Skip to main content
School of Economics and Finance

No. 916: Estimation of time-varying covariance matrices for large datasets

Yiannis Dendramis , Athens University of Economics and Business
Luidas Giraitis , Queen Mary University of London
George Kapetanios , King's College London

November 9, 2020

Download full paper

Abstract

Time variation is a fundamental problem in statistical and econometric analysis of macroeconomic and financial data. Recently there has been considerable focus on developing econometric modelling that enables stochastic structural change in model parameters and on model estimation by Bayesian or non-parametric kernel methods. In the context of the estimation of covariance matrices of large dimensional panels, such data requires taking into account time variation, possible dependence and heavy-tailed distributions. In this paper we introduce a non-parametric version of regularisation techniques for sparse large covariance matrices, developed by Bickel and Levina (2008) and others. We focus on the robustness of such a procedure to time variation, dependence and heavy-tailedness of distributions. The paper includes a set of results on Bernstein type inequalities for dependent unbounded variables which are expected to be applicable in econometric analysis beyond estimation of large covariance matrices. We discuss the utility of the robust thresholding method, comparing it with other estimators in simulations and an empirical application on the design of minimum variance portfolios.

J.E.L classification codes: C13; C22; C51

Keywords:covariance matrix estimation, large dataset, regularization, thresholding, shrinkage, exponential inequalities, minimum variance portfolio

Back to top