基于KMV模型的国内信用违约互换定价及实证研究

发布时间：2018-06-27 02:57

  本文选题：信用违约互换 + KMV模型　； 参考：《南京大学》2017年硕士论文

【摘要】：近年来,随着不良贷款的增多和债市违约风险的逐渐暴露,国内金融机构面临着愈加剧烈的信用风险挑战。为有效缓释市场信用风险,丰富信用风险管理工具,中国交易商协会于2016年9月正式推出了信用违约互换业务。信用违约互换是如今全球市场中交易最广泛的信用衍生产品,能够有效分离和转移信用风险。在业务发展初期,建立适用于国内市场的信用违约互换定价模型至关重要。当前学术界关于信用违约互换定价的主流模型可分为结构化模型和简约化模型。结构化模型认为公司自身资本结构和财务状况的恶化是导致违约事件发生的根本原因,并将公司股权看作一个标的为公司资产,行权价格为公司负债水平的欧式看涨期权,进而可以通过B-S期权定价模型估算出公司的预期违约概率;而简约化模型则假设违约事件发生的概率为外生,其违约强度可通过外部信息获得。本文首先结合中国债市违约事件发生的本质原因,对结构化模型和简约化模型的理论原理及所需数据的可获得性进行对比,得出结论认为运用结构化对信用违约互换产品进行定价更适用于国内市场环境。同时考虑模型的可操作性,选择KMV模型作为本文的理论定价模型。随后根据国内金融市场特征以及信用风险缓释工具的付息规则等对KMV模型部分参数的估计方法以及信用违约互换价差的计算公式进行修正。在实证部分,由于国内信用违约互换产品交易活跃度低导致相应价格数据获取的难度偏大,本文选取信用风险缓释凭证作为替代研究对象,首先探究了影响其价格变动的主要因素,并将一段连续时间内通过修正后定价模型估算出的理论价格与CBIC公布的每日报价数据趋势进行对比。此外还选择了 3只曾发生实质违约的债券,研究违约事件发生前一段时间内债券发行人预期违约概率的变化趋势。得出结论认为:(1)修正后的信用违约互换定价模型在对于剩余期限较长的CRMW产品定价方面的适用性较好,但其对临近到日的CRMW价格估算可能存在偏差;(2)对于剩余期限较短的CRMW而言,剩余期限是影响其价格变动的重要因素;(3)修正后的KMV模型能够较好地识别债券违约事件的发生,但对债券评级下调事件的识别能力一般。
[Abstract]:In recent years, with the increase of non-performing loans and the exposure of default risk, domestic financial institutions are facing more and more severe credit risk challenges. In order to effectively mitigate the market credit risk and enrich the credit risk management tools, the China Dealers Association officially launched the credit default swap business in September 2016. Credit default swaps (CDS) are the most widely traded credit derivatives in the global market, which can effectively separate and transfer credit risk. In the early stage of business development, it is very important to establish a credit default swap pricing model suitable for domestic market. The current mainstream model of credit default swap pricing can be divided into structured model and simplified model. The structured model holds that the deterioration of the company's capital structure and financial situation is the fundamental cause of the default, and regards the company's equity as a European call option whose target is the company's assets and the exercise price is the level of the company's liabilities. Furthermore, the expected default probability can be estimated by B-S option pricing model, while the reduced model assumes that the probability of default event is exogenous, and the default intensity can be obtained by external information. Based on the essential reasons of default events in China's bond market, this paper compares the theory principle of structured model and simplified model and the availability of required data. The conclusion is that structured pricing of credit default swaps is more suitable for domestic market environment. At the same time, considering the maneuverability of the model, the KMV model is chosen as the theoretical pricing model. Then, according to the characteristics of domestic financial market and the rules of interest payment of credit risk mitigation tools, the estimation method of KMV model parameters and the formula for calculating the spread of credit default swaps are revised. In the empirical part, due to the low transaction activity of domestic credit default swap products, it is difficult to obtain the corresponding price data. Firstly, the main factors affecting the price change of CBIC are explored, and the theoretical price estimated by the modified pricing model in a continuous period is compared with the trend of daily quotation data published by CBIC. In addition, three bonds with material default were selected to study the trend of expected default probability of bond issuers in the period before default. The conclusions are as follows: (1) the modified credit default swap pricing model is more suitable for CRMW products with longer remaining term. However, there may be some deviation in the price estimation of CRMW near to Japan; (2) for CRMW with short residual period, the residual period is an important factor affecting the price change of CRMW; (3) the modified KMV model can better identify the occurrence of bond default events. But the ability to identify bond downgrades is mediocre.
【学位授予单位】：南京大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F224;F832.51
，

本文编号：2072405