Humboldt-Universität zu Berlin >> Interdisziplinäre Zentren

Humboldt Logo

Workshop on ''Copulae: Theory and Practice''

Date: 7-8. December, 2007

Organization Committee

organizer institute
Enzo Giacomini Humboldt-Universität zu Berlin
Prof. Dr. Wolfgang Härdle Humboldt-Universität zu Berlin
Ostap Okhrin European University Viadrina, Frankfurt (Oder)
Dr. Yarema Okhrin European University Viadrina, Frankfurt (Oder)

Despite of its numerous advantages, the Gaussian distribution is widely criticized as an appropriate way of modeling data in finance. Recently the need for non-Gaussian alternatives was recognized not only for modeling the univariate marginal distributions but also for modeling the dependencies within multivariate data sets. The theory of copulae provides a convenient and flexible solution to this problem. The copulae have numerous advantages. First, they allow the separation of the margins from the pure dependency structure. Second, they generate a much more rich family of distributions compared to the elliptical distributions, for example. And third, in most cases an explicit expression for the distribution function is available. Due to this, copulae found numerous applications in finance. Among the most important fields are modeling the asymmetries in dependencies, value-at-risk, portfolio selection under copula based distribution of asset returns, etc. The importance of copulae is reflected by an increasing number of papers published in top journals both in finance and in statistics.

The workshop is aimed to get together researchers working on copulae and to stimulate the exchange of ideas and current achievements. A two-day workshop provides an ideal opportunity for tight and intensive discussions. The workshop consists of an invited talk and 8-10 contributed talks forty five minutes each. Around half of the contributed talks will deal with the theoretical problems and the other half with the empirical applications.

For participants with no talk the participation fee is 100 euros and it is payable upon arrival to the workshop. It includes the workshop dinner. The registration form is downloadable from here

The workshop will take place in the Weierstrass Institute for Applied Analysis and Stochastics (Berlin, Mohrenstrasse, 39. Not far away from the Gendarmenmarkt)


Friday, 7 December 2007

Opening Talk by

Prof. Dr. Wolfgang Härdle
Humboldt-Universität zu Berlin
Qua de causa copulae me placent

Invited Talk by

Dr. Johanna Neslehova
ETH Zürich, Switzerland
A primer on copulas for count data
10:15-10:30 coffee break
10:30-11:15 Prof. Dr. Friedrich Schmid
University of Cologne

Measuring Conditional Dependance of Stock

Returns in Bull and Bear Markets

11:15-12:00 Kjersti Aas
Norwegian Computing Center
Pair-copula constructions - an introduction
12:00-14:00 lunch
14:00-14:45 Prof. Dr. Claudia Czado
Technical University Munich

Bayesian estimation of multivariate copulas using pair-copula constructions

14:45-15:30 Dr. Gabriel Frahm
University of Cologne

On the extremal dependence coefficient of multivariate distributions

15:30-15:45 coffee break
15:45-16:30 Barbara Choros
Humboldt-Universität zu Berlin
Modelling Dependence of Time Series with Copulae
16:30-17:15 Ostap Okhrin
European University Viadrina

HAC: Theory and Applications

19:00- workshop dinner
Saturday, 8 December 2007
9:00-9:45 Marius Hofert
Universitaet Ulm
Sampling Archimedean copulas
9:45-10:30 PD Dr. Matthias Fischer
University of Erlangen-Nürnberg

An empirical analysis of multivariate copula models

10:30-10:45 coffee break
10:45-11:30 Prof. Dr. Mark Trede
University of Münster
Testing Archimedeanity
11:30-12:15 Prof. Dr. Ray-Bing Chen
National University of Kaohsiung,Taiwan

Independent Component Analysis via Copula Techniques

12:15-13:00 Closing Talk by Enzo Giacomini
Humboldt-Universität zu Berlin
Inhomogenous Dependence Modelling with Time Varying Copulae


CASE - Center for Applied Statistics and Economics
Humboldt-Universität zu Berlin
Wirtschaftswissenschaftliche Fakultät
Spandauer Str. 1
10178 Berlin
Phone: +49(0)30-2093-5630
Fax: +49(0)30-2093-5649