基于困境概率的我国公司债信用利差研究

发布时间：2018-03-18 22:39

本文选题：信用利差　切入点：困境概率　出处：《复旦大学》2014年硕士论文　论文类型：学位论文

【摘要】：我国的信用债市场近几年发展迅速,上市公司发行的公司债逐渐增多,大量中低等级债券陆续出现,债券的信用违约风险开始显现,信用利差分析逐渐获得越来越多的关注。本文从对信用利差定价经典的Merton (1974)公司债务模型出发,根据Leland (1994)对公司资本结构的扩展和KMV公司的信用风险实务操作模型以及Atkeson, Eisfeldt and Weill (2013)最新的研究成果,对公司的财务困境概率进行简单而有效的计算,以此作为公司债信用违约风险的代理变量。在充分借鉴国内外学者对信用利差研究的基础上,考虑到我国的公司债市场发展现状,本文基于信用风险分解理论,从信用风险、流动性风险、宏观经济风险等角度出发,仔细分析和选取了适合我国信用债市场情况的影响信用利差因素的解释变量,用来对信用利差进行实证研究分析。本文搜集整理了我国公司债2011年9月初到2013年12月底之间存续的所有上市公司的公司债数据,剔除异常样本后共得到17990组横跨113个横截面的周度数据样本,进行混合横截面OLS回归分析。本文接着按照信用等级进行了分组回归分析,并对困境概率的解释力进行了讨论。通过对比分析,本文发现信用风险对低等级的债券影响最大,而流动性风险对中等级的债券影响最强,宏观经济风险中主要是货币政策相关因素对公司债的影响较为明显,并且2013年的信用利差有一个整体缩窄的过程,降幅随等级降低而增大,本文还研究发现中等级的债券拥有独有的特点,其受票面利率和国企因素的影响最大。对于上述统计分析得到的结论,本文对造成该情况的可能的原因进行了解释。通过对比分析本文发现使用股票的特质性波动率相较总波动率计算得到的困境概率对信用利差的影响更大,而且与直接使用股票波动率进行线性回归相比对不同等级的债券的解释力更一致,是较好的违约风险代理变量。最后,本文分析了研究的不足之处,并对进一步工作的方向进行了展望。
[Abstract]:In recent years, the credit bond market of our country has developed rapidly, the corporate bonds issued by listed companies have gradually increased, a large number of low and medium grade bonds have appeared one after another, and the risk of credit default of bonds has begun to appear. Credit spread analysis has gradually attracted more and more attention. This paper starts from the classic Merton debt model of credit spread pricing. According to the latest research results of Leland 1994, the expansion of the company's capital structure, the practical operating model of credit risk of KMV Company and the latest research results of Atkeson, Eisfeldt and Weill / 2013), the probability of financial distress of the company is calculated simply and effectively. Taking this as the proxy variable of corporate bond credit default risk, based on the domestic and foreign scholars' research on credit interest margin, and considering the present situation of our country's corporate bond market, this paper based on the theory of credit risk decomposition, from the credit risk, From the point of view of liquidity risk and macroeconomic risk, this paper carefully analyzes and selects the explanatory variables that are suitable for the credit bond market in China to influence the credit spreads. This paper collects and collates the data of corporate bonds of all listed companies that exist between early September 2011 and the end of December 2013 in our country. After eliminating abnormal samples, a total of 17990 sets of circumference data across 113 cross sections were obtained, and then mixed cross section OLS regression analysis was carried out. Through the comparative analysis, we find that the credit risk has the greatest influence on the low grade bond, while the liquidity risk has the strongest influence on the middle grade bond. The influence of monetary policy factors on corporate debt is obvious in macroeconomic risk, and the credit spread in 2013 has an overall narrowing process, the decline increases with the decrease of grade. This paper also finds that the mid-grade bonds have unique characteristics, which are most affected by the par interest rate and state-owned enterprise factors. In this paper, the possible reasons for this situation are explained. Through comparative analysis, it is found that the probability of distress obtained by using the idiosyncratic volatility of stocks is more important to the credit spread than that of the total volatility. And compared with the direct use of stock volatility for linear regression, the explanatory power of different grades of bonds is more consistent, it is a better proxy variable of default risk. Finally, this paper analyzes the shortcomings of the research. The direction of further work is prospected.
【学位授予单位】：复旦大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F832.51

【相似文献】