The Annals of Statistics

Efficient detection of random coefficients in autoregressive models

Abdelhadi Akharif and Marc Hallin

Full-text: Open access


The problem of detecting randomness in the coefficients of an AR(p) model, that is, the problem of testing ordinary AR(p) dependence against the alternative of a random coefficient autoregressive [RCAR(p)] model is considered. A nonstandard LAN property is established for RCAR(p) models in the vicinity of AR(p) ones. Two main problems arise in this context. The first problem is related to the statistical model itself: Gaussian assumptions are highly unrealistic in a nonlinear context, and innovation densities should be treated as nuisance parameters. The resulting semiparametric model however appears to be severely nonadaptive. In contrast with the linear ARMA case, pseudo-Gaussian likelihood methods here are invalid under non-Gaussian densities; even the innovation variance cannot be estimated without a strict loss of efficiency. This problem is solved using a general result by Hallin and Werker, which provides semiparametrically efficient central sequences without going through explicit tangent space calculations. The second problem is related to the fact that the testing problem under study is intrinsically one-sided, while the case of multiparameter one-sided alternatives is not covered by classical asymptotic theory under LAN. A concept of locally asymptotically most stringent somewhere efficient test is proposed in order to cope with this one-sided nature of the problem.

Article information

Ann. Statist., Volume 31, Number 2 (2003), 675-704.

First available in Project Euclid: 22 April 2003

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

Primary: 62G10: Hypothesis testing 62M10: Time series, auto-correlation, regression, etc. [See also 91B84]

Random coefficient autoregressive models local asymptotic normality one-sided multiparameter tests one-sided multiparameter alternatives


Akharif, Abdelhadi; Hallin, Marc. Efficient detection of random coefficients in autoregressive models. Ann. Statist. 31 (2003), no. 2, 675--704. doi:10.1214/aos/1051027885.

