| Risk management is an essential issue concerns the financial institutions, and banks must also actively respond to this unavoidable question. With the rapid development of the world economy, the reform of financial markets and innovative financial derivatives, the risk exposure of commercial banks is increasing, especially in recent years constantly frequent operational risk, covering almost all business sectors, business type and processes. So how to measure and manage operational risk, has great significance for the banking industry. At present, many studies have focused on the domestic operation qualitative risk management, quantitative research on operational risk management is still in its infancy, and lack of a system and complete and more suitable operational risk measurement methods for Chinese banks. Accurately measure the operational risk is the premise and basis of effective management of operational risk, therefore, to deep research the operational risk measurement and develop the establishment of our country’s bank operational risk measurement methods, is one of China’s banking practical and urgent task.Loss distribution approach is currently main method of measure the operational risk, and is also more realistic and reasonable to measure the operational risk capital for China’s banking industry. In this paper, we use the Agricultural Bank of China operational risk loss data for the2000-2009sample, select the loss distribution approach as the research object, in the case under consideration of concealing the actual loss after the operational risk, the Commercial bank would safeguard their reputation, we proposed the model based on compound Poisson-Geometric of loss distribution approach to measure the bank’s operational risk. Through this study, we provide useful lessons in the quantitative analysis of operational risks for our bank, and improve and develop the loss distribution in operational risk measurement, that has important theoretical and practical significance.First, When fitting the frequency of loss function,we compared the compound Poisson-Geometric Poisson distribution with model of Poisson distribution which is widely used in the industry,and through the generalized likelihood ratio test, showed that the frequency of the Agricultural Bank of China lost operational risk distributions obey compound Poisson-Geometric distribution losses instead of the traditional distribution method of the Poisson distribution, indicating that the loss distribution approach model based on compound Poisson-Geometric distribution will be able to measure operational risk Agricultural Bank of China more accurately.Second.When fitting the loss intensity distribution function,we use P-P Figure fitting, K-S test to find that the loss intensity distribution obey the lognormal distribution; Finally, we use the Monte Carlo method to simulate the Agricultural Bank of China2000-2009in operational risk loss data’s VaR, and use this model to estimate the Agricultural Bank of China operational risk economic capital, found that the loss distribution approach based on Poisson distribution models underestimate the operational risk. |
