【作者】 乌画；

【导师】 邓超；

【作者基本信息】 中南大学 ， 管理科学与工程， 2012， 博士

【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。更多还原

【Abstract】 CDO,as a credit derivative,is the most important innovation technology in the asset securitization field recently,which is paid much attention by market. The basic idea of CDO is that through asset securitization technology, bonds, loans and other financial assets to set up the pool of assets on the basis of mortgage debt credit, re-partition the return on investment and risk, and thus designed to meet the needs of different investors but also to improve the innovation of the risk of bank assets, earnings the nature of derivative securities finished.It is perfectly combined the quality of asset securitization and credit derivatives,whose underlying asset is not only to bank loans, bonds, and to ABS, MBS.With the development of credit derivative products, the new credit derivatives of CDO are appeared.The U.S. subprime mortgage crisis erupted in2008, maked the effect of the risk of collateralized debt obligations were enlarged,and the financial risks of collateralized debt obligations to avoid recurrence of tight bell sounded. China’s collateralized debt obligations and practices are still in the preliminary trial basis with the exploratory stage. So the classification mechanism and changing the type of collateralized debt obligations has very broad application prospects in China’s financial reform process. At present, China’s major commercial banks were actively designing collateralised financial products, in order to effectively deal with a large number of bad loans.This paper will be based on the financial characteristics of collateralized debt obligations, systematic detection of financial pricing models, build and simulate their credit risk identification, pricing and Prevention. In the research process, we will make use of collateralized debt obligations on financial derivatives and hedging the underlying asset prices, risk characteristics, as well as between the volatility of the match simulation analysis in order to seek appropriate derivatives hedging rate and its associated hedging the underlying asset; in addition, the reference to the existing historical data and price behavior,collateralized debt obligations pricing simulation,empirical investigation of collateralized debt obligations of historical and empirical data to measure the pricing model of collateralized debt obligations feasibility and reliability. Specific studies are as follows:First,through the discusses of collateralized debt market development, market structure, market efficiency, this article can grasp the characteristics and development trend of the market of collateralized debt obligations for the pricing of CDO products lay the foundation of theory and practice. Second, this paper is based on credit derivatives pricing model and pricing mechanism to axplore some points. From the assumptions of the model, the simulation techniques boundary conditions and the initial state to determine the value,we will introduce GARCH effect to construct the pricing model. Steps as follows:(1) estimate asymmetric effect parameters of our country interest rate level;(2) use DTSM model to produce simulation data sequence and got the simulation rates to calculate the range of actual assets-interest rates fluctuation;(3) through the initial condition value and data generation algorithm, simulation of the upper limit of the interest rates cross time combination, and get the CDO based assets-interest rates fluctuate;(4) Shibor price series products of the statistical features of especially wave characteristics and maturity structure is analyzed, and the analysis of the model with the QTMS of China’s collateralised debt obligations product market pricing boundary conditions and initial conditions for a statistical description;(5) use multiple GARCH prediction technology in Eviews for Shibor’s future price behavior prediction, so as to ensure that our country credit derivatives market simulation initial input value validity;(6) use our country interest rate market data to analyse the CDO data in our country market. Based on the analysis of underlying asset price behavior,the paper was used DTSM model to obtain the boundary conditions and initial state input value of our credit derivatives market, then it was used the pricing kernel model and Monte Carlo simulation technology on China CDO product price process simulation to measure the applicability and feasibility of simulation method.Through the analysis of foundation assets (shibor) price behavior, this paper uses Gauss-Coupla model for our country collateralised debt obligations product market pricing boundary conditions and the initial state input value, then uses positive asymmetric effects pricing model and monte carlo simulation technology to our country CDO the price of the product process simulation, measure simulation method the applicability and feasibility. Finally, synthetic CDO in paper to real estate mortgage financial products, for example, analysis of the debt risk characteristics of mortgage products, and on this basis, using forward pricing of asymmetric effects of kernels for simulation and Monte Carlo models on the price of financial products. Research results indicates that:debt mortgage bond equity and sandwich part is lever role of underlying, a sandwich part of risk and lever role depends on its credit of strengthening degree, and equity part of risk transfer is limited of; mortgage bond and other innovation credit products is CDO of associated trading; investors if can correctly using default correlation, on can created trade opportunities, for related of risk management, and accordingly implementation commercial cycle of measure.This article mainly took the methods of comparative research, interpretation of modeling, simulation and empirical statistical inference. Morever, simulation technology was used to creat pricing theory by creatly using Monte Carlo simulation method.At the same time,this paper was combined with historical tradition of the credit derivatives market in China,considering differences between Chinese and western cultural background. In addition it was first time introduced stochastic discount factor behavior to simulate pricing kernel function simulation effect evaluation.更多还原