Durbin-Watson Type Unit Root Test Statistics

  • Kim, Byung-Soo (Department of Applied Statistics, Inje University, Kimhae 621-749, Korea, E-mail : kbs@stat.inje.ac.kr) ;
  • Cho, Sin-Sup (Department of Statistics, Seoul National University, Seoul 151-742, Korea, E-mail : sinsupch@plaza.snu.ac.kr)
  • Published : 1998.03.01

Abstract

In the analysis of time series it is an important issue to determine whether a time series under study is stationary. For the test of the stationary of the time series the Dickey-Fuller (DF) type tests have been mainly used. In this paper, we consider the regular unit root tests and seasonal unit root tests based on the generalized Durbin-Watson (DW) statistics when the errors are independent. The limiting distributions of the proposed DW-type test statistics are the functionals of standard Brownian motions. We also obtain the finite distributions and powers of the DW-type test statistics and compare the performances with the DF-type tests. It is observed that the DW-type test statistics have good behaviors against the DF-type test statistics especially in the nonzero (seasonal) mean model.

Keywords

References

  1. Review of Economic Studies v.53 On the theory of testing for unit roots in observed time series Bhargava, A.
  2. Journal of the American Statistical Association v.74 Distribution of the Estimators for Autoregressive Time Series with a Unit Root Dickey, D. A.;Fuller, W. A.
  3. Econometrica v.49 Likelihood Ratio Test Statistics for Autoregressive Time Series with a Unit Root Dickey, D. A.;Fuller, W. A.
  4. Journal of the American Statistical Association v.79 Testing for Unit Roots in Seasonal Time Series Dickey, D. A.;Hasza, D. P.;Fuller, W. A.
  5. Biometrika v.37 Testing for serial correlation in least squares regression I Durbin, J.;Watson, G. S.
  6. Biometrika v.38 Testing for serial correlation in least squares regression Ⅱ Durbin, J.;Watson, G. S.
  7. Biometrika v.48 Computing the distribution of quadratic forms in normal variables Imhof, J. P.
  8. Journal of Econometrics v.46 Limiting power of unit-root tests in time series regression Nabeya, S.;Tanaka, K.
  9. Journal of Business and Economic Statistics v.12 A comparison of the unit-root test criteria Pantula, S. G.;Gonzalez-Farias, G.;Fuller, W. A.
  10. Econometrica v.51 Testing residuals form least squares regression for being generated by a Gaussian random work Sargan, J. D.;Bhargava, A.
  11. Time series analysis Tanaka, K.
  12. Communication in Statistics v.2 Generalization of the Durbin-Watson statistic for higher order autoregressive processes Vinod, H. D.