Selection of weak VARMA models by Akaïke's information criteria

Boubacar Mainassara, Yacouba (2010): Selection of weak VARMA models by Akaïke's information criteria.

Abstract

This article considers the problem of orders selections of vector autoregressive moving-average (VARMA) models and the sub-class of vector autoregressive (VAR) models under the assumption that the errors are uncorrelated but not necessarily independent. We relax the standard independence assumption to extend the range of application of the VARMA models, and allow to cover linear representations of general nonlinear processes. We propose a modified criterion to the corrected AIC (Akaïke information criterion) version (AICc) introduced by Tsai and Hurvich (1989). This modified criterion is an approximately unbiased estimator of the Kullback-Leibler discrepancy, originally used to derive AIC-based criteria. Moreover, this criterion requires the estimation of the matrice involved in the asymptotic variance of the quasi-maximum likelihood (QML) estimator of the models, which provide an additional information about models. Monte carlo experiments show that the proposed modified criterion estimates the models orders more accurately than the standard AIC and AICc in large samples and often in small samples.

• Item Type: MPRA Paper
• Language: English
• Keywords: AIC, discrepancy, Kullback-Leibler information, QMLE/LSE, order selection, structural representation, weak VARMA models.
• Subjects: C - Mathematical and Quantitative Methods > C5 - Econometric Modeling > C52 - Model Evaluation, Validation, and Selection; C - Mathematical and Quantitative Methods > C2 - Single Equation Models ; Single Variables > C22 - Time-Series Models ; Dynamic Quantile Regressions ; Dynamic Treatment Effect Models ; Diffusion Processes; C - Mathematical and Quantitative Methods > C0 - General > C01 - Econometrics

References

Akaïke, H. (1973) Information theory and an extension of the maximum likelihood principle. In {\em 2nd International Symposium on Information Theory,} Eds. B. N. Petrov and F. Cs\'{a}ki, pp. 267--281. Budapest: Akad\'{e}mia Kiado.

Anderson, T. W. (2003) {\em An Introduction to Multivariate Statistical Analysis 3nd Edn.} New Jersey, Wiley.

\item[ Boubacar Mainassara, Y. ] (2009a) Multivariate portmanteau test for structural VARMA models with uncorrelated but non-independent error terms. \emph{Working Papers,} http://perso.univ-lille3.fr/$\sim$yboubacarmai/yac/VARMAPORTM.pdf

\item[ Boubacar Mainassara, Y.] (2009b) Estimating the asymptotic variance matrix of structural VARMA models with uncorrelated but non-independent error terms. \emph{Working Papers,} http://perso.univ-lille3.fr/$\sim$yboubacarmai/yac/VARMAEstCOV2.pdf

\item[ Boubacar Mainassara, Y. and Francq, C.] (2009) Estimating structural VARMA models with uncorrelated but non-independent error terms. \emph{Working Papers,} http://perso.univ-lille3.fr/$\sim$cfrancq/pub.html.

\item[ Dufour, J-M., and Pelletier, D.] (2005) Practical methods for modelling weak VARMA processes: identification, estimation and specification with a macroeconomic application. \emph{ Technical report, D\'epartement de sciences \'economiques and CIREQ, Universit\'e de Montr\'eal, Montr\'eal, Canada.}

\item[ Findley, D.F.] (1993) The overfitting principles supporting AIC, Statistical Research Division Report RR 93/04, Bureau of the Census.

\item[ Francq, C. and Ra\"{\i}ssi, H.] (2007) Multivariate Portmanteau Test for Autoregressive Models with Uncorrelated but Nonindependent Errors, {\em Journal of Time Series Analysis} 28, 454--470.

\item[ Francq, C., Roy, R. and Zako\"{\i}an, J-M.] (2005) Diagnostic checking in ARMA Models with Uncorrelated Errors, {\em Journal of the American Statistical Association} 100, 532--544.

\item[ Francq, and Zako\"{\i}an, J-M.] (1998) Estimating linear representations of nonlinear processes, {\em Journal of Statistical Planning and Inference} 68, 145--165.

\item[ Francq, and Zako\"{\i}an, J-M.] (2000) Covariance matrix estimation of mixing weak ARMA models, {\em Journal of Statistical Planning and Inference} 83, 369--394.

\item[ Francq, and Zako\"{\i}an, J-M.] (2005) Recent results for linear time series models with non independent innovations. In {\em Statistical Modeling and Analysis for Complex Data Problems,} Chap. 12 (eds P. {\sc Duchesne} and B. {\sc R\'{e}millard}). New York: Springer Verlag, 241--265.

\item[ Francq, and Zako\"{\i}an, J-M.] (2007) HAC estimation and strong linearity testing in weak ARMA models, {\em Journal of Multivariate Analysis} 98, 114--144.

\item[ Hannan, E. J. and Rissanen] (1982) Recursive estimation of mixed of Autoregressive Moving Average order, {\em Biometrika} 69, 81--94.

\item[ Hurvich, C. M. and Tsai, C-L.] (1989) Regression and time series model selection in small samples. {\em Biometrika} 76, 297--307.

\item[ Hurvich, C. M. and Tsai, C-L.] (1993) A corrected Akaïke information criterion for vector autoregressive model selection. {\em Journal of Time Series Analysis} 14, 271--279.

\item[ L\"utkepohl, H.] (1993) {\em Introduction to multiple time series analysis.} Springer Verlag, Berlin.

\item[ L\"utkepohl, H.] (2005) {\em New introduction to multiple time series analysis.} Springer Verlag, Berlin.

\item[ Magnus, J.R. and H. Neudecker] (1988) {\em Matrix Differential Calculus with Application in Statistics and Econometrics.} New-York, Wiley.

\item[ Muirhead, R.J.] (1982) {\em Aspects of Multivariate Statistical Theory.} Wiley, New-York, Wiley.

\item[ Reinsel, G. C.] (1997) {\em Elements of multivariate time series Analysis.} Second edition. Springer Verlag, New York.

\item[ Romano, J. L. and Thombs, L. A.] (1996) Inference for autocorrelations under weak assumptions, {\em Journal of the American Statistical Association} 91, 590--600.

\item[ Wei, W.H. ] (1994) {\em Time Series Analysis: Univariate and Multivariate Methods 2nd Edn.} New York, Addison-Wesley.