我国商业银行操作风险实证研究

发布时间：2018-05-27 10:45

本文选题：操作风险 + 收入模型　；参考：《西南财经大学》2013年硕士论文

【摘要】：操作风险是一项自银行开展经营活动以来就存在的古老风险,但长期都没有以一种独立存在的风险形式纳入银行风险管理体系中。20世纪90年代以来,随着金融深化的逐渐演进银行规模越来越大,业务日趋繁杂,操作风险带来的危害愈发彰显。操作风险案例频频暴发给银行业带来了重大的资金损失和信誉危机,至此人们才意识到对操作风险进行有效监管的重要性。在此背景下巴塞尔银行监管委员会积极应对操作风险的研究和监管,2004年6月,巴塞尔委员会公布了新协议定稿“资本计量和资本标准的国际协议：修订框架”。新协议首次提出操作风险资本金的计算方法,并出台更先进、更贴近市场的监管条例督促银行增强操作风险的管理措施。新协议一出台,全球商业银行开始有意识地从各个经营环节改善对操作风险的管控措施,如实施强有力的内部控制制度：建立独立、专用的操作风险管理后台部门；制定全行统一的操作风险的战略部署；倾入人力、物力开发实际可用的操作风险管理工具和度量模型、加强操作风险文档建设、建立内部损失数据库等。国际上一系列的举措让操作风险的研究和监管得到了快速的发展。相比之下国内商业银行在操作风险管理研究方面起步比较晚,甚至至今都没有有效的操作风险度量模型,也没有充足完备的操作风险内部损失数据库。由于金融行业发展历史不长,市场化程度不够,信息存在短缺,国外较精准的计量操作风险的模型并不能直接移植。现阶段国内学者的研究更多的只停留在对《新巴塞尔协议》操作风险管理框架的介绍,并沿用已有的模型对我国商业银行做实证分析,还没有涉及太多基于中国实际情况的操作风险度量模型和框架的开发。从国内宏观经济环境分析操作风险的现状特征及其计量管理的存在的障碍。现阶段我国商业银行正处在转型时期,很多必备的机制都在查漏补缺之中。人员道德问题和信息披露不精确、技术滞后、电子化运作漏洞、制度建设和系统流程管理实施不善等统统都导致了操作风险成为威胁我国银行业健壮发展的最大隐患。另外在操作风险量化管理方面也障碍重重,首先面临的问题是量化管理操作风险的意识薄弱,即银行业根本就没有利用建立内部数据库累积损失文档的方式对操作风险进行电子化的管理的经验。其次由于我国商业银行传统的产权归属和信息披露不透明的原因,内部数据的收集是相当大的难题。而根据新协议的规定利用高级的、精确的操作风险回归模型或计量方法的前提是有必要的充足的内部数据做基础,数据的匮乏意味着对操作风险的量化只能采取最老套、原始的方式。由此得到的数据不精准,不能以此为据计提相应的经济资本,而没有计提操作风险损失准备金作为抗击风险的保证操作风险监管始终不能有效到位,而低效无保障的管理又会进一步加剧操作风险的潜在隐患,如此一来形成一个监控纰漏的恶性循环,操作风险带来的冲击将愈演愈烈。再次,由于中国商业银行实行行长负责制,稽核部门根本不能与决策部门抗衡,换句话说国内银行业实行的行长责任制将权利过多赋予内部高层,稽核审计部门不能独立地完成对全行整体业务和风险管理评审的工作,因而其搜集的操作风险监管一手资料存在信息不全或者信息偏颇的问题,后续的量化过程根部无从做起。鉴于操作风险的度量环境不佳、内部数据不充足的情况,笔者只能借“由上至下”的收入模型对国内操作风险做试探性的研究。收入模型的基本思路是以目标银行的净利润为应变量,净利润的波动性代表银行面临的总风险,把波动能够被度量的部分视为有信用风险和市场风险引起,而其不能被度量的部分则是由操作风险导致的。具体的步骤是先确定度量各类风险大小的解释变量,包括代表市场波动、监管部门的约束、放贷企业的还款能力、银行自身的经营情况的指标,然后基于目标变量建立模型,反应出各种风险指标和净利润的关系。计算目标变量的方差,将方差中能被度量的部分看做是信用风险和市场风险的大小,而不能被度量的部分则解释成操作风险的大小。建模之初笔者选择了真实GDP增速、上证综指、法定存款准备金率等七个指标代表市场风险和信用风险,并且将10家国有商业银行从2009年第一季度至2012年第三季度的财务数据做样本分别进行时间序列模型回归,经过层层筛选最终得到了以上证综指、不良贷款率、存贷比率为解释变量的模型以及九家商业银行各自的操作风险占比和损失金额,为其下一步计提操作风险资本金作一个参考。为了进一步说明问题笔者做了更深入的计量研究,将国有银行和股份制银行的数据分别建立时间序列模型,得到的结果是国有银行的操作风险占总风险的比例约是股份制银行操作风险总占比的两倍,说明股份制银行治理结构和内控管理有更高的效率,原因在于股份制银行的营业网点和员工数量较少,管理者能够更容易地对日常运营中的各个环节实施监控。而国有银行由于资产规模大,存在产权结构、信息系统、内控机制等各方面的缺漏,而且基于中国的特色国有银行存在政府强制干预,其独立性不足,加之因经营时间较长,网店分布较广,报告监管等滞后,管理效率相比更低,操作风险占比自然更大。实证结果与第二章国内银行业操作风险现状尤其是国有商业银行存在的问题描述是一致的,基于此国有商业银行应该加快制度改革步伐、改善产权结构和激励约束机制、强化信息透明度和系统配备、转变揽储为立行之本的经营方式加大发展中间业务的力度分散操作风险,积极推进风险管理制度改革。本文的框架：第一章,绪论。该部分首先介绍随着全球商业银行操作风险损失事件的频繁发生,有必要设计一套可靠的操作风险度量模型、实施量化管理。然后概述了本文的研究思路和行文脉络。最后对国内外与本文研究主题相关的文献进行了回顾。第二章,操作风险的概念介绍。该部分首先描述了操作风险的定义、分类和研究现状。然后介绍了国内银行操作风险管理的现状,并指出了度量操作风险时遇到的障碍。第三章,操作风险的度量方法及模型介绍。该部分详细介绍了各种操作风险的定量模型,对国际上商业银行的定量研究成果进行综述。第四章,操作风险的实证分析。该部分是本文的核心,利用收入法对我国上市商业银行操作风险进行实证研究,根据10家上市银行的数据做回归分析求出操作风险占的占比和涉及的损失金额,并对回归结果加以验证和说明。第五章,针对国内操作风险现状和实证回归结合管理建议。介绍一些有利于我国商业银行实现操作风险量化管理的政策建议。