EViews.com

Hi,
I am doing a panel VAR analysis with 6 endogeneous variables. I have quaterly data for 11 countries from 1991Q1 - 2011Q4. I also include dummy variable which takes value 1 for dates later than 1999, in order to control for change in currency Euro. Now I want add country fixed effects. To do that I include '@expand(@crossid)' in the box for exogeneous variable in VAR. I also include the euro dummy variable here. But when I am trying to estimate, its showing 'near singular matrix', even if I exclude the constant. If I exclude the euro dummy and only include '@expand(@crossid)', then I am getting result. What is the problem here? Why I cannot include both the dummies? Is it a 'dummy variable trap' problem? If yes, how can I solve it??? ...PLEASE HELP!!!
Thank you.

It probably is the dummy variable trap. You can't include two full sets of dummies. You have to drop one dummy from one category.

Thank you for your reply. Can you please tell me how to drop one dummy from one category? There is a command called '@dropfirst', but how should I use it? If I write @expand(@crossid,@dropfirst), its still showing the same message 'near singular matrix' !!

That looks right, so you might want to post your entire command.

Let me tell you what I am doing. I am running a panel VAR with 6 endogeneous variables. I have data for 11 countries from 1991Q1 to 2011Q4. I also have a dummy variable for tranmission of currency to euro, which takes value 1 after 1999Q1 and zero otherwise. I have pooled the data for all countries to get the 6 endogeneous variable and the dummy variable. Now I also want to add country fixed effects. When I am running the VAR, I go to Quick> Estimate VAR. Then in the box I put the endogeneous variables under the box 'endogeneous variables' and in the box for exogeneous variable I write 'dummy @expand(@crossid,@dropfirst)'. Here dummy is the dummy variable for shift in euro and @expand(@crossid,@dropfirst) is to control for country fixed effects. Then when I am clicking OK, its saying 'near singular matrix'!! I get the same message even if I add a constant. What should I do?

Let me tell you what I am doing. I am running a panel VAR with 6 endogeneous variables. I have data for 11 countries from 1991Q1 to 2011Q4. I also have a dummy variable for tranmission of currency to euro, which takes value 1 after 1999Q1 and zero otherwise. I have pooled the data for all countries to get the 6 endogeneous variable and the dummy variable. Now I also want to add country fixed effects. When I am running the VAR, I go to Quick> Estimate VAR. Then in the box I put the endogeneous variables under the box 'endogeneous variables' and in the box for exogeneous variable I write 'dummy @expand(@crossid,@dropfirst)'. Here dummy is the dummy variable for shift in euro and @expand(@crossid,@dropfirst) is to control for country fixed effects. Then when I am clicking OK, its saying 'near singular matrix'!! I get the same message even if I add a constant. What should I do?

Everything you are doing sounds right. If you post your workfile, including the VAR object, perhaps someone will spot a problem.