| ![](/includes/contimages/spacer.gif) |
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
RE: st: GMM
Janine,
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of
> Janine Darfield
> Sent: 28 May 2006 22:34
> To: [email protected]
> Subject: Re: st: GMM
>
> Hi,
>
> I posted the following question last week but didn't get any
> replies, I am trying again in case anyone can help. Basically
> I am trying to do GMM analysis in STATA and am trying to
> understand the difference between
> - xtabond/xtabond2;
> - ivreg, gmm / xtivreg, gmm; and
> - ivgmm0
>
> Is there any guidance as to under what circumstances one such
> method is better than the other? How (if at
> all) could one get equivalent results with all 3 methods? I
> realise this might be a broad question but even if someone
> could help with part of it I would appreciate it a lot.
Your question is a combination of big, important questions and little,
not-so-important ones.
An example of the latter is the difference between ivreg2, gmm (not
ivreg as you say) and ivgmm0. As I recall, at one point the latter
iterates once more to get an updated set of residuals, but the
difference is inconsequential; both report 2-step feasible efficient GMM
estimates.
The big questions are too big to answer in a posting to Statalist, but
maybe I can rephrase some of them:
- Does the estimator use the first-differences transformation?
(xtabond/2, or xtivreg/2 with the fd option)
- Does the estimator use the within-transformation? (xtivreg/2 with the
fe option)
- Does the estimator not transform the data at all? (ivreg/2 or ivgmm0)
- Does the estimator use clever and hard-to-describe orthogonality
conditions? (xtabond/2)
Hope this helps.
Cheers,
Mark
Prof. Mark Schaffer
Director, CERT
Department of Economics
School of Management & Languages
Heriot-Watt University, Edinburgh EH14 4AS
tel +44-131-451-3494 / fax +44-131-451-3296
email: [email protected]
web: http://www.sml.hw.ac.uk/ecomes
> (I have already read the help files as well as the paper by
> Baum Schaffer and Stillman, and Stephen Bond's paper on
> Dynamic Panel Data Models; but am still confused on this point.)
>
> Thanks in advance,
> Janine
>
> --- Janine Darfield <[email protected]>
> wrote:
>
> > Hi,
> >
> > I need to do some GMM estimation on panel data (which has about 700
> > companies and about 40 quarters). I have found 3 stata commands for
> > doing GMM:
> > xtabond/xtabond2; ivreg, gmm / xtivreg, gmm; and ivgmm0 (I
> hope I am
> > correct that these are all ways of estimating with GMM). Can anyone
> > tell me what the difference between the 3 methods is and which of
> > these is the "best" to use? And is there a way of getting an
> > equivalent result in these 3 methods (for example what would be an
> > equivalent specification, if any?)
> >
> > I would be terribly grateful for any guidance as I have
> tried to read
> > the help files and other literature but am still very
> confused about
> > this.
> >
> > Janine
> >
> >
> >
> >
> >
> >
> >
> ___________________________________________________________
> >
> > All new Yahoo! Mail "The new Interface is stunning in its
> simplicity
> > and ease of use." - PC Magazine
> > http://uk.docs.yahoo.com/nowyoucan.html
> > *
> > * For searches and help try:
> > *
> > http://www.stata.com/support/faqs/res/findit.html
> > * http://www.stata.com/support/statalist/faq
> > * http://www.ats.ucla.edu/stat/stata/
> >
>
>
>
>
> ___________________________________________________________
> Try the all-new Yahoo! Mail. "The New Version is radically
> easier to use" - The Wall Street Journal
> http://uk.docs.yahoo.com/nowyoucan.html
> *
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/