On the Marčenko-Pastur law for linear time series

Abstract

This paper is concerned with extensions of the classical Marčenko-Pastur law to time series. Specifically, $p$-dimensional linear processes are considered which are built from innovation vectors with independent, identically distributed (real- or complex-valued) entries possessing zero mean, unit variance and finite fourth moments. The coefficient matrices of the linear process are assumed to be simultaneously diagonalizable. In this setting, the limiting behavior of the empirical spectral distribution of both sample covariance and symmetrized sample autocovariance matrices is determined in the high-dimensional setting $p/n\to c\in (0,\infty )$ for which dimension $p$ and sample size $n$ diverge to infinity at the same rate. The results extend existing contributions available in the literature for the covariance case and are one of the first of their kind for the autocovariance case.

Published at http://dx.doi.org/10.1214/14-AOS1294 in the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org)

References in corpus (1)

• Spectral Analysis of High-dimensional Time Series

Cited by in corpus (6)

• An adaptable generalization of Hotelling's $T^2$ test in high dimension
• Large sample behaviour of high dimensional autocovariance matrices
• Estimation of Large Covariance and Precision Matrices from Temporally Dependent Observations
• Moment approach for singular values distribution of a large auto-covariance matrix
• On singular values distribution of a large auto-covariance matrix in the ultra-dimensional regime
• Singular value distribution of dense random matrices with block Markovian dependence