Error message in GMM estimation

[facosta]

Hello there (by the way thanks a lot Gareth for the help with my last question).
I am doing a Panel estimation with GMM.
I am trying to estimate an equation with three lags of the dependant variable and other exogenous variables by GMM and fixed effects. However, when I estimate the equation, an error message appears. It says something like "Order condition violated-Insufficient instruments in GMM(..)..."

I´m thinking maybe I write the code in a wrong way, I am regressing:

gmm(cx=f, per=f) y y(-1) y(-2) y(-3) x1 x2 x3 x4 x5

Any idea will be appreciated.
Thanks!

[facosta]

Well, I was asking because I was trying to reply the STATA command "xtabond", where I regress:
xtabond y x1 x2 x3, lags(3) robust
Since in Eviews there are many options I am little confused.
Thank you.

[trubador]

You should include a list of instrumental variables. Just put a "@" symbol after the regressors and add your instrumental variables. Something like the following:

Code: Select all

gmm(cx=f, per=f) y x1 x2 x3 x4 @ y(-1 to -3) x5

Please do not forget that, there must be at least as many instrumental variables as there are coefficients to be estimated. EViews automatically adds a constant as an instrument.

[facosta]

Thank you very much trubador.
I tried the following:

"gmm(cx=f, per=f) y y(-1) y(-2) y(-3) x1 x2 x3 x4 x5 x6 x7 @ @dyn(y) x1(-1) x2(-1) x3(-1) x4(-1) x5(-1) x6(-1) x7(-1)"

But i still get the same message. Looking at this I really don´t know what can I change since I believe I am using all the possible variables as instruments.
I am sorry the questions, it´s just I am used to estimate panel equations with STATA.

Regards.