Re: st: RE: dfuller: why do I get different results?

[Yuval Arbel <yuval.arbel@gmail.com>]

It took me two hours to run the -xtunit- command, but here is what I get:

. xtunitroot fisher reduct_per,dfuller lag(1)
could not compute test for panel 119
could not compute test for panel 125
could not compute test for panel 154
could not compute test for panel 782
could not compute test for panel 814
could not compute test for panel 1050
could not compute test for panel 2707
could not compute test for panel 3587
could not compute test for panel 3894
could not compute test for panel 4269
could not compute test for panel 7289
could not compute test for panel 8104
could not compute test for panel 9155
could not compute test for panel 9340
could not compute test for panel 9391
could not compute test for panel 9542
could not compute test for panel 9547

Fisher-type unit-root test for reduct_per
Based on augmented Dickey-Fuller tests
-----------------------------------------
Ho: All panels contain unit roots           Number of panels       =   9547
Ha: At least one panel is stationary        Avg. number of periods =  53.19

AR parameter: Panel-specific                Asymptotics: T -> Infinity
Panel means:  Included
Time trend:   Not included
Drift term:   Not included                  ADF regressions: 1 lag
------------------------------------------------------------------------------
                                  Statistic      p-value
------------------------------------------------------------------------------
 Inverse chi-squared(19060)P      8814.2739       1.0000
 Inverse normal            Z        60.7097       1.0000
 Inverse logit t(46659)    L*       55.5908       1.0000
 Modified inv. chi-squared Pm      -52.4767       1.0000
------------------------------------------------------------------------------
 P statistic requires number of panels to be finite.
 Other statistics are suitable for finite or infinite number of panels.
------------------------------------------------------------------------------

.
I'm happy with the results, because they show that tenants could not
anticipate a long-run mean reduction rates.

Many thanks again anyway
Yuval


On Fri, Nov 18, 2011 at 7:08 PM, Yuval Arbel <yuval.arbel@gmail.com> wrote:
> Note that it did not work because these commands need balanced panel,
> and I have unbalanced panel.
>
> I'm trying now the command:
>
> xtunitroot fisher reduct_per,dfuller lag(1)
>
> according to the help file:   xtunitroot fisher does not require
> strongly balanced data, and the individual series can have gaps.
>
> It takes stata a long time to run this command (I have 9547 panels) -
> but it suppose to do exactly what I need: to run the dfuller on each
> panel separately and construct average statistic.
>
> I let you know whether it worked eventually
>
> many thanks,
> Yuval
>
> On Fri, Nov 18, 2011 at 6:50 PM, Nick Cox <n.j.cox@durham.ac.uk> wrote:
>> Others may be able to give constructive comments. I note only that your -xtset- reveals that your data are gappy, which underlines why using [_n-1] to get previous values did not work for you.
>>
>> Nick
>> n.j.cox@durham.ac.uk
>>
>> Yuval Arbel
>>
>> Nick,
>>
>> Note that I looked very carefully at the material Austin Nicols sent
>> me. His implication was that I should run a unit root test on a panel
>> dataset. I tried to run the following commands, but they did not seem
>> to work:
>>
>> . xtset appt time_index
>>       panel variable:  appt (unbalanced)
>>        time variable:  time_index, 0 to 114, but with gaps
>>                delta:  1 unit
>>
>> . xtunitroot llc reduct_per
>> Levin-Lin-Chiu test requires strongly balanced data
>> r(498);
>>
>> . xtunitroot ht reduct_per
>> Harris-Tzavalis test requires strongly balanced data
>> r(498);
>>
>> . xtunitroot breitung reduct_per
>> Breitung test requires strongly balanced data
>> r(498);
>>
>>
>> On Fri, Nov 18, 2011 at 3:08 PM, Nick Cox <n.j.cox@durham.ac.uk> wrote:
>>> See my earlier answer on using -foreach- with an FAQ as reference. I don't know if it will work, in the sense of doing what you want.
>>>
>>> By the way, I have a horrible feeling that you are in econometric peril here in some sense, and I am not endorsing your choices.
>>>
>>> You've already chosen to ignore the advice of Austin Nichols. If I were doing econometrics, I would want a very good reason to ignore Austin Nichols.
>>>
>>
>> *
>> *   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/
>>
>
>
>
> --
> Dr. Yuval Arbel
> School of Business
> Carmel Academic Center
> 4 Shaar Palmer Street, Haifa, Israel
> e-mail: yuval.arbel@gmail.com
>



-- 
Dr. Yuval Arbel
School of Business
Carmel Academic Center
4 Shaar Palmer Street, Haifa, Israel
e-mail: yuval.arbel@gmail.com

*
*   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/