商业银行流动性风险衡量及相关问题研究
|
摘要
人们对于银行流动性风险的衡量方法基本上反映了在一定的历史背景下,银行对流动性管理的认识过程。同时,对流动性风险进行衡量也是银行有效管理流动性风险的前提和基础。从目前看来,金融体系的迅速发展和金融全球化已成为推动全球金融风险管理发展的现实原因和直接动力。各国银行界都在依靠产品和服务的创新与不断深化混业经营模式来适应新的变化。由此,巴塞尔资本协议(Ⅱ)为银行和金融监管当局提供了更多的衡量资本充足的计量方法,其基本目的是如何使监管原则更为灵敏的反映银行经营环境的变化,使银行的风险监控始终能适应金融市场的风险变动。
遵循巴塞尔资本协议(Ⅱ)框架中的基本思想,在本论文中,我们对衡量商业银行流动性风险的风险测度进行了研究。在借鉴国内外已有研究成果的基础上结合我国银行体系风险管理和监管的实际,特别是针对当前我国银行业“流动性过剩”的风险计量和监管问题,提出一些新的思路用以开发具有高度风险敏感性的流动性风险管理方法,用以验证我国银行体系“流动性过剩”的真实性,同时用来解决银行流动性风险管理和监管中的实际问题。在当前银行界的实践中,VaR是度量缓冲各种风险所需资本金的主要风险测度。长期以来,VaR受到学术界的批评,批评者认为风险度量应使用以ES为代表的相容风险测度(还有SpectralRisk Measure、Concave Distortion Risk Measure等等)。然而,相容风险测度CRM虽然克服了VaR的局限,但其只能保证不违反零阶随机占优(即绝对占优)规则。在传统随机占优理论下,人们无法得到既满足传统随机占优单调性公理又满足CRM(Coherent Risk Measure)另外三条公理的风险测度。因此无法在符合绝大多数金融机构和监管机构(站在监管者和公众的立场上)共同价值判断的条件下实现风险测度的四大功能(1、风险排序和信息披露;2、风险预防和控制;3、风险监管;4、投资决策和业绩评价)。本论文充分利用我国安徽省境内主要金融机构的实际数据,探讨运用满足广义随机占优单调一致风险测度的高阶期望损失风险测度ES~((n))来解决银行流动性风险管理中的实际问题。ES~((n))不但能保证风险测度不违反广义随机占优规则,还可以保证满足CRM另外三条公理。本论文的主要内容构架和创新贡献概括如下:
1、文献综述工作,综述了银行流动性风险的特征及其衡量方法。首先对巴塞尔银行监管委员会在流动性计量与管理方面的相关研究成果做了一个详细的介绍。其次对国内外银行体系流动性风险管理和监管的实践进行了总结。最后对学术理论界的相关研究背景和现状进行了综述。
2、在(Franck,R.& Krausz,M.,2004)的基础上对银行流动性风险进行扩展的经济学分析,增加了三个方面的内容:在2.2节中,引入(Kyle,1985)中的市场流动性风险模型分析资本市场对银行流动性风险及银行最优资产配置的影响,并做了相关实证分析。在2.3节中,增加了银行间市场的相关模型分析。在2.4节中,增加了存款保险制度的相关模型分析。这样分别从资本市场、银行间市场、存款保险制度和最后贷款人制度几方面(透过经济学视角,这几方面与银行的流动性提供功能关系十分密切)入手,扩展并分析了这几方面与银行流动性风险及银行最优资产配置的关系。最后我们证明了:银行间市场和存款保险制度同资本市场和最后贷款人制度一样,都能够改善银行的流动性,提高银行的预期利润。
3、第三章和第四章是本论文的核心工作,主要的创新性贡献有两点:
第一、运用用于理解多元随机变量之间关系的统计工具Copula与广义随机占优理论中的ES~((n))测度对银行流动性风险进行阶段性的衡量。这是一种能够更加准确度量商业银行清偿能力和“流动性过剩”程度的方法,在此领域具有首创性和新颖性的。然后,我们对我国安徽省境内各主要商业银行1997-2006年间的流动性统计数据进行实证分析,以验证我国银行体系“流动性过剩”的真实性(我们验证了当前银行体系确实存在着流动性的“普遍过剩”,并且精确度量了银行体系流动性的“过剩程度”,对这一风险进行了科学的排序)。通过这种方法可以增强监管当局流动性风险监管的识别能力,提高金融机构自身风险控制的有效性。在模型中加入风险因子之间的Copula相依结构之后计算得到的风险值与直接计算得到的风险值在排序上有很大的差异。因此使用Copula方法对于银行流动性风险度量问题的研究具有十分重要的意义。通过寻求风险因子之间的Copula相依结构,不但能成功的反映我省当前金融的实际形势,而且改进了度量流动性风险资本金充足程度的方法,研究结果对金融监管和银行加强自身的流动性安全具有重要的意义。
同时运用ES~((n))测度进行实证研究发现:对于本文实证研究的问题,VaR、ES两种低阶期望损失风险测度在一阶以上广义随机占优的情况下将发生风险排序错误。而ES~((n))测度是对ES测度的高阶推广,它不但具有一阶ES测度的全部优良性质,而且是0～(n+1)阶广义随机占优单调一致风险测度,可以保证在高阶广义随机占优的情况下不发生风险判断错误,因此可以更好的实现风险测度的四大功能,而且ES~((n))测度的阶次越高,其提供的风险控制措施就越安全。因此,金融机构或监管机构使用ES~((n))测度来度量流动性风险可以增强风险监管的有效性,提高风险控制的可靠性,减少风险决策错误的可能性。
第二、运用QAR(Quantile Autoregression)分位数自回归方法对银行某一时点的流动性风险进行实时衡量和预测。QAR分位数自回归模型对于研究时间序列问题中的不对称波动和局部平稳十分有用;模型可以有效反映出时间序列数据中的相似单位根趋势甚至还能解释局部的一些突变行为;其主要特征是其回归系数随着更新过程中分位数的改变而相应改变,因此较常系数自回归模型来说更为合理。我们分析得出当前安徽省金融机构存在流动性过剩的风险。此方法与第三章运用的方法不同,能动态的、实时的衡量某一时点的流动性风险,并且能对下一时刻的风险进行预测。与VaR和ES测度相比较,此方法对于在某个时点上风险的排序更加符合实际情况。
通过目前我们对这两种方法的实证检验,显示其不但能准确地反映一定时期我省区域经济金融运行的基本态势,而且可以准确度量出我省区域内各家银行“流动性过剩”的程度及其自身纠偏能力的大小。据此监管当局可以识别出某些问题严重的银行,并在依赖这些银行自身的纠偏能力和对其实施外部干预之间做出选择。进而为中央银行宏观调控与货币政策实施以及金融监管当局把握整个银行体系的流动性风险提供了可操作性的标准,因此对提升我国银行体系的流动性风险管理水平和防范系统性金融风险有着重要的现实意义。因此,第三章与第四章共同构筑起了衡量商业银行流动性风险的一套新颖的、科学的测度体系。该体系不仅能衡量银行体系阶段性的流动性风险,也能动态的、实时的衡量银行体系某一时点的流动性风险,进而能够对下一时期的风险进行预测。实证检验该体系能够对区域性银行体系流动性风险进行科学有效的度量,故可以应用于对全国银行体系流动性风险的度量及监管实践。因此,对货币和监管当局来说均具有更高的实际应用价值。这部分内容是本论文的核心贡献所在。
4、鉴于流动性缺口管理在商业银行有效管理流动性风险上的重要性,我们运用一种统计分析方法——最长链统计量,对商业银行流动性缺口统计数据进行了持续性分析,得到安徽省内各家商业银行持续维持适度流动性缺口能力的持续性强度。得到的结论是:安徽省银行系统总体流动性状况十分良好,其P值达到0.94960。安徽省内主要商业银行运行稳健,维持适度流动性缺口能力的排序为:建设银行、中信银行、光大银行、农业银行、招商银行、工商＆交通银行、浦发银行、中国银行。由此,可以看出4家国有商业银行和5家股份制商业银行的排序结果呈犬牙交错的状态,不存在一方超越另一方的情况。因此,银行的所有制体系结构与银行流动性风险状况并不存在显著的相关关系。换句话说,四大国有商业银行的资金计划管理水平与股份制商业银行相比毫不逊色。
5、运用灰色系统和非线性协整分析的方法分析了影响安徽省商业银行流动性风险的各种宏观风险因素。主要是想找出对安徽省商业银行流动性风险影响程度最高的宏观风险因素,并得到安徽省商业银行流动性风险与安徽省宏观经济风险因素之间长期稳定的均衡关系。进而对安徽省商业银行规避宏观经济风险的能力进行检验。得到的结论是:安徽省金融机构流动性风险受到政府财政因素、外贸因素、消费因素和投资因素的重大影响。充分验证了安徽省经济“正走向”政府主导的投资增长型、外向型市场经济,与当前安徽省经济的实际情况比较吻合。根据银行的实践经验,财政存款涉及面广,资金沉淀量大,是各银行营销的重点。而外贸与投资离不开银行的信贷支持,对银行流动性需求有着重大的影响。而物价因素对银行流动性需求和供给两方面都有着重要的影响,原材料、能源、运输价格和人力资源成本的提高必然增强企业对银行信贷资金的需求。而消费价格上升必然导致人们可支配收入的相对减少,进而对银行流动性供给方面产生影响。
最后是全文的结束语,对本文的研究工作进行总结,并提出了需要进一步完善的地方。
