Empirical likelihood based testing for multivariate regular variation
Open Access
- 1 February 2025
- journal article
- Published by Institute of Mathematical Statistics in The Annals of Statistics
- Vol. 53 (1) , 352-373
Abstract
Multivariate regular variation is a common assumption in the statistics literature and needs to be verified in real-data applications. We develop a novel hypothesis test for multivariate regular variation, employing localized empirical likelihood. We establish the weak convergence of the test statistic to a nonstandard, distribution-free limit and hence can provide universal critical values for the test. We show the very good finite-sample behavior of the procedure through simulations and apply the test to several real-data examples.Keywords
This publication has 26 references indexed in Scilit:
- Risk contagion under regular variation and asymptotic tail independenceJournal of Multivariate Analysis, 2018
- Extreme geometric quantiles in a multivariate regular variation frameworkExtremes, 2015
- Portfolio optimization for heavy-tailed assets: Extreme Risk Index vs. MarkowitzJournal of Empirical Finance, 2015
- Heavy tailed time series with extremal independenceExtremes, 2015
- Tail Risk of Multivariate Regular VariationMethodology and Computing in Applied Probability, 2010
- On optimal portfolio diversification with respect to extreme risksFinance and Stochastics, 2010
- Regularly varying multivariate time seriesStochastic Processes and their Applications, 2009
- Poisson and Gaussian approximation of weighted local empirical processesStochastic Processes and their Applications, 1997
- Rates of growth and sample moduli for weighted empirical processes indexed by setsProbability Theory and Related Fields, 1987
- Extreme Values, Regular Variation and Point Processes1987