F-Vine Copula : Profiting from Asset Price Heterogeneity to Estimate Systematic Risk

In this paper we introduce a new model, based on the synthesis of conditional copulas and Gaussian graphical models under a copula -- vine framework. The use of the copula vine permits each pair between the market and a stock to have their own dynamics. In that case the asset keeps its statistical properties allowing for more realistic estimate of systematic risk. It turns out that this important feature of the model along with the consideration of tail dependent structures can be proved extremely meaningful when asset dynamics are driven by heterogeneous agent trading. Our empirical findings are in line with those of Chiarella et al. (2011) who find that taking the diversity of agents' beliefs into account reduces the market risk