Résumé en anglais

Understanding the relationships among multivariate assets would help one greatly about how best to position one’s investments and enhance one’s financial risk protection. We present a new method to model parametrically the dependence structure of stock index returns through a continuous distribution function, which links an n‐dimensional density to its one‐dimensional margins. The resulting multivariate model could be used in a wide range of financial applications. Focusing on risk management, we show that a misspecification of the dependence structure introduces, on average, an error in Value‐at‐Risk estimates.