PPP test for Asian countries and regions: new evidence from a wild bootstrap AESTAR test

ABSTRACT

This article investigates the effects of time-varying variance on the asymmetric exponential smooth transition autoregressive (AESTAR) unit root test. We propose a wild bootstrap-based implementation of the test, which is asymptotically valid under time-varying variance. We apply our proposed method to test the Purchasing Power Parity (PPP) hypothesis for Asian countries and regions, and find that our proposed test provides stronger evidence against the PPP hypothesis than the conventional AESTAR test.

KEYWORDS:

• Purchasing Power Parity (PPP)
• wild bootstrap
• AESTAR
• time-varying variance

JEL CLASSIFICATION:

• C22
• F31

Acknowledgements

The authors thank an anonymous referee and the co-editor (David Peel) of this journal for their helpful comments, suggestions and time spent in reading this article. Any remaining errors are ours. This work was supported by the National Natural Science Foundation of China (71671070).

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ The notation in this article follows Cavaliere and Taylor (2007).

² Detailed proofs of the asymptotic theory in this article are available upon request.

³ In the simulation study, we consider models with two or four volatility breaks. The volatility break fractions are selected so that the high volatility periods are relatively short. The results confirm that there exist similar abrupt breaks in the simulated data and the test is robust. The detailed results are available upon request. We restrict attention to the PPP hypothesis with constant mean and the ‘quasi PPP’ with changing mean (Hegwood and Papell 1998) is beyond the scope of this article. Thus, we employ volatility break model rather than mean break model.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [71671070].