基于极值理论和Copula的灾难风险建模研究

Study on synthetical models for catastrophe based on EVT-Copula method

利用极值理论中的广义极值分布(GEVD)对单一灾难事件的特性进行刻画,然后利用Copula函数把不同的灾难风险组合起来,构造其联合分布,并提出了适合模型的蒙特卡罗算法.最后利用实例说明了GEVD-Copula组合建模的应用,得出了有益结论.

Generalized extreme value distribution is introduced to describe the single disaster event. The different risk features are combined with Copula to build the joint distribution. An appropriate MC method is also developed for the GEVD-Copula model. It is applied in a practical instance. A meaningful conclusion is obtained.