债券组合信用风险模型及应用研究

发布时间：2018-04-29 17:17

本文选题：组合信用风险 + 结构模型　；参考：《山东科技大学》2017年硕士论文

【摘要】：现代经济是一种信用经济,交易双方都必须依靠信用进行合作。由于金融危机的波及范围越来越大,以及市场上普遍存在信息不对称等种种情况,对信用风险管理的要求越来越高。如何建立与实际情况更相符的组合信用风险模型,已经成为人们关注和研究的课题之一。本文针对莫顿模型中的不足,主要工作如下:首先,建立一个离散的首次通过时间的信用风险模型。在结构模型的基础之上,建立一个新的违约障碍。以公司的年度或半年最大负收益率作为状态变量,利用极值理论得到各状态变量的分布函数。鉴于有些公司属于不同行业,选用分层Gumbel关联函数来构造状态变量的联合分布函数。该模型是离散的首次通过时间的信用风险模型,改进了莫顿模型的缺点。通过实证研究表明,与不分层的Gumbel Copula模型相比较,测得的信用违约损失尾部要更厚一点,相对而言,效果更为保守。其次,建立一个动态的债券组合信用风险模型。已知公司在t时刻及其之前信息的基础上,来度量公司在未来某一时刻的信用风险情况。利用KPMG风险中性模型,在一定假设条件之下得到债券i在t时刻的违约概率与违约损失比例之间的关系。以债券i在t时刻的对数收益率为状态变量,选用GARCH模型来描述状态变量的边缘分布;利用Pair-Copula-GARCH模型来构造状态变量的联合分布函数。较其他模型而言,此模型具有动态性和前瞻性。最后,对本文进行了总结,指出了不足之处,并对下一步的研究方向进行了展望。
[Abstract]:Modern economy is a kind of credit economy, both sides must rely on credit to cooperate. Because the financial crisis is spreading more and more widely, and the information asymmetry exists in the market, the requirement of credit risk management is higher and higher. How to establish a combination credit risk model which is more consistent with the actual situation has become one of the topics that people pay attention to and study. The main work of this paper is as follows: firstly, a discrete credit risk model of first-pass time is established. Based on the structural model, a new default obstacle is established. Taking the annual or semi-annual maximum negative return rate of the company as the state variable, the distribution function of each state variable is obtained by using the extreme value theory. In view of the fact that some companies belong to different industries, hierarchical Gumbel correlation function is used to construct the joint distribution function of state variables. This model is a discrete credit risk model with first passing time, which improves the shortcomings of the Morton model. The empirical study shows that compared with the unstratified Gumbel Copula model, the measured credit default loss has a thicker tail and a more conservative effect. Secondly, establish a dynamic bond portfolio credit risk model. The company is known to measure the credit risk of the company at a certain time in the future based on the information of t moment and its prior information. Using KPMG risk neutral model, the relationship between default probability and default loss ratio of bond I at t moment is obtained under certain assumptions. Taking the logarithmic yield of bond I at t as the state variable, the GARCH model is used to describe the edge distribution of the state variable, and the Pair-Copula-GARCH model is used to construct the joint distribution function of the state variable. Compared with other models, this model is dynamic and forward-looking. Finally, this paper summarizes, points out the inadequacies, and looks forward to the next research direction.
【学位授予单位】：山东科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：O211.67;F832.51

【参考文献】