【作者】 谢静；

【导师】 王之怡；

【作者基本信息】 西南财经大学 ， 计算机应用技术， 2012， 硕士

【摘要】 中国经济的迅猛发展,对现代化大额实时支付系统的流动性提出了更高要求,优化管理系统流动性不仅可以提高支付结算效率,还能降低支付系统参与者的流动性成本。本文以中国现代化大额支付系统为背景,结合金融机构流动性管理所面临的各种支付环境,通过对支付数据的分析,运用仿真模拟等手段,构建适用于一般流动性优化管理的原型系统,并通过研究该系统在不同制度设计下的反应,为中央银行大额支付系统的流动性管理提供政策建议。论文共分为七章,引言部分主要对国内外支付系统流动性方面的研究做了一个综述,国外主要研究三类流动性问题：(1)流动性节约机制,(2)流动性与速度,(3)流动性与风险。国内研究相对滞后,主要分为两类：第一类主要是通过与其它国家的大额支付系统进行对比或通过分析现行系统的制度,从政策上给出建议；另一类主要利用我国大额支付系统的历史数据进行实证分析。通过文献综述发现国外并没有专门针对货币政策影响支付系统流动性方面的研究,国内也只是有学者用历史支付数据进行过简单分析。第二章先介绍中国现代支付系统的体系结构；接着介绍了我国大额支付系统的参与者和系统运行现况,以及对每个营业日日末流动性不足的参与者进行管理的清算窗口,重点分析我国大额支付系统采用的RTGS(实时全额支付结算)机制,指出RTGS系统在效率和规避风险等方面的优势是以参与者的高流动性成本为代价的；最后介绍文章的研究主题——大额支付系统流动性管理,对支付系统中的流动性风险给出了明确的定义,详细地介绍了我国大额支付系统目前在流动性管理方面所采取的手段和措施,特别是第二代支付系统中优化大额支付系统流动性管理的“一揽子”实时流动性查询、排队业务撮合、日末自动拆借等功能。本文的核心理论部分是第三章的内容,研究存款准备金率和利率政策的调整对大额支付系统参与者流动性的影响。人民银行采用的主要货币政策工具有四种：法定存款准备金政策、央行贷款、公开市场操作和利率政策。货币政策在国内主要通过信贷渠道传导的,存款准备金率的调整短期内直接影响银行超额存款准备金额度,其长期效应是会改变商业银行的资金结构的,本文将大额支付系统参与者的日间来账、往账、日初余额以及存款准备金率和超额存款准备金率的关系用数学模型推理出来。利率政策是通过信贷渠道对参与者的存贷款产生影响,进而影响支付系统的流动性的。笔者根据已有文献提出的息差与银行存贷款的模型,变换得出存贷款跟基准利率的关系,再与支付系统的交易数据换算匹配,得出最终的数学模型。通过对2010年相关统计数据进行两步系统GMM统计得出货币政策贷款供给函数的估计结果,在其中能找到部分参数,为了增加模型的准确性,本文利用同年的存贷款数据、存贷款利率、银行间国债回购利率求得其他参数。第四章讲述流动性优化管理原型系统的设计与实现。CNAPS (China National Advanced Payment System)作为研究的目标系统,是中国的核心循环系统,对迅速发展的经济的和谐顺利起着至关重要的作用。中央银行高度关注它在国内经济稳定和金融安全方面的重要性。本文先介绍基于CHAPS创建流动性优化管理系统的基本思路,由于它的复杂性,安全性和不确定性,不可能直接测试CNAPS产品系统来获取研究目的。所以在芬兰央行的模拟支付软件BoF-PSS2上具体搭建LOMS(流动性优化管理系统),LOMS主要在排队优化和研究货币政策对流动性影响两方面对大额支付系统进行流动性优化管理,包含排队优化算法和货币政策分析模块。本文的重点是货币政策分析模块。第五章介绍支付数据的模拟生成。先说明使用模拟数据的优缺点,对比几个国家支付交易数据呈现的不同特征,然后针对我国大额支付系统,研究得出支付数据服从泊松-对数正态的随机过程分布,利用真实统计数据估计模型参数,最后用Matlab编程产生支付数据。第六章进行支付仿真实验。先说明实验使用的数据表,然后根据系统之间的关系和实验目的安排实验流程,最后在LOMS原型系统中进行模拟实验,研究支付系统不同的货币政策制度设计对参与者流动性的影响。本文主要结论如下：第一、随着法定存款准备金率的不断上调,银行的流动性压力越来越大,出现流动性不足的参与者个数不断增加。如果下调准备金率,银行则会拥有更多的流动资金。但并非存款准备金率调得越低越好,参与者超额准备金变多,代表着流动性成本增高,为维持自身利益,银行会转移部分资金,这样就降低清算账户上的日初余额,从而影响后期参与者的日间流动性并易造成系统性风险。存款准备金率是人民银行控制各个参与者的流动性的指挥棒。当支付系统流动性过剩时,适当上调,冻结部分资金；当支付系统参与者流动性压力过大时,适当下调,释放部分资金。过度上调或下调存款准备金率都会对支付系统带来流动性风险,找到调整的平衡点,即合适的存款准备金率是至关重要的。第二、本文通过仿真实验得出贷款基准利率与参与者的日末余额是呈正相关性的,但利率政策的调整对支付系统流动性影响较小。这一结论符合我国的经济现状。随着利率市场化的不断推进,我国利率调控由直接方式转向主要依靠间接方式。利率市场化改革进程中货币政策利率传导渠道效应出现了体制转换特征。转换之后贷款基准利率调整对产出和物价水平的效应都降低了。实验结果表明利率作为货币政策对支付系统参与者存贷款的影响越来越小。这也从另外一个角度证明了：我国货币政策传导渠道的中介目标逐渐由M2(货币供应量)转变为利率是一种趋势,利率将随金融市场的变动而改变,而强制性的使用利率政策对市场经济已不再起明显作用。更多还原

