Re: [Gretl-users] Fama and French using GMM in Grelt
Hello Allin
I am teaching a course in time series and will use GRETL.
Does any one has the code for any of the following models?
* estimating time varying covariances using GMM (Harvey 1991);
* testing mean/variance efficiency using GMM (Mackinlay & Richardson 1991).;
*factor modelling for the interest rate term structure (Brennan & Schwartz 1982).
Regards
Professor Kostas Giannopoulos
Dean, Business School
Neapolis University
2 Danais Avenue
Pafos 8042
Cyprus
tel +357 26843372
www.nup.ac.cy
-----Original Message-----
From: gretl-users-bounces(a)lists.wfu.edu [mailto:gretl-users-bounces@lists.wfu.edu] On
Behalf Of Allin Cottrell
Sent: Wednesday, February 08, 2012 4:01 PM
To: Anutechia Asongu; Gretl list
Subject: Re: [Gretl-users] Fama and French using GMM in Grelt
On Wed, 8 Feb 2012, Anutechia Asongu wrote:
I have a qualm applying a four-factor Fama and French model with
Grelt.
The name is "gretl".
The reference article I''m using applies GMM. As far as I
have perused
GMM goes with instruments(moment conditions).
But in this article, no reference is made to choice of instruments. Is
there a GMM application on Grelt that doesn't require instruments for
its application?.
No, not in gretl and not anywhere. The principle of GMM is to choose parameter values that
satisfy (or come as close as possible to satisfying) a set of orthogonality conditions.
Each orthogonality condition takes the form
u is orthogonal to z
where the z's are the "instruments". In a very simple case (where GMM as
such is not really required) the set of instruments may coincide with the set of
regressors.
See chapter 20 of the Gretl User's Guide for details.
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users