Covariate unit root tests under structural change and asymmetric STAR dynamics

摘要

In this paper, we intend to develop a new unit root testing procedure. The novelty of this methodology includes (1) accommodating possible trend breaks of unknown number, unknown dates, and unknown form by employing the Fourier form without directly estimating such breaks; (2) considering possible asymmetric STAR adjustments under the alterative; and (3) utilizing related covariates to boost the testing power. The limiting distribution of the test is derived, and the asymptotic critical values are tabulated. Simulation experiments show that the test can deliver robust size for various breaks commonly seen in economic analysis and enjoy high power property, even in small sample sizes encountered in empirical studies. The usefulness of the test is illustrated in an empirical study on the issue of debt sustainability in 18 OECD countries.