dynamic panel estimation for system of equations

[kelaynak]

Hello,

As far as I understood from previous discussions, Eviews 7 currently can not support dynamic panel estimation for a system equations with GMM.
This issue is related with Eviews or is it theoretically is impossible and therefore, there is no solution in other statistical program?
Is it possible to see support for it in new updates of eviews 7?

Is there any suggestion what I can do for that kind of system equations? Instead of GMM; can I use 3SLS or 2SLS?

My panel sytstem equation is like that
A(t,i)= A(t-1,i)+X1(t,i)+X2(t,i)+X3(t,i) + B(t,i) + C(t,i)
B(t,i)= B(t-1,i) + X3(t,i)+X4(t,i)+X5(t,i)+ A(t,i) + C(t,i)
C(t,i)= A(t,i)+ B(t,i) + X5(t,i)+X6(t,i)

Thank you for your help

[EViews Gareth]

I think you need to be more clear with your definitions. EViews does not support the A-B Dynamic Panel System Estimator.

That is completely separate from estimating a system of equations in a panel workfile, which is what your post appears to be about. Systems of equations can be estimated in a panel workfile (although there is no inbuilt support for fixed or random effects).

[kelaynak]

Gareth,
Thank you for answer.
I dont have eviews 7. But, if the following dynamic panel simultaneous system equations can be solved with it, I will use Eviews 7.

But I am abit confused. Could you please confirm that Eviews can solve the following simultaneous equations in dynamic panel?
As there are lag term of dependent variables in right terms of my system of equations, I am saying it is "dynamic panel system equations".

My panel simultenous sytstem equations are like that, there are lagged terms in right sides
A(t,i)= A(t-1,i)+X1(t,i)+X2(t,i)+X3(t,i) + B(t,i) + C(t,i)
B(t,i)= B(t-1,i) + X3(t,i)+X4(t,i)+X5(t,i)+ A(t,i) + C(t,i)
C(t,i)= A(t,i)+ B(t,i) + X5(t,i)+X6(t,i)

As far as I understood from your reply, it is possible with GMM, but not using A-B Dynamic panel estimator (Arellano&Bond and Arellano&Bover).
I will create system panel data for my simultaneous system equations, and estimate with GMM (cross section) option for my system dynamic panel data. There is no bias problem for lagged term in simulatneous panel data.
Did I understood right?

What about EC3sls? Does eviews7 use EC3SLS for estimation of simultaneous system equation in dynamic panel?

Thanks in advance

[EViews Gareth]

Nothing has been added to 7 with regards to panel data estimation. Your specification looks like you're need to estimate fixed effects models, and the system object in EViews does not have options for fixed effects, although you could, possibly, always do them yourself with dummy variables.

[kelaynak]

Gareth,
Thank you for your help.

1) Does Eviews support dynamic panel for simultaneous equations? I am planning to use lagged values of dependent variables in right side of equation for simultaneous equations in panel data . Do you think It is not bias?

Or it is just suitable for static panel for simulataneous equation system (no any lagged values of dependent variables in right side of equations as exogenous)?

2) If it is not suitable for dynamic panel, then in that case I will use static panel for simultaneous equatins and delete all lagged dependent variables A((t-1),i) and B((t-1),i) in right side of equtaions.
In that static panel data for simulatneous equations, can I use lagged values of dependent variables as instruments in GMM or 3sls ? Is there any bias problem?

Thanks in advance
Regards

[EViews Gareth]

You can put lagged dependent variables on the right hand side inside a system equation on a panel data set in EViews.