中文摘要：

本文以1996-2011年中国地震巨灾损失数据为样本,以对数正态分布拟合地震巨灾损失分布,运用VaR估计地震巨灾再保险最优自留比例,并运用蒙特卡罗模拟一定置信度下的地震巨灾损失的条件在险价值（CVaR）,以CVaR为风险度量指标进行地震巨灾基金规模的测算。本文实证结果表明,在其他条件相同的情况下,基于CVaR测算的地震基金规模要高于基于VaR测算的地震巨灾基金规模,且CVaR对超大规模的地震损失更敏感。建议在设置巨灾基金规模时,以VaR与CVaR作为风险度量指标进行风险分层。

英文摘要：

Selecting the data of Chinese earthquake catastrophe losses in 1996-2011 as a sample,this paper uses logarithm normal distribution to fit the earthquake catastrophe loss distribution and uses VaR to estimate the optimal retention ratio of earthquake catastrophe reinsurance.Monte Carlo method is used to simulate the conditional value at risk（ CVaR） of the earthquake catastrophe losses under a certain confidence level and CVaR is used as a risk measurement index to calculate the fund size of earthquake catastrophe. The empirical results show that,under the case that other conditions are the same,the earthquake catastrophe fund size measured by CVaR is higher than that measured by VaR and CVaRis more sensitive to the losses of large scale earthquakes. It is recommended that we can use VaR and CVaR as the risk measurement indexes to classify risk when setting the catastrophe fund size.