A Monte Carlo Simulation Analysis of Panel Stationarity Tests under a Single Framework

Abstract

A unified framework, stringency criterion have been used to compare the six panel unit root tests having the null hypothesis of stationary and to find the best performer test/tests. Simulated critical values, instead of asymptotic critical values, have been used to keep the size of all tests around nominal size of 5%. Our findings suggest HD and HL tests as better performing tests as compared to other panel stationarity tests.