The eigenstructure of the sample covariance matrices of high-dimensional stochastic volatility models with heavy tails (Q2325386)

A high-dimensional stochastic volatility time series is considered for the case when dimension grows with the sample size. The authors focuse on analysis of the dependence structure of observations from time series based on the spectral properties of the sample covariance matrix. Both the stationary case, i.e. the marginal distribution does not change over time, and the nonstationary case of an iid stochastic volatility field with time-varying marginal distribution are considered. The main results of this paper concern convergence results for the stochastic volatility model.