Re: st: Pesaran CD Test - The panel is highly unbalanced

[Nick Cox <njcoxstata@gmail.com>]

Your claim of balance is contradicted by your own results which show
varying group sizes and a number of observations much less than your N
multiplied by your T.

You refer to user-written -xtcd- and -xtcsd- without explaining that
they are user-written or where they come from.

The authors of these programs are not necessarily Statalist members,
so if you do not get a more detailed answer, you may need to contact
the authors directly for support.

The last error message indicating that Stata can't find -sqrt()- will
be a side-effect of some other problem. Stata does have a -sqrt()-
function; the problem evidently lies in how it is called within one of
the programs you are using, but I can't say more on how that arises.

Please read the Statalist FAQ for advice on how to post questions.

Nick

On Mon, Jan 2, 2012 at 12:12 AM, Lim Boon Leong <lboonl@streamyx.com> wrote:

> I have a balanced panel data set of N=191 and T=6 but when I performed the
> pesaran CD test by using the following command in Stata12,
>
> . xtreg lbonus ownexr ownex2r ownex3r ceodual ined remcom1 exblock1 L.lmktbk
> L.roa L.ret L.debt L.ltotasset i.year i.sector2, re
>
> . xtcd, pesaran abs show
>
> I got the error message r(2001): The panel is highly unbalance. Not enough
> common observations across panel to perform Pesaran's test.
>
> I am quite puzzled as in the RE model, the followings are reported:
>
> Number of obs = 526, number of groups = 121, obs per group: min=1, avg=4.3,
> max=5
>
> Based on the above, my question is: Are 526 observations still insufficient
> to carry out the Pesaran CD test?  And may I know what ate the minimum
> observations required for Pesaran test?
>
> For the other balanced panel data set with the same N(191) and T (6), when I
> run the Pesaran Test I obtained the error message as shown below after the
> following command:
>
> .xtreg lsalary ownexr ownex2r ownex3r L.ldiv ceodual ined remcom1 extblock1
> L.lmktbk L.roa L.ret L.debt L.ltotasset i.year, fe
>
> .xtcsd, pesaran abs
>
> unkown function*sqrt () r(133)
>
> From the error message does it mean that for this data set I have sufficient
> common observations across panel to carry out Pesaran test but because of
> the unknown function *sqrt - Stata could not carry out the test?  May I know
> how to solve this problem?
>

*
*   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/