Measures of liquidity risk of commercial bank basically reflects the knowing course of bank to liquidity management. And Liquidity measure is precondition & foundation of liquidity risk management of commercial bank. At present, Rapidly development of financial system and financial globalization trend already becomes the cause and drive to evolvement of global financial risk management. Eevery country's bank group depends on product & service innovation and mix-work to adapt to new change. Thereout, New Basel Accord offers more computation methods to banks and there supervisor. In order to make supervisory principle more flexible to reflect the change of bank working environment, and make bank risk control keeps all the time sensitivity to risk movement of financial market.
Follow the basic idea of New Basel Accord in this thesis, we use the known foreign &domestic research fruit for reference and we also think of risk management practice of bank system of China to mainly discuss bank liquidity risk's computation problem, especially "liquidity overplus" problem. Then we could bring forward some new idea to study risk computation methods with high risk sensitivity. So we can validate the authenticity of "the liquidity overplus" of bank system of China, and slove the actual problem on bank liquidity risk management and supervision. In bank current practice, Var is the main risk measure for computing capital to cover with all kinds of risk. But Var is criticized by academia, they recommend Coherent Risk Measure include ES、Spectral Risk Measure、Concave Distortion Risk Measure etc. Although CRM overcome shortcoming of Var, it just meet zero-rank stochastic domination rule. In traditional stochastic domination theory, people can not get the risk measure which meets both monotonicity axiom of traditional stochastic domination theory and other three axioms of CRM. So people can not come true risk measure's four functions under the condition of according with value judgement of most finance and supervision institutions which stand by public side. The four functions are: 1、risk compositor and information issuance, 2、risk prevention and control, 3、risk supervision, 4、investment decision and achievement evaluation. In this paper, we made a demonstration analysis about main commercial banks in AnHui province. We use the high-rank lossexpectation risk measure ES~((n)) of general stochastic domination theory to solve the factual problem of bank liquidity riskmanagement. ES~((n)) can meet both general stochastic domination ruleand other three axioms of CRM. The primary content and contribution in this thesis as follow:
The first chapter is literature summary. First of all, we make a detailed introduction of correlative research fruit of Basel committee about bank liquidity computation and management. And secondly, we make a all around summary and comment about the Measures of liquidity risk which a commercial bank generally adopting and it's up to date development. Finally, we summarize the research background and status quo of academia.
On the basis of (R. Franck & M. Krausz, 2004), the second chapter makes a patulous analysis of three economic aspects (Capital Market、Interbank Market、Deposit insurance and Lender of Last Resort of Central Bank system) which have colse relation to bank liquidity provision. In 2.2 section, we analyse the influence of capital market on bank liquidity risk and optimal assets collocation by introduce market liqiudity risk model of (Kyle, 1985), and we make some relevant demonstration analysis.. In 2.3 section, we add interbank market model.. In 2. 4 section, we add deposit insurance model. Finally, we testify: Interbank Market、Deposit insurance system, the same as Capital Market、Lender of Last Resort of Central Bank system, can improve bank liquidity status and increase bank expect profit.