本文的创新之处在于试图寻找出最能反应我国商业银行操作风险现状的数据和模型,为了使结果公正和客观,笔者尽最大努力保证原始数据的充分,样本要够大,本文选择了10家上市银行15个季度的数据作为研究对象。考虑到不同银行的操作风险的特殊性,并且为了保证模型和估计结果的准确度和可靠性,笔者建立了属于各家银行的时间序列模型,通过模型的多次修改和数据的回归筛选最后选取了上证综指、存贷比、不良贷款率这三个指标,并在统一的模型下求出九家各自的操作风险估计值,为下一步银行计提操作风险准备金提供依据。此外,笔者还沿用此模型将股份制银行和国有银行分别进行实证分析,对比两类银行的回归结果提出我国操作风险的管理对策。本文的不足在于收集到的都是公开的财务报表数据,不能反映银行所有的经营信息。再加上数量模型本身不可能绝对精确地解释现实经济状况。因此计量结果会跟银行的实际情况有出入。除此之外操作风险的量化值被收入模型定义为引起净利润波动的因素中排除市场和信用风险剩下的部分,显得太过绝对了,与现实情况存在一定的差距,最后得到的估计结果也只是一个大概的取值并不精确,不能直接为银行提供计提操作风险损失准备金的依据。事实上巴塞尔委员会的要求,衡量操作风险首先应该收集银行内部的数据资料,然后评估风险事件的发生概率和预期损失,而收入模型忽略了银行的内部经营情况,也没有利用风险事件的发生概率和预期损失计算操作风险所需配置的资本,仅仅从外部数据估计处了操作风险的大致占比和涉及的损失金额。未来,我国商业银行应当同时利用“由上至下”和“由下至上”的方式对操作风险实施量化管理,前提是要得到更符合实际的结果还需要数据的大范畴收纳和结合中国银行业的切实考察,这是笔者下一步的努力方向。
[Abstract]:Operational risk is an ancient risk since the bank has carried out its business activities, but it has not been incorporated into the bank risk management system in the form of an independent risk for a long time since the 90s.20 century. With the gradual evolution of the financial deepening, the scale of the bank has become increasingly large, the business is increasingly complex, and the harm caused by operational risk is more and more serious. The frequent outbreaks of operational risk cases have brought great financial loss and credit crisis to the banking industry. At this point, people have realized the importance of effective supervision of operational risks. In this context, the Basel Banking Regulatory Commission actively deals with the research and supervision of operational risks. In June 2004, the Basel committee published a new report. The agreement finalized the international agreement on capital measurement and capital standards: the revised framework. The new agreement first proposed the calculation method of operating risk capital, and introduced more advanced, closer to the market regulatory regulations to urge banks to strengthen operational risk management measures. The new agreement was introduced, and the global commercial banks began to operate consciously from each operation. Steps to improve control measures for operational risks, such as implementing strong internal control systems: establishing independent, dedicated operational risk management backstage departments; formulating a strategic deployment of operational risk throughout the whole line; leaning into manpower, material resources to develop practical operational risk management tools and measurement models, and strengthening operational risk documents. Set up an internal loss database and so on. A series of international initiatives have made rapid development of operational risk research and regulation.
