Lag selection for panel data using pvarsoc command - long runtime / convergence not achieved

[Hjalmar Jernstrom]

I am new to Stata and have issues with using the pvarsoc function that is contained in the pVAR package created by Inessa Love, available at: https://sites.google.com/a/hawaii.edu/inessalove/home

I have searched this topic but found no solution. I want to select optimum lag from balanced panel data with 9 companies, 30 days and variables of return and trading volume. I have done all necessary data management such as including trading calendar. I typed the following command in Stata:

. pvarsoc return volume, maxlag (7)

Where 7 is arbitrarly selected. The problem is that the program runs for about 20 mins, then I get the error message convergence not achieved. Does anyone have any idea why this is? Can anyone suggest other methods to select optimum lag for my data? I'd be very grateful for any help or input!

[Hjalmar Jernstrom]

The issue is resolved. If anyone encounter the same problem, I adjusted the variables as the absolute values of returns (expressed in decimals) and trade volumes (expressed in single units). Adjusting the values to be closer to eachother in absolute terms solved the issue, and I can run a pvarsoc.

[Franck ZIKE]

Bonjour Monsieur, Je rencontre actuellement le même problème que vous avez rencontré il y a des années. Je vous remercie pour la solution. Cependant, je ne comprends pas quand vous parlez d’ajustement des valeurs. Pouvez-vous me donner plus de détails sur ce qu'il faut faire pour résoudre ce problème? Je vous remercie