The third chapter and the fourth chapter are the core part of this thesis. There are two primary contribution as follow:The first is in the third chapter: For the first time in thisfield, we use Stat. instrument Copula which used to understandingrelationships among multivariate outcomes & High-Level ES Measurein General Stochastic Dominance Theory to measure commercial bankliquidity risk. This is a method which can measure bank solvencyand "liquidity overplus" degree more accurately. We can use thismethod to validate the authenticity of "the liquidity overplus"of bank system of China. Then we made a demonstration researchduring 1997-2006 on liquidity data of commercial banks in Anhuiprovince of China. We find the liquidity of current bank systemof China is strictly superfluous and also eaccurately measure it's degree, then we get a scientific taxis. This method can increasethe discernment of supervision institution and validity of riskcontrol of finance institution. In our model, if use Copula or notmay greatly impact the result of taxis. So Copula have a importantsignificance to the research on bank liquidity computation problem.Through seeking Copula dependence structure among the risk factors,it can not only successfully reflect the actual finance situationof AnHui province, but also improve the method for computingcapital to cover with liquidity risk. The conclusion is very helpful to finance supervision and bank liquidity safety.
During demonstration research We also find: For the problem we study, VaR、ES these two low-rank loss expectation risk measures will be wrong under the condition of general stochasticdomination. Whereas ES~((n)) measure is the high-rank spread of ES measure, it has all the fine character of ES measure, and also is 0～(n + 1) -rank general stochastic domination monotone &consistent risk measure. It can be always right under the condition of general stochastic domination. So it can come true better risk measure's four functions, and the rank more higher, the risk control measure more safer. If finance institution or supervision institution adopt this method, they can increase the validity of risk supervision and enhance the reliability of risk control and decrease the possibility of decision error.
The second is in the fourth chapter : We use Quantile Autoregression method to measure and forecast bank liquidity risk on some time point. The leading character of QAR is it's coefficient changing along with Quantile's change. So compare to constant coefficient model, it is more reasonable. This method differs from the third chapter's, which can make a Real-time measure of liquidity risk on some time point and make a forecast on the next point. Compare with Var & Es measure, this method is more logical for risk sort order on some time point.
Through demonstration test on these two methods, we can see that they can not only reflect regional economy & finance movement well and truly, but also accurately compute the degree of "liquidity overplus" and "correcting capability" of banks of AnHui province. So supervision institution can find out some banks which problem are very serious, then supervisor will make a choose between keep silence and external interference. Thereout, it provides a operable standard for central bank actualizing monetary policy and for finance supervision institution holding liquidity risk of the whole bank system. So it has a significance for enhancing liquidity risk manament level and keeping away systemic finance risk of bank system of China. Therefore, the third and fourth chapter construct a new scientific system together, which can make a accurate measure of a commercial bank's liquidity risk. This system can not only measure bank liquidity risk in a long time, but also on some time point, even can forecast on the next time point. Demonstration analysis testifies to the validity of the regional applications, this system can be lightly applied to the whole country. So it is highly valuable to money & supervision institutions. This part of content contains the core contribution of this thesis.
Because of liquidity gap management is very important to commercial bank effectively manage it's liquidity risk. In the fifth chapter, we use the Longest Run Statistics to research on financial data of commercial banks in Anhui province of China. Then we get hold of durative intensity about capacity to maintain moderation liquidity gap of commercial banks in AnHui province. Our conclusion is: As a whole, the bank system liquidity status of AnHui province is all right. It's P value is 0.94960. So we can consider that the banks of AnHui province run steadily, and the taxis of their ability to keep appropriate liquidity gap is: China Construction Bank、China Citic Bank、China Everbright Bank、Agricultural Bank of China、China Merchants Bank、Industrial and Commercial Bank of China & Bank of Communications、ShangHai PuDong Development Bank、Bank of China. Therefore, we can see that the taxis of the four state-owned banks and the five joint-stock banks is jumbly, there is no circumstance which one side exceeds another side. So there no correlative relationship between proprietary system and bank liquidity risk status. In other words, the capital capital management level of the four state-owned banks is as good as state-owned bank's.