In contrast, the domestic commercial banks have a relatively late start in the study of operational risk management, and even now there are no effective operational risk measurement models, and there is not enough complete operational risk internal loss database. Because the history of the development of the financial industry is not long, the degree of marketization is not enough, the information is short, and the foreign more accurate measurement exercises. The model of risk making can not be transplanted directly. At present, more research of domestic scholars is only on the introduction of the framework of the operation risk management of the new Basel agreement and the empirical analysis of the commercial banks in China along with the existing models, and there are not too many operational risk measurement models and frameworks based on the actual situation in China.
From the domestic macro economic environment, the current characteristics of operational risk and the obstacles of measurement management are analyzed. At the present stage, the commercial banks in China are in the period of transition, and many necessary mechanisms are in the missing and missing. The problems of personnel ethics and information disclosure are inaccurate, technology lag, electronic operation loopholes, system construction and system process. The problem of operational risk has become the biggest threat to the robust development of China's banking industry. In addition, there are many obstacles in the quantitative management of operational risk. The first problem is that the awareness of the operational risk of quantitative management is weak. That is, the banking industry has not made use of the accumulated loss documents of the internal database. The collection of internal data is a considerable problem. The premise of using advanced, accurate operation risk regression model or measurement method is necessary according to the regulations of the new agreement. As the basis of adequate internal data, the lack of data means that the quantification of operational risk can only be used in the most old-fashioned, original way. The resulting data is not accurate, and the corresponding economic capital can not be taken as the basis for the operation of the risk. In place, the low efficiency and unguaranteed management will further exacerbate the potential risks of operational risk, so as to form a vicious cycle of monitoring errors, the impact of operational risk will be intensified. Again, because the Bank of China is responsible for the system, the audit department can not compete with the decision-making department at all, in other words domestic The executive responsibility system implemented by the banking industry gives excessive rights to the internal high-level, and the audit audit department can not perform the work on the overall business and risk management review independently. Therefore, there is a problem of incomplete information or biased information on the first-hand data of operational risk supervision, and the root of the follow-up quantitative process can not be done.
