Statistical Analysis of Copulas

What it is about

You learn to model multivariate dependency structures using Copulas and to understand, estimate, and apply Vine-Copula models (C-, D- and R-Vines). In the end you can select suitable vine models, implement them in R with VineCopula/CDVine, simulate, interpret and assess results.

What you will be able to do

• Knowledge of important bivariate Copula families and their properties
• Understanding of dependence measures (including tail dependence)
• Construction and structuring of C-, D- and R-Vines
• Simulation from vine models
• Parameter estimation and model selection for vines (IFM, MPL, Bayes)
• Application of the R packages CDVine() and VineCopula()
• Conducting goodness-of-fit tests for copulas
• Understanding of special vine models (e.g., time varying, Markov switching, factor vines)

What the module consists of

• LecturePresentation of content, demonstrative examples and discussions
• Übung / PraxisExercise sheets, solutions and practical data analysis for deepening understanding

Teaching method

• LectureIntroduction of concepts and examples, encouragement of independent literature work
• Theoretical and data exercises for self studyDeepening through exercises and practical applications
• Practice sessionsConducting exercises, availability of tasks and solutions for self-check