Comparing of Unit Root with and without Structural Breaks: Monte Carlo Evaluation
Author: Muhammad Aslam

The unit root became the most important feature that directed to the construction of new time series econometrics, and study of time series structural breaks was a specific area of unit root research. Conventional procedures assume the break and apply a test accordingly. This leads to identification of spurious breaks, and therefore biased results, Lee and Strazicich, (2001). We suggest an alternative strategy where we propose to test for structural breaks before applying unit root test. The debates of Structural breaks in unit root testing starts with Perron (1989). Nelson and Plossoro(1982) found unit roots in 1 out of 14 macroeconomic time series of US economy and Perron (1989) taking the Nelson and Plossor’s data set, reversed the findings for 11 out of 14 series. The later development in unit roots with structural breaks developed procedures for endogenizing structural breaks (Zivot and Andrew, 1992; Christianoo1992 etc). However, the original Perron’s Procedures and the later development in unit root testing with structural breaks, assume that there is a structural break. The studies endogenizing structural breaks also assume the break and determine the break date endogenously. We propose that structural breaks should be tested for existence. The purpose of this study is to compare the existing strategy with the proposed strategy using Monte Carlo experiments. Our results are indicating that existing strategy is significantly suffering in power problems but the proposed strategy is better and significantly perform as compare to conventional or existing strategy. Supervisor:- Dr. Atiq-ur-Rehman

Meta Data

Supervisor: Atiq Ur Rahman

Related Thesis​