On Stationarity and the Interpretation of the ADF Statistic
The paper considers the nature of stationarity of a time series or signal, and how it may be quantified. It is argued that a subjective assessment is as effective as one based on mathematical definitions, if one actually has finite samples of data, and that such an assessment is fundamentally based on the number of cycles of the dominant periodic component visible in the sample. It is shown by dimensional analysis that one of the most often-used measures of stationarity – the Augmented Dickey-Fuller (ADF) statistic – supports this hypothesis. The paper should be of interest not just to engineers, but also to econometricians, or anyone concerned with time series analysis and the impact of nonstationarity.
KeywordsTime series Stationarity Augmented Dickey-Fuller (ADF) statistic Dimensional analysis
KW would like to acknowledge the support of the UK Engineering and Physical Sciences Research Council (EPSRC) through grant reference numbers EP/J016942/1 and EP/K003836/2.
- 1.Ellis, A.K.: Teaching and Learning Elementary Social Studies, 9th edn. Pearson, Boston (2010)Google Scholar
- 3.Fuller, W.A.: Introduction to Statistical Time Series. Wiley Interscience, New York (1976)Google Scholar
- 9.Silverman, B.W.: Density Estimation for Statistics and Data Analysis. Monographs on Statistics and Applied Probability. Chapman and Hall, London (1986)Google Scholar
- 10.Scott, D.: Multivariate Density Estimation: Theory, Practice and Visualization. Wiley, New York (1992)Google Scholar
- 11.Montgomery, D.C.: Introduction to Statistical Quality Control. Wiley, Hoboken (2009)Google Scholar
- 12.Hurd, H.L., Miamee, A.: Periodically Correlated Random Sequences: Spectral Theory and Practice. Wiley, Hoboken (2007)Google Scholar
- 13.Kramer, C., de Smet, C.A.M., de Roeck, G.: Z24 Bridge damage detection tests. In: Proceedings of IMAC 17, the International Modal Analysis Conference, Kissimmee, pp. 1023–1029 (1999)Google Scholar
- 16.Perman, R.: Cointegration: an introduction to the literature. J. Econ. Stud. 18, 3–30 (1993)Google Scholar
- 17.DeJong, D., Nankervis, J., Savin, N., Whiteman, C.: Integration versus trend-stationarity in macroeconomic time series. Technical report working paper, Department of Economics, University of Iowa (1988)Google Scholar
- 18.Nelson, C.R., Prosser, C.R.: Trends and random walks in macroeconomic time series: some evidence and implications. J. Monet. Econ. 10, 139–162 (1982)Google Scholar
- 21.Shiller, R.J., Perron, P.: Testing the random walk hypothesis: power versus frequency of observation. Technical report, National Bureau of Economic Research, Cambridge (1985)Google Scholar