The Kalman filter method for break point estimation in unit root tests

Abstract

In this study, in addition to Zivot–Andrews (1992), Perron (1997) and Schwarz Bayesian Criteria (SBC) approaches of the true break point estimation performances of the Kalman filter method is examined using Monte Carlo simulation experiments. Our simulation results show that the SBC and Kalman filter methods both exhibit a good performance in estimating true break point.