Good morning,
I want to come up with a Monte Carlo design of Fixed Effects Heteroskedastic Panel regression. The model is
Yit = B*Xit + Ai + Eit, i = 1, 2..N; t = 1, 2, ... T
where parameters B and the regressors Xit might be vectors (then "*" would be the inner product); Ai is the fixed by individual effect allowed to be correlated with the regressors, and Eit is the idiosyncratic error.
Can you propose
1) design to generate the regressors Xit
2) Heteroskedastic function making E(Eit^2 | Xi1, Xi2... XiT) = f(Xit)
?
I want to come up with a Monte Carlo design of Fixed Effects Heteroskedastic Panel regression. The model is
Yit = B*Xit + Ai + Eit, i = 1, 2..N; t = 1, 2, ... T
where parameters B and the regressors Xit might be vectors (then "*" would be the inner product); Ai is the fixed by individual effect allowed to be correlated with the regressors, and Eit is the idiosyncratic error.
Can you propose
1) design to generate the regressors Xit
2) Heteroskedastic function making E(Eit^2 | Xi1, Xi2... XiT) = f(Xit)
?