The sixth chapter is about to analyzes diversified macro-risk factors which have significant influence on liquidity risk of commercial bank of AnHui province. We will use Gray system method & Non-linear co-integration theory and it's methods to solve this problem. We hope to find the most significant macro-risk factor to it, and We also hope to find the long-range equilibrium relation between macro-risk factors and liquidity risk of commercial bank of AnHui province. Therefore, we can make a test on capacity of eluding macro-economy risk for commercial bank of AnHui province. Our conclusion is: Factors of government finance、foreign trade、consumption、investment have significant influence on liquidity risk of commercial bank of AnHui province. It adequately validates that economy of AnHui province becomes more and more extroverted and it's development largely depend on the investment by government. This do accord with the fact of AnHui province.
In the end, the tag is summary on research work of this thesis. And however it have many places need to perfect.
遵循巴塞尔资本协议(Ⅱ)框架中的基本思想,在本论文中,我们对衡量商业银行流动性风险的风险测度进行了研究。在借鉴国内外已有研究成果的基础上结合我国银行体系风险管理和监管的实际,特别是针对当前我国银行业“流动性过剩”的风险计量和监管问题,提出一些新的思路用以开发具有高度风险敏感性的流动性风险管理方法,用以验证我国银行体系“流动性过剩”的真实性,同时用来解决银行流动性风险管理和监管中的实际问题。在当前银行界的实践中,VaR是度量缓冲各种风险所需资本金的主要风险测度。长期以来,VaR受到学术界的批评,批评者认为风险度量应使用以ES为代表的相容风险测度(还有SpectralRisk Measure、Concave Distortion Risk Measure等等)。然而,相容风险测度CRM虽然克服了VaR的局限,但其只能保证不违反零阶随机占优(即绝对占优)规则。在传统随机占优理论下,人们无法得到既满足传统随机占优单调性公理又满足CRM(Coherent Risk Measure)另外三条公理的风险测度。因此无法在符合绝大多数金融机构和监管机构(站在监管者和公众的立场上)共同价值判断的条件下实现风险测度的四大功能(1、风险排序和信息披露;2、风险预防和控制;3、风险监管;4、投资决策和业绩评价)。本论文充分利用我国安徽省境内主要金融机构的实际数据,探讨运用满足广义随机占优单调一致风险测度的高阶期望损失风险测度ES~((n))来解决银行流动性风险管理中的实际问题。ES~((n))不但能保证风险测度不违反广义随机占优规则,还可以保证满足CRM另外三条公理。本论文的主要内容构架和创新贡献概括如下:
1、文献综述工作,综述了银行流动性风险的特征及其衡量方法。首先对巴塞尔银行监管委员会在流动性计量与管理方面的相关研究成果做了一个详细的介绍。其次对国内外银行体系流动性风险管理和监管的实践进行了总结。最后对学术理论界的相关研究背景和现状进行了综述。
2、在(Franck,R.& Krausz,M.,2004)的基础上对银行流动性风险进行扩展的经济学分析,增加了三个方面的内容:在2.2节中,引入(Kyle,1985)中的市场流动性风险模型分析资本市场对银行流动性风险及银行最优资产配置的影响,并做了相关实证分析。在2.3节中,增加了银行间市场的相关模型分析。在2.4节中,增加了存款保险制度的相关模型分析。这样分别从资本市场、银行间市场、存款保险制度和最后贷款人制度几方面(透过经济学视角,这几方面与银行的流动性提供功能关系十分密切)入手,扩展并分析了这几方面与银行流动性风险及银行最优资产配置的关系。最后我们证明了:银行间市场和存款保险制度同资本市场和最后贷款人制度一样,都能够改善银行的流动性,提高银行的预期利润。
3、第三章和第四章是本论文的核心工作,主要的创新性贡献有两点:
第一、运用用于理解多元随机变量之间关系的统计工具Copula与广义随机占优理论中的ES~((n))测度对银行流动性风险进行阶段性的衡量。这是一种能够更加准确度量商业银行清偿能力和“流动性过剩”程度的方法,在此领域具有首创性和新颖性的。然后,我们对我国安徽省境内各主要商业银行1997-2006年间的流动性统计数据进行实证分析,以验证我国银行体系“流动性过剩”的真实性(我们验证了当前银行体系确实存在着流动性的“普遍过剩”,并且精确度量了银行体系流动性的“过剩程度”,对这一风险进行了科学的排序)。通过这种方法可以增强监管当局流动性风险监管的识别能力,提高金融机构自身风险控制的有效性。在模型中加入风险因子之间的Copula相依结构之后计算得到的风险值与直接计算得到的风险值在排序上有很大的差异。因此使用Copula方法对于银行流动性风险度量问题的研究具有十分重要的意义。通过寻求风险因子之间的Copula相依结构,不但能成功的反映我省当前金融的实际形势,而且改进了度量流动性风险资本金充足程度的方法,研究结果对金融监管和银行加强自身的流动性安全具有重要的意义。
同时运用ES~((n))测度进行实证研究发现:对于本文实证研究的问题,VaR、ES两种低阶期望损失风险测度在一阶以上广义随机占优的情况下将发生风险排序错误。而ES~((n))测度是对ES测度的高阶推广,它不但具有一阶ES测度的全部优良性质,而且是0～(n+1)阶广义随机占优单调一致风险测度,可以保证在高阶广义随机占优的情况下不发生风险判断错误,因此可以更好的实现风险测度的四大功能,而且ES~((n))测度的阶次越高,其提供的风险控制措施就越安全。因此,金融机构或监管机构使用ES~((n))测度来度量流动性风险可以增强风险监管的有效性,提高风险控制的可靠性,减少风险决策错误的可能性。
第二、运用QAR(Quantile Autoregression)分位数自回归方法对银行某一时点的流动性风险进行实时衡量和预测。QAR分位数自回归模型对于研究时间序列问题中的不对称波动和局部平稳十分有用;模型可以有效反映出时间序列数据中的相似单位根趋势甚至还能解释局部的一些突变行为;其主要特征是其回归系数随着更新过程中分位数的改变而相应改变,因此较常系数自回归模型来说更为合理。我们分析得出当前安徽省金融机构存在流动性过剩的风险。此方法与第三章运用的方法不同,能动态的、实时的衡量某一时点的流动性风险,并且能对下一时刻的风险进行预测。与VaR和ES测度相比较,此方法对于在某个时点上风险的排序更加符合实际情况。
通过目前我们对这两种方法的实证检验,显示其不但能准确地反映一定时期我省区域经济金融运行的基本态势,而且可以准确度量出我省区域内各家银行“流动性过剩”的程度及其自身纠偏能力的大小。据此监管当局可以识别出某些问题严重的银行,并在依赖这些银行自身的纠偏能力和对其实施外部干预之间做出选择。进而为中央银行宏观调控与货币政策实施以及金融监管当局把握整个银行体系的流动性风险提供了可操作性的标准,因此对提升我国银行体系的流动性风险管理水平和防范系统性金融风险有着重要的现实意义。因此,第三章与第四章共同构筑起了衡量商业银行流动性风险的一套新颖的、科学的测度体系。该体系不仅能衡量银行体系阶段性的流动性风险,也能动态的、实时的衡量银行体系某一时点的流动性风险,进而能够对下一时期的风险进行预测。实证检验该体系能够对区域性银行体系流动性风险进行科学有效的度量,故可以应用于对全国银行体系流动性风险的度量及监管实践。因此,对货币和监管当局来说均具有更高的实际应用价值。这部分内容是本论文的核心贡献所在。
4、鉴于流动性缺口管理在商业银行有效管理流动性风险上的重要性,我们运用一种统计分析方法——最长链统计量,对商业银行流动性缺口统计数据进行了持续性分析,得到安徽省内各家商业银行持续维持适度流动性缺口能力的持续性强度。得到的结论是:安徽省银行系统总体流动性状况十分良好,其P值达到0.94960。安徽省内主要商业银行运行稳健,维持适度流动性缺口能力的排序为:建设银行、中信银行、光大银行、农业银行、招商银行、工商＆交通银行、浦发银行、中国银行。由此,可以看出4家国有商业银行和5家股份制商业银行的排序结果呈犬牙交错的状态,不存在一方超越另一方的情况。因此,银行的所有制体系结构与银行流动性风险状况并不存在显著的相关关系。换句话说,四大国有商业银行的资金计划管理水平与股份制商业银行相比毫不逊色。
5、运用灰色系统和非线性协整分析的方法分析了影响安徽省商业银行流动性风险的各种宏观风险因素。主要是想找出对安徽省商业银行流动性风险影响程度最高的宏观风险因素,并得到安徽省商业银行流动性风险与安徽省宏观经济风险因素之间长期稳定的均衡关系。进而对安徽省商业银行规避宏观经济风险的能力进行检验。得到的结论是:安徽省金融机构流动性风险受到政府财政因素、外贸因素、消费因素和投资因素的重大影响。充分验证了安徽省经济“正走向”政府主导的投资增长型、外向型市场经济,与当前安徽省经济的实际情况比较吻合。根据银行的实践经验,财政存款涉及面广,资金沉淀量大,是各银行营销的重点。而外贸与投资离不开银行的信贷支持,对银行流动性需求有着重大的影响。而物价因素对银行流动性需求和供给两方面都有着重要的影响,原材料、能源、运输价格和人力资源成本的提高必然增强企业对银行信贷资金的需求。而消费价格上升必然导致人们可支配收入的相对减少,进而对银行流动性供给方面产生影响。
最后是全文的结束语,对本文的研究工作进行总结,并提出了需要进一步完善的地方。
Measures of liquidity risk of commercial bank basically reflects the knowing course of bank to liquidity management. And Liquidity measure is precondition & foundation of liquidity risk management of commercial bank. At present, Rapidly development of financial system and financial globalization trend already becomes the cause and drive to evolvement of global financial risk management. Eevery country's bank group depends on product & service innovation and mix-work to adapt to new change. Thereout, New Basel Accord offers more computation methods to banks and there supervisor. In order to make supervisory principle more flexible to reflect the change of bank working environment, and make bank risk control keeps all the time sensitivity to risk movement of financial market.
