Bayesian Model Selection of Regular Vine Copulas

Regular vine copulas are a novel and very flexible class of dependence models. This paper presents a reversible jump MCMC strategy for Bayesian model selection and inference of regular vine copulas, which can select all levels of a regular vine copula simultaneously. This is a substantial improvement over existing frequentist and Bayesian strategies, which can only select in a sequential, level-by-level scheme, which is known to induce selection bias. A simulation study demonstrates the feasibility of our strategy and shows that it combines superior selection and reduced computation time compared to sequential Bayesian selection. In a real data example, we forecast the daily expected tail loss of a portfolio of nine exchangetraded funds to illustrate the viability of our proposed copula model for risk estimation.