Abstract
This article investigates the relative small sample performance of some robust unit root tests by means of a simulation study. We propose to investigate the stability of unit root test under various kinds of contamination models, and then the article proposes two new statistics that perform better than the celebrated Dickey-Fuller statistic. We illustrate the performances of the tests and their discrepancies with the Dickey-Fuller unit root test through an empirical example based on the US/Finland real exchange rate series.
Acknowledgments
We thank the referees for their helpful comments. This research was supported by grants from the Spanish Agency for International Development Cooperation (AECID) and MTM2009-14087-C04-01, 2010–2012, ECO2011-25706 projects.
Notes
To obtain the Kernel density estimates, we use the so-called triweight Kernel with bandwidth of 0.08.