st: Re: levin-lin-chu

[Kit Baum <baum@bc.edu>]

There is obviously something wrong with the way in which your time 
variable is identified when you examine lnQ. Once you have the panel 
properly tsset, the test should work fine.

Thanks
Kit

Kit Baum, Boston College Economics
http://ideas.repec.org/e/pba1.html

On Nov 29, 2004, at 2:33 AM, Naceur wrote:


> lnQ, lnREus, lnREjp, lnREmq, lnREsk, lnREhk, lnREau are respectively 
> the =
> canada capacity constraint and the real exchange rate between canada 
> and =
> USA, Japon, Mexico, South-Korea, Hong-Kong and Australia (independent =
> variables).
> At the begining I have a Panel and I want to identify each unit's time 
> =
> series, in order to achieve this goal, having the result of  tsset =
> country datevar to define both a panel variable and a time variable.
> I tested lnq variable  Ihave this result (there is no problem):
> =20
> levinlin lnq, lag(0)
> =20
> Levin-Lin-Chu test for lnq        Deterministics chosen: constant
> Pooled ADF test, N,T =3D (6,191)    Obs =3D 1140  =20
> Augmented by 0 lags (average)     Truncation: 18 lags
> coefficient         t-value        t-star            P > t
>  -0.19554      -11.273      -9.85790      0.0000
> =20
> But When I tested lnQ, I had this results:
> =20
> levinlin lnQ, lag(0)
> Levin-Lin-Chu test for lnQ        Deterministics chosen: constant
> no observations
> r(2000);
> =20
> In the end the tsset of this variable give this resuts:
> =20
> tsset lnQ
> time variable must contain only integer values
> r(451);
> =20
*
*   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/