This thesis analyses the dependence structure of 11 global stock markets to estimate portfolio risk measures. A parametric copula model from Huser and Wadsworth (forthcoming 2018) is applied to model the co-exceedances over a threshold. The model allows a smooth transition between asymptotic dependence and asymptotic independence. We find that the strongest spillover effect exists for countries within the European Union as opposed to more geographically diverse countries. The risk estimates based on the Huser and Wadsworth (forthcoming 2018) model outperform the benchmark estimates based on conventional copula models.

Additional Metadata
Keywords Keywords : asymptotic (in)dependence, copula, Value-at-Risk
Thesis Advisor Zhou, C.
Persistent URL
Series Econometrie
Kroondijk, K.R.R. (2019, June 11). Portfolio risk management: a new model for asymptotic (in)dependence. Econometrie. Retrieved from