【Abstract】 As China’s economy develops rapidly and vigorously, it put forward higher requirements on the liquidity of the existing of large value real-time payment system. Optimize the management of system liquidity can not only improve the efficiency of payment and settlement, but also reduce the liquidity cost of payment system participants. This article focused on Chinese High-Value Payment System, built up prototype system of liquidity optimization management with the payment environments those financial institutions faced in their liquidity management, the analysis of payment data, and simulation researches. And studied the responses of the system under different monetary policies, in order to provide policy recommendations for the Central Bank of the large-value payment system liquidity management.The thesis is divided into seven chapters. Chapter1introduces the background and significance of this research, and the study status of domestic and international payment system liquidity management, as well as the research method, content, innovation and papers framework. Chapter2gives an overview of China’s large-value payment systems and liquidity management, including the architecture of China’s modern payment system, the participants and system operating mode, as well as payment and settlement mechanism. Chapter3theoretically studied the change of deposit reserve ratio and interest rates influence on the liquidity of LVPS, and established the relevant mathematical model. Chapter4illustrated the design and implementation of the prototype system to optimize the management of liquidity in LVPS. Mainly about the basic idea to create liquidity optimization management system based on the CHAPS, and how to develop LOMS based on BoF-PSS2. Chapter5is about how to generate simulated payment data. Firstly explain the advantages and disadvantages of using simulation data. Then through literature study and practical investigation, proved that payment values are Poisson-Lognormal distributed. Finally generate payment data by programming on Matlab. In chapter6, there’re simulation experiments. Through simulation testing of the prototype system performance, point out the balance of liquidity needs and payment and settlement efficiency. By using the prototype system, analysis the liquidity needs of system participants under different monetary policies. Chapter7is summary and prospect.The main conclusions are as follows:First, if the deposit reserve ratio continues to increase, the bank will face more and more liquidity pressure, the number of participants that lack of liquidity also grows. If central bank cut down the reserve ratio, banks will have more liquidity. But the more excess reserves the participants hold, the higher costs of liquidity they have. In order to maximize their income, the banks will transfer part of the funds, thus reducing the reserve account in the beginning of day, then affecting the daily liquidity for participants and could easily lead to systemic risk in later period. The deposit reserve ratio is the baton of the People’s Bank to control the liquidity of the various participants.Second, through the simulation experiments, it turns out that the benchmark loan rate and the participants’end of day balance is in a positive correlation, but the adjustment of interest rate policy affects liquidity in payment system very slightly. This conclusion is consistent with China’s economic situation. With the marketization of interest rates getting forward, China’s interest rates regulation change from the direct way to rely on an indirect way. In the market-oriented interest rate reform process, the monetary policy transmission channel effect result in structural transformation characteristics. The adjustment of basic rate has less effect on output and price levels. The experimental results show that the interest rate as monetary policy has less influence on the deposits and loans for payment system participants.更多还原