Annales de l'Institut Henri Poincaré, Probabilités et Statistiques

Limit theorems for stationary Markov processes with L2-spectral gap

Déborah Ferré, Loïc Hervé, and James Ledoux

Full-text: Open access

Abstract

Let $(X_{t},Y_{t})_{t\in \mathbb {T}}$ be a discrete or continuous-time Markov process with state space $\mathbb {X}\times \mathbb {R}^{d}$ where $\mathbb {X}$ is an arbitrary measurable set. Its transition semigroup is assumed to be additive with respect to the second component, i.e. $(X_{t},Y_{t})_{t\in \mathbb {T}}$ is assumed to be a Markov additive process. In particular, this implies that the first component $(X_{t})_{t\in \mathbb {T}}$ is also a Markov process. Markov random walks or additive functionals of a Markov process are special instances of Markov additive processes. In this paper, the process $(Y_{t})_{t\in \mathbb {T}}$ is shown to satisfy the following classical limit theorems:

(a) the central limit theorem,

(b) the local limit theorem,

(c) the one-dimensional Berry–Esseen theorem,

(d) the one-dimensional first-order Edgeworth expansion,

provided that we have $\sup_{t\in(0,1]\cap \mathbb {T}}\mathbb {E}_{\pi,0}[|Y_{t}|^{\alpha}]\textless\infty$ with the expected order α with respect to the independent case (up to some ε > 0 for (c) and (d)). For the statements (b) and (d), a Markov nonlattice condition is also assumed as in the independent case. All the results are derived under the assumption that the Markov process $(X_{t})_{t\in \mathbb {T}}$ has an invariant probability distribution π, is stationary and has the $\mathbb {L}^{2}(\pi)$-spectral gap property (that is, (Xt)t∈ℕ is ρ-mixing in the discrete-time case). The case where $(X_{t})_{t\in \mathbb {T}}$ is non-stationary is briefly discussed. As an application, we derive a Berry–Esseen bound for the M-estimators associated with ρ-mixing Markov chains.

Résumé

Soit $(X_{t},Y_{t})_{t\in \mathbb {T}}$ un processus de Markov en temps discret ou continu et d’espace d’état $\mathbb {X}\times \mathbb {R}^{d}$ où $\mathbb {X}$ est un ensemble mesurable quelconque. Son semi-groupe de transition est supposé additif suivant la seconde composante, i.e. $(X_{t},Y_{t})_{t\in\mathbb {T}}$ est un processus additif Markovien. En particulier, ceci implique que la première composante $(X_{t})_{t\in \mathbb {T}}$ est également un processus de Markov. Les marches aléatoires Markoviennes ou les fonctionnelles additives d’un processus de Markov sont des exemples de processus additifs Markoviens. Dans cet article, on montre que le processus $(Y_{t})_{t\in \mathbb {T}}$ satisfait les théorèmes limites classiques suivants :

(a) le théorème de la limite centrale,

(b) le théorème limite local,

(c) le théorème uniforme de Berry–Esseen en dimension un,

(d) le développement d’Edgeworth d’ordre un en dimension un,

pourvu que la condition de moment $\sup_{t\in(0,1]\cap \mathbb {T}}\mathbb {E}_{\pi,0}[|Y_{t}|^{\alpha}]\textless\infty$ soit satisfaite, avec l’ordre attendu α du cas indépendant (à un ε > 0 près pour (c) et (d)). Pour les énoncés (b) et (d), il faut ajouter une condition nonlattice comme dans le cas indépendant. Tous les résultats sont obtenus sous l’hypothèse d’un processus de Markov $(X_{t})_{t\in \mathbb {T}}$ admettant une mesure de probabilité invariante π et possédant la propriété de trou spectral sur $\mathbb {L}^{2}(\pi)$ (c’est à dire, (Xt)t∈ℕ est ρ-mélangeante dans le cas du temps discret). Le cas où $(X_{t})_{t\in \mathbb {T}}$ est non-stationnaire est brièvement abordé. Nous appliquons nos résultats pour obtenir une borne de Berry–Esseen pour les M-estimateurs associés aux chaînes de Markov ρ-mélangeantes.

