Hi, I want to perform White heteroskedasticity test, using a VAR model, but when I click on that test, it appears ”near singular matrix” error. I have 6 variables, annual data from 1995 to 2017. How could I manage this problem? Thank you!
alinavoda
you mean that despite the fact that we have an insignificant variable educ*black because prob=26% if coeficient is close to 0 our estimation presision is quite large;
