Re: st: Re: estimation of nonlinear simultaneous equations system

["Brian P. Poi" <bpoi@stata.com>]

On Mon, 4 Jun 2007, Hyeok Jeong wrote:

> Thank you for the kind replies from Nick Cox and Kit Baum. Somehow I 
> received your messages just now (not only yours but also all messages 
> from statalist). The estimation method that I have in mind is the 
> "minimum distance estimation" since I do not want to assume a parametric 
> form for the error terms. So, neither "ml" nor "nl" seem to be helpful. 
> Are there any other ways?
And earlier he wrote

> I would like to estimate a nonlinear simultaneous equation
> system using
> Stata, for example:
> Y1(t) = F[X1(t),X2(t); p] +U1(t)
> Y2(t) = G[X1(t),X2(t); p] +U2(t),
> where Y1 and Y2 are dependent variables,  X1 and X2 are explanatory
> variables, U1 and U2 are possibly correlated error terms, and p
> is the vector of parameters to be estimated.
Stata 10 has a new command -nlsur- that fits nonlinear seemingly unrelated 
regression (NLSUR) models and is designed specifically for this type of 
problem.  The syntax is very similar to -nl-, except that you can specify 
multiple equations.

  -- Brian Poi
  -- bpoi@stata.com
*
*   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/