Page 1 of 1

Syntax for the Granger causality test

Posted: Fri Apr 01, 2016 6:34 pm
by Kavorka
Why does
cause(4) y x
work, but not
equation eq1.cause(4) y x

I get the error message "Equation does not have estimates" when I click at "eq1"

I would like to save the result (with a loop), so therefore I am interesting in saving in equation eq1 (which I will put in {!i} with a loop. I could of course use OLS manually and test using a e.g. an F-test, and save it, but I would like to utilize the canned routines for e.g. AIC etc.

Re: Syntax for the Granger causality test

Posted: Fri Apr 01, 2016 6:46 pm
by EViews Gareth
I've spent some time trying to figure out what you're trying to do, but I have failed.

Granger causality works on two series objects (or, more properly, a group object). It has nothing to go with equations objects.

I don't know what you're trying to achieve by trying to estimate an equation object using the Granger causality test. Granger causality is not an estimation method. It is a test.