Cannot run Pvarsoc command(Problem)

[Deanna Brown]

Dear
I hope you can help me in what the wrong in pvarsoc STATA to determine the optimal lags for my independent variables before run the regressions in STATA

pvarsoc depvar indep1 indep2 indep3
Running panel VAR lag order selection on estimation sample
....

Selection order criteria
Sample:  2007 - 2018                              No. of obs    =    11607
No. of panels    =    1115
Ave. no. of T    =    10.410

        
lag     CD          J      J pvalue     MBIC       MAIC        MQIC    
        -
1   .7278297          .          .          .          .        . 
2   .6544439          .          .          .          .        . 
3   .6921387          .          .          .          .        . 
4          1          .          .          .          .        . 

[Deanna Brown]

Dear Andrew Musau
Could you help me please

[Andrew Musau]

pvarsoc is from the Stata Journal, as you are asked to explain (FAQ Advice #12). Look at the option -pinstlag(numlist)- explained in

help pvarsoc

For example, you can have

pvarsoc depvar indep1 indep2 indep3, pinstl(1/5)

to use the first to fifth lags from the highest lag used in the panel VAR model as instruments. This may allow computation of the statistics. Also, read the accompanying SJ article for more details.

[Deanna Brown]

Dear Dr Andrew Musau
Thanks for valuable response
It works but the code menations that the convergence not achived , also could you please explain it for for ?The results? how I can find the optimal lag length regarding to these results?

. pvarsoc depvar indep1 indep2 indep3 , pinstl(1/5)
Running panel VAR lag order selection on estimation sample
convergence not achieved
.convergence not achieved
...

 Selection order criteria
 Sample:  2012 - 2018                              No. of obs      =      6172
                                                   No. of panels   =       989
                                                   Ave. no. of T   =     6.241

  +--------------------------------------------------------------------------+
  |   lag |    CD          J      J pvalue     MBIC       MAIC       MQIC    |
  |-------+------------------------------------------------------------------|
  |     1 |  .8385454   5040.342          0   4778.509   4980.342   4910.351 |
  |     2 |         1   6.641714   .5757351  -63.18051  -9.358286  -28.02246 |
  |     3 |  .9999919   .0101724          1  -69.81205  -15.98983    -34.654 |
  |     4 |   .999993   .0019994          1  -69.82023    -15.998  -34.66217 |
  +--------------------------------------------------------------------------+ 

[George Ford]

Add another lag (very little difference between 3/4), but you're getting into the small T. Might try on a balanced panel. You using differences or level?

[Deanna Brown]

Dear Dr.George Ford,

Thanks for your response.
Actually, I use level variables and my data is unbalanced panel.
I want to understand the output first what does it mean?
Why do you recommend to try another lag?

Thanks
Deanna