基于分箱策略的巨灾债券风险息差定价模型

Catastrophe Bonds Spread Pricing Model Based on Binning Strategy

在现有文献中,由于巨灾债券的数据有限,考虑的影响因素较少,关于巨灾债券定价方法的研究存在一定局限性。通过查找多个数据库增加研究样本,基于Logit风险度量构建新的定价因子,采用基于进化树的分箱策略构建广义线性模型,兼顾定价模型的预测能力和经济解释能力。分析结果表明,影响巨灾债券风险息差的主要因素是期望损失、风险附加、债券信用等级、债券发行时间、自然环境状况、再保险市场状况和金融市场状况。基于风箱策略构建的定价模型对巨灾债券定价和产品推广具有现实参考价值。

Due to the scarce data of catastrophe bonds,very few pricing factors have been considered in the existing literature.As a result,the research conclusions may have certain limitations.The data of catastrophe bonds are applied by searching multiple databases,and considers multiple pricing indicators are considered in the pricing model,so that the model can catch the main factors influencing catastrophe bond price.We consider more pricing factors,build new pricing factor based on Logit risk measurement,use evolutionary tree-based binning strategies to build a Generalized Linear Model,which improves flaw of the Generalized Additive Model.The results show that the key factors affecting the cat bond price is the expected loss,risk measure,credit rating,issuing time,environment situation,reinsurance market situation and financial market situation.The model established here can be used to price the catastrophe bonds issued in the primary market.