Dependence modeling and portfolio risk estimation using GARCH-copula approach

Ruzanna Ab Razak, Noriszura Ismail

Research output: Contribution to journalArticle

Abstract

Past studies have shown that linear correlation measure may result in misleading interpretations and implications of dependency when financial variables are involved. The copula approach can be adopted as an alternative for measuring dependence as it provides the solution to fat tail problems in multivariate cases which arises from the probability of large or extreme co-movements. Due to limited studies on copulas using Islamic financial data, this study set outs to obtain a clear picture on the dependence between Islamic and conventional stock markets in Malaysia. Firstly, we model the dependence between Islamic and conventional returns data using the copula-ARMA-GARCH models with normal and non-normal error distributions, and secondly, we evaluate the portfolios of Islamic and conventional indices using recent risk measures. This paper shows that, by using the copula approach for measuring the dependency between two financial variables while maintaining their true nature as described by the ARMA-GARCH models, meaningful interpretation can be made about the association of the financial variables which reflects the real association between markets. Furthermore, this study proposes a set of procedures on how portfolio risks can be estimated using VaR based on the ARMA(p,q)GARCH(1,1)-t-copula models including backtesting via simulation.

Original languageEnglish
Pages (from-to)1547-1555
Number of pages9
JournalSains Malaysiana
Volume48
Issue number7
DOIs
Publication statusPublished - 1 Jan 2019

Fingerprint

Portfolio risk
Modeling
Generalized autoregressive conditional heteroscedasticity
Copula
Financial variables
Autoregressive moving average
GARCH model
Backtesting
Risk measures
Simulation
Comovement
Fat tails
Financial data
Malaysia
Stock market

Keywords

  • Copula
  • GARCH
  • Risk
  • Stock returns

ASJC Scopus subject areas

  • General

Cite this

Dependence modeling and portfolio risk estimation using GARCH-copula approach. / Ab Razak, Ruzanna; Ismail, Noriszura.

In: Sains Malaysiana, Vol. 48, No. 7, 01.01.2019, p. 1547-1555.

Research output: Contribution to journalArticle

@article{27ab4784b904485e8b81d86aa638e04e,
title = "Dependence modeling and portfolio risk estimation using GARCH-copula approach",
abstract = "Past studies have shown that linear correlation measure may result in misleading interpretations and implications of dependency when financial variables are involved. The copula approach can be adopted as an alternative for measuring dependence as it provides the solution to fat tail problems in multivariate cases which arises from the probability of large or extreme co-movements. Due to limited studies on copulas using Islamic financial data, this study set outs to obtain a clear picture on the dependence between Islamic and conventional stock markets in Malaysia. Firstly, we model the dependence between Islamic and conventional returns data using the copula-ARMA-GARCH models with normal and non-normal error distributions, and secondly, we evaluate the portfolios of Islamic and conventional indices using recent risk measures. This paper shows that, by using the copula approach for measuring the dependency between two financial variables while maintaining their true nature as described by the ARMA-GARCH models, meaningful interpretation can be made about the association of the financial variables which reflects the real association between markets. Furthermore, this study proposes a set of procedures on how portfolio risks can be estimated using VaR based on the ARMA(p,q)GARCH(1,1)-t-copula models including backtesting via simulation.",
keywords = "Copula, GARCH, Risk, Stock returns",
author = "{Ab Razak}, Ruzanna and Noriszura Ismail",
year = "2019",
month = "1",
day = "1",
doi = "10.17576/jsm-2019-4807-24",
language = "English",
volume = "48",
pages = "1547--1555",
journal = "Sains Malaysiana",
issn = "0126-6039",
publisher = "Penerbit Universiti Kebangsaan Malaysia",
number = "7",

}

TY - JOUR

T1 - Dependence modeling and portfolio risk estimation using GARCH-copula approach

AU - Ab Razak, Ruzanna

AU - Ismail, Noriszura

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Past studies have shown that linear correlation measure may result in misleading interpretations and implications of dependency when financial variables are involved. The copula approach can be adopted as an alternative for measuring dependence as it provides the solution to fat tail problems in multivariate cases which arises from the probability of large or extreme co-movements. Due to limited studies on copulas using Islamic financial data, this study set outs to obtain a clear picture on the dependence between Islamic and conventional stock markets in Malaysia. Firstly, we model the dependence between Islamic and conventional returns data using the copula-ARMA-GARCH models with normal and non-normal error distributions, and secondly, we evaluate the portfolios of Islamic and conventional indices using recent risk measures. This paper shows that, by using the copula approach for measuring the dependency between two financial variables while maintaining their true nature as described by the ARMA-GARCH models, meaningful interpretation can be made about the association of the financial variables which reflects the real association between markets. Furthermore, this study proposes a set of procedures on how portfolio risks can be estimated using VaR based on the ARMA(p,q)GARCH(1,1)-t-copula models including backtesting via simulation.

AB - Past studies have shown that linear correlation measure may result in misleading interpretations and implications of dependency when financial variables are involved. The copula approach can be adopted as an alternative for measuring dependence as it provides the solution to fat tail problems in multivariate cases which arises from the probability of large or extreme co-movements. Due to limited studies on copulas using Islamic financial data, this study set outs to obtain a clear picture on the dependence between Islamic and conventional stock markets in Malaysia. Firstly, we model the dependence between Islamic and conventional returns data using the copula-ARMA-GARCH models with normal and non-normal error distributions, and secondly, we evaluate the portfolios of Islamic and conventional indices using recent risk measures. This paper shows that, by using the copula approach for measuring the dependency between two financial variables while maintaining their true nature as described by the ARMA-GARCH models, meaningful interpretation can be made about the association of the financial variables which reflects the real association between markets. Furthermore, this study proposes a set of procedures on how portfolio risks can be estimated using VaR based on the ARMA(p,q)GARCH(1,1)-t-copula models including backtesting via simulation.

KW - Copula

KW - GARCH

KW - Risk

KW - Stock returns

UR - http://www.scopus.com/inward/record.url?scp=85071511473&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85071511473&partnerID=8YFLogxK

U2 - 10.17576/jsm-2019-4807-24

DO - 10.17576/jsm-2019-4807-24

M3 - Article

AN - SCOPUS:85071511473

VL - 48

SP - 1547

EP - 1555

JO - Sains Malaysiana

JF - Sains Malaysiana

SN - 0126-6039

IS - 7

ER -