Misspecification of the breaking date in series with a change in the growth rate. Effect on the LBI test for stationarity

Abstract

We study the asymptotic behaviour of the LBI statistic for testing stationarity around a trend with a change in the growth rate when the breaking date is erroneously positioned, showing its divergence. The analysis is completed with Monte Carlo simulations.

Notes

¹ Lee and Strazicich (2001b) analyse the tests proposed by Zivot and Andrews (1992 and Perron (1997, concluding that they tend to estimate the break point erroneously and the greater the increase of the magnitude of the change, the greater the error.