A new bootstrap test for multiple assets joint risk testing

David Ardia, Lukasz Gatarek, Lennart F. Hoogerheide

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

In this paper, a novel simulation-based methodology is proposed to test the validity of a set of marginal time series models, where the dependence structure between the time series is taken directly from the observed data. The procedure is useful when one wants to summarize the test results for several time series in one joint test statistic and p value. The proposed test method can have higher statistical power than a test for a univariate time series, especially for short time series. Therefore, our test for multiple time series is particularly useful if one wants to assess value-at-risk (or expected shortfall) predictions within a small time frame (eg, a crisis period). We apply our method to test generalized autoregressive conditional heteroscedasticity (GARCH) model specifications for a large panel data set of stock returns.

Original languageEnglish
Pages (from-to)1-22
Number of pages22
JournalJournal of Risk
Volume19
Issue number4
DOIs
Publication statusPublished - 2017

Fingerprint

Bootstrap test
Testing
Assets
Dependence structure
Value at risk
Multiple time series
Stock returns
Prediction
Statistical power
P value
Expected shortfall
Methodology
Test statistic
Time series models
Panel data
Test methods
Simulation
Model specification
Generalized autoregressive conditional heteroscedasticity

Keywords

  • Bootstrap test
  • Generalized autoregressive conditional heteroscedasticity (GARCH)
  • Marginal models
  • Multiple time series
  • Value-at-risk (VaR)

Cite this

Ardia, David ; Gatarek, Lukasz ; Hoogerheide, Lennart F. / A new bootstrap test for multiple assets joint risk testing. In: Journal of Risk. 2017 ; Vol. 19, No. 4. pp. 1-22.
@article{549066efff974e49be733576a69d862b,
title = "A new bootstrap test for multiple assets joint risk testing",
abstract = "In this paper, a novel simulation-based methodology is proposed to test the validity of a set of marginal time series models, where the dependence structure between the time series is taken directly from the observed data. The procedure is useful when one wants to summarize the test results for several time series in one joint test statistic and p value. The proposed test method can have higher statistical power than a test for a univariate time series, especially for short time series. Therefore, our test for multiple time series is particularly useful if one wants to assess value-at-risk (or expected shortfall) predictions within a small time frame (eg, a crisis period). We apply our method to test generalized autoregressive conditional heteroscedasticity (GARCH) model specifications for a large panel data set of stock returns.",
keywords = "Bootstrap test, Generalized autoregressive conditional heteroscedasticity (GARCH), Marginal models, Multiple time series, Value-at-risk (VaR)",
author = "David Ardia and Lukasz Gatarek and Hoogerheide, {Lennart F.}",
year = "2017",
doi = "10.21314/JOR.2017.356",
language = "English",
volume = "19",
pages = "1--22",
journal = "Journal of Risk",
issn = "1465-1211",
publisher = "Incisive Media Ltd.",
number = "4",

}

A new bootstrap test for multiple assets joint risk testing. / Ardia, David; Gatarek, Lukasz; Hoogerheide, Lennart F.

In: Journal of Risk, Vol. 19, No. 4, 2017, p. 1-22.

Research output: Contribution to JournalArticleAcademicpeer-review

TY - JOUR

T1 - A new bootstrap test for multiple assets joint risk testing

AU - Ardia, David

AU - Gatarek, Lukasz

AU - Hoogerheide, Lennart F.

PY - 2017

Y1 - 2017

N2 - In this paper, a novel simulation-based methodology is proposed to test the validity of a set of marginal time series models, where the dependence structure between the time series is taken directly from the observed data. The procedure is useful when one wants to summarize the test results for several time series in one joint test statistic and p value. The proposed test method can have higher statistical power than a test for a univariate time series, especially for short time series. Therefore, our test for multiple time series is particularly useful if one wants to assess value-at-risk (or expected shortfall) predictions within a small time frame (eg, a crisis period). We apply our method to test generalized autoregressive conditional heteroscedasticity (GARCH) model specifications for a large panel data set of stock returns.

AB - In this paper, a novel simulation-based methodology is proposed to test the validity of a set of marginal time series models, where the dependence structure between the time series is taken directly from the observed data. The procedure is useful when one wants to summarize the test results for several time series in one joint test statistic and p value. The proposed test method can have higher statistical power than a test for a univariate time series, especially for short time series. Therefore, our test for multiple time series is particularly useful if one wants to assess value-at-risk (or expected shortfall) predictions within a small time frame (eg, a crisis period). We apply our method to test generalized autoregressive conditional heteroscedasticity (GARCH) model specifications for a large panel data set of stock returns.

KW - Bootstrap test

KW - Generalized autoregressive conditional heteroscedasticity (GARCH)

KW - Marginal models

KW - Multiple time series

KW - Value-at-risk (VaR)

UR - http://www.scopus.com/inward/record.url?scp=85022229766&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85022229766&partnerID=8YFLogxK

U2 - 10.21314/JOR.2017.356

DO - 10.21314/JOR.2017.356

M3 - Article

VL - 19

SP - 1

EP - 22

JO - Journal of Risk

JF - Journal of Risk

SN - 1465-1211

IS - 4

ER -