A Study On Operational Risk Of Chinese Banking Industry: Based On Extreme Value Theory

Whether the South China Sea Branch of ICBC’s7.4billion yuan cheat case in2004, or the latest Everbright Securities "Oolong finger" event–7.5billion self funding upset the Chinese entire stock market, are telling us how to manage operational risk is so important. Only by strengthening research on operational risk management can we minimize the possibility of such financial turmoil’s outburst. with the advance of China’s financial system reform, commercialbanks operating business types and transaction volume will rapidly increase, which will increased the likelihood of operational risk.In addition, Opening bank sector to personal capital will bring a great impact on the financial industrywhich will aggravates the likelihood of bank operational risk.It is an important issue about How to manage operational risk, especially for the studies on how to measure the operational risk.At first, the paper overviewed the definition of basic operational risk andthe frame of classification.Secondly, we apply the Payoff Matrix Method to the Seven Categories of Operational Risk Events in order to solve the vaguely defined issue. Thirdly, we compare the domestic operational risk with the internationally active banking sector, noting that the current China’s commercial banks operational risk has its own feature. On this basis, this paper presents usingthe extreme theory to study China’s operational risk is reasonable and necessary.The steps of using Extreme Value Theory to measure China’s commercialbanks operational risk as follows: First, through public media we collected621cases about China’s banking operational risk during1987-2013. Second, we established POT model based on China’s operational risk, then we apply changepoint theory to the HILL point estimation in order to acquire more accurate threshold.Subsequently we use two kinds of goodness of fit test (Chi-square value and the KS value) to select the optimal threshold. Third, based on POT model of China’s commercial banks operational risk we can acquire the over-threshold generalized Pareto distribution function, according to the over-thresholdGPD function we can get the distribution function of operational risk which ca n help us calculate the VaR and ES value in China’s commercial banks operational risk at a certain level of confidence.at the same time we compare the ES values obtained in this study with that obtained by scholars before and found that our conclusion have much more empirical advantage.The results of this study showed that:(1) using extreme value theory tomeasure China’s banking operational risk is more appropriate.(2) the subjectivity in choosing threshold can be solved by goodness of fit test and change point theory,which make the estimation of VaR and ES value of operational riskof commercial banks (the commercial banks as a whole) more accurate.Innovations of this paper are as follows:(1)There are some flaws in previous study, such as, inadequate and less rigorous date, vague risk cases classification and so on. In view of these, this paper make improvement as follows: we apply payoff matrix method to the Basel’s seven risk cases classification, which will help us gain more precise classification of the loss events. While weexpand the range of sample selection, the data collected covers25years’ operational risk loss cases amounting to621valid data which may ensure the validity of the conclusions.(2) Based on change point theory we solve out the subjectivity in choosing threshold and realized quantitative-threshold selection In addition, we use Pearson’ Chi-square value and the KS value to select the optimal threshold from the candidate thresholds.Inadequacies of this study: when we apply the extreme value theory intochina’s commercial banks operational risk measurement, we found that china’sbanks have not established operational risk database. In view of this, we lookinto china’s commercial banks as a whole in order to get enough sample data,thereby we can calculate the VaR and ES. However, the result achieved by this mean is not very purposeful. The ES and VaR maybe not so precise to aspecial bank.Significance of this paper is as follows:(1) With the operational risk loss database of China’s commercial banks being established, the paper will advance an usefull method to help every commercial bank to measure its own operational risk (2) the results of this paper show us that China’s commercialbanks need to configure the operational risk capital funds around18.3billion.wedivide the above capital funds by the average of the four major state-owned banks’ net capital, then we will obtain a proportion which will be taked as astandard to weigh the controllability of operational risk. Accordingly, the regulatory authorities can make the policy more accurate and tendentious.