Credit Card Risk Management Techniques And Profit Forecasting Model

In recent years, the rapid development of credit card business has brought tremendous business opportunities to banks and other credit card issuers, but also exposes the business of asymmetric information, moral risk to credit risk, transition of consumption, high rate of sleep card. In order to effectively control the risk, the credit card risk management technology is essentialed. Overseas risk management technology is relatively mature, but the domestic development is relatively backward. Applying the foreign techniques directly is not applicable. In this paper, basing on the perspective of domestic credit card, we research risk management techniques of credit card for reducing the loss of credit card issuers. The main objective of this paper is the classification of credit card customers as accurately as possible, through using of risk management, so that the probability of error is as small as possible.The article combinates the advantages of two methods, which are decision tree model and the logistic regression model. It puts forward the two orders of logistic regression model and the mixture model of decision tree and logistic regression.Then the paper uses real credit card data to do the empirical study. In addition, in order to improve the accuracy of classification of credit card customer, we conduct the secondary discrimination. To predict the risk and the income of credit card business, the article establishes a predictive model of losses and earnings. Then we use an example to compare with the actual gain.