03-22-2014 02:48 PM
I hope you are having a nice day.
I have been trying to run a 2-step GMM with SAS to estimate my dynamic panel data but it seems that I am doing something wrong.
I attached a picture of what my data looks like.
So basically I have intraday observations over a number days for different contracts. My time series are set by the "start" variable which goes 08:05- 08:10 - 08:15-... My cross-section is between different contracts.
This is the procedure that I used but it is not working:
proc panel data=basvoldum_short_lond;
model BASw = tod1-tod105 / gmm nolevels twostep;
id cnt start;
Sorry for taking your time and thank you for your tremendous help.
03-24-2014 09:15 AM
So the first question is how is it "not working"? Are there error messages in the log, are the results un-interprebable, or does the procedure just not run?
I would also recommend reposting this in the SAS Forecasting and Econometrics forum, as the folks that follow that group will see this sooner than here in the general procedures forum.
03-24-2014 07:08 PM
Thanks for your answer.
The SAS error reported says that the time series is not properly sorted, although I sorted the data before.
Thank you for the tip. I did not know there was a SAS Forecasting and Econometrics forum. I will post my question there.