Size and power properties of structural break unit root tests

Abstract

In this article, we compare the small sample size and power properties of a newly developed endogenous structural break unit root test of Narayan and Popp (NP, 2010) with the existing two break unit root tests, namely the Lumsdaine and Papell (LP, 1997) and the Lee and Strazicich (LS, 2003) tests. In contrast to the widely used LP and LS tests, the NP test chooses the break date by maximizing the significance of the break dummy coefficient. Using Monte Carlo simulations, we show that the NP test has better size and high power, and identifies the structural breaks accurately. Power and size comparisons of the NP test with the LP and LS tests reveal that the NP test is significantly superior.

Keywords:

• unit root test
• structural breaks
• sample size
• Monte Carlo

JEL Classification:

• C12
• C13

Notes

¹ Thanks to Junsoo Lee for providing the GAUSS codes of the LP and LS tests.

² A situation in which one always identifies the break date correctly, i.e. , is like knowing the break date.

³ The Perron test is a Dickey–Fuller-type test. As stated by Perron (1989), the exogeneous break test is invariant to the break date if we account for the break. In case of an unknown break date, the test is invariant if it is possible to identify the break date accurately. So, for an endogenous break unit root test, it is important to meet these preconditions. The advantage of the NP test is that it identifies the break date very accurately, even in the case of very small breaks. The reason for this is the slightly different approach for estimating the break date. One indicator of good properties of the NP test is that the critical values of the endogenous and exogenous tests are very similar in finite samples, which is not the case with the other tests considered in this article.