Autoregressive processes with data‐driven regime switching

We develop a switching-regime vector autoregressive model in which changes in regimes are governed by an underlying Markov process. In contrast to the typical hidden Markov approach, we allow the transition probabilities of the underlying Markov process to depend on past values of the time series and exogenous variables. Such processes have potential applications in finance and neuroscience. In the latter, the brain activity at time t (measured by electroencephalograms) will be modelled as a function of both its past values as well as exogenous variables (such as visual or somatosensory stimuli). In this article, we establish stationarity, geometric ergodicity and existence of moments for these processes under suitable conditions on the parameters of the model. Such properties are important for understanding the stability properties of the model as well as for deriving the asymptotic behaviour of various statistics and model parameter estimators. Copyright 2009 Blackwell Publishing Ltd

[1]  Y. Davydov.  Mixing Conditions for Markov Chains , 1974 .

[2]  Zongwu Cai,et al.  Adaptive varying‐coefficient linear models , 2000 .

[3]  Richard L. Tweedie,et al.  Markov Chains and Stochastic Stability , 1993, Communications and Control Engineering Series.

[4]  M. B. Priestley,et al.  Design Relations for Non‐Stationary Processes , 1966 .

[5]  R. Bhattacharya,et al.  On geometric ergodicity of nonlinear autoregressive models , 1995 .

[6]  R. Dahlhaus.  Fitting time series models to nonstationary processes , 1997 .

[7]  B. M. Pötscher,et al.  Dynamic Nonlinear Econometric Models: Asymptotic Theory , 1997 .

[8]  G. Nason,et al.  Wavelet processes and adaptive estimation of the evolutionary wavelet spectrum , 2000 .

[9]  Thomas Mikosch,et al.  Quasi-maximum-likelihood estimation in conditionally heteroscedastic time series: A stochastic recurrence equations approach , 2006 .

[10]  Pascal Massart,et al.  The functional central limit theorem for strongly mixing processes , 1994 .

[11]  T. Lai,et al.  Stochastic Neural Networks With Applications to Nonlinear Time Series , 2001 .

[12]  J. Zakoian,et al.  Stationarity of Multivariate Markov-Switching ARMA Models , 2001 .

[13]  M. Priestley.  Evolutionary Spectra and Non‐Stationary Processes , 1965 .

[14]  Michael A. West,et al.  Evaluation and Comparison of EEG Traces: Latent Structure in Nonstationary Time Series , 1999 .

[15]  B. M. Pötscher,et al.  Dynamic Nonlinear Econometric Models , 1997 .

[16]  Joseph Tadjuidje Kamgaing.  Competing Neural Networks as Models for Non Stationary Financial Time Series -Changepoint Analysis- , 2005 .

[17]  Richard A. Davis,et al.  Structural Break Estimation for Nonstationary Time Series Models , 2006 .

[18]  J. Franke,et al.  A note on the identifiability of the conditional expectation for the mixtures of neural networks , 2008 .

[19]  Ruey S. Tsay,et al.  Functional-Coefficient Autoregressive Models , 1993 .

[20]  G. Lindgren.  Markov regime models for mixed distributions and switching regressions , 1978 .

[21]  E. Hannan,et al.  The statistical theory of linear systems , 1989 .

[22]  K. Gordon,et al.  Modeling and Monitoring Biomedical Time Series , 1990 .

[23]  James D. Hamilton.  A New Approach to the Economic Analysis of Nonstationary Time Series and the Business Cycle , 1989 .

[24]  A F Smith,et al.  Monitoring renal transplants: an application of the multiprocess Kalman filter. , 1983, Biometrics.

[25]  R. Rao.  Relations between Weak and Uniform Convergence of Measures with Applications , 1962 .

[26]  Paul D. Feigin,et al.  RANDOM COEFFICIENT AUTOREGRESSIVE PROCESSES:A MARKOV CHAIN ANALYSIS OF STATIONARITY AND FINITENESS OF MOMENTS , 1985 .

[27]  Jianqing Fan,et al.  Functional-Coefficient Regression Models for Nonlinear Time Series , 2000 .

[28]  S. H. Hsieh,et al.  THRESHOLD MODELS FOR NONLINEAR TIME SERIES ANALYSIS. , 1987 .

[29]  H. Ombao,et al.  SLEX Analysis of Multivariate Nonstationary Time Series , 2005 .

[30]  Hung Man Tong,et al.  Threshold models in non-linear time series analysis. Lecture notes in statistics, No.21 , 1983 .