Credit risk is the oldest and the most important risk for banks. It's a very difficult problem to measure quantitative credit risk for banks all over the world, especially for banks of China, which have joined WTO recently. Monte Carlo simulation is a kind of method of statistical simulation experiment. With the development of operation power of computer, this method is more and more used in varies scientific fields. This article attempt to research how to measure credit risk of banks using Monte Carlo simulation.Firstly, this article summarizes the fundamental idea and general steps of Monte Carlo simulation. The key point is how to generate random numbers or random variables. This article uses the professional software, Crystal Ball, to generate them.Secondly, this article summarizes the definition, features and measurement models of credit risk. With expensively reading and research, the writer sums up the following contents: the inputs, process and outputs of measurement models of credit risk; the different methods of identification of risk factors in different models.Thirdly, this article sum up the aspects of using Monte Carlo method in credit risk measurement models: generating random numbers of risk factors, simulating distribution of combinatorial losses and computing VaR, prove parameterized models, using in single-variable time series model.Finally, the last chapter is the empirical part of the article. The difficulty of the empirical research is the lack of historical data. Based on the analysis of foreign credit risk models, the writer use the simplified time series model to forecast the ratio of non-performing Loan of Minsheng Bank. In the end of the chapter, the article gives some advises on further researches. |