Are US macroeconomic series difference stationary or trend-break stationary?

Abstract

This article tests for the presence of a unit-root in all time series included in the extended Nelson–Plosser data set using the statistics devised by Zivot and Andrews, Perron and Murray and Zivot. It specifies the mixed model characterization of the trend-break stationary alternative that allows for a simultaneous break in both the intercept and slope of the trend-function. It rejects the unit-root null hypothesis for real GNP, nominal GNP, real per capita GNP, industrial production, employment, GNP deflator, nominal wages, interest rate and common stock prices. Use of appropriate critical values to assess the significance of the trend-function coefficients reveals that the slope-break should be included in real GNP, nominal GNP, real per capita GNP, nominal wages, interest rate and common stock prices. The results indicate that there is less evidence against the unit-root hypothesis with the extended Nelson–Plosser data compared to the original Nelson–Plosser data.

Acknowledgements

Part of this research was supported by a Summer Fellowship grant at Xavier University, and the Williams Intellectual Contributions Development Summer Grant, William College of Business, Xavier University. The author thanks Herman Bierens for kindly providing the extended Nelson–Plosser data set.