Header

UZH-Logo

Maintenance Infos

Quadratic shrinkage for large covariance matrices


Ledoit, Olivier; Wolf, Michael (2020). Quadratic shrinkage for large covariance matrices. Working paper series / Department of Economics 335, University of Zurich.

Abstract

This paper constructs a new estimator for large covariance matrices by drawing a bridge between the classic Stein (1975) estimator in finite samples and recent progress under large-dimensional asymptotics. The estimator keeps the eigenvectors of the sample covariance matrix and applies shrinkage to the inverse sample eigenvalues. The corresponding formula is quadratic: it has two shrinkage targets weighted by quadratic functions of the concentration (that is, matrix dimension divided by sample size). The first target dominates mid-level concentrations and the second one higher levels. This extra degree of freedom enables us to outperform linear shrinkage when optimal shrinkage is not linear (which is the general case). Both of our targets are based on what we term the “Stein shrinker”, a local attraction operator that pulls sample covariance matrix eigenvalues towards their nearest neighbors, but whose force diminishes with distance, like gravitation. We prove that no cubic or higher-order nonlinearities beat quadratic with respect to Frobenius loss under large-dimensional asymptotics. Non-normality and the case where the matrix dimension exceeds the sample size are accommodated. Monte Carlo simulations confirm state-of-the-art performance in terms of accuracy, speed, and scalability.

Abstract

This paper constructs a new estimator for large covariance matrices by drawing a bridge between the classic Stein (1975) estimator in finite samples and recent progress under large-dimensional asymptotics. The estimator keeps the eigenvectors of the sample covariance matrix and applies shrinkage to the inverse sample eigenvalues. The corresponding formula is quadratic: it has two shrinkage targets weighted by quadratic functions of the concentration (that is, matrix dimension divided by sample size). The first target dominates mid-level concentrations and the second one higher levels. This extra degree of freedom enables us to outperform linear shrinkage when optimal shrinkage is not linear (which is the general case). Both of our targets are based on what we term the “Stein shrinker”, a local attraction operator that pulls sample covariance matrix eigenvalues towards their nearest neighbors, but whose force diminishes with distance, like gravitation. We prove that no cubic or higher-order nonlinearities beat quadratic with respect to Frobenius loss under large-dimensional asymptotics. Non-normality and the case where the matrix dimension exceeds the sample size are accommodated. Monte Carlo simulations confirm state-of-the-art performance in terms of accuracy, speed, and scalability.

Statistics

Downloads

454 downloads since deposited on 13 Nov 2019
187 downloads since 12 months
Detailed statistics

Additional indexing

Item Type:Working Paper
Communities & Collections:03 Faculty of Economics > Department of Economics
Working Paper Series > Department of Economics
Dewey Decimal Classification:330 Economics
JEL Classification:C13
Uncontrolled Keywords:Inverse shrinkage, Hilbert transform, large-dimensional asymptotics, signal amplitude, Stein shrinkage, Kovarianzmatrix, Schätzfunktion, Monte-Carlo-Simulation
Scope:Discipline-based scholarship (basic research)
Language:English
Date:December 2020
Deposited On:13 Nov 2019 15:34
Last Modified:06 Mar 2024 14:31
Series Name:Working paper series / Department of Economics
Number of Pages:73
ISSN:1664-705X
Additional Information:Revised version
OA Status:Green
Other Identification Number:merlin-id:18787
  • Content: Updated Version
  • Language: English
  • Description: Revised version December 2020
  • Content: Published Version
  • Language: English
  • Permission: Download for registered users
  • Description: Version November 2019