Page 1 of 1

How to backout parameters underlying estimated coefficients?

Posted: Thu Apr 16, 2015 10:07 am
by mguennew
Dear EViews forum users and admins,

I would like estimate the following equation:

C(t) = λ0 + λ1*C(t-1) + λ2*Y(t-1) + λ3*Y(t-2) + λ4*G(t-1) + λ5*G(t-2) + v(t).

However, I am not interested in the obtained estimations for the lambdas. There is a system of nonlinear equations underlying the regression coefficients, they are functions of my model's structural parameters, i.e. λ0 = ƒ(α,δ,θ,κ,μ,π) and λ1 = g(α,δ,θ,κ,μ,π) etc.
I would like to back out the structural parameters of interest and test them for significance (whether they significantly deviate from 0 or 1 respectively).

This should be feasible in EViews when using FIML estimation methods. Does anybody know how to do this?

Thank you very much in advance, I appreciate any help.

Best regards,
Maximilian

Re: How to backout parameters underlying estimated coefficie

Posted: Thu Apr 16, 2015 12:09 pm
by startz
You may be able to do this simply by substituting f(), g(), etc. for the coefficients and doing nonlinear regression.