult and not currently supported. Support will almost certainly have to be using simulation methods, which are likely to be computationally intensive. I'll put this on the list of things to consider for future development.