Stata ARMA multiple analysis

Hi all,

I am writing my thesis but i am having some issues by estimating the variance of the income.

i have a big dataset (Panel data: 4 years, 2600 householders - every household has 1 observation for every year).

There are two ways in which i would like to proceed:

1. I would like to regress their income with a VARMA (1,1).
- how can i do it with one command instead of 2600 single regressions?
the model will be: log(income)= age + sex + year + educ + country + log(income_t-1) + e_t + e_t-1
- after that i would like to insert into a graph the coefficients MA and AR (so i will have 2600 points in the graph)

2. I would like to regress the income with a ARX(1)
income_t,h= a + cx_t,h +b(income_t-1,h)+ e_t,h with e_t,h =e'_t,h + e"_t where h is the household and t is the time. x is a vector containing all the individual variables(sex, educ, age...)
i would like to estimate e"_t so i would like to remove the individual effects and estimate (e"_t)^2


the models are more or less the same, it does not change so much but i am not able to find the proper command to write on stata.

THANKS in advance for your help, please ask me if it is not clear

BR,
Andrea