Estimation of the parameters for unstable AR models
- 48 Downloads
This paper is concerned with the unstable autoregressive process which satisfies the unstable autoregressive (AR) modelU(B)G(B)x t =ε t , where all the roots of the polynomialsU(z) andG(z) lie on and outside the unit circle respectively. We propose several procedures to estimate the coefficients ofU(z) andG(z) separately, in order to guarantee that the estimated polynomials ofU(z) andG(z) have all the roots lying on and outside the unit circle respectively. The estimators of the coefficients ofU(z) andG(z) are shown to be of strong consistency. The limiting distribution of the estimators of the coefficients ofU(B)G(B) are obtained for some special cases.
Key wordsUnstable AR model estimation parameters strong consistency asymptotic distribution
Unable to display preview. Download preview PDF.
- Ahtola, J.A. and Tiao, G.C. Distributions of Least Squares Estimators of Autoregressive Parameters for a Process with Complex Roots on the Unit Circle.J. Time Ser. Anal., 1987, 8: 1–14.Google Scholar
- An, H.Z, Chen, Z.G. and Hannan, E.J. Autocorrelation, Autoregressive and Autoregressive Approximation.Ann. Statist., 1982, 10: 926–936.Google Scholar
- Chan, N.H. and Wei C.Z. Limiting Distribution of Least Squares Estimates of Unstable Autoregressive Process.Ann. Statist., 1988, 16: 367–401.Google Scholar
- Huang, D. Selecting Order for General Autoregressive Models by Minimum Description Length.J. Time Ser. Anal., 1990, 11: 107–119.Google Scholar
- Lai, T.L. and Wei C.Z. Asymptotic Properties of General Autoregressive Models and Strong Consistency of Least-squares Estimates of Their Parameters.J. Multivariate Anal., 1983, 13: 1–23.Google Scholar
- Li G. Estimates of Orders and Parameters in General ARMA Models in Prob. & Statist., ed. by Jiang et al. World Sci. Pub. Co. Pte. Ltd., Singapore, 1992, 111–137.Google Scholar
- Tiao, G.C. and Tsay, R.S. Consistency Properties of Least Squares Estimates of Autoregressive Parameters in ARMA Models.Ann. Statist., 1983, 11: 856–871.Google Scholar