Value at Risk estimation using Extreme Value Theory

Abhay K. Singh, David E. Allen, Robert J. Powell

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contributionResearchpeer-review

Abstract

A common assumption in quantitative financial risk modelling is the distributional assumption of normality in the asset's return series, which makes modelling easy but proves to be inefficient if the data exhibit extreme tails. When dealing with extreme financial events like the Global Financial Crisis of 2007-2008 while quantifying extreme market risk, Extreme Value Theory (EVT) proves to be a natural statistical modelling technique of interest. Extreme Value Theory provides well established statistical models for the computation of extreme risk measures like the Return Level, Value at Risk and Expected Shortfall. In this paper we apply Univariate Extreme Value Theory to model extreme market risk for the ASX-All Ordinaries (Australian) index and the S&P-500 (USA) Index. We demonstrate that EVT can be successfully applied to Australian stock market return series for predicting next day VaR by using a GARCH(1,1) based dynamic EVT approach. We also show with backtesting results that EVT based method outperforms GARCH(1,1) and RiskMetrics™ based forecasts.

LanguageEnglish
Title of host publicationMODSIM 2011
Subtitle of host publication19th International Congress on Modelling and Simulation: proceedings
EditorsF. Chan, D. Marinova, R. S. Anderssen
Place of PublicationCanberra
Pages1478-1484
Number of pages7
Publication statusPublished - 2011
Externally publishedYes
Event19th International Congress on Modelling and Simulation - Sustaining Our Future: Understanding and Living with Uncertainty, MODSIM2011 - Perth, WA, Australia
Duration: 12 Dec 201116 Dec 2011

Conference

Conference19th International Congress on Modelling and Simulation - Sustaining Our Future: Understanding and Living with Uncertainty, MODSIM2011
CountryAustralia
CityPerth, WA
Period12/12/1116/12/11

Fingerprint

Estimation Theory
Extreme Value Theory
Value at Risk
Extremes
Generalized Autoregressive Conditional Heteroscedasticity
Expected Shortfall
Financial Risk
Financial Crisis
Series
Risk Measures
Statistical Modeling
Stock Market
Modeling
Normality
Statistical Model
Univariate
Forecast
Tail
Demonstrate

Keywords

  • Extreme Value Theory
  • GARCH
  • Risk modelling
  • RiskMetrics™
  • Value at Risk

Cite this

Singh, A. K., Allen, D. E., & Powell, R. J. (2011). Value at Risk estimation using Extreme Value Theory. In F. Chan, D. Marinova, & R. S. Anderssen (Eds.), MODSIM 2011: 19th International Congress on Modelling and Simulation: proceedings (pp. 1478-1484). Canberra.
Singh, Abhay K. ; Allen, David E. ; Powell, Robert J. / Value at Risk estimation using Extreme Value Theory. MODSIM 2011: 19th International Congress on Modelling and Simulation: proceedings. editor / F. Chan ; D. Marinova ; R. S. Anderssen. Canberra, 2011. pp. 1478-1484
@inproceedings{058ae73957a34396875f0e95b1b40d5f,
title = "Value at Risk estimation using Extreme Value Theory",
abstract = "A common assumption in quantitative financial risk modelling is the distributional assumption of normality in the asset's return series, which makes modelling easy but proves to be inefficient if the data exhibit extreme tails. When dealing with extreme financial events like the Global Financial Crisis of 2007-2008 while quantifying extreme market risk, Extreme Value Theory (EVT) proves to be a natural statistical modelling technique of interest. Extreme Value Theory provides well established statistical models for the computation of extreme risk measures like the Return Level, Value at Risk and Expected Shortfall. In this paper we apply Univariate Extreme Value Theory to model extreme market risk for the ASX-All Ordinaries (Australian) index and the S&P-500 (USA) Index. We demonstrate that EVT can be successfully applied to Australian stock market return series for predicting next day VaR by using a GARCH(1,1) based dynamic EVT approach. We also show with backtesting results that EVT based method outperforms GARCH(1,1) and RiskMetrics™ based forecasts.",
keywords = "Extreme Value Theory, GARCH, Risk modelling, RiskMetrics™, Value at Risk",
author = "Singh, {Abhay K.} and Allen, {David E.} and Powell, {Robert J.}",
year = "2011",
language = "English",
isbn = "9780987214317",
pages = "1478--1484",
editor = "F. Chan and D. Marinova and Anderssen, {R. S.}",
booktitle = "MODSIM 2011",

