Re: st: cointegration, johans command

[Christopher Baum <kit.baum@bc.edu>]

<>
Stephanie,

You should be very careful in working with a VECM which contains arbitrary exogenous variables. Quoting from [TS] p. 481:

> Johansen (1995, chap. 11) shows that inference about the
> number of cointegrating equations is based on nonstandard distributions and that the addition of any
> term that generalizes the deterministic specification in (1) changes the asymptotic distributions of the
> statistics used for inference on the number of cointegrating equations and the asymptotic distribution
> of the ML estimator of the cointegrating equations. In fact, Johansen (1995, 84) notes that including
> event indicators causes the statistics used for inference on the number of cointegrating equations to
> have asymptotic distributions that must be computed case by case. For this reason, event indicators
> may not be specified in the present version of vec.


I would rely on Stata's vec if I were you. The vec apparatus contains commands to test for appropriate lag length, dynamic stability, etc., which I do not believe are available in the older user-written software.

Kit

On Jul 28, 2010, at 9:54 AM, Stéphanie Combes wrote:

> I was recommended that command so I kind of thought it had some abilities the command vec didn't have but I'm relatively new at stata so I may have been mistaken. As a matter of fact, with johans and vececm I can add exogeneous variables to my model, something I don't manage to do with vec.. 
> 
> 2010/7/28 Christopher Baum <kit.baum@bc.edu>
> On Jul 28, 2010, at 2:33 AM, Stephanie wrote:
> 
> > I'm trying to use the command johans for cointegration but I fail to
> > understand how to get the estimated cointegrated relations (it is apparently
> > not given as a non optional output, but I can't find which option would supply
> > them). I tried to get them next through the vececm command but I couldn't
> > either.
> > If someone uses regularly the command, I would be very grateful for your
> > help. Would you give any piece of advice for the choice of lags?
> >
> > Bonus question : if cointegration rank is above 2, do I have to check that
> > both elasticities to long-run equations are negative in my estimated models?
> > I must confess that I have a hard time interpretating when there is more
> > than one cointegrated equation.
> 
> Out of curiosity, why would you want to use this user-written command, dating from 2003 and depending
> on software written for Stata version 5, when official Stata has had the ability to do Johansen-Juselius tests for cointegration
> for several versions now?
> 
> Kit
> 
> 
> 
> Kit Baum   |   Boston College Economics & DIW Berlin   |   http://ideas.repec.org/e/pba1.html
>                              An Introduction to Stata Programming  |   http://www.stata-press.com/books/isp.html
>   An Introduction to Modern Econometrics Using Stata  |   http://www.stata-press.com/books/imeus.html
> 
> 
> *
> *   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/
> 




Kit Baum   |   Boston College Economics & DIW Berlin   |   http://ideas.repec.org/e/pba1.html
                              An Introduction to Stata Programming  |   http://www.stata-press.com/books/isp.html
   An Introduction to Modern Econometrics Using Stata  |   http://www.stata-press.com/books/imeus.html


*
*   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/