Empirical Credit Risk Ratings of Individual Corporate Bonds and Derivation of Term Structures of Default Probabilities
Empirical Credit Risk Ratings of Individual Corporate Bonds and Derivation of Term Structures of Default Probabilities
复制标题
DOI：
10.3390/jrfm12030124
复制
发表时间：
2019-07
影响因子：
--
通讯作者：
T. Kariya;Y. Yamamura;Koji Inui
中科院分区：
文献类型：
--
作者：
T. Kariya;Y. Yamamura;Koji Inui
毫无疑问，在金融业、企业甚至政府的决策中，拥有一种经验上有效的信用风险评级方法是很重要的。在我们的方法中，对于每个公司债券（CB）及其发行人，我们首先提出了一个信用风险评级（Crisk-rating）系统，该系统具有卡里耶等人（2015）提出的标准化信用风险价格价差（S-CRiPS）度量的评级区间，其中信用信息基于CRiPS度量，该度量是CB价格与其政府债券（GB）等价CB价格之间的差异。其次，对于通过风险评级系统获得的每个风险同质类，通过卡里耶（2013）提出的可转债定价模型推导出违约概率的期限结构（TSDP），该模型将每个类的Crisk水平转换为违约概率，显示未来时间范围内的违约可能性，其中分析了截至2010年8月的1545个日本可转债价格。为了实现这一目标，需要建立具有高实证绩效的政府债券定价横截面模型，以获得高精度的CRiPS和S-CRiPS测度。我们的GB模型和S-CRiPS措施的有效性已经在我们的论文中用日本和美国的GB价格以及卡里耶等人（2016 a，B）对欧盟5个国家的GB和美国能源公司发行的CB的信用风险进行了评估。我们的评级间隔的风险评级系统进行了测试与分布的1545 CB的评级，一个特定的机构的信用评级，并通过三阶段聚类分析获得的群体的评级。
Undoubtedly, it is important to have an empirically effective credit risk rating method for decision-making in the financial industry, business, and even government. In our approach, for each corporate bond (CB) and its issuer, we first propose a credit risk rating (Crisk-rating) system with rating intervals for the standardized credit risk price spread (S-CRiPS) measure presented by Kariya et al. (2015), where credit information is based on the CRiPS measure, which is the difference between the CB price and its government bond (GB)-equivalent CB price. Second, for each Crisk-homogeneous class obtained through the Crisk-rating system, a term structure of default probability (TSDP) is derived via the CB-pricing model proposed in Kariya (2013), which transforms the Crisk level of each class into a default probability, showing the default likelihood over a future time horizon, in which 1545 Japanese CB prices, as of August 2010, are analyzed. To carry it out, the cross-sectional model of pricing government bonds with high empirical performance is required to get high-precision CRiPS and S-CRiPS measures. The effectiveness of our GB model and the S-CRiPS measure have been demonstrated with Japanese and United States GB prices in our papers and with an evaluation of the credit risk of the GBs of five countries in the EU and CBs issued by US energy firms in Kariya et al. (2016a, b). Our Crisk-rating system with rating intervals is tested with the distribution of the ratings of the 1545 CBs, a specific agency’s credit rating, and the ratings of groups obtained via a three-stage cluster analysis.