Article information

Source
Ann. Inst. H. Poincaré Probab. Statist., Volume 48, Number 2 (2012), 396-423.

Dates
First available in Project Euclid: 11 April 2012

Permanent link to this document
https://projecteuclid.org/euclid.aihp/1334148205

Digital Object Identifier
doi:10.1214/11-AIHP413

Mathematical Reviews number (MathSciNet)
MR2954261

Zentralblatt MATH identifier
1245.60068

Subjects
Primary: 60J05: Discrete-time Markov processes on general state spaces 60F05: Central limit and other weak theorems 60J25: Continuous-time Markov processes on general state spaces 60J55: Local time and additive functionals 37A30: Ergodic theorems, spectral theory, Markov operators {For operator ergodic theory, see mainly 47A35} 62M05: Markov processes: estimation

Keywords
Markov additive process Central limit theorems Berry–Esseen bound Edgeworth expansion Spectral method ρ-mixing M-estimator

Citation

Ferré, Déborah; Hervé, Loïc; Ledoux, James. Limit theorems for stationary Markov processes with L 2 -spectral gap. Ann. Inst. H. Poincaré Probab. Statist. 48 (2012), no. 2, 396--423. doi:10.1214/11-AIHP413. https://projecteuclid.org/euclid.aihp/1334148205