In view of the poor measurement environment of operational risk and insufficient internal data, the author can only use the "from top to bottom" income model to study the operational risk in China. The basic idea of the income model is that the net profit of the target bank is the corresponding variable, the wave of net profit represents the total risk that the bank faces, and the fluctuation can be used. The part that is measured is caused by credit risk and market risk, and the part that can not be measured is caused by operational risk. The specific step is to determine the explanatory variables that measure the size of all kinds of risks, including the market volatility, the regulatory authority, the repayment ability of the lending enterprises, and the management of the banks themselves. And then the model is built based on the target variable, and the relationship between the risk index and the net profit is reflected. The variance of the target variable is calculated. The part of the variance can be measured as the size of the credit risk and the market risk, but the part that can not be measured is interpreted as the size of the operation style risk. Speed, Shanghai Composite Index, legal deposit reserve ratio and other seven indicators represent market risk and credit risk, and 10 state-owned commercial banks from the first quarter of 2009 to the third quarter of the third quarter of the financial data to do the time series model regression model, after layers of screening finally got the above index, the rate of non-performing loans, deposit and loan. The ratio is the model of the explanatory variable and the respective operational risk ratio and the loss amount of the nine commercial banks. For the next step, we make a reference for the operation of the venture capital. In order to further explain the problem, the author made a more in-depth measurement and study, and established the time series model of the data of the state-owned banks and the joint-stock bank respectively. The result is that the proportion of the operational risk of the state-owned banks is about two times the total operational risk of the joint-stock banks, which indicates that the governance structure and internal control management of the joint-stock banks have higher efficiency. The reason is that the business outlets and the number of employees in the joint-stock banks are less, and the managers can more easily to each of the daily operations. Because of the large scale of the assets, the state-owned banks have the lack of property right structure, information system, internal control mechanism and so on. Moreover, the state owned banks have the government compulsory intervention based on the Chinese characteristic, and their independence is insufficient. In addition, because of the long operation time, the network stores are distributed widely, the report supervision is lagging behind, and the management efficiency is lower, The operational risk is greater than that of nature. The empirical results are consistent with the current situation of the operational risk of domestic banks in the second chapter, especially the existing problems of state-owned commercial banks. Based on this, the state-owned commercial banks should accelerate the pace of system reform, improve the structure of property rights and incentive and restraint mechanisms, strengthen the transparency and system of information, and change the storage of the banks. The way of running this business is to increase the intensity of developing intermediate business and decentralization of operational risks and actively promote the reform of risk management system.
The framework of this article:
The first chapter, introduction. This section first introduces the frequent occurrence of operational risk loss events in the global commercial banks. It is necessary to design a reliable operational risk measurement model and implement quantitative management. Then it summarizes the research ideas and lines of research in this paper. Finally, it reviews the literature related to the subject of this paper at home and abroad.
The second chapter introduces the concept of operational risk. This section first describes the definition, classification and research status of operational risk, then introduces the current situation of operational risk management in domestic banks, and points out obstacles encountered in measuring operational risk.
The third chapter is the measurement and model introduction of operational risk. This section introduces the quantitative model of various operational risks in detail, and summarizes the quantitative research results of commercial banks in the world.
The fourth chapter is the empirical analysis of operational risk. This part is the core of this article. Using the income method, the operational risk of the listed commercial banks in China is empirically studied. According to the data of 10 listed banks, the proportion of operational risk and the amount involved are calculated, and the results of the regression are verified and explained.
The fifth chapter, in view of the domestic operational risk status and empirical regression combined with management suggestions, introduces some policy suggestions for commercial banks to achieve quantitative management of operational risk in China.
In order to make the result fair and objective, the author tries to make the best effort to ensure the full and large data of the original data, and the sample should be large enough. This paper chooses 15 quarterly data of 10 listed banks as the research object. In order to ensure the accuracy and reliability of the model and the estimated results, the author set up a time series model which belongs to all the banks, and finally selected the three indexes of the Shanghai Composite Index, the loan ratio and the bad loan rate through the multiple modification of the model and the regression screening of the data. The nine respective operational risk estimates provide the basis for the next bank to raise the operational risk reserve. In addition, the author also uses this model to carry out an empirical analysis of the joint-stock banks and the state-owned banks respectively, and compares the regression results of the two types of banks to put forward the management countermeasures of operational risk in China.
The shortage of this article is that all the collected financial statements are collected and do not reflect all the operating information of the bank. And the quantitative model itself can not explain the real economic situation absolutely. Therefore, the measurement results will be different from the actual situation of the bank. The factors that cause the net profit fluctuation to exclude the remaining part of the market and credit risk appear too absolute, and there is a certain gap with the reality. The final results are only an inaccurate estimate of the value, which can not provide the basis for the bank to provide the financial loss reserve. In fact, the Basel committee member is in fact. In order to measure operational risk, it should first collect data from the bank, then evaluate the probability and expected loss of the risk events, and the income model ignores the internal operation of the bank, and does not use the capital to calculate the operational risk of the occurrence probability and expected loss of the risk events, only from the external number. It is estimated that the general proportion of operational risk and the amount of losses involved. In the future, the commercial banks of China should use the "from top to bottom" and "from the bottom to the highest" to implement quantitative management of operational risk. The premise is to get more conforming to the actual results and need the large category of data to receive and combine the Chinese banking sector. In fact, this is the next step of the author's efforts.
【学位授予单位】：西南财经大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.33

【参考文献】