Rolling Cointegration

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

pgupta
Posts: 7
Joined: Mon Apr 01, 2013 10:50 pm

Rolling Cointegration

Postby pgupta » Fri Nov 22, 2013 3:45 am

Dear Gareth,

I tried to create a code to achieve three purposes:
1. Estimate rolling trace statistic
2) Estimate rolling "adjustment coefficients" of a VEC model.
3) Estimate rolling LR statistic for testing Weak Exogeneity of the ECM term in the VEC model.

Sir, while the code is working fine for the above stated requirements, I have one more requirement:
To store the standard errors and t-stat of these "rolling adjustment coefficients". I am not able to do this. I will be very obliged if you could please help me with this.

Thanks a ton!
Attachments
rolling estimation.prg
(1.05 KiB) Downloaded 542 times

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13306
Joined: Tue Sep 16, 2008 5:38 pm

Re: Rolling Cointegration

Postby EViews Gareth » Sat Nov 23, 2013 11:42 am

What coefficients?
Follow us on Twitter @IHSEViews

pgupta
Posts: 7
Joined: Mon Apr 01, 2013 10:50 pm

Re: Rolling Cointegration

Postby pgupta » Sun Nov 24, 2013 7:15 am

The adjustment coefficient (alpha) (A(1,1) in my case). As defined on Pg. 575 of Eviews 8 Users Guide II.

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13306
Joined: Tue Sep 16, 2008 5:38 pm

Re: Rolling Cointegration

Postby EViews Gareth » Sun Nov 24, 2013 5:17 pm

You can probably get the standard errors from the coefficient standard errors matrix, v.@coefcov. You'll have to calculate the t-stats manually.
Follow us on Twitter @IHSEViews

pgupta
Posts: 7
Joined: Mon Apr 01, 2013 10:50 pm

Re: Rolling Cointegration

Postby pgupta » Sun Nov 24, 2013 9:40 pm

Thank you Gareth! :)

fcb85
Posts: 2
Joined: Thu Jul 06, 2017 9:16 am

Re: Rolling Cointegration

Postby fcb85 » Thu Jul 06, 2017 9:48 am

Dear Gareth,
I'm using similar code but for a larger sample (4157 observations, 8 variables) and smaller rolling window (500). Unfortunately the program does not run till the end of the sample, it turns "near singular matrix" error after 1100 replications. Do I have to define the matrix in a different manner? The same error is occurred when I use "pgupta" code
Attachments
rolling cointegration.prg
(603 Bytes) Downloaded 357 times

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13306
Joined: Tue Sep 16, 2008 5:38 pm

Re: Rolling Cointegration

Postby EViews Gareth » Thu Jul 06, 2017 10:31 am

Looks good to me.
Follow us on Twitter @IHSEViews

fcb85
Posts: 2
Joined: Thu Jul 06, 2017 9:16 am

Re: Rolling Cointegration

Postby fcb85 » Fri Jul 07, 2017 6:38 am

Then why it breaks out the code after 1100 replications, when normally it should perform 3657 replications. However, I've noticed that the break it varies if I change either number of variables or size of rolling window. Any support is more than welcome!

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13306
Joined: Tue Sep 16, 2008 5:38 pm

Re: Rolling Cointegration

Postby EViews Gareth » Fri Jul 07, 2017 7:31 am

Your data probably results in a singular matrix.
Follow us on Twitter @IHSEViews


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 32 guests