Using the R package VineCopula
Ulf Schepsmeier, Technical University of Munich
Vine copulas are a flexible class of dependence models consisting of bivariate copulas as building blocks. The R package VineCopula is primarily made for the statistical analysis of vine copula models. The package comprehensively offers parameter estimation, model selection, simulation, goodness-of-fit tests, and visualization. Additionally, methods for estimation, selection and exploratory data analysis of bivariate copula models are also provided. In the workshop we will introduce the functionality of the VineCopula package by giving illustrative data examples and code chunks. Data analysis, model selection and estimation will be explained step by step. The theoretical explanations are paired with practical exercises.
Requirements:
- Basic knowledge about (vine) copulas, a comprehensive list of publications and other materials can be found on vine-copula.org.
- basic R knowledge
- Laptop with R and the package installed. Resources are:
- the stable release on CRAN:
install.pacakges("VineCopula") or
cran.r-project.org/web/packages/VineCopula/index.html or - the latest development version:
devtools::install_github("tnagler/VineCopula") or
github.com/tnagler/VineCopula/
- the stable release on CRAN:
The Rearrangement Algorithm: theory and practice.
Giovanni Puccetti, University of Milan
The Rearrangement Algorithm is a numerical tool developed to compute dependence uncertainty bounds on the VaR/ES of a joint risk portfolio. In this workshop, we illustrate the theory and the practice of the algorithm by applying it to a real risk portfolio. Various case studies, sample codes and extensions will be further studied through a live implementation in R.
Sample codes will be obtained from the website.
Requirements:
- basic R knowledge
- Laptop with R installed. A version of R can be found at https://www.r-project.org