Am 19.04.2023 um 18:57 schrieb JOSE FRANCISCO PERLES RIBES:

Dear Nikola:

I'm not sure, but probably it would be possible to test for non-Granger causality test as David Giles explains for Eviews using your option  VEC Granger Causality Test/ Block Exogeneity Wald Test  with the VAR options in Gretl.

https://davegiles.blogspot.com/2011/04/testing-for-granger-causality.html

Hi José, this is an interesting idea (declaring the p+1-th lag as an exogenous variable) - but first it didn't work for me, because gretl is apparently "too clever" with its lag handling and doesn't directly allow you to put a lag of an endogenous variable into the list of exogenous ones.

However, what did work is to manually create an extra list before running the VAR. Example in script form:

<hansl>

open denmark

order = 4    # let's assume this is correct

list endogenous = LRM LRY IBO # just an arbitrary example

list extralag = lags({order+1}, endogenous)    # note the curly braces

var order endogenous ; extralag

</hansl>

And then you have your relevant and appropriate Granger test results in the standard VAR output indeed. The same thing can be achieved via the menus.

thanks

sven

_______________________________________________
Gretl-users mailing list -- gretl-users@gretlml.univpm.it
To unsubscribe send an email to gretl-users-leave@gretlml.univpm.it
Website: https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/