Export citation

References

  • [1] S. Asmussen. Ruin Probabilities. World Sci. Publishing Co. Inc., River Edge, NJ, 2000.
  • [2] S. Asmussen. Applied Probability and Queues, Vol. 51, 2nd edition. Springer-Verlag, New York, 2003.
  • [3] S. Asmussen, F. Avram and M. R. Pistorius. Russian and American put options under exponential phase-type Lévy models. Stochastic Process. Appl. 109 (2004) 79–111.
  • [4] M. Babillot. Théorie du renouvellement pour des chaînes semi-markoviennes transientes. Ann. Inst. H. Poincaré Probab. Statist. 24 (1988) 507–569.
  • [5] A. Benveniste and J. Jacod. Systèmes de Lévy des processus de Markov. Invent. Math. 21 (1973) 183–198.
  • [6] J. Bergh and J. Löfström. Interpolation Spaces. An Introduction. Springer-Verlag, Berlin, 1976.
  • [7] R. N. Bhattacharya. On the functional central limit theorem and the law of the iterated logarithm for Markov processes. Probab. Theory Related Fields 60 (1982) 185–201.
  • [8] P. Billingsley. Probability and Measure, 3rd edition. John Wiley & Sons Inc., New York, 1995.
  • [9] M. Bladt, B. Meini, M. F. Neuts and B. Sericola. Distributions of reward functions on continuous-time Markov chains. In Matrix-Analytic Methods 39–62. World Sci. Publishing, Adelaide, 2002.
  • [10] R. C. Bradley. Basic properties of strong mixing conditions. a survey and some open questions. Probab. Surv. 2 (2005) 107–144.
  • [11] R. C. Bradley. Introduction to strong mixing conditions (Volume I). Technical report, Indiana Univ., 2005.
  • [12] L. Breiman. Probability. SIAM, Philadelphia, PA, 1993.
  • [13] S. Campanato. Proprietà di una famiglia di spazi funzionali. Ann. Scuola Norm. Sup. Pisa 18 (1964) 137–160.
  • [14] O. Cappé, E. Moulines and T. Rydén. Inference in Hidden Markov Models. Springer, New York, 2005.
  • [15] E. Çinlar. Markov additive processes Part II. Probab. Theory Related Fields 24 (1972) 95–121.
  • [16] E. Çinlar. Introduction to Stochastic Processes. Prentice-Hall, Inc., Englewood Cliffs, NJ, 1975.
  • [17] E. Çinlar. Shock and wear models and Markov additive processes. In The Theory and Applications of Reliability, with Emphasis on Bayesian and Nonparametric Methods, Vol. I 193–214. Academic Press, New York, 1977.
  • [18] M.-F. Chen. From Markov Chains to Non-equilibrium Particle Systems, 2nd edition. World Sci. Publishing Co. Inc., River Edge, NJ, 2004.
  • [19] D. Dehay and J.-F. Yao. On likelihood estimation for discretely observed Markov jump processes. Aust. N. Z. J. Stat. 49 (2007) 93–107.
  • [20] J. L. Doob. Stochastic Processes. John Wiley & Sons, New York, 1953.
  • [21] Ī. Ī. Ezhov and A. V. Skorohod. Markov processes which are homogeneous in the second component. I. Theory Probab. Appl. 14 (1969) 1–13.
  • [22] Ī. Ī. Ezhov and A. V. Skorohod. Markov processes which are homogeneous in the second component. II. Theory Probab. Appl. 14 (1969) 652–667.
  • [23] W. Feller. An Introduction to Probability Theory and Its Applications, Vol. II. John Wiley and Sons, New York, 1971.
  • [24] D. Ferré. Développement d’Edgeworth d’ordre 1 pour des M-estimateurs dans le cas de chaînes V-géométriquement ergodiques. CRAS 348 (2010) 331–334.
  • [25] G. Fort, E. Moulines, G. O. Roberts and J. S. Rosenthal. On the geometric ergodicity of hybrid samplers. J. Appl. Probab. 40 (2003) 123–146.
  • [26] C.-D. Fuh and T. L. Lai. Asymptotic expansions in multidimensional Markov renewal theory and first passage times for Markov random walks. Adv. in Appl. Probab. 33 (2001) 652–673.
  • [27] M. Fukushima and M. Hitsuda. On a class of Markov processes taking values on lines and the central limit theorem. Nagoya Math. J. 30 (1967) 47–56.
  • [28] H. Ganidis, B. Roynette and F. Simonot. Convergence rate of some semi-groups to their invariant probability. Stochastic Process. Appl. 79 (1999) 243–263.
  • [29] V. Genon-Catalot, T. Jeantheau and C. Larédo. Stochastic volatility models as hidden Markov models and statistical applications. Bernoulli 6 (2000) 1051–1079.
  • [30] P. W. Glynn and W. Whitt. Limit theorems for cumulative processes. Stochastic Process. Appl. 47 (1993) 299–314.
  • [31] P. W. Glynn and W. Whitt. Necessary conditions in limit theorems for cumulative processes. Stochastic Process. Appl. 98 (2002) 199–209.
