【作者】 韩铁；

【导师】 詹原瑞；

【作者基本信息】 天津大学 ， 管理科学与工程， 2009， 博士

【摘要】 在国际信用衍生品市场快速发展和我国银行业迫切的现实需求背景下,发展信用衍生品并设计合理的信用衍生品价格成为当前理论和实践研究的热点。虽然美国次贷危机暂缓了国际信用衍生品市场的发展,但究其深层次原因与信用衍生品本身无关,而是美国金融监管当局忽视或有意忽视了次贷产品的潜在风险造成的,所以在加强信息披露和金融监管措施后,信用衍生品仍将是未来我们银行业解决信贷风险集中和不良贷款比例高等问题的有效工具。信用衍生品中最重要的是单资产信用违约互换,信用违约互换组合是单资产信用违约互换的发展,是将不同信用等级的基础资产打包,再以一篮子基础资产为标的签订互换合约,其优势在于降低合约成本,并促进信用风险顺利转移。本文对信用违约互换组合定价方法进行了深入探讨,主要研究成果概括为:第一,本文针对违约强度具有水平期限结构的情况,开发了一种基于具有动态特征的CreditRisk+模型计算信用违约互换组合公平利差的方法,给出了计算公平利差的解析表达式,并结合算例,研究了具有动态特征的CreditRisk+模型参数估计方法和求解公平利差的计算步骤,算例结果显示该方法的有效性。第二,本文针对违约强度为扩散过程的情况,开发了基于偏微分方程计算信用违约互换组合公平利差的方法,同样给出了公平利差的解析表达式,结合算例给出了求解公平利差的计算步骤,通过算例参数设置验证了偏微分方程方法与CreditRisk+模型方法求解信用违约互换组合公平利差具有相近的结果。第三,本文利用Copula函数构建违约时间的相关关系,将Copula函数族引入到信用违约互换组合定价研究中。在模拟违约时间的过程中:本文给出了一种基于核密度函数估计边缘分布,极大似然法估计Copula参数的两阶段参数估计方法;并且给出多元Copula函数随机数的模拟技术。具体算例结果显示相对于正态Copula函数和t-Copula函数而言,Clayton Copula函数模拟的第一次违约概率更高,意味着基于Clayton Copula函数定价得到一个保守的定价价格。本论文是国家自然科学基金资助项目《一致性风险量度在信用风险的度量与管理中的应用》(No.70573076)和高等学校博士学科点专项科研基金资助项目《一致性风险量度的理论与应用研究》(No. 20050056057)的组成部分。更多还原

【Abstract】 Under the background of the rapid development in the international credit derivatives market and the urgently practical demand with Chinese banks, it becomes a hot spot of current theoretical and empirical studies to exploit credit derivatives and design a reasonable deposit credit derivatives price. Although the market capacity of the international credit derivatives market is less than before because of the current Subprime Lending Crisis in US, after deep analysis it is found that there is no relationship of credit derivatives and reduction of market capacity. The real reason is the overlooking of American Financial Supervision Authority in potential risk of Subprime Lending. As long as information disclosure and financial supervise are strengthened, the credit derivatives will be the effective tool for the problem of credit concentration risk and non-performing loans for Chinese banks.Single Credit Default Swaps (CDS) is the most important form of credit derivatives. Basket Credit Default Swaps (BCDS) is the extension of CDS. It packs several reference assets from different credit rating, and concludes the contrat with the object of reference portfolio. The advantages of the BCDS are cost reduction of concluding contrat and smooth transfer of the credit risk. This dissertation systemically discusses the pricing problem of BCDS. The core and achievements of this dissertation can be generalized as follows:Firstly, under situation of the constant default intensity, this dissertation develops a method of calculating the fair credit spread of BCDS based on CreditRisk+ model with dynamic characteristic, and the analytical expression for the fair credit spread of BCDS is also given out. In this dissertation though a numerical example, the calculation method for parameters of CreditRisk+ model with dynamic characteristic is studied and the calculation procedures of fair credit spread is also established. The result of the numerical example shows the effectiveness of the new model.Secondly, under the situation that the default intensities are diffusion process, this dissertation exploit a method of calculating the fair credit spread of BCDS based on partial differential equation (PDE), and the analytical expression for the fair credit spread of BCDS is also given out. In this dissertation though a numerical example, the calculation procedures of fair credit spread is established. The result of the numerical example shows equivalence of CreditRisk+ model and PDE method, as long as the parameters of two models are equivalence.Finanly, this dissertation establishes the correlation of default time based on the Copula function, and though this processes the Copula function is used in the pricing study of BCDS. In this dissertation the multivariate Copula is used to analyze the asymmetric dependence structure among financial asset returns, whose marginal processes are captured by nonparametric kernel density estimation, and parameters of Copula function are estimated by maximum likelihood estimation. This dissertation also introduces the simulation technology of the multivariate Copula. The result of the numerical example shows that compared with normal Copula and t-Copula, Clayton Copula gives the bigger probability of the first default time of BCDS, so it means a conservative price is given out based on the Clayton Copula.The research is sponsored by the National Natural Science Foundation of China (Grant No. 70576076) and Research Foundation of the Doctoral Program of Higher Education (Grant No. 20050056057)更多还原