Testing for instability in covariance structures

We propose a test for the stability over time of the covariance matrix of multivariate time series. The analysis is extended to the eigensystem to ascertain changes due to instability in the eigenvalues and/or eigenvectors. Using strong Invariance Principles and Law of Large Numbers, we normalise th...

Full description

Bibliographic Details
Main Authors: Kao, Chihwa, Trapani, Lorenzo, Urga, Giovanni
Format: Article
Published: Bernoulli Society for Mathematical Statistics and Probability 2017
Subjects:
Online Access:https://eprints.nottingham.ac.uk/46942/
_version_ 1848797433613516800
author Kao, Chihwa
Trapani, Lorenzo
Urga, Giovanni
author_facet Kao, Chihwa
Trapani, Lorenzo
Urga, Giovanni
author_sort Kao, Chihwa
building Nottingham Research Data Repository
collection Online Access
description We propose a test for the stability over time of the covariance matrix of multivariate time series. The analysis is extended to the eigensystem to ascertain changes due to instability in the eigenvalues and/or eigenvectors. Using strong Invariance Principles and Law of Large Numbers, we normalise the CUSUM-type statistics to calculate their supremum over the whole sample. The power properties of the test versus alternative hypotheses, including also the case of breaks close to the beginning/end of sample are investigated theoretically and via simulation. We extend our theory to test for the stability of the covariance matrix of a multivariate regression model. The testing procedures are illustrated by studying the stability of the principal components of the term structure of 18 US interest rates.
first_indexed 2025-11-14T20:03:48Z
format Article
id nottingham-46942
institution University of Nottingham Malaysia Campus
institution_category Local University
last_indexed 2025-11-14T20:03:48Z
publishDate 2017
publisher Bernoulli Society for Mathematical Statistics and Probability
recordtype eprints
repository_type Digital Repository
spelling nottingham-469422020-05-04T18:57:06Z https://eprints.nottingham.ac.uk/46942/ Testing for instability in covariance structures Kao, Chihwa Trapani, Lorenzo Urga, Giovanni We propose a test for the stability over time of the covariance matrix of multivariate time series. The analysis is extended to the eigensystem to ascertain changes due to instability in the eigenvalues and/or eigenvectors. Using strong Invariance Principles and Law of Large Numbers, we normalise the CUSUM-type statistics to calculate their supremum over the whole sample. The power properties of the test versus alternative hypotheses, including also the case of breaks close to the beginning/end of sample are investigated theoretically and via simulation. We extend our theory to test for the stability of the covariance matrix of a multivariate regression model. The testing procedures are illustrated by studying the stability of the principal components of the term structure of 18 US interest rates. Bernoulli Society for Mathematical Statistics and Probability 2017-07-27 Article PeerReviewed Kao, Chihwa, Trapani, Lorenzo and Urga, Giovanni (2017) Testing for instability in covariance structures. Bernoulli, 24 (1). pp. 740-771. ISSN 1573-9759 changepoint; covariance matrix; CUSUM statistic; eigensystem https://projecteuclid.org/euclid.bj/1501142461 doi:10.3150/16-BEJ894 doi:10.3150/16-BEJ894
spellingShingle changepoint; covariance matrix; CUSUM statistic; eigensystem
Kao, Chihwa
Trapani, Lorenzo
Urga, Giovanni
Testing for instability in covariance structures
title Testing for instability in covariance structures
title_full Testing for instability in covariance structures
title_fullStr Testing for instability in covariance structures
title_full_unstemmed Testing for instability in covariance structures
title_short Testing for instability in covariance structures
title_sort testing for instability in covariance structures
topic changepoint; covariance matrix; CUSUM statistic; eigensystem
url https://eprints.nottingham.ac.uk/46942/
https://eprints.nottingham.ac.uk/46942/
https://eprints.nottingham.ac.uk/46942/