  • [32] B. Goldys and B. Maslowski. Exponential ergodicity for stochastic reaction-diffusion equations. In Stochastic Partial Differential Equations and Applications – VII 115–131. Chapman & Hall/CRC, Boca Raton, FL, 2006.
  • [33] B. Goldys and B. Maslowski. Lower estimates of transition densities and bounds on exponential ergodicity for stochastic PDE’s. Ann. Probab. 34 (2006) 1451–1496.
  • [34] M. I. Gordin. On the central limit theorem for stationary Markov processes. Soviet Math. Dokl. 19 (1978) 392–394.
  • [35] S. Gouëzel. Characterization of weak convergence of Birkhoff sums for Gibbs–Markov maps. Preprint, 2008.
  • [36] S. Gouëzel and C. Liverani. Banach spaces adapted to Anosov systems. Ergodic Theory Dynam. Systems 26 (2006) 189–217.
  • [37] J.-B. Gravereaux and J. Ledoux. Poisson approximation for some point processes in reliability. Adv. in Appl. Probab. 36 (2004) 455–470.
  • [38] S. Grigorescu and G. Opriçan. Limit theorems for JX processes with a general state space. Probab. Theory Related Fields 35 (1976) 65–73.
  • [39] D. Guibourg and L. Hervé. A renewal theorem for strongly ergodic Markov chains in dimension d≥3 and in the centered case. Potential Anal. 34 (2011) 385–410.
  • [40] Y. Guivarc’h. Application d’un théorème limite local à la transcience et à la récurrence de marches aléatoires. In Théorie du potentiel (Orsay, 1983) 301–332. Lecture Notes in Math. 1096. Springer, Berlin, 1984.
  • [41] Y. Guivarc’h. Limit theorems for random walks and products of random matrices. In Proceedings of the CIMPA-TIFR School on Probability Measures on Groups, Mumbai 2002 257–332. TIFR Studies in Mathematics Series. Tata Institute of Fundamental Research, Mumbai, India, 2002.
  • [42] Y. Guivarc’h and J. Hardy. Théorèmes limites pour une classe de chaînes de Markov et applications aux difféomorphismes d’Anosov. Ann. Inst. H. Poincaré Probab. Statist. 24 (1988) 73–98.
  • [43] O. Häggström. Acknowledgement of priority concerning “On the central limit theorem for geometrically ergodic Markov chains.” Probab. Theory Related Fields 135 (2006) 470.
  • [44] H. Hennion and L. Hervé. Limit Theorems for Markov Chains and Stochastic Properties of Dynamical Systems by Quasi-Compactness. Lecture Notes in Math. 1766. Springer, Berlin, 2001.
  • [45] H. Hennion and L. Hervé. Central limit theorems for iterated random Lipschitz mappings. Ann. Probab. 32 (2004) 1934–1984.
  • [46] L. Hervé. Théorème local pour chaînes de Markov de probabilité de transition quasi-compacte. Applications aux chaînes v-géométriquement ergodiques et aux modèles itératifs. Ann. Inst. H. Poincaré Probab. Statist. 41 (2005) 179–196.
  • [47] L. Hervé. Vitesse de convergence dans le théorème limite central pour des chaînes de Markov fortement ergodiques. Ann. Inst. H. Poincaré Probab. Statist. 44 (2008) 280–292.
  • [48] L. Hervé, J. Ledoux and V. Patilea. A Berry–Esseen theorem on M-estimators for geometrically ergodic Markov chains. Bernoulli (2012). To appear.
  • [49] L. Hervé and F. Pène. The Nagaev–Guivarc’h method via the Keller–Liverani theorem. Bull. Soc. Math. France 138 (2010) 415–489.
  • [50] M. Hitsuda and A. Shimizu. The central limit theorem for additive functionals of Markov processes and the weak convergence to Wiener measure. J. Math. Soc. Japan 22 (1970) 551–566.
  • [51] H. Holzmann. Martingale approximations for continuous-time and discrete-time stationary Markov processes. Stochastic Process. Appl. 115 (2005) 1518–1529.
  • [52] I. A. Ibragimov. A note on the central limit theorem for dependent random variables. Theory Probab. Appl. 20 (1975) 135–141.
  • [53] I. A. Ibragimov and Y. V. Linnik. Independent and Stationary Sequences of Random Variables. Walters-Noordhoff, the Netherlands, 1971.
  • [54] M. Jara, T. Komorowski and S. Olla. Limit theorems for additive functionals of a Markov chain. Ann. Appl. Probab. 19 (2009) 2270–2300.
  • [55] S. F. Jarner and E. Hansen. Geometric ergodicity of Metropolis algorithms. Stochastic Process. Appl. 85 (2000) 341–361.
  • [56] A. Jobert and L. C. G. Rogers. Option pricing with Markov-modulated dynamics. SIAM J. Control Optim. 44 (2006) 2063–2078.