Follow the basic idea of New Basel Accord in this thesis, we use the known foreign &domestic research fruit for reference and we also think of risk management practice of bank system of China to mainly discuss bank liquidity risk's computation problem, especially "liquidity overplus" problem. Then we could bring forward some new idea to study risk computation methods with high risk sensitivity. So we can validate the authenticity of "the liquidity overplus" of bank system of China, and slove the actual problem on bank liquidity risk management and supervision. In bank current practice, Var is the main risk measure for computing capital to cover with all kinds of risk. But Var is criticized by academia, they recommend Coherent Risk Measure include ES、Spectral Risk Measure、Concave Distortion Risk Measure etc. Although CRM overcome shortcoming of Var, it just meet zero-rank stochastic domination rule. In traditional stochastic domination theory, people can not get the risk measure which meets both monotonicity axiom of traditional stochastic domination theory and other three axioms of CRM. So people can not come true risk measure's four functions under the condition of according with value judgement of most finance and supervision institutions which stand by public side. The four functions are: 1、risk compositor and information issuance, 2、risk prevention and control, 3、risk supervision, 4、investment decision and achievement evaluation. In this paper, we made a demonstration analysis about main commercial banks in AnHui province. We use the high-rank lossexpectation risk measure ES~((n)) of general stochastic domination theory to solve the factual problem of bank liquidity riskmanagement. ES~((n)) can meet both general stochastic domination ruleand other three axioms of CRM. The primary content and contribution in this thesis as follow:
The first chapter is literature summary. First of all, we make a detailed introduction of correlative research fruit of Basel committee about bank liquidity computation and management. And secondly, we make a all around summary and comment about the Measures of liquidity risk which a commercial bank generally adopting and it's up to date development. Finally, we summarize the research background and status quo of academia.
On the basis of (R. Franck & M. Krausz, 2004), the second chapter makes a patulous analysis of three economic aspects (Capital Market、Interbank Market、Deposit insurance and Lender of Last Resort of Central Bank system) which have colse relation to bank liquidity provision. In 2.2 section, we analyse the influence of capital market on bank liquidity risk and optimal assets collocation by introduce market liqiudity risk model of (Kyle, 1985), and we make some relevant demonstration analysis.. In 2.3 section, we add interbank market model.. In 2. 4 section, we add deposit insurance model. Finally, we testify: Interbank Market、Deposit insurance system, the same as Capital Market、Lender of Last Resort of Central Bank system, can improve bank liquidity status and increase bank expect profit.
The third chapter and the fourth chapter are the core part of this thesis. There are two primary contribution as follow:The first is in the third chapter: For the first time in thisfield, we use Stat. instrument Copula which used to understandingrelationships among multivariate outcomes & High-Level ES Measurein General Stochastic Dominance Theory to measure commercial bankliquidity risk. This is a method which can measure bank solvencyand "liquidity overplus" degree more accurately. We can use thismethod to validate the authenticity of "the liquidity overplus"of bank system of China. Then we made a demonstration researchduring 1997-2006 on liquidity data of commercial banks in Anhuiprovince of China. We find the liquidity of current bank systemof China is strictly superfluous and also eaccurately measure it's degree, then we get a scientific taxis. This method can increasethe discernment of supervision institution and validity of riskcontrol of finance institution. In our model, if use Copula or notmay greatly impact the result of taxis. So Copula have a importantsignificance to the research on bank liquidity computation problem.Through seeking Copula dependence structure among the risk factors,it can not only successfully reflect the actual finance situationof AnHui province, but also improve the method for computingcapital to cover with liquidity risk. The conclusion is very helpful to finance supervision and bank liquidity safety.
During demonstration research We also find: For the problem we study, VaR、ES these two low-rank loss expectation risk measures will be wrong under the condition of general stochasticdomination. Whereas ES~((n)) measure is the high-rank spread of ES measure, it has all the fine character of ES measure, and also is 0～(n + 1) -rank general stochastic domination monotone &consistent risk measure. It can be always right under the condition of general stochastic domination. So it can come true better risk measure's four functions, and the rank more higher, the risk control measure more safer. If finance institution or supervision institution adopt this method, they can increase the validity of risk supervision and enhance the reliability of risk control and decrease the possibility of decision error.
The second is in the fourth chapter : We use Quantile Autoregression method to measure and forecast bank liquidity risk on some time point. The leading character of QAR is it's coefficient changing along with Quantile's change. So compare to constant coefficient model, it is more reasonable. This method differs from the third chapter's, which can make a Real-time measure of liquidity risk on some time point and make a forecast on the next point. Compare with Var & Es measure, this method is more logical for risk sort order on some time point.
Through demonstration test on these two methods, we can see that they can not only reflect regional economy & finance movement well and truly, but also accurately compute the degree of "liquidity overplus" and "correcting capability" of banks of AnHui province. So supervision institution can find out some banks which problem are very serious, then supervisor will make a choose between keep silence and external interference. Thereout, it provides a operable standard for central bank actualizing monetary policy and for finance supervision institution holding liquidity risk of the whole bank system. So it has a significance for enhancing liquidity risk manament level and keeping away systemic finance risk of bank system of China. Therefore, the third and fourth chapter construct a new scientific system together, which can make a accurate measure of a commercial bank's liquidity risk. This system can not only measure bank liquidity risk in a long time, but also on some time point, even can forecast on the next time point. Demonstration analysis testifies to the validity of the regional applications, this system can be lightly applied to the whole country. So it is highly valuable to money & supervision institutions. This part of content contains the core contribution of this thesis.
