Abstract
Using the language of copulas, we generalize the famous Fisher-Tippett Theorem of extreme value theory to the case with sequences of dependent random variables. The dependence structure is modelled using archimedean copulas. This generalization enables to study the behaviour of the maxima of dependent random sequences.
Keywords: Archimedean Copula, Dependent Risks, Extreme Value Theory, Maximum Domain Of Attraction, Fisher-Tippett Theorem
Volume
No. 3
Page
211-228
Year
2004
Categories
Financial and Statistical Methods
Simulation
Copulas/Multi-Variate Distributions
Financial and Statistical Methods
Loss Distributions
Extreme Values
Financial and Statistical Methods
Extreme Event Modeling
Publications
Scandinavian Actuarial Journal
