Application Research Of CreditMetrics Model Based On Copula Improvement

Credit risk is main risk faced by commercial banks. Since frequent occurance of contemporary financial crisis, commercial banks, as the important link in the financial field, have gradually shifted focus from traditional balance of risk management to comprehensive risk management on quantitative and optimized risk. Credit risk management as an important component of comprehensive credit has become increasingly concerned by international financial community. Credit risk management of commercial bank in China is largely stayed in the qualitative analysis, and lags behind the international advanced level. After learning from some related research and application, and considering the specific characteristics of domestic commercial banks, writer will attempt to introduce quantitative credit risk management to commercial banks in China.Beginning with the development process of risk management theory, this thesis introduces requirements on commercial banks quantifying credit risk from comprehensive risk management theory according to "The New Basel Capital Accord", which is compose of three pillars include minimum capital requirements, regulatory supervision and market discipline. It also analyzes current situation of commercial banks through introduction of four traditional credit risk management, and notes that the trend of modern credit risk models.In respect of the model, taken principle of calculation, classification standards, data availability and market maturity as the main reference,this thesis compares four models include CreditMetrics, KMV, Credit Portfolio View, CreditRisk in detail and analyzes their applicability in accordance with usage of IRB when you need to build econometric models and related technical procedures.It recommends that CreditMetrics model, at this stage, is more applicable to the credit risk management of commercial banks than the other three.Introduced the basic framework and input parameters of CreditMetrics, this thesis discusses inadequacy and improved way of the model. This thesis chooses Copula function to improve correlation module of CreditMetrics model in order that the model will have better flexibility and maneuverability. Given the characteristics of the data, writer selects a variety of bivariate two parameters Copula for simulation. In addition, considering the actual development conditions of financial market in China, writer adjusts input parameters include loss given default, credit migration matrix and the forward yield curve.The thesis selects the specific credit data of commercial bank as the object of empirical research, and makes the empirical research of single loan and portfolio loans respectively through improved CreditMetrics model. It also makes comparison on value at risk between improved empirical research and unimproved. Moreover, the thesis tells not only advantage of improved model but also suggestions on quantified credit risk management of commercial bank in China.In summary, this thesis conducts in-depth analysis on the feasibility of CreditMetrics Model in China, overall credit risk measurement framework of model, improvement of model parameters, selection and introduction of Copula, specific application of model and related measures of credit risk management of commercial banks, etc. in order to make a useful exploration for quantified credit risk management of commercial banks in China.