Design and Implementation of Retail Assets Credit Risk Measurement System in Commercial Banks
|Keywords||Retail Assets Credit Risks Decision Tree Model|
Due to the high profit margin and risk diversification, retail credit business has become the focus of commercial banks’business development. However, with the rapid development of business, various risk controls faced by retail credit products (credit risks in particular) has become the important issue of commercial banks.Retail credit business has the following characteristics:a large amount of customer data, small volume of a single customer’s business, high product standardization, long-term potentiality, and with many uncertain factors. Therefore for statistical purpose, banks may use pool assignments of risks to measure retail credit risk exposure.This paper calculates the risk characteristic variables based on the theories of data warehousing and data mining, and the scorecards results (such as application scoring and behavior scoring) and the basic customer information. By using the risk characteristic variables calculated and the decision tree model, this paper conducts pool assignments of risks of retail assets. According to the retail assets in each risk pool, the paper consolidates the statistics, and calculates the corresponding probability of default (PD), loss given default (LGD) and exposure at default (EAD) of each risk pool. The system utilizes TERADATA as the data warehouse platform and adopts J2EE, the enterprise application architecture, to ensure the system has strong data storage and processing capacity and good scalability.This paper focuses on the discussion of the objectives of commercial banks’ credit risk measurement systems. Based on the model of pool assignment of risks, the paper demonstrates the logical structure, technical structure, physical structure and data structure of the system. The system design describes in detail the calculation of risk characteristic variables of personal loans, credit cards, and small and micro businesses’loans, as well as the division of risk pools and the management of the model of pool assignments of risks. The system realizes the establishment of the retail credit asset risk measurement model, and the risk measurement of retail assets by implementing the Internal Ratings-Based Approach introduced by the New Basel Capital Accord. By achieving the management of credit risk measurement models, variable parameter and model parameter of retail assets, and the dynamic changes of the measurement models and the results, the retail asset credit risk measurement system fulfills commercial banks’requirement of credit risk measurement of retail assets.