Because of liquidity gap management is very important to commercial bank effectively manage it's liquidity risk. In the fifth chapter, we use the Longest Run Statistics to research on financial data of commercial banks in Anhui province of China. Then we get hold of durative intensity about capacity to maintain moderation liquidity gap of commercial banks in AnHui province. Our conclusion is: As a whole, the bank system liquidity status of AnHui province is all right. It's P value is 0.94960. So we can consider that the banks of AnHui province run steadily, and the taxis of their ability to keep appropriate liquidity gap is: China Construction Bank、China Citic Bank、China Everbright Bank、Agricultural Bank of China、China Merchants Bank、Industrial and Commercial Bank of China & Bank of Communications、ShangHai PuDong Development Bank、Bank of China. Therefore, we can see that the taxis of the four state-owned banks and the five joint-stock banks is jumbly, there is no circumstance which one side exceeds another side. So there no correlative relationship between proprietary system and bank liquidity risk status. In other words, the capital capital management level of the four state-owned banks is as good as state-owned bank's.
The sixth chapter is about to analyzes diversified macro-risk factors which have significant influence on liquidity risk of commercial bank of AnHui province. We will use Gray system method & Non-linear co-integration theory and it's methods to solve this problem. We hope to find the most significant macro-risk factor to it, and We also hope to find the long-range equilibrium relation between macro-risk factors and liquidity risk of commercial bank of AnHui province. Therefore, we can make a test on capacity of eluding macro-economy risk for commercial bank of AnHui province. Our conclusion is: Factors of government finance、foreign trade、consumption、investment have significant influence on liquidity risk of commercial bank of AnHui province. It adequately validates that economy of AnHui province becomes more and more extroverted and it's development largely depend on the investment by government. This do accord with the fact of AnHui province.
In the end, the tag is summary on research work of this thesis. And however it have many places need to perfect.
引文
[1]Betaman,G.I.,1948."On The Power Function of The Longest Run As a Test For Randomness In a Sequence ofAlternatives[J]",Biometrika,35,97-112.
[2]BIS Committee on the Global Financial System,1999."A review of financial market events in autumn 1998"
[3]Kahn,C.M.& Shantos,J.A.,"Allocating bank regulatory powerslender of last resort,deposit insurance and supervision",BIS Working Papers No.102.
[4]Dewatripont,M.& Tirole,J.,1994."The prudential regulation of banks",Cambridge,MA:MIT Press
[5]Diamond,D.K & Dybvig,P.H.,1983."Bank Runs,Deposit Insurance,and Liquidity",Journal of Political Economy June
[6]Diamond,D.W.,1997."Liquidity,Banks,and Markets",Journal of Political Economy,vol.105 no.5
[7]Diamond,D.W.& Rajah,R.G.,1999."Liquidity risk,Liquidity Creation and Financial Fragility:A Theory of Banking",NBER Working Paper No.7430,December
[8]Gatev,E.& Schuermann,T.& Strahan,P.E.,2005."How do Banks Manage Liquidity Risk? Evidence from The Equity and Deposit Markets in The Fall of 1998"
[9]Karsak,E.E.,1998."Measures of liquidity risk supplementing fuzzy discounted cash flow analysis",The engineering economist,summer,Volume 43.No.4
[10]Carletti,P.H.& Spagnolo,G.,2003.11."Bank mergers,competition and liquidity",ECB Working papers No292.
[11]Engle,R.F.& Granger,C.W.J.,1987."Co-integration and Error Correction" Representation,Estimation,and Testing",Econometric,55" 251-276.
[12]Freixas,X.& Rochet,J.C.,1997."Microeconomics of Banking",Cambridge,MA:MIT Pressa
[13]Glaz,Joseph,Joseph I.Naus,and Sylvan Wallenstein,"Scan Statistics"[M],(2001).New York:Springer-Verlag.
[14]Green.E.J & Lin.P.,2003."Implementing efficient allocations in a model of financial intermediation",Journal of Economic theory,109(1),1-23
[15]Hurvich C M,Shumway R,Tsai CL.(1990)."Improved estimators of Kullback-Leibler information for autoregressive model selection in small samples",Biometrika,77:709-719."
[16]Ericsson,J.& Renault,0.,2006."Liquidity and Credit Risk"
[17]Tapking,J.,2006."Multiple equilibrium overnight rates in a dynamic interbank market game",Games and Economic Behavior,56(2),pp.350-370.
[18]Horowitz,J.L.,1998.Department of Economics University of lowa iowa City,《Semi-parametric Methods in Econometrics》,Springer.
[19]Mitusch,K.,2002."Debt contracts,banks,and aggregate liquidity",economic letters 74.145-150
[20]Konenker,R.& Bassett,G.,1978."Regression Quantiles",Econometrica,No.46 33-49.
[21]Konenker,R.& Xiao,Z.,2006."Quantile Autoregression",Journal of the American Statistical Association,Vol.101,No.475 980-1006.
[22]Lawrence Saez,Xianwen Shi,"Liquidity Pools,Risk Sharing,and Financial Contagion",Journal of Financial Services Research,25:15-23,2004
[23]Peck,J.& Shell,K.,2003."Equilibrium bank runs",Journal of Political Economy,111(1),103-123
[24]Brockwell,P.J.& Oavis,R.A.,2002."Introduction to Time Series and Forecasting",Springer-Verlag New York,Inc.
[25]Raphael Franck & Miriam Krausz,"Liquidity risk and bank portfolio allocation",International Review of Economics & Finance,September 2004
[26]Barr,R.S.& Siems,T.F.,1996.7."Bank failure prediction using DEA to measure management quality".
[27]Dubil,R.,2003."How to include liquidity in a market VaR statistic",Journal of Applied Finance.
[28]Roberto de Matteis,"Fitting Copula to Date",Diploma thesis,Institute of Mathematics of the University of Zurich,2001.
[29]Little,R.J.A.& Rubin,D.B.,2004.9."Statistical Analysis with Missing Data",中国统计出版社.
[30]Samartin,M.,2003."Should bank runs be prevented?",Journal oY Banking and Finance,27(5),977-1000
[31]Sebastian Alberto Rey,Javier Ignacio Garcia Fronti,Maria Teresa Casparri,"Liquidity risk estimation using fuzzy measure theory",Working papers,0504012,EconWPA,2005
[32]Silverman,B.W.,1986."Density Estimation for Statistics and Data Analysis",Chapman and Holl,London.
[33]Zhao,L.C.& Krishnaiah,P.R.& Bai,Z.D.,1987." Remarks on certain criteria for detection of number of signals." IEEE Trans.AcousticsSpeech Sienal Process.ASSP-35,No.2,129-132.
