non-linear GMM

[red2]

Dear All,
Actually, I’m working on the topic of estimating the budget function based on the Eviews. Below is some few details on the method I’m working with and I was wondering if you would like to support me in addressing some issues.
The function below will be solved by non-linear GMM method:
S_t=PS_(t-1)+(1-P)βd_(t-1)+(1-P)γx_t+ e_t

S = Government budget balance
P = the current value of the fiscal policy instrument partially adjusts
d= Public Debt
X=output Gap
e= shock
Instrumental variables: Inflation rate, interest rate and real and effective exchange rate.

Basic articles tackling this function had worked with the non-linear GMM method.
The estimation of this method is mainly obtained by following two steps:
Fist step:
Non-linear 2sls are used to obtain the initial values of the above parameters.
Second step :
These parameters are used to build an optimal weighting matrix that is used to estimate the same function again.


If you don’t mind, could you please give more explanation on the obove steps? And How can we use these steps in Eviews?

I will be waiting for your feedback which will be highly appreciated and please let me know if you need any further information.


Many thanks and best regards.

[EViews Gareth]

Specify the equation as:

Code: Select all

s=c(1)*s(-1)+(1-c(1))*c(2)*dd(-1)+(1-c(1))*yx(-1) @ i int r er

Choose GMM as the estimation method, with weight updating set to Seq. N-step Iterative, with 1 step.

You don't say what sort of weighting matrix you want to use, but set it to what you want it to be. That should be it.

[trubador]

Gareth is right. But, I guess the third term should be as follows:
...+(1-c(1))*(1-c(3))*x(-1)...