first difference and system GMM

[siti_ng]

Hi there,

I am new with Eviews, esp EViews 7. I am currently doing a research and might use a dynamic model.

My question is: does EViews 7 able to estimate first difference and system GMM?

I know EViews 6 can only estimate for Arrelano and Bond Dynamic GMM (one-step and two-step)

Thanks.

[EViews Gareth]

Nothing has changed - EViews 7 cannot estimate system DPD

[samijo]

Gareth, this is a side note;
On the webpage of TimberLake (a partner of your company), for a course they offer on EViews, it says that the course covers:
An overview of estimator approaches: Anderson and Hsiao (1982), Blundell and Bond (1998), Arelano and Bover (1995) .
http://www.timberlake.co.uk/Training/EViews-112010.php

I wonder how they do that!

[EViews Gareth]

EViews does perform those estimators.

[samijo]

System-GMM in dpd framework is called awkwardly so in literature, though it's not a system in the real meaning of systems. It just augments a level equation to the differenced equation and estimate them together. I guess EViews does provide with this.

[EViews Gareth]

System DPD is generally thought of as being the Arellano-Bover estimator of DPD. EViews does not do that. However EViews does do the Arrelano-Bond estimators. I misread the Timberlake quote. I'm not sure how they can say that EViews does Arellano-Bover estimation.

[paul84]

Dear Members,
Dear Gareth,
it is not clear to me what is done in EViews if one specifies “orthogonal deviations” instead of “first differences” (Arellano Bond) as the transformation method in estimating a dynamic panel data model. Arellano and Bover (1995) describe “forward orthogonal deviations”, which is the paper cited in User Guide II (e.g. p.575). Of what kind are those “orthogonal deviations” in EViews 6?
Thank you,
Paul