Export citation


  • ABELSON, R. P. and TUKEY, J. W. (1963). Efficient utilization of non-numerical information in quatitative analysis: General theory and the case of simple order. Ann. Math. Statist. 34 1347-1369.
  • AND EL, J. (1976). Autoregressive series with random parameters. Math. Oper. Statist. 7 735-741.
  • BARTHOLOMEW, D. J. (1961). A test of homogeneity of means under restricted alternatives. J. Roy. Statist. Soc. Ser. B 23 239-281.
  • BENGHABRIT, Y. and HALLIN, M. (1996). Locally asy mptotically optimal tests for autoregressive against bilinear serial dependence. Statist. Sinica 6 147-169.
  • BENGHABRIT, Y. and HALLIN, M. (1998). Locally asy mptotically optimal tests for AR(p) against diagonal bilinear dependence. J. Statist. Plann. Inference 68 47-63.
  • BERGER, R. L. (1989). Uniformly more powerful tests for hy potheses concerning linear inequalities and normal means. J. Amer. Statist. Assoc. 84 192-199.
  • BICKEL, P. J., KLAASSEN, C. A. J., RITOV, Y. and WELLNER, J. A. (1993). Efficient and Adaptive Estimation for Semiparametric Models. Johns Hopkins Univ. Press, Baltimore, MD.
  • CHACKO, V. J. (1963). Testing homogeneity against ordered alternatives. Ann. Math. Statist. 34 945-956.
  • CHOI, S., HALL, W. J. and SCHICK, A. (1996). Asy mptotically uniformly most powerful tests in parametric and semiparametric models. Ann. Statist. 24 841-861.
  • CONLISK, J. (1976). A further note on stability in a random coefficient model. Internat. Econom. Rev. 17 759-764.
  • DROST, F. C., KLAASSEN, C. A. J. and WERKER, B. J. M. (1997). Adaptive estimation in timeseries models. Ann. Statist. 25 786-817.
  • EHM, W., MAMMEN, E. and MÜLLER, D. W. (1995). Power robustification of approximately linear tests. J. Amer. Statist. Assoc. 90 1025-1033.
  • FEIGIN, P. D. and TWEEDIE, R. L. (1985). Random coefficient autoregressive processes: A Markov chain analysis of stationarity and finiteness of moments. J. Time Ser. Anal. 6 1-14.
  • GAREL, B. and HALLIN, M. (1995). Local asy mptotic normality of multivariate ARMA processes with a linear trend. Ann. Inst. Statist. Math. 47 551-579.
  • GRANGER, C. W. J. and ANDERSEN, A. P. (1978). An Introduction to Bilinear Time Series Models. Vandenhoeck and Ruprecht, G ottingen.
  • GUÉGAN, D. (1994). Séries chronologiques non linéaires à temps discret. Economica, Paris.
  • GUTMANN, S. (1987). Tests uniformly more powerful than uniformly most powerful monotone tests. J. Statist. Plann. Inference 17 279-292.
  • GUy TON, D. A., ZHANG, N. F. and FOUTZ, R. V. (1986). A random parameter process for modeling and forecasting time series. J. Time Ser. Anal. 7 105-115.
  • HÁJEK, J. (1972). Local asy mptotic minimax and admissibility in estimation. Proc. Sixth Berkeley Sy mp. Math. Statist. Probab. 1 175-194. Univ. California Press, Berkeley.
  • HÁJEK, J. and SIDÁK, Z. (1967). Theory of Rank Tests. Academic Press, New York.
  • HÁJEK, J., SIDÁK, Z. and SEN, P. K. (1999). Theory of Rank Tests, 2nd ed. Academic Press, New York.
  • HALLIN, M. and PURI, M. L. (1994). Aligned rank tests for linear models with autocorrelated error terms. J. Multivariate Anal. 50 175-237.
  • HALLIN, M. and WERKER, B. (1999). Optimal testing for semiparametric AR models: From Gaussian Lagrange multipliers to autoregression rank scores and adaptive tests. In Asy mptotics, Nonparametrics and Time Series (S. Ghosh, ed.) 295-350. Dekker, New York.
  • HALLIN, M. and WERKER, B. (2003). Semi-parametric efficiency, distribution-freeness, and invariance. Bernoulli 9 137-165.
  • HWANG, S. Y. and BASAWA, I. V. (1993a). Asy mptotic optimal inference for a class of nonlinear time series models. Stochastic Process. Appl. 46 91-113.
  • HWANG, S. Y. and BASAWA, I. V. (1993b). Parameter estimation in a regression model with random coefficient autoregressive errors. J. Statist. Plann. Inference 36 57-67.
  • HWANG, S. Y., BASAWA, I. V. and REEVES, J. (1994). The asy mptotic distribution of residual autocorrelations and related tests of fit for a class of nonlinear time series models. Statist. Sinica 4 107-125.
  • HWANG, S. Y. and BASAWA, I. V. (1998). Parameter estimation for generalized random coefficient autoregressive processes. J. Statist. Plann. Inference 68 323-337.
  • KOUL, H. L. and SCHICK, A. (1996). Adaptive estimation in a random coefficient autoregressive model. Ann. Statist. 24 1025-1052.
  • KOUL, H. L. and SCHICK, A. (1997). Efficient estimation in nonlinear autoregressive time series models. Bernoulli 3 247-277.
  • KREISS, J. P. (1987). On adaptive estimation in stationary ARMA processes. Ann. Statist. 15 112- 133.
  • KUDÔ, A. (1963). A multivariate analogue of the one-sided test. Biometrika 50 403-418.
  • LE CAM, L. (1986). Asy mptotic Methods in Statistical Decision Theory. Springer, New York.
  • LE CAM, L. and YANG, G. L. (1990). Asy mptotics in Statistics: Some Basic Concepts. Springer, New York.
  • LEE, S. (1998). Coefficient constancy test in a random coefficient autoregressive model. J. Statist. Plann. Inference 74 93-101.
  • MENÉNDEZ, J. A., RUEDA, C. and SALVADOR, B. (1992). Dominance of likelihood ratio tests under cone constraints. Ann. Statist. 20 2087-2099.
  • MENÉNDEZ, J. A. and SALVADOR, B. (1991). Anomalies of the likelihood ratio test for testing restricted hy potheses. Ann. Statist. 19 889-898.
  • NICHOLLS, D. F. and QUINN, B. G. (1980). The estimation of random coefficient autoregressive models. I. J. Time Ser. Anal. 1 37-46.
  • NICHOLLS, D. F. and QUINN, B. G. (1981). The estimation of random coefficient autoregressive models. II. J. Time Ser. Anal. 2 185-203.
  • NICHOLLS, D. F. and QUINN, B. G. (1982). Random Coefficient Autoregressive Models: An Introduction. Lecture Notes in Statist. 11. Springer, New York.
  • NÜESCH, P. E. (1966). On the problem of testing location in multivariate populations for restricted alternatives. Ann. Math. Statist. 37 113-119.
  • PAGAN, A. R. (1980). Some identification and estimation results for regression models with stochastically varying coefficients. J. Econometrics 13 341-363.
  • PERLMAN, M. D. (1969). One-sided testing problems in multivariate analysis. Ann. Math. Statist. 40 549-567. [Correction (1971) 42 1777.]
  • PRIESTLEY, M. B. (1988). Nonlinear and Nonstationary Time Series Analy sis. Academic Press, London.
  • RAMANATHAN, T. V. and RAJARSHI, M. B. (1994). Rank tests for testing the randomness of autoregressive coefficients. Statist. Probab. Lett. 21 115-120.
  • RIEDER, H. (2000). One-sided confidence about functionals over tangent cones. Dept. Mathematics, Univ. Bay reuth. Available at www.uni-bay rieder/publications.html.
  • ROBINSON, P. M. (1978). Statistical inference for a random coefficient autoregressive model. Scand. J. Statist. 5 163-168.
  • SCHAAFSMA, W. and SMID, L. J. (1966). Most stringent somewhere most powerful tests against alternatives restricted by a number of linear alternatives. Ann. Math. Statist. 37 1161- 1172.
  • SCHICK, A. (1996). n-consistent estimation in a random coefficient autoregressive model. Austral. J. Statist. 38 155-160.
  • SHORACK, G. R. (1967). Testing against ordered alternatives in model I analysis of variance: Normal theory and nonparametric. Ann. Math. Statist. 38 1740-1752.
  • SWENSEN, A. R. (1985). The asy mptotic distribution of the likelihood ratio for autoregressive time series with a regression trend. J. Multivariate Anal. 16 54-70.
  • TANIGUCHI, M. and KAKIZAWA, Y. (2000). Asy mptotic Theory of Statistical Inference for Time Series. Springer, New York.
  • TONG, H. (1990). Nonlinear Time Series: A Dy namical Approach. Oxford Univ. Press.
  • VAN DER VAART, A. (1988). Statistical Estimation in Large Parameter Spaces. CWI, Amsterdam.
  • VAN DER VAART, A. (1989). On the asy mptotic information bound. Ann. Statist. 17 1487-1500.
  • VERVAAT, W. (1979). On a stochastic difference equation and a representation of non-negative infinitely divisible random variables. Adv. in Appl. Probab. 11 750-783.
  • WEISS, A. A. (1985). The stability of the AR(1) process with an AR(1) coefficient. J. Time Ser. Anal. 6 181-186.