Characterizing Correlation Matrices that Admit a Clustered Factor Representation

Chen Tong; Peter Reinhard  Hansen

doi:10.1016/j.econlet.2023.111433

Characterizing Correlation Matrices that Admit a Clustered Factor Representation

Chen Tong, Peter Reinhard Hansen^*

^*Corresponding author af dette arbejde

Institut for Finansiering

Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › peer review

Abstract

The Clustered Factor (CF) model is commonly used to parametrize block correlation matrices. We show that the CF model imposes additional superfluous restrictions. This can be avoided by a different parametrization, based on the logarithmic block correlation matrix.

Originalsprog	Engelsk
Artikelnummer	111433
Tidsskrift	Economics Letters
Vol/bind	233
Antal sider	4
ISSN	0165-1765
DOI	https://doi.org/10.1016/j.econlet.2023.111433
Status	Udgivet - dec. 2023

Emneord

Block correlation matrix
Copula
Clustering
Factor models

Adgang til dokumentet

10.1016/j.econlet.2023.111433Licens: Ikke-specificeret

Handle.net

Persistent link

Citationsformater

@article{c5db9837101743a3aee32c2f3ec9ba4b,

title = "Characterizing Correlation Matrices that Admit a Clustered Factor Representation",

abstract = "The Clustered Factor (CF) model is commonly used to parametrize block correlation matrices. We show that the CF model imposes additional superfluous restrictions. This can be avoided by a different parametrization, based on the logarithmic block correlation matrix.",

keywords = "Block correlation matrix, Copula, Clustering, Factor models, Block correlation matrix, Copula, Clustering, Factor models",

author = "Chen Tong and Hansen, {Peter Reinhard}",

year = "2023",

month = dec,

doi = "10.1016/j.econlet.2023.111433",

language = "English",

volume = "233",

journal = "Economics Letters",

issn = "0165-1765",

publisher = "Elsevier",

}

TY - JOUR

T1 - Characterizing Correlation Matrices that Admit a Clustered Factor Representation

AU - Tong, Chen

AU - Hansen, Peter Reinhard

PY - 2023/12

Y1 - 2023/12

N2 - The Clustered Factor (CF) model is commonly used to parametrize block correlation matrices. We show that the CF model imposes additional superfluous restrictions. This can be avoided by a different parametrization, based on the logarithmic block correlation matrix.

AB - The Clustered Factor (CF) model is commonly used to parametrize block correlation matrices. We show that the CF model imposes additional superfluous restrictions. This can be avoided by a different parametrization, based on the logarithmic block correlation matrix.

KW - Block correlation matrix

KW - Copula

KW - Clustering

KW - Factor models

KW - Block correlation matrix

KW - Copula

KW - Clustering

KW - Factor models

U2 - 10.1016/j.econlet.2023.111433

DO - 10.1016/j.econlet.2023.111433

M3 - Journal article

SN - 0165-1765

VL - 233

JO - Economics Letters

JF - Economics Letters

M1 - 111433

ER -