04 April 2014 1 882 Report

I am running the unit-root test of Im et al. (2003) for panel data using Stata (xtunitroot ips). I started with the least restrictive model including a constant, a trend, and the maximum possible lags (chosen by AIC). The problem is that Stata does not provide the parameter estimates and the residuals. Following the time series literature, one should test for the significance of the trend and the constant and test the autocorrelation of residuals before selecting the best model.

Do you have any idea how to get the parameter estimates and the residuals from xtunitroot in Stata?

Similar questions and discussions