[34](美)约瑟夫.F.辛基,2002年5月,《商业银行财务管理》,中国
[35](美)埃里克,班克斯,2005年6月,《流动性风险——资产管理和筹资风险》,经济管理出版社。
[36](美)Peter S.Rose[著],刘园[译],2005年,《商业银行管理》,第5版,机械工业出版社。
[37](英)马克.洛尔,列夫.博罗多夫斯基,2002年11月,《金融风险管理手册》,机械工业出版社。
[38]补冯林,张为国,何伟,2005年,“基于超高频数据的股票流动性度量研究”,统计与决策,第24-26页。
[39]陈晖,谢赤,2004年4月,“中国银行间同业拆借市场利率结构转换研究”,管理科学,第65-70页。
[40]陈希孺,1996,《概率论与数理统计》,中国科学技术大学出版社。
[41]邓聚龙,《灰色系统理论教程》,华中理工大学出版社,55页。
[42]董文泉,高铁梅,姜诗章,陈磊,1998,《经济周期波动的分析与预测方法》,吉林大学出版社,第12章。
[43]范建军,2004年6月,“现代中央银行最后贷款人制度的演进”,经济评论,第86-95页。
[44]房振明,王春峰,曹媛媛,2005年,“上海证券市场流动性模式的研究”,管理工程学报,第33-39页。
[45]高铁梅主编,2006年,《计量经济分析方法与建模》,清华大学出版社,第155-157页。
[46]葛奇,霍团结,黄小军,2001年1月,《美国商业银行流动性风险和外汇风险管理》,中国经济出版社。
[47]韩贵英,李尧,2004年11月,“当前我国银行间市场发展的机遇和挑战”,西南民族大学学报,第203-205页。
[48]姜占英,刘浏,2005年2月,“防止银行挤兑:存款保险与最后贷款人制度的比较研究”,吉林财税高等专科学校学报,第28-31页。
[49]李德辉,方兆本,2006年1月,“扫描统计量——检测基金业绩持续性的新方法”.运筹与管理,第82-87页。
[50]李世宏,2004年2月,“香港最后贷款人制度分析及启示”,西南金融,第54-55页。
[51]栗智,1996年1月,“DBMS在正态样本异常值判断和处理中的应用”,新疆钢铁,第38-43页。
[52]刘思峰等,《灰色系统理论及其应用》,科学出版社,第49-70页。
[53]刘逖,“如何衡量流动性:理论与文献综述”,上海证券交易所
[54]刘印旭、张世英,2005年8月,“基于门限协整系统的预测方法研究”,数量经济技术经济研究,第129-136页。
[55]马晓兰,2005年1月,“从银行间货币市场看中国利率的市场化过程”,贵州财经学院学报,第15-18页。
[56]唐爱国,2005年,《广义随机占优理论:群体决策、收入分配与风险管理》,北京大学出版社。
[57]王春峰,2001年2月,《金融市场风险管理》,天津文学出版社。
[58]温珂,2003年3月,“银行流动性管理中外比较研究”,重庆工商大学学报,社会科学版,第9-11页。
[59]吴磊,陈日华,2005年2月,“证券市场流动性研究评述”,同济大学学报:社会科学版,114-120页。
[60]吴振翔,2004年,“Copula与组合投资风险分析”,博士学位论文,中国科学技术大学,合肥。
[61]徐枫,2004年9月,“我国银行间同业拆放利率影响因素的实证分析”,金融论坛,第57-63页。
[62]徐晓明,张松,赵海焰,“金融市场微观结构理论综述”,西南财大中国金融研究中心。
[63]叶阿忠,2002年,“非线性协整的非参数检验方法”,预测,第79-80页。
[64]应展宇,2003年10月,“流动性提供中的银行与市场”,资本市场评论。
[65]詹场,胡星阳,2001年7月,“流动性衡量方法之综合评论”,国家科学委员会研究期刊(台湾省):人文及社会科学,十一卷三期,第205-221页。
[66]张川,陈桂云,潘德惠,2003年3月,“银行资产流动性模型及预测和优化控制”,控制与决策18卷第2期。
[67]张喜彬、孙青华、张世英,1999年1月,“非线性协整关系及其检验方法研究”,系统工程学报,第57-68页。
[68]张正平,何广文,2005年6月,“存款保险制度在全球的最新发展、运行绩效及其启示”。国际金融研究。
[69]周理,1997年4月,“1996年全国银行间拆借市场报告”,上海金融,第28页。
[2]BIS Committee on the Global Financial System,1999."A review of financial market events in autumn 1998"
[3]Kahn,C.M.& Shantos,J.A.,"Allocating bank regulatory powerslender of last resort,deposit insurance and supervision",BIS Working Papers No.102.
[4]Dewatripont,M.& Tirole,J.,1994."The prudential regulation of banks",Cambridge,MA:MIT Press
[5]Diamond,D.K & Dybvig,P.H.,1983."Bank Runs,Deposit Insurance,and Liquidity",Journal of Political Economy June
[6]Diamond,D.W.,1997."Liquidity,Banks,and Markets",Journal of Political Economy,vol.105 no.5
[7]Diamond,D.W.& Rajah,R.G.,1999."Liquidity risk,Liquidity Creation and Financial Fragility:A Theory of Banking",NBER Working Paper No.7430,December
[8]Gatev,E.& Schuermann,T.& Strahan,P.E.,2005."How do Banks Manage Liquidity Risk? Evidence from The Equity and Deposit Markets in The Fall of 1998"
[9]Karsak,E.E.,1998."Measures of liquidity risk supplementing fuzzy discounted cash flow analysis",The engineering economist,summer,Volume 43.No.4
[10]Carletti,P.H.& Spagnolo,G.,2003.11."Bank mergers,competition and liquidity",ECB Working papers No292.
[11]Engle,R.F.& Granger,C.W.J.,1987."Co-integration and Error Correction" Representation,Estimation,and Testing",Econometric,55" 251-276.
[12]Freixas,X.& Rochet,J.C.,1997."Microeconomics of Banking",Cambridge,MA:MIT Pressa
[13]Glaz,Joseph,Joseph I.Naus,and Sylvan Wallenstein,"Scan Statistics"[M],(2001).New York:Springer-Verlag.
[14]Green.E.J & Lin.P.,2003."Implementing efficient allocations in a model of financial intermediation",Journal of Economic theory,109(1),1-23
[15]Hurvich C M,Shumway R,Tsai CL.(1990)."Improved estimators of Kullback-Leibler information for autoregressive model selection in small samples",Biometrika,77:709-719."
[16]Ericsson,J.& Renault,0.,2006."Liquidity and Credit Risk"
[17]Tapking,J.,2006."Multiple equilibrium overnight rates in a dynamic interbank market game",Games and Economic Behavior,56(2),pp.350-370.
[18]Horowitz,J.L.,1998.Department of Economics University of lowa iowa City,《Semi-parametric Methods in Econometrics》,Springer.
