A test for parameter change in general causal time series using quasi-likelihood estimator

William Charky Kengne

doi:10.1016/j.crma.2012.03.001

Statistics

A test for parameter change in general causal time series using quasi-likelihood estimator
[Utilisation de la quasi-vraisemblance pour un test de détection de rupture dans les paramètres des processus causaux]

Présenté par : the Editorial Board

William Charky Kengne ^1,²

¹ SAMM, Université Paris 1 Panthéon-Sorbonne, 90 rue de Tolbiac, 75013 Paris, France
² ENSP, Université de Yaoundé I, BP 8390, Yaoundé, Cameroon

Comptes Rendus. Mathématique, Volume 350 (2012) no. 5-6, pp. 307-312.

Résumé
Abstract

Dans cette Note, nous proposons un nouveau test de détection de rupture dans le paramètre dʼun processus $X = {(X_{t})}_{t \in Z}$ appartenant à une classe de processus causaux contenant les modèles AR(∞), ARCH(∞), TARCH(∞), … . Deux statistiques ${\hat{Q}}_{n}^{(1)}$ et ${\hat{Q}}_{n}^{(2)}$ sont construites en utilisant lʼestimateur du maximum de quasi-vraisemblance du paramètre. Sous lʼhypothèse nulle selon laquelle aucun changement nʼintervient dans le paramètre, chacune de ces statistiques converge vers une distribution connue et le maximum diverge vers lʼinfini sous lʼhypothèse alternative dʼune rupture. Quelques résultats de simulations sont présentés.

In this Note, we propose a new procedure to test a change in the parameter of a process $X = {(X_{t})}_{t \in Z}$ belonging to a class of causal models including AR(∞), ARCH(∞), TARCH(∞), … models. Two statistics ${\hat{Q}}_{n}^{(1)}$ and ${\hat{Q}}_{n}^{(2)}$ are constructed using the quasi-likelihood estimator (QMLE) of the parameter. Under the null hypothesis that there is no change, each of these statistics converges weakly to a well-known distribution and the maximum diverges to infinity under the alternative of one change. Some simulation results are reported.

Reçu le : 2011-07-11
Accepté le : 2012-03-01
Publié le : 2012-03-01

DOI : 10.1016/j.crma.2012.03.001

Affiliations des auteurs :

William Charky Kengne ^{1, 2}

¹ SAMM, Université Paris 1 Panthéon-Sorbonne, 90 rue de Tolbiac, 75013 Paris, France
² ENSP, Université de Yaoundé I, BP 8390, Yaoundé, Cameroon

@article{CRMATH_2012__350_5-6_307_0,
     author = {William Charky Kengne},
     title = {A test for parameter change in general causal time series using quasi-likelihood estimator},
     journal = {Comptes Rendus. Math\'ematique},
     pages = {307--312},
     publisher = {Elsevier},
     volume = {350},
     number = {5-6},
     year = {2012},
     doi = {10.1016/j.crma.2012.03.001},
     language = {en},
}

TY  - JOUR
AU  - William Charky Kengne
TI  - A test for parameter change in general causal time series using quasi-likelihood estimator
JO  - Comptes Rendus. Mathématique
PY  - 2012
SP  - 307
EP  - 312
VL  - 350
IS  - 5-6
PB  - Elsevier
DO  - 10.1016/j.crma.2012.03.001
LA  - en
ID  - CRMATH_2012__350_5-6_307_0
ER  -

%0 Journal Article
%A William Charky Kengne
%T A test for parameter change in general causal time series using quasi-likelihood estimator
%J Comptes Rendus. Mathématique
%D 2012
%P 307-312
%V 350
%N 5-6
%I Elsevier
%R 10.1016/j.crma.2012.03.001
%G en
%F CRMATH_2012__350_5-6_307_0

William Charky Kengne. A test for parameter change in general causal time series using quasi-likelihood estimator. Comptes Rendus. Mathématique, Volume 350 (2012) no. 5-6, pp. 307-312. doi : 10.1016/j.crma.2012.03.001. https://comptes-rendus.academie-sciences.fr/mathematique/articles/10.1016/j.crma.2012.03.001/

Bibliographie
Cité par

[1] A. Aue; S. Hörmann; L. Horváth; M. Reimherr Break detection in the covariance structure of multivariate time series models, Ann. Statist., Volume 37 (2009), pp. 4046-4087

[2] J.-M. Bardet; O. Wintenberger Asymptotic normality of the quasi-maximum likelihood estimator for multidimensional causal processes, Ann. Statist., Volume 37 (2009), pp. 2730-2759

[3] J.-M. Bardet; W. Kengne; O. Wintenberger Detecting multiple change-points in general causal time series using penalized quasi-likelihood (Preprint available on) | arXiv

[4] I. Berkes; L. Horváth; P. Kokoszka Testing for parameter constancy in GARCH $(p; q)$ models, Statist. Probab. Lett., Volume 70 (2004), pp. 263-273

[5] L. Horváth Change in autoregressive processes, Stochastic Process. Appl., Volume 44 (1993), pp. 221-242

[6] C. Inclan; G.C. Tiao Use of cumulative sums of squares for retrospective detection of changes of variance, J. Amer. Statist. Assoc., Volume 89 (1994), pp. 913-923

[7] W. Kengne, Testing for parameter constancy in general causal time series models, Preprint Hal-00561015, 2011.

[8] J. Kierfer K-sample analogues of the Kolmogorov–Smirnov and Cramér–von Mises tests, Ann. Math. Statist., Volume 30 (1959), pp. 420-447

[9] S. Kim; S. Cho; S. Lee On the CUSUM test for parameter changes in GARCH $(1, 1)$ models, Comm. Statist. Theory Methods, Volume 29 (2000), pp. 445-462

[10] P. Kokoszka; R. Leipus Testing for parameter changes in ARCH models, Lith. Math. J., Volume 39 (1999), pp. 182-195

[11] O. Kouamo; E. Moulines; F. Roueff Testing for homogeneity of variance in the wavelet domain (P. Doukhan; G. Lang; D. Surgailis; G. Teyssiere, eds.), Dependence in Probability and Statistics, Lecture Notes in Statist., vol. 200, Springer-Verlag, 2010, pp. 420-447

[12] R. Kulperger; H. Yu High moment partial sum processes of residuals in GARCH models and their applications, Ann. Statist., Volume 33 (2005), pp. 2395-2422

[13] S. Lee; O. Na Test for parameter change in stochastic processes based on conditional least-squares estimator, J. Multivariate Anal., Volume 93 (2005), pp. 375-393

[14] S. Lee; J. Song Test for parameter change in ARMA models with GARCH innovations, Statist. Probab. Lett., Volume 78 (2008), pp. 1990-1998

[15] E.S. Page A test for a change in a parameter occurring at an unknown point, Biometrika, Volume 42 (1955), pp. 523-526

Cité par Sources :

Commentaires - Politique