Pair-copulas modeling in finance

This paper concerns itself with applications of pair-copulas in finance, and bridges the gap between theory and application. We provide a broad view of the problem of modeling multivariate financial log-returns using pair-copulas, gathering together for this purpose theoretical and computational results from the literature on canonical vines. From the practitioner’s viewpoint, the paper shows the advantages of modeling through pair-copulas and makes clear that it is possible to implement this methodology on a daily basis. All the necessary steps (model selection, estimation, validation, simulations, and applications) are discussed at a level easily understood by all data analysts.