  • [57] G. L. Jones. On the Markov chain central limit theorem. Probab. Surv. 1 (2004) 299–320.
  • [58] N. V. Kartashov. Determination of the spectral ergodicity exponent for the birth and death process. Ukrain. Math. J. 52 (2000) 1018–1028.
  • [59] J. Keilson and D. M. G. Wishart. A central limit theorem for processes defined on a finite Markov chain. Proc. Cambridge Philos. Soc. 60 (1964) 547–567.
  • [60] G. Keller and C. Liverani. Stability of the spectrum for transfer operators. Ann. Scuola Norm. Sup. Pisa Cl. Sci. (4) 28 (1999) 141–152.
  • [61] C. Kipnis and S. R. S. Varadhan. Central limit theorem for additive functionals of reversible Markov processes and applications to simple exclusions. Comm. Math. Phys. 104 (1986) 1–19.
  • [62] P. Lezaud. Chernoff and Berry–Esseen inequalities for Markov processes. ESAIM Probab. Stat. 5 (2001) 183–201.
  • [63] T. M. Liggett. Exponential L2 convergence of attractive reversible nearest particle systems. Ann. Probab. 17 (1989) 403–432.
  • [64] N. Limnios and G. Opriçan. Semi-Markov Processes and Reliability. Birkhauser Boston Inc., Boston, MA, 2001.
  • [65] N. Maigret. Théorème de limite centrale fonctionnel pour une chaî ne de Markov récurrente au sens de Harris et positive. Ann. Inst. H. Poincaré Probab. Statist. 14 (1978) 425–440.
  • [66] M. Maxwell and M. Woodroofe. A local limit theorem for hidden Markov chains. Statist. Probab. Lett. 32 (1997) 125–131.
  • [67] S. P. Meyn and R. L. Tweedie. Markov Chains and Stochastic Stability. Springer-Verlag, London, 1993.
  • [68] S. V. Nagaev. Some limit theorems for stationary Markov chains. Theory Probab. Appl. 11 (1957) 378–406.
  • [69] J. Neveu. Une généralisation des processus à accroissements positifs indépendants. Abh. Math. Sem. Univ. Hambourg 25 (1961) 36–61.
  • [70] S. Özekici and R. Soyer. Reliability modeling and analysis in random environments. In Mathematical Reliability: An Expository Perspective 249–273. Kluwer Acad. Publ., Boston, MA, 2004.
  • [71] A. Pacheco and N. U. Prabhu. Markov-additive processes of arrivals. In Advances in Queueing 167–194. CRC, Boca Raton, FL, 1995.
  • [72] A. Pacheco, L. C. Tang and N. U. Prabhu. Markov-Modulated Processes & Semiregenerative Phenomena. World Sci. Publishing, Hackensack, NJ, 2009.
  • [73] M. Peligrad. On the central limit theorem for ρ-mixing sequences of random variables. Ann. Probab. 15 (1987) 1387–1394.
  • [74] J. Pfanzagl. The Berry–Esseen bound for minimum contrast estimates. Metrika 17 (1971) 81–91.
  • [75] M. Pinsky. Differential equations with a small parameter and the central limit theorem for functions defined on a finite Markov chain. Probab. Theory Related Fields 9 (1968) 101–111.
  • [76] B. L. S. P. Rao. On the rate of convergence of estimators for Markov processes. Probab. Theory Related Fields 26 (1973) 141–152.
  • [77] D. Revuz and M. Yor. Continuous Martingales and Brownian Motion, 3rd edition. Springer-Verlag, Berlin, 1999.
  • [78] G. O. Roberts and J. S. Rosenthal. Geometric ergodicity and hybrid Markov chains. Electron. Commun. Probab. 2 (1997) 13–25.
  • [79] G. O. Roberts and J. S. Rosenthal. General state space Markov chains and MCMC algorithms. Probab. Surv. 1 (2004) 20–71.
  • [80] G. O. Roberts and R. L. Tweedie. Geometric L2 and L1 convergence are equivalent for reversible Markov chains. J. Appl. Probab. 38A (2001) 37–41.
  • [81] M. Rosenblatt. Markov Processes. Structure and Asymptotic Behavior. Springer-Verlag, New York, 1971.
  • [82] V. T. Stefanov. Exact distributions for reward functions on semi-Markov and Markov additive processes. J. Appl. Probab. 43 (2006) 1053–1065.
  • [83] J. L. Steichen. A functional central limit theorem for Markov additive processes with an application to the closed Lu–Kumar network. Stoch. Models 17 (2001) 459–489.
  • [84] A. Touati. Théorèmes de limite centrale fonctionnels pour les processus de Markov. Ann. Inst. H. Poincaré Probab. Statist. 19 (1983) 43–55.
  • [85] A. W. van der Vaart. Asymptotic Statistics. Cambridge Univ. Press, Cambridge, 1998.
  • [86] L. Wu. Essential spectral radius for Markov semigroups. I. Discrete time case. Probab. Theory Related Fields 128 (2004) 255–321.