Home > Research > Publications & Outputs > Two Cholesky-log-GARCH models for multivariate ...

Associated organisational unit

Links

Text available via DOI:

View graph of relations

Two Cholesky-log-GARCH models for multivariate volatilities

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

Two Cholesky-log-GARCH models for multivariate volatilities. / Pedeli, X.; Fokianos, K.; Pourahmadi, M.
In: Statistical Modelling, Vol. 15, No. 3, 2015, p. 233-255.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

Pedeli, X, Fokianos, K & Pourahmadi, M 2015, 'Two Cholesky-log-GARCH models for multivariate volatilities', Statistical Modelling, vol. 15, no. 3, pp. 233-255. https://doi.org/10.1177/1471082X14551246

APA

Pedeli, X., Fokianos, K., & Pourahmadi, M. (2015). Two Cholesky-log-GARCH models for multivariate volatilities. Statistical Modelling, 15(3), 233-255. https://doi.org/10.1177/1471082X14551246

Vancouver

Pedeli X, Fokianos K, Pourahmadi M. Two Cholesky-log-GARCH models for multivariate volatilities. Statistical Modelling. 2015;15(3):233-255. Epub 2014 Nov 27. doi: 10.1177/1471082X14551246

Author

Pedeli, X. ; Fokianos, K. ; Pourahmadi, M. / Two Cholesky-log-GARCH models for multivariate volatilities. In: Statistical Modelling. 2015 ; Vol. 15, No. 3. pp. 233-255.

Bibtex

@article{0000bafeab9245539193ed1d94dca227,
title = "Two Cholesky-log-GARCH models for multivariate volatilities",
abstract = "Parsimonious estimation of high-dimensional covariance matrices is of fundamental importance in multivariate statistics. Typical examples occur in finance, where the instantaneous dependence among several asset returns should be taken into account. Multivariate GARCH processes have been established as a standard approach for modelling such data. However, the majority of GARCH-type models are either based on strong assumptions that may not be realistic or require restrictions that are often too hard to be satisfied in practice. We consider two alternative decompositions of time-varying covariance matrices Σt. The first is based on the modified Cholesky decomposition of the covariance matrices and second relies on the hyperspherical parametrization of the standard Cholesky factor of their correlation matrices Rt. Then, we combine each Cholesky factor with the log-GARCH models for the corresponding time–varying volatilities and use a quasi maximum likelihood approach to estimate the parameters. Using log-GARCH models is quite natural for achieving the positive definiteness of Σt and this is a novelty of this work. Application of the proposed methodologies to two real financial datasets reveals their usefulness in terms of parsimony, ease of implementation and stresses the choice of the appropriate models using familiar data-driven processes such as various forms of the exploratory data analysis and regression.",
keywords = "Cholesky decomposition, Covariance matrix, GARCH model, hyperspherical coordinates, volatility",
author = "X. Pedeli and K. Fokianos and M. Pourahmadi",
year = "2015",
doi = "10.1177/1471082X14551246",
language = "English",
volume = "15",
pages = "233--255",
journal = "Statistical Modelling",
issn = "1471-082X",
publisher = "SAGE Publications Ltd",
number = "3",

}

RIS

TY - JOUR

T1 - Two Cholesky-log-GARCH models for multivariate volatilities

AU - Pedeli, X.

AU - Fokianos, K.

AU - Pourahmadi, M.

PY - 2015

Y1 - 2015

N2 - Parsimonious estimation of high-dimensional covariance matrices is of fundamental importance in multivariate statistics. Typical examples occur in finance, where the instantaneous dependence among several asset returns should be taken into account. Multivariate GARCH processes have been established as a standard approach for modelling such data. However, the majority of GARCH-type models are either based on strong assumptions that may not be realistic or require restrictions that are often too hard to be satisfied in practice. We consider two alternative decompositions of time-varying covariance matrices Σt. The first is based on the modified Cholesky decomposition of the covariance matrices and second relies on the hyperspherical parametrization of the standard Cholesky factor of their correlation matrices Rt. Then, we combine each Cholesky factor with the log-GARCH models for the corresponding time–varying volatilities and use a quasi maximum likelihood approach to estimate the parameters. Using log-GARCH models is quite natural for achieving the positive definiteness of Σt and this is a novelty of this work. Application of the proposed methodologies to two real financial datasets reveals their usefulness in terms of parsimony, ease of implementation and stresses the choice of the appropriate models using familiar data-driven processes such as various forms of the exploratory data analysis and regression.

AB - Parsimonious estimation of high-dimensional covariance matrices is of fundamental importance in multivariate statistics. Typical examples occur in finance, where the instantaneous dependence among several asset returns should be taken into account. Multivariate GARCH processes have been established as a standard approach for modelling such data. However, the majority of GARCH-type models are either based on strong assumptions that may not be realistic or require restrictions that are often too hard to be satisfied in practice. We consider two alternative decompositions of time-varying covariance matrices Σt. The first is based on the modified Cholesky decomposition of the covariance matrices and second relies on the hyperspherical parametrization of the standard Cholesky factor of their correlation matrices Rt. Then, we combine each Cholesky factor with the log-GARCH models for the corresponding time–varying volatilities and use a quasi maximum likelihood approach to estimate the parameters. Using log-GARCH models is quite natural for achieving the positive definiteness of Σt and this is a novelty of this work. Application of the proposed methodologies to two real financial datasets reveals their usefulness in terms of parsimony, ease of implementation and stresses the choice of the appropriate models using familiar data-driven processes such as various forms of the exploratory data analysis and regression.

KW - Cholesky decomposition

KW - Covariance matrix

KW - GARCH model

KW - hyperspherical coordinates

KW - volatility

U2 - 10.1177/1471082X14551246

DO - 10.1177/1471082X14551246

M3 - Journal article

VL - 15

SP - 233

EP - 255

JO - Statistical Modelling

JF - Statistical Modelling

SN - 1471-082X

IS - 3

ER -