[19]Mitusch,K.,2002."Debt contracts,banks,and aggregate liquidity",economic letters 74.145-150
[20]Konenker,R.& Bassett,G.,1978."Regression Quantiles",Econometrica,No.46 33-49.
[21]Konenker,R.& Xiao,Z.,2006."Quantile Autoregression",Journal of the American Statistical Association,Vol.101,No.475 980-1006.
[22]Lawrence Saez,Xianwen Shi,"Liquidity Pools,Risk Sharing,and Financial Contagion",Journal of Financial Services Research,25:15-23,2004
[23]Peck,J.& Shell,K.,2003."Equilibrium bank runs",Journal of Political Economy,111(1),103-123
[24]Brockwell,P.J.& Oavis,R.A.,2002."Introduction to Time Series and Forecasting",Springer-Verlag New York,Inc.
[25]Raphael Franck & Miriam Krausz,"Liquidity risk and bank portfolio allocation",International Review of Economics & Finance,September 2004
[26]Barr,R.S.& Siems,T.F.,1996.7."Bank failure prediction using DEA to measure management quality".
[27]Dubil,R.,2003."How to include liquidity in a market VaR statistic",Journal of Applied Finance.
[28]Roberto de Matteis,"Fitting Copula to Date",Diploma thesis,Institute of Mathematics of the University of Zurich,2001.
[29]Little,R.J.A.& Rubin,D.B.,2004.9."Statistical Analysis with Missing Data",中国统计出版社.
[30]Samartin,M.,2003."Should bank runs be prevented?",Journal oY Banking and Finance,27(5),977-1000
[31]Sebastian Alberto Rey,Javier Ignacio Garcia Fronti,Maria Teresa Casparri,"Liquidity risk estimation using fuzzy measure theory",Working papers,0504012,EconWPA,2005
[32]Silverman,B.W.,1986."Density Estimation for Statistics and Data Analysis",Chapman and Holl,London.
[33]Zhao,L.C.& Krishnaiah,P.R.& Bai,Z.D.,1987." Remarks on certain criteria for detection of number of signals." IEEE Trans.AcousticsSpeech Sienal Process.ASSP-35,No.2,129-132.
[34](美)约瑟夫.F.辛基,2002年5月,《商业银行财务管理》,中国
[35](美)埃里克,班克斯,2005年6月,《流动性风险——资产管理和筹资风险》,经济管理出版社。
[36](美)Peter S.Rose[著],刘园[译],2005年,《商业银行管理》,第5版,机械工业出版社。
[37](英)马克.洛尔,列夫.博罗多夫斯基,2002年11月,《金融风险管理手册》,机械工业出版社。
[38]补冯林,张为国,何伟,2005年,“基于超高频数据的股票流动性度量研究”,统计与决策,第24-26页。
[39]陈晖,谢赤,2004年4月,“中国银行间同业拆借市场利率结构转换研究”,管理科学,第65-70页。
[40]陈希孺,1996,《概率论与数理统计》,中国科学技术大学出版社。
[41]邓聚龙,《灰色系统理论教程》,华中理工大学出版社,55页。
[42]董文泉,高铁梅,姜诗章,陈磊,1998,《经济周期波动的分析与预测方法》,吉林大学出版社,第12章。
[43]范建军,2004年6月,“现代中央银行最后贷款人制度的演进”,经济评论,第86-95页。
[44]房振明,王春峰,曹媛媛,2005年,“上海证券市场流动性模式的研究”,管理工程学报,第33-39页。
[45]高铁梅主编,2006年,《计量经济分析方法与建模》,清华大学出版社,第155-157页。
[46]葛奇,霍团结,黄小军,2001年1月,《美国商业银行流动性风险和外汇风险管理》,中国经济出版社。
[47]韩贵英,李尧,2004年11月,“当前我国银行间市场发展的机遇和挑战”,西南民族大学学报,第203-205页。
[48]姜占英,刘浏,2005年2月,“防止银行挤兑:存款保险与最后贷款人制度的比较研究”,吉林财税高等专科学校学报,第28-31页。
[49]李德辉,方兆本,2006年1月,“扫描统计量——检测基金业绩持续性的新方法”.运筹与管理,第82-87页。
[50]李世宏,2004年2月,“香港最后贷款人制度分析及启示”,西南金融,第54-55页。
[51]栗智,1996年1月,“DBMS在正态样本异常值判断和处理中的应用”,新疆钢铁,第38-43页。
[52]刘思峰等,《灰色系统理论及其应用》,科学出版社,第49-70页。
[53]刘逖,“如何衡量流动性:理论与文献综述”,上海证券交易所
[54]刘印旭、张世英,2005年8月,“基于门限协整系统的预测方法研究”,数量经济技术经济研究,第129-136页。
[55]马晓兰,2005年1月,“从银行间货币市场看中国利率的市场化过程”,贵州财经学院学报,第15-18页。
[56]唐爱国,2005年,《广义随机占优理论:群体决策、收入分配与风险管理》,北京大学出版社。
[57]王春峰,2001年2月,《金融市场风险管理》,天津文学出版社。
[58]温珂,2003年3月,“银行流动性管理中外比较研究”,重庆工商大学学报,社会科学版,第9-11页。
[59]吴磊,陈日华,2005年2月,“证券市场流动性研究评述”,同济大学学报:社会科学版,114-120页。
[60]吴振翔,2004年,“Copula与组合投资风险分析”,博士学位论文,中国科学技术大学,合肥。
[61]徐枫,2004年9月,“我国银行间同业拆放利率影响因素的实证分析”,金融论坛,第57-63页。
[62]徐晓明,张松,赵海焰,“金融市场微观结构理论综述”,西南财大中国金融研究中心。
[63]叶阿忠,2002年,“非线性协整的非参数检验方法”,预测,第79-80页。
[64]应展宇,2003年10月,“流动性提供中的银行与市场”,资本市场评论。
[65]詹场,胡星阳,2001年7月,“流动性衡量方法之综合评论”,国家科学委员会研究期刊(台湾省):人文及社会科学,十一卷三期,第205-221页。
[66]张川,陈桂云,潘德惠,2003年3月,“银行资产流动性模型及预测和优化控制”,控制与决策18卷第2期。
[67]张喜彬、孙青华、张世英,1999年1月,“非线性协整关系及其检验方法研究”,系统工程学报,第57-68页。
[68]张正平,何广文,2005年6月,“存款保险制度在全球的最新发展、运行绩效及其启示”。国际金融研究。
[69]周理,1997年4月,“1996年全国银行间拆借市场报告”,上海金融,第28页。