Regime Switching

[athens]

I am using Eviews 8 (62 bit) version and I have just updated it. When I run a regime switching model with 5 lags of the dependent and AR(1) error I get the following message "Eviews 8 has stopped working". I do not get the same message if I estimate a regime switching model with e.g. 4 lags of the dependent and AR(1) error.
My code is as follows:

Code: Select all

equation eq1.switchreg(type=markov, heterr) y c y(-1) y(-2) y(-3) y(-4) y(-5) ar(1) @nv x1 x2

Please inform me if this is a bug or if I am doing something wrong. Thanks.

[EViews Gareth]

Almost certainly a bug. Could you provide your workfile?

[athens]

regswitch_ex.wf1

(158.77 KiB) Downloaded 356 times

[EViews Glenn]

Thanks. We've tracked this one down and after a bit more testing I'll check it into the main code. It should appear in the next update.

[edit: just to follow-up, this bugfix was posted in the August 1 update]