Asymptotically distribution-free goodness-of-fit testing for tail copulas
S.U. Can – University of Amsterdam, J.H.J. Einmahl – Tilburg University,
E.V. Khmaladze – Victoria University of Wellington & R.J.A. Laeven – University of Amsterdam

Tail copulas are an important tool in modeling the dependence of extreme values of random variables. Checking the goodness-of-fit of a given parametric family of tail copulas to the observed data is therefore an important problem. We propose a methodology that will allow practitioners to construct goodness-of-fit tests for tail copulas. The tests based on our methodology are asymptotically distribution-free, which means that for large enough sample sizes, the distribution of the test statistics under the null hypothesis do not depend on the underlying data or the particular model being tested.

Sami Umut Can

Umut Can was born in 1980 in Istanbul, Turkey. He received his Bachelor’s degree in Mathematics from Cornell University in 2003 and his Ph.D. in Applied Mathematics from Cornell University in 2010. After a two-year post-doctoral position at Tilburg University, he joined the Actuarial Science section of the Faculty of Economics and Business at University of Amsterdam, as an assistant professor. To know more…


“There are many systems in the real world, in finance, insurance, engineering, etc, that are vulnerable to extreme shocks. Understanding the extreme behavior ofrandom quantities is crucial to properly model and protect such systems. Plus, the mathematical theory of extreme values is simply beautiful.”

Leave a Reply

Your email address will not be published. Required fields are marked *