System GMM

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

econwork
Posts: 2
Joined: Thu May 11, 2017 8:27 am

System GMM

Postby econwork » Thu May 11, 2017 9:19 am

Hi,
My goal is to analyze the effect of macroeconomic policy on agricultural export and production under a simultaneous equations model for 8 countries. As export and production are endogenous to each other and macro-policy effect on both of them, the best method is to apply a system GMM (GMM estimation of a system of non-panel equations), in which:
eq1: dependent variable is production (independents are export and macro-policy variables)
eq2: dependent variable is export (independents are production and macro-policy variables)
and use macro-policy indicators as IV.

I want to estimate the system of equations for each of countries separately, however I recognize that under a panel workfile also the system estimations are available! may I ask you if I use the panel workfile and estimate a system GMM (a system including two equations of production and export), how eviews reads the data?
I know the difference between panel GMM, panel system GMM and system GMM, and I know the two firs cases are out of discussion here, however I am not sure estimating a system of equations using GMM under a panel workfile is econometrically correct? or maybe Eviews reads the data as stacked and not panel?
Thank you in advance
Regards

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 11868
Joined: Tue Sep 16, 2008 5:38 pm

Re: System GMM

Postby EViews Gareth » Thu May 11, 2017 9:23 am

The system object is not panel aware, so will estimate assuming the data is stacked data. The exception being that lags/difference etc... will work correctly (they will not cross across cross sections [that's a mouthful!]).

I won't comment on whether using such an estimator on your panel data makes sense.
Follow us on Twitter @IHSEViews


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 10 guests