EViews.com

Hi, I am trying to make a model with restrictions on coefficients.
I have defined one system for each country with the following restrictions :
system sys_USA
sys_USA.append LR_USA = C(1) + C(2) * SR_USA + C(3) * INF_USA + C(4) *GDP_USA + + [AR(1) = C(5)] @ INSTR_LIST_USA
sys_USA.append C(5)=0.85 @ INSTR_LIST_USA
sys_USA.append C(2)=0 @ INSTR_LIST_USA

Then I estimate each system using TSLS and that works perfectly.

The problem is when I am trying to make a model ( proc: makemodel) or to merge this system into an existing model (proc:merge).
It gives me the following error:
Syntax error in '' sys_usa''.
How can I make a model with this kind of system?
Thanks in advance for your help.

You cannot apply restrictions like that. You'll have to replace the coefficients in the first line.

Ok thanks for the reply.
When I replace the coefficient, it works for C(2) = 0
sys_USA.append LR_USA = C(1) + 0 * SR_USA + C(3) * INF_USA + C(4) *GDP_USA + [AR(1) = C(5)] @ INSTR_LIST_USA
but it doesn't work for C(5) = 0.85
sys_USA.append LR_USA = C(1) + 0 * SR_USA + C(3) * INF_USA + C(4) *GDP_USA + [AR(1) = 0.85] @ INSTR_LIST_USA

Any idea how to replace the AR(1) coefficient ?

Ah, right. You cannot.

Ok! Thanks a lot for your answers !
I also want to run endogeneity tests for a set of regressors.
Do I need to test each regressor individually for endogeneity or all at once?
I saw that Eviews allows to test one at the time or all at once, but I am wondering what is the right way to do the test (it is more a theorical question).