[R] GMM estimation

[Markus Jäntti]

On Tue, 2008-05-27 at 13:38 -0700, Horace Tso wrote: 
> Leo, do you mean GMM as in Generalized Method of Moment? If so i've posted a similar
>  question here before and the answer appears to be none at this time.
> 

Generalized Method of Moments (GMM) is an estimation method that can be
applied to estimate the parameters of a statistical model in very many
different situations, including Generalized Least Squares and Systems of
Regression Equations. It is in fact quite possible that some R packages
do, in fact, use GMM to estimate the parameters. However, any GMM
estimation problem can be set up in R by estimating the empirical
moments and setting up the quadratic function in the moment conditions
as a function of the parameters to be estimated. The resulting function
can then be estimated using one of the optimisers, for instance optim.
See ?optim. In some cases, GMM can be estimated by applying OLS
(?lm; if the moment conditions are linear in the parameters) to the
suitably transformed problem, or, non-linear regression (nls). 

Thus, there is no general answer to that question other than that in any
GMM problem, you can use general optimization routines to estimate the
parameters. 

Markus

> Horace
> 
> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf Of Leo Correia
> Sent: Tuesday, May 27, 2008 11:53 AM
> To: r-help at r-project.org
> Subject: [R] GMM estimation
> 
> Hello there!!!
> 
> Sorry to bother you all with such question and difficulties that I have been
> facing on.
> 
> Recently I have been searching for packages to run GMM estimatives with R.
> 
> I have been searching for such packages for a while, but since I am a new
> user of R system,
> my quest so far was unsucessful.
> 
> That´s why I had decided to ask to this forum.  Hope that anyone could help
> me!
> 
> I know that such emails might bother you guys at all... but I dont know
> where to search for help about this issue.
> 
> Many thanks!!!
> hope to hear from you soon!
> 
> Leo!
> 
>         [[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

-- 
Markus Jantti
Abo Akademi University
mjantti at abo.fi
http://www.iki.fi/~mjantti