Page 1 of 1

GMM - questions: dummy variables and large Ts

Posted: Tue Apr 05, 2016 1:39 pm
by larchalr14
I am running regression on bank profitability and different variables.

My question is why won't GMM let me use my dummy variable? Is there a different way I can specify my dummy variable? (whether or not a bank is consider a community bank).

Second, I have 82 banks with 4 years. Is this a large enough sample size and observation size?

Thank you so much for your help. I've been looking all over the internet and in books for the answer.

Re: GMM - questions: dummy variables and large Ts

Posted: Tue Apr 05, 2016 3:25 pm
by startz
What does GMM do when you try, i.e. what do you mean by GMM won't let you?

Re: GMM - questions: dummy variables and large Ts

Posted: Wed Apr 06, 2016 12:34 pm
by larchalr14
It gives me "Near singular matrix" and as soon as I get rid of the dummy variable it runs.

Re: GMM - questions: dummy variables and large Ts

Posted: Wed Apr 06, 2016 12:36 pm
by startz
Check whether you are in the dummy variable trap.