ABSTRACT
Recently there is a worldwide increase in the operational risk frequency of the commercial banks, leading the bank industry and the academic field to pay great attention to the operational risk management. Measurement of operational risk is the core of the operational risk management work of commercial banks. Therefore, many financial institutions are sparing no eort to develop the tools and methods for operational risk measurement. The Basel committee formally divided the calculating method of operational risk into Basic Index Approach (BIA), Standard Approach (SA), and Advanced Measuring Approach (AMA) in the New Capital Accord issued in 2014. However, the practice has proved that the BIA and SA have lots of deficiencies. So the recent researches on the calculating of operational risk have been focusing on the typical AMA, giving rise to a quantity of operational risk measurement models. Peaks over threshold (POT), among all these methods, becomes an important approach for its extraordinary data sensitivity and expansibility. POT methods have many applications on measuring operational risk. Song and Turk [1] pointed out that when lacking the data and its heavy-tailed characteristic, POT methods can easily and reliably measure, predict, and manage the relative risk elements. Bin and Wu [2] used POT to simulate the loss distribution in the operational risks of Chinese commercial banks and diagnosed the simulation, and tested the feasibility of the operational risk by using POT methods.