}

Singh, AK, Allen, DE & Powell, RJ 2011, Value at Risk estimation using Extreme Value Theory. in F Chan, D Marinova & RS Anderssen (eds), MODSIM 2011: 19th International Congress on Modelling and Simulation: proceedings. Canberra, pp. 1478-1484, 19th International Congress on Modelling and Simulation - Sustaining Our Future: Understanding and Living with Uncertainty, MODSIM2011, Perth, WA, Australia, 12/12/11.

Value at Risk estimation using Extreme Value Theory. / Singh, Abhay K.; Allen, David E.; Powell, Robert J.

MODSIM 2011: 19th International Congress on Modelling and Simulation: proceedings. ed. / F. Chan; D. Marinova; R. S. Anderssen. Canberra, 2011. p. 1478-1484.

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contributionResearchpeer-review

TY - GEN

T1 - Value at Risk estimation using Extreme Value Theory

AU - Singh, Abhay K.

AU - Allen, David E.

AU - Powell, Robert J.

PY - 2011

Y1 - 2011

N2 - A common assumption in quantitative financial risk modelling is the distributional assumption of normality in the asset's return series, which makes modelling easy but proves to be inefficient if the data exhibit extreme tails. When dealing with extreme financial events like the Global Financial Crisis of 2007-2008 while quantifying extreme market risk, Extreme Value Theory (EVT) proves to be a natural statistical modelling technique of interest. Extreme Value Theory provides well established statistical models for the computation of extreme risk measures like the Return Level, Value at Risk and Expected Shortfall. In this paper we apply Univariate Extreme Value Theory to model extreme market risk for the ASX-All Ordinaries (Australian) index and the S&P-500 (USA) Index. We demonstrate that EVT can be successfully applied to Australian stock market return series for predicting next day VaR by using a GARCH(1,1) based dynamic EVT approach. We also show with backtesting results that EVT based method outperforms GARCH(1,1) and RiskMetrics™ based forecasts.

AB - A common assumption in quantitative financial risk modelling is the distributional assumption of normality in the asset's return series, which makes modelling easy but proves to be inefficient if the data exhibit extreme tails. When dealing with extreme financial events like the Global Financial Crisis of 2007-2008 while quantifying extreme market risk, Extreme Value Theory (EVT) proves to be a natural statistical modelling technique of interest. Extreme Value Theory provides well established statistical models for the computation of extreme risk measures like the Return Level, Value at Risk and Expected Shortfall. In this paper we apply Univariate Extreme Value Theory to model extreme market risk for the ASX-All Ordinaries (Australian) index and the S&P-500 (USA) Index. We demonstrate that EVT can be successfully applied to Australian stock market return series for predicting next day VaR by using a GARCH(1,1) based dynamic EVT approach. We also show with backtesting results that EVT based method outperforms GARCH(1,1) and RiskMetrics™ based forecasts.

KW - Extreme Value Theory

KW - GARCH

KW - Risk modelling

KW - RiskMetrics™

KW - Value at Risk

UR - http://www.scopus.com/inward/record.url?scp=84863405515&partnerID=8YFLogxK

M3 - Conference proceeding contribution

SN - 9780987214317

SP - 1478

EP - 1484

BT - MODSIM 2011

A2 - Chan, F.

A2 - Marinova, D.

A2 - Anderssen, R. S.

CY - Canberra

ER -

Singh AK, Allen DE, Powell RJ. Value at Risk estimation using Extreme Value Theory. In Chan F, Marinova D, Anderssen RS, editors, MODSIM 2011: 19th International Congress on Modelling and Simulation: proceedings. Canberra. 2011. p. 1478-1484