【作者】 刘湘云；

【导师】 杜金岷；

【作者基本信息】 暨南大学 ， 金融学， 2007， 博士

【摘要】 随着我国利率市场化进程的加快，商业银行将面临日益严重的利率风险：由于我国商业银行是金融体系的重要组成部分，是中央银行货币政策传导机制中的重要环节，且涉及千家万户；因此，研究商业银行利率风险计量与管理问题已成为金融界十分关注的重要课题之一。首先，本文提出商业银行利率风险的形成机理存在收入效应、市值效应和动态效应等三种途径，从而银行利率风险计量必须全面考虑这三种效应。其次，基于利率敏感性标准重新划分商业银行业务，并遵循动态、精确和理性原则对商业银行所有的利率敏感性业务面临的利率风险进行识别：研究表明，不同的银行利率敏感性业务所面临的利率风险形式有所区别。从操作上讲，商业银行利率风险可以分为单一利率风险和综合利率风险；相应地，商业银行利率风险计量也包括单一利率风险计量和综合利率风险计量两个层次。本文采取“先单一后综合”的利率风险计量路径，先运用各种调整久期分析研究一些特殊银行业务的利率风险计量，如有隐含期权的可赎回债券、含有违约风险的逾期贷款等。其三，从收益曲线非平移和利率期限结构等利率动态行为角度考察商业银行利率风险动态计量问题；本文提出，若要测度利率变化对银行未来净值和预期收益的影响，还必须考虑利率期限结构和收益率曲线非平移等利率动态行为，来构建相应的随机久期模型。其四，构建商业银行利率风险动态综合计量的基本框架，根据民生银行2005年年报进行实例计算和效率分析，实证研究表明：期权调整久期、违约调整久期、考虑收益率曲线非平移的随机久期和基于利率期限结构的随机久期都在一定程度上提高了利率风险计量精确度。最后，提出了商业银行利率风险动态随机久期免疫模型，结合中国民生银行实例分析得出，相对传统久期免疫策略来说，动态随机久期免疫策略的有效性大大提高了。研究中采用了宏观分析与微观分析相结合、规范分析与实证分析相结合、一般分析与特殊分析相结合、定性分析与定量分析相结合等方法，尤其是为了构建一个客观实用的商业银行利率风险动态综合计量模型，本文在定量分析中非常注重随机积分、随机微分、线性规划及数理统计工具和计算机先进技术的有机结合，并对模型进行实证检验。具体的定量分析方法包括：①建模方法主要采用动态随机模拟法；②参数估计采用最大似然估计法(MLE)和广义矩分析法(GMM)：③模型检验采用历史数据拟合法。更多还原

【Abstract】 With the speeding process of the interest rate marketization, commercial banks face increasingly serious interest rate risk. How to measure and manage it draws most attention in finance industry because commercial banks are an important part in our finance system and an important chain in transmission mechanisms of monetary policy for central bank.Firstly, this paper puts forward that there are three methods named as income effect, market value effect and dynamic effect, which can explain the forming mechanism of interest rate risk in commercial banks. Therefore, these effects must be considered roundly in interest rate risk measurement. Secondly, bank operations will be re-partitioned on the basis of interest rate sensitivity and all interest rate risk of sensitive operation will be identified on dynamic, precise and rational rule. Studies show that the form of interest rate risk of different interest rate sensitive bank operation is different. Interest rate risk in commercial banks can be divided into single interest rate risk and integrated interest rate risk on operation. Accordingly, interest rate risk measurements also include two layers as single interest rate risk measurement and integrated interest rate risk measurement. This paper measures interest rate risk in commercial banks in accordance to the path, which single risk measurement will be doing before integrated risk measurement. Thus, this thesis will first measure by all kinds of adjusted duration interest rate risk in some special bank operations such as Callable Bonds with embedded options, overdue loans with credit risk and so on. Thirdly, the thesis analyzes dynamic measurement interest rate risk in commercial banks with point of interest rate dynamic behavior such as yield curve non-parallel shifts, term structure of interest rate and so on, and suggest that, these interest rate dynamic behavior must be considered to design stochastic duration models based on it, if have to measure the effect of interest rate changes to net future value and expected return of bank. Fourthly, build up a framework for dynamic and integrated measurement on interest rate risk in commercial banks, and calculate bank integrated risk on example of China Minsheng Banking Ltd Corp. Then, analyze efficiency of dynamic and integrated modeling. Positive studies show that all kinds of adjusted duration(such as option-adjusted duration, breaking-promise-adjusted duration, stochastic duration considered yield curve non-parallel shifts, stochastic duration based on term structure of interest rate and so on) promote measuring precision in interest rate risk to a certain extent. Lastly, put forward dynamic stochastic duration immunization modeling interest rate risk in commercial banks. The thesis will analyze a example of China Minsheng Banking Ltd Corp. and show that compared to traditional duration immunization strategy, efficiency of dynamic stochastic duration immunization strategy is increased largely.In the paper, research ways adopted include macro analysis combined with micro analysis, theoretical analysis combined with empirical analysis and so on. Especially, stochastic process, differential equation, linear programming and computer techniques connecting with mathematical statistics are highly emphasized to create a practical dynamic and integrated interest rate risk measuring model in commercial banks. Concrete quantitative analysis methods include:①Dynamic stochastic simulating method in creating model;②Maximum likelihood Estimation(MLE) method and General Matrix Method(GMM) in parameter estimating;③Historical data simulation method in model verifying.更多还原