债务抵押债券（CDO）定价模型及其仿真研究

发布时间：2018-01-10 08:22

本文关键词：债务抵押债券（CDO）定价模型及其仿真研究　出处：《中南大学》2012年博士论文　论文类型：学位论文

【摘要】：债务抵押债券(Collateralized Debt Obligation, CDO)是近年来国际金融市场上资产证券化领域重要的创新产品,颇受市场关注。作为近十年来增长速度最快的金融产品之一,债务抵押债券的规模扩张对国际金融市场形成了很大影响力。债务抵押债券是以抵押债务信用为基础,选择债券贷款等金融资产组建资产池,并重新分割投资回报和风险,通过资产证券化技术设计出既可满足不同投资者需求又能改善银行资产风险收益状况的创新性衍生证券成品。债务抵押债券产品标的资产不仅可以是银行贷款、债券,还包括ABS、MBS等,随着债务抵押债券产品的进一步发展,目前,还出现了将资产证券化技术与债务抵押债券产品相结合的新型信用衍生工具CDO。 2008年美国爆发次贷危机,债务抵押债券的风险效应被放大,债务抵押债券的金融风险防范问题引发深层思考。而我国的债务抵押债券研究与实践尚处于探索阶段,债务抵押债券的分级机制和多变的种类在我国的金融改革进程中亦备具广阔的应用前景。我国各大商业银行大都在积极设计债务抵押债券金融产品,试图对数目庞大的银行不良贷款进行有效处理。本文根据债务抵押债券产品的金融特性,对其定价模型进行系统性分析、构建与模拟,且对其信用风险进行识别与防范研究；同时,利用债务抵押债券对金融衍生工具及避险标的资产价格、风险特性以及两者的波动性匹配情况进行模拟仿真分析,以寻求合适的衍生工具套头率及其相关的避险标的资产；另外,参照已有的历史数据和价格行为资料,在对债务抵押债券进行定价模拟仿真的同时,实证考察债务抵押债券的历史经验数据,以测度债务抵押债券定价模型的可行性与可靠性。本文的主要研究路线与结论如下：本研究首先通过对债务抵押债券产品市场的发展现状、市场结构、市场功效等进行较全面的分析,以把握债务抵押债券的市场特性与发展趋势,为债务抵押债券产品的定价奠定理论和实践基础。然后,本文围绕债务抵押债券产品的定价模型和定价机理展开深入研究。从模型假设条件的提出、模型求解的推导以及仿真技术边界条件和初始状态值的确定等方面入手,引入非对称性GARCH效应来构建债务抵押债券定价模型。主要工作有：(1)估计我国利率水平上资产市场的正向非对称效应参数；(2)确定模型初始状态值(风险识别和期限结构分析),运用DTSM模型产生模拟数据序列而构建的算法模型；(3)通过初始状态值和数据生成算法,模拟利率上限的的跨期组合,从而得到CDO的基础资产---利率变化的波动范围；(4)对shibor产品价格序列的统计特性特别是波动特性和期限结构进行分析,并借助前述分析的QTMS模型对我国债务抵押债券产品市场定价的边界条件和初始条件进行统计描述；(5)利用Eviews多元GARCH预测技术对shibor的未来价格行为进行预测,从而确保我国债务抵押债券产品市场的模拟初始输入值的有效性；(6)利用我国利率市场数据,对CDO在我国市场的数据进行分析：通过对基础资产(shibor)价格行为分析,利用Gauss-Coupla模型获得我国债务抵押债券产品市场定价的边界条件和初始状态输入值,再运用正向非对称效应定价模型以及蒙特卡洛模拟技术对我国CDO产品的价格过程进行模拟仿真,测度模拟仿真方法的适用性和可行性。最后,论文以房地产抵押贷款的合成CDO金融产品为例,分析了债务抵押债券产品的风险特征,并在此基础上,运用正向非对称效应定价核函数和蒙特卡洛模型对该金融产品价格进行仿真。研究结果表明：债务抵押债券股权和夹层部分是杠杆作用的底层,一个夹层部分的风险和杠杆作用取决于其信用的强化程度,而股权部分的风险转移是有限的；抵押债券和其他创新信贷产品是CDO的关联交易；投资者如果能正确利用违约相关性,就可以创造交易机会,进行相关的风险管理,并据此实施商业周期的衡量。本研究主要采用了比较研究、演绎建模、模拟仿真及实证统计推断等研究方法,以模拟仿真技术,对债务抵押债券产品在现有定价理论基础上进行定价模拟。并结合我国债务抵押债券产品市场的历史文化传统,在债务抵押债券产品定价的模拟分布函数中首次引入行为随机折现因子变量,对模拟定价核函数的模拟效果进行评价,以期为债务抵押债券产品在我国证券市场的应用提供理论和实践上的指导。
[Abstract]:Debt - backed securities ( CDOs ) is an important innovative product in the field of asset securitization in the international financial market in recent years . As one of the fastest - growing financial products in recent years , the expansion of debt - backed securities has exerted great influence on international financial markets . The assets of debt - backed securities can not only be bank loans , bonds , but also ABS , MBS , etc . This paper makes a systematic analysis , construction and simulation of the pricing model of the debt - backed securities based on the financial characteristics of the debt - backed securities . In addition , based on the existing historical data and price - behavior data , the paper makes an empirical study on the historical empirical data of the debt - backed securities to measure the feasibility and reliability of the debt - backed securities pricing model . The main research routes and conclusions are as follows : First of all , through the analysis of the present situation , market structure and market efficiency of the debt - backed securities market , the market characteristics and development trend of the debt - backed securities are grasped to lay a theoretical and practical basis for the pricing of debt - backed securities products . This paper studies the pricing model and the pricing mechanism of the debt mortgage bond products . The paper introduces the pricing model of the debt mortgage bond products by introducing the model assumption conditions , the derivation of the model solution and the determination of the boundary condition and initial state value of the simulation technology . Finally , taking the CDO financial product as an example , the paper analyzes the risk characteristics of the debt - backed securities product , and simulates the price of the financial product by using the forward asymmetric effect pricing kernel function and the Monte Carlo model . The research results show that the risk and leverage of the equity part depend on the degree of credit enhancement , and the risk transfer of the equity part is limited ; the risk and leverage of the equity part is limited ; and if the investor can use the default correlation correctly , the transaction opportunity can be created , the relevant risk management is carried out , and the measurement of the business cycle can be carried out accordingly . This paper mainly adopts the research methods such as comparative research , deduction modeling , simulation simulation and empirical statistical inference . Based on the traditional pricing theory , the paper makes a pricing simulation on the existing pricing theory based on the simulation technology and the historical and cultural tradition of our debt mortgage bond product market . The simulation effect of the simulated pricing kernel function is evaluated in order to provide theoretical and practical guidance for the application of the debt mortgage bond product in our security market .

【学位授予单位】：中南大学
【学位级别】：博士
【学位授予年份】：2012
【分类号】：F224;F830.91

【参考文献】