摘要
通过考察2个非独立担保债务凭证进行再证券化生成的复杂资产组合,研究了该资产组合的违约相关性测度及其与资产组合总体风险的关系,组合间的相关性定义为总风险因素的Kendall秩相关系数。通过构建Copula函数模型,并运用蒙特卡罗模拟技术对违约相关性与总体风险的关系进行仿真模拟。实验结果表明:当总风险的相关性增大时,投资组合内证券的预期违约比率会降低;同时,该实验也为担保债务凭证(CDO)发行人如何确定股权级比例的下边界提供了参考。只有在该下边界以上,CDO发行人才有可能盈利,并且投资者获得正收益。
This paper deals with the default correlation between two simple portfolios （CDO） in more complex portfolios. We structure the Gaussian Copula Model and define the correlation between two CDO portfolios as the Kendall＇s Tau between their total risk factors. In addition, the interesting results which regard the influences of changes in the correlation between CDO portfolio performances are investigated with the Monte Carlo Simulation. We conclude that the expected value of the default ratio of the securities in the two CDO portfolios considered jointly decreases with the increase of the correlation between the total risk factors. With this conclusion, we can develop equity trenches. an approach for making the decisions on the quantity of
出处
《系统管理学报》
CSSCI
CSCD
北大核心
2017年第4期713-721,共9页
Journal of Systems & Management
基金
国家社会科学基金资助项目(16CGL015)
