Bayesian Networks For Managing Operational Risk In China

This paper chooses the operational risk of commercial bank in China as research object, because comparing to credit risk and market risk the research of operational risk is becoming start. The lose from operational risk in China has became bigger and bigger recent years. So it is necessary for us to study operational risk management now. From the view of the result of research in China, most papers are about qualitative. Most papers are about lose distributing theory and extreme theory in quantitative. Because we are lack of losing data in operational risk, we can't use these methods to manage operational risk now. Based on this case, this paper uses Bayesian network with scene analysis in it to study it. Bayesian network is a kind of statistical model, and is an important tool to deal with uncertain information with lack of data.This paper proves the feasibility of using Bayesian network to research operational risk of commercial bank in China from the characters of operational risk and Bayesian network. Then I get the lose data which range is from 2000 to June 2006 from public media. I analyzed the main influencing factors to operational risk of commercial bank in China based on the data and designed the Bayesian network model. I put the result of data to the model, run the model getting the range of lose distribution. According to the characters of Bayesian network, this paper introduce the risk hypothesis analysis, speculating the estate probability of main influencing factors from result, dissertating using the Bayesian network method to manage operational risk by systemic. In the end, this paper gives some suggestions on operational risk management from Staff quality and bank control.