Vine copula models for multivariate time series
In a multivariate time series, there are two types of dependence: cross-sectional, serial. Copulas can be used to model both types of dependence. Recently, several vine copula models have been proposed that capture both types in the same framework. I will review and generalize these approaches and show how this viewpoint provides new insights into other models, not necessarily based on vines. Lastly, I address some open problems and ongoing work in this area.