A Research On Integrate Measurement Of Operational Risk Of Chinese Commercial Bank Based On Copula Methods

The outbreak of the subprime mortgage crisis in the United States, completely changed the financial structure of Wall Street, and it also had a profound impact on the global financial system. The enormous destruction of this crisis was beyond the expectation of many people. Alan Greenspan, Former Federal Reserve Chairman also believes that the subprime mortgage crisis was unprecedented in the recent century, government can do act better in risk regulation. Many foreign banks involved in real estate mortgage loans suffered huge losses, alarming Chinese commercial banks of how to manage risk well.Operational risk is one of risks the commercial bank faces. Bank staffs have not strict credit check to these real estate mortgage loan applicants, led to excessive lending, this behavior belongs to the scope of operational risk. Ba(2003) analyzed the characteristics of operational risk and the evolution of New Basel Capital Accord on requirements for operational risk, and discussed the operational risk measurement methods which the current international financial community is usually use. Zhao (2008) discussed the standard method of alternative forms in the measurement of operational risk capital of commercial banks. The research literatures about operational risk is based mainly on Basel Committee which assume perfect independence between the loss events, but the actual situation is not so.The various operational sections of commercial bank as a whole, there is interaction between them. If the internal staff of the bank perform illegal operation procedures, after it was disclosed, commercial bank through internal bulletin,corresponding penalties and other measures, would run impact to behavior that may exist in the other sections, such as internal fraud, etc. This increased vigilance of the internal staff, will strengthen the emphasis on external risks, to an extent, it will weaken some hidden risks, even eliminate. New Basel Capital Accord think that all loss events are completely independent, it is partial in conservatively and does not conform to reality. This paper use t-copula approach to describe the correlation structure of operational risk of commercial bank based on loss event types, instead of the correlation structure assumption between the events type which is more conservative and independent in New Basel Capital Accord. Using t-copula method to estimate the correlation of the loss events of operational risk for commercial banks, we can better understand the operational risk and take correct and effective response to it.In theory aspects, this paper summarizes the academic progress on operational risk of risk management scholar, especially the researches and conclusions on operational risk measurement of commercial bank. Simultaneously has briefly described the researches on the control and release of operational risk for commercial bank. This is helpful for establishing our own operational risk management system and measurement model based on the actual situation of China's commercial banks.From a practical point of view, Basel Committee requires every commercial bank to provide the operational risk reserve. Provision for operational risk reserve is based on the more accurate assessment of operational risk that commercial bank faces. Basel Committee encourages the commercial banks in various regions according to their own country's economic and financial laws and the actual situation to develop internal operational risk measurement model. This provides a good chance and conditions for China's commercial banks to establish our operational risk measurement model. In this way, it will don't underestimate the loss of operational risk and cause the potential hazards for direct copying foreign operational risk management regulations, at the meanwhile not have low-efficiency capacity to finance because of high operational risk reserve requirement.In order to more accurately measure the size of operational risk of China's commercial bank, this paper arranged four parts to analysis and research:Chapter 1 is introduction, including writing background,writing significance,domestic and international literature review of operational risk management.Chapter 2 is introducing the integration theory of operational risk of commercial bank based on the copula methods. Usually we use to describe the relationship between things by the linear correlation, this is not particularly appropriate in the measurement of operational risk. In this paper, we use copula methods to describe the tail dependence of variables to measure the structure of operational risk events of commercial bank.Chapter 3 is the empirical content of this paper; we use copula methods to integrate the operational risk of commercial bank. This article selected China's 14 commercial banks during 1987-2006 operational risk loss data as a data pool simulate the relationship between the operational risk loss event types of commercial banks based on EVT and t-copula methods, to get the size of the total operational risk loss that the commercial banks face.Chapter 4 is an end section of this paper, summarizes the challenges when China's commercial banks to carry out the construction of operational risk management system, meanwhile, we put a outlook to the operational risk management of china's commercial banks.Through the analysis and research above, the following conclusion can be safely drawn:(1)The correlation among the loss events of operational risk of commercial bank based on copula methods is obvious, which is remarkably different from the completely independent correlation suggestions by New Basel Capital Accord (Correlation coefficients see Table 3.7).(2)Fitting the severity of the operational risk of commercial bank by the generalized Pareto distribution is superior to the log-normal distribution suggested by traditional methods. (See graph3.1-graph3.3).(3)Compared with the traditional methods suggested by the New Capital Adequacy Framework (Basel II), the copula-EVT methods can help commercial banks save averagely 10% of provisions of operation risk.Overall, the major innovation of this paper is the following aspects:First of all, In Basel II, the measurement of operation risk of commercial banks is based on the simple arithmetic sum of such risk in various business lines or departments. It assumes that the operation risk of one business line/department is completely independent from that of another. However, this assumption is not consistent with the reality. As proposed in this article, the extreme-value theory (EVT) is used to model the loss severities, and the t-copula is used to estimate the correlation coefficient of operation risk between any two business lines. As a result, the measurement of operation risk in commercial banks can be optimized so that it can better match the real-world management and regulation in various commercial banks.Second of all, a Poisson distribution is usually used to model the frequency of operation risk. However, in the description of statistical characteristics of loss data, it's found that the loss data of event types is not consistent with the statistical characteristics of the Poisson distribution. Therefore, we use the negative binomial distribution here to model the frequency of loss event type.