Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Second derivative for GMM


From   John Bates <[email protected]>
To   "[email protected]" <[email protected]>
Subject   Re: st: Second derivative for GMM
Date   Mon, 11 Jul 2011 11:33:23 -0700 (PDT)

Maybe I'm just confused, but I always thought the problem was to minimize the Objective Function E[xu] (where x are covariates and u is the residual).  It is only equal to zero in the just-identified case.  To minimize it, you would take the first derivative with respect to the parameters and set those equal to 0.  But then you would need the second derivative for the Hessian, right?



----- Original Message -----
From: Stas Kolenikov <[email protected]>
To: [email protected]
Cc: 
Sent: Thursday, July 7, 2011 7:24 PM
Subject: Re: st: Second derivative for GMM

On Thu, Jul 7, 2011 at 10:43 AM, John Bates <[email protected]> wrote:
>   Stata's GMM command allows the user to specify analytic first derivatives. Is there any way to specify the second derivatives?  (Or will this functionality be added in future versions?)  I would think that would speed the command up.

I did not think the second derivatives play any role either in the
theory of GMM or in numerical procedures. You are minimizing the GMM
objective function; the Jacobian is a linear combination of the moment
conditions, and the Hessian is the linear combination of the first
order derivatives. Where do the second derivatives come into play?

-- 
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.

*
*   For searches and help try:
*  http://www.stata.com/help.cgi?searchhttp://www.stata.com/support/statalist/faqhttp://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index