Extending the CAPM using pair copulas: The Regular Vine Market Sector model

The demand for accurate models involving larger numbers of assets is strong not only in view of the financial crisis of 2007-2009. In particular dependencies among assets have not been captured adequately. While standard multivariate copulas have added some flexibility, this flexibility is insufficient in higher dimensional applications. Regular vines can fill this gap by benefiting from the rich class of existing bivariate parametric copula families. Exploiting this in combination with GARCH models for margins, we develop a regular vine copula based factor model for asset returns that extends the classical CAPM to non-linearity and non-normality. In a substantial application to European stock market returns we demonstrate its excellent performance in comparison to relevant benchmark models. In particular we show its superiority to the CAVA model proposed by Heinen and Valdesogo (2009).