estimation Switching regression

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

lagrace
Posts: 2
Joined: Sat Sep 23, 2017 7:42 am

estimation Switching regression

Postby lagrace » Sun Sep 24, 2017 12:47 am

Goodmorning to you all,
I'm a new user of EViews and I'm doing a series of switching regressions.
I've noticed that every time I repeat the estimation my results change. Could someone give me an explanation about this issue? The software inverts the regimes results so I'm not able to understand if an estimation belongs to regime one or regime two. Sometimes it says to me that a number belongs to regime 1, the subsequent estimation says that the same number belongs to regime 2!
Have I to leave Eviews and use Gauss?
Please help me!!!

startz
Non-normality and collinearity are NOT problems!
Posts: 3139
Joined: Wed Sep 17, 2008 2:25 pm

Re: estimation Switching regression

Postby startz » Sun Sep 24, 2017 7:40 am

Regime names are not identified. It's perfectly normal for all the regimes to be swapped. Nothing to do with EViews.

lagrace
Posts: 2
Joined: Sat Sep 23, 2017 7:42 am

Re: estimation Switching regression

Postby lagrace » Sun Sep 24, 2017 8:27 am

ok.. thanks.. very kind.. my fault.. i'm not able to establish the high regime in the case of a regression.
I just know the formula for the autoregressive case..


Return to “Estimation”

Who is online

Users browsing this forum: Yahoo [Bot] and 6 guests