I have a quantile regression in Stata which I'd like to replicate in eViews. While the coefficients are the same, the standard errors are different (irrespective of the method I use to compute the covariance matrix).
In Stata I use default options, while I've varied the cov-option in eViews.
The commands are:
Stata -> qreg dependent.variable independent.variables, quantile(.9)
eViews -> equation eq_name.qreg(quant=0.9, cov=boot, btmethod=mcmba) dependent_variable c independent_variables
As said, for the cov-option in eViews I've already tried all possible values.