The Reconstruction Of CreditRisk+ Model Based On Variaion Of Macro-economy And Its Application In Commercial Bank

Credit risk is the main risks that confront commercial banks. Currently, credit loan business is the main businesses of Chinese commercial banks; net interest income occupies the major proportion of total incomes. By introspection of the recent international financial crisis, prevention of systemic risk has become the key point that banking pays attention to. Thus, from the viewpoint of macro-prudential supervision, it has significant theoretical and practical value to study on the credit risk of Chinese commercial banks. Owing to the CreditRisk+ model being the important financial models of credit risk measurement and economic capital management, and aiming to prevent systemic risk, by analyzing the major drawbacks of CreditRisk+ model, this dissertation intends to reconstruct this model, and study on implication of the new CreditRisk+ model to the economic capital management in Chinese commercial banks.First, this dissertation reviews relevant literature on basic principle of the CreditRisk+ model, and carefully compares the two methods of using CreditRisk+ model to calculate economic capital from multiple angles, so we acquire a comprehensive understanding of the basic principles of CreditRisk+ model. This dissertation also analyzes the internal relationship between CreditRisk+ model and economic capital management as well as the internal relationship between CreditRisk+ model and Basel Capital Accord.In order to prevent systemic risk, it is necessary to involve the fluctuation of macro-economy into the framework of CreditRisk+ model. Therefore, this dissertation analyzes three drawbacks of this model. First, because of affected by the fluctuation of macro-economy, the default probability of obligor may change on the duration of credit loan. However, it didn't give a definitely feasible interval of the value of default probability in the originally technical documents of CreditRisk+ model, so the applicability and reliability of this model can not been judged. Second, different sectors are correlated because of macro-economy, but the original CreditRisk+ model assumed that sectors are independent, which is inconsistent with reality. Third, the CreditRisk+ model assumes that the loss given default is a constant, which has not considered the variation of the collateral value in the wake of macro economy. Those two hypotheses reduced the calculation accuracy of CreditRisk+ model in credit risk measurement in consideration of macro-economy fluctuation.In response to the three drawbacks of Credit Risk+ model in the condition of macro-economy fluctation, this dissertation proposes revised method respectively.This dissertation analyzes the feasible interval of the value of default probability in the CreditRisk+ model. This dissertation gives a rigorous mathematical proof that the adoption of Poisson distribution as the debtor default even distribution in CreditRisk + model will lead to the calculated economic capital been overestimated compared with the actual level of risk of the loan portfolio. By means of Monte Carlo method and the sensitivity tests, this dissertation finds out the relationship between the value of the probability of default and the error ratio of the economic capital calculated by the CreditRisk + model, and obtains the important conclusion that in order to control the economic capital measurement error ratio at a certain range, the value of the default probability of debtor should not exceed a certain threshold.This dissertation discards the assumption that the sector risk factors are independent and admits that they are correlated. Considering that sector risk factors are affected by the fluctuation of macro-economy, this dissertation introduces multi-system risk factors, and denotes the shape parameter of sector risk factor by a product of linear combination of systematic risk factors and a parameter that reflects inner character of sector risk factor, and adopts general covariance matrix of sector risk factors into the framework of CreditRisk+ model, and ultimately proposes the multi-systematic risk factors CreditRisk+ model based on sector character (abbreviate MS-CreditRisk+ model).This dissertation discards the assumption that the loss given default is a constant. On the basis of MS-CreditRisk+ model, by bringing the variation of loss given default into the model, the dissertation proposes a new model (abbreviate VL-MS-CreditRisk+ model) which can adopt the correlation of the sector risk factors, and also considered the variation of the loss given default, so this new model is consistent with the reality of commercial banks, and provides a feasible method for economic capital calculation of loan portfolio in the condition of the macro-economy fluctuation.To explain the application of VL-MS-CreditRisk+ in economic capital management of Chinese commercial banks, this dissertation gives a case study used loan data of a city commercial bank as a sample. According to the character of this model, this paper makes further research from four aspects on how to prevent system risk.