EViews.com

Hello. Please help me with this problem.

I'm estimating a equation with 4 coefficients using GMM [with the following options: (1) HAC (Newey-West); (2) Iterate to Convergence;] and using 7 instrumens (in especific i'm using lags of variables like instruments).
I have more instruments that coeficients and Eviews send me the error message "Order condition violated-Insufficient instrumens", I don't know what is happend with this?

And in some cases, if I change the order of the instrument (the same instruments but in different order) Eviews can do the estimation.
The order of the instruments it's important?
Can you help me? Thanks

Post the workfile, along with the equation specification.

Hello, here are more specific details and the Workfile in Eviews 8.

Specification:
GMM(METHOD=CONVERGE,INSTWGT=HAC,NOCINST) INF=C(1)*INF(1)+C(2)*GAP+C(3)*PT+(1-C(1)-C(3))*INF(-1)+C(4)*D2010M1

1) So, with this instruments order i can't estimate, because Eviews send me the error "Order condition violated-Insufficient instrumens";
@ infus(-7 to -10) inf(-1 to -1) gap(-0 to -0) dner(-7 to -7) d(d2010m1)

2) when i change the order of instruments Eviews can do it without a problem
@ inf(-1 to -1) infus(-7 to -10) gap(-0 to -0) dner(-7 to -7) d(d2010m1)

I'm confused about this problem. This is happening me with differents sets of instruments.
Thank you.

Looks like the weighting matrix is very close to singular, and is just failing the singularity check with one order, and just passing with another order.

OK, but then the error message is incorrect, didn't?
There are a problem with set of instruments but isn't this "Order Condition Violated - Insufficient Instruments".
Thank you

When the weighting matrix is singular, EViews drops instruments so that estimation can occur. If it drops too many, eventually you hit insufficient instruments.

It has been a great help. Greetings and thank you very much.