Engle Granger Cointegration Test

For questions regarding programming in the EViews programming language.

Moderators: EViews Gareth, EViews Jason, EViews Moderator, EViews Matt

nicolesaba
Posts: 2
Joined: Wed Mar 11, 2015 6:57 pm

Engle Granger Cointegration Test

Postby nicolesaba » Wed Mar 11, 2015 7:05 pm

Hi,
I have several series on which I would like apply Engle-Granger test.
But I am having trouble saving the results as a table and then moving on to apply the same test on other series.

I tried the freeze command, but all it does is come up with an empty table.

This is what I am using:

coint(method="eg", trend="const") ser01 ser02
freeze(table1)

If I try freeze(table1) coint(method="eg", trend="const") ser01 ser02, I get an error message.

Thanks in advance.

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 12540
Joined: Tue Sep 16, 2008 5:38 pm

Re: Engle Granger Cointegration Test

Postby EViews Gareth » Wed Mar 11, 2015 7:35 pm

Code: Select all

group mygrp ser01 ser02
freeze(table1) mygrp.coint(method="eg", trend="const)
Follow us on Twitter @IHSEViews

nicolesaba
Posts: 2
Joined: Wed Mar 11, 2015 6:57 pm

Re: Engle Granger Cointegration Test

Postby nicolesaba » Wed Mar 11, 2015 7:48 pm

Hi again. Thanks for the help.
I just copied and ran this code above, but I got an error message "Unmatched quotes in FREEZE......."
Do you know what it is and how can I fix it?
Thanks again!

trubador
Did you use forum search?
Posts: 1518
Joined: Thu Nov 20, 2008 12:04 pm

Re: Engle Granger Cointegration Test

Postby trubador » Thu Mar 12, 2015 12:39 am

As the message indicates, there is a missing quote in Freeze command: ...,trend="const")

eisamabodian
Posts: 13
Joined: Tue Aug 05, 2014 3:50 am

Re: Engle Granger Cointegration Test

Postby eisamabodian » Wed Mar 25, 2015 10:50 am

In the first step you must discern integration degree of the variables.
the variables should be I(1).
In the second step Run your regression, for example:
y c x1 x2 x3

in the third step you should run unit root test (ADF,PP,...)on the residuals of the regression.
Residuals should be stationary or I(0)
the other method is using SARGAN & BARGHAVA table.
run your regression and then check it's Durbin Watson value.
If your regression's DW be higher than critical values in the table,you have cointegration among variables.


Return to “Programming”

Who is online

Users browsing this forum: No registered users and 23 guests