CMStatistics 2018: Start Registration
View Submission - CFE
Title: Nonparametric dynamic copula modelling to analyze dependence structures between domestic indexes Authors:  Jone Ascorbebeitia - University of the Basque Country (Spain) [presenting]
Eva Ferreira - University of the Basque Country (Spain)
Susan Orbe - University of the Basque Country (Spain)
Abstract: The analysis between portfolio comovements is of great interest in economics and finance in order to have as much as possible power over the risk. The time varying asset dynamics make more difficult to control those comovements and require much more accuracy and more sophisticated estimation models able to capture the dynamics. Unfortunately, financial variables are non-gaussian distributed and present more complicated structures. So, to measure the dependence between them it is necessary to consider dependence measures beyond Pearson's linear correlation. To overcome this fact, we suggest the use of time-varying copulas to analyze the relationship between European domestic index dynamics. As it is already known, the use of copulas allows us to model the dependence better than elliptic distributions do. In this context, we propose a nonparametric time-varying dependence estimator based on Kendall's tau to analyze the dependence between index dynamics and we derive its asymptotic properties. A simulation study investigates the performance of the estimator and compares with the performance of other dependence estimation methods existing in the literature. Finally, we provide a statistic to test for Kendall's tau significance.