Development And Use Of Fraud Prevention Strategies And Bankcard Credit Scoring Models Within The Taiwan Financial Market

This paper applied logistic regression model to cross check between fraud case and credit card scoring system. Results show that several variables have significance influence to normal and default cards, those variables include sex, age, income, education, marriage, occupation etc.Significant financial events in recent years globally, especially financial tsunami occurring in2008forced so many financial institutions facing bankruptcy or merger and acquisition mostly due to the huge bad debt losses. In order to achieve the economic scale as well as to reduce the operating cost resulting in higher profit, the banks purposely relaxed the audit of the financial condition of the credit card applicants. However, the more relaxed review of credit conditions effectively improved card approving rates, and so, it will seriously affect the credit quality. As a result, the banks were easily to endure adversities in the increase in overdue delinquent loans. Operating costs are too expensive if the basis for the Bank Nonperforming loans, delinquent loans and doubtful way of handling procedures for handling, and resulting in higher collection costs, it is not cost-effective to financial institutions. Making the original high issuance volume using additional high card-holder spending amount to bring high profit is impracticable. The consequence is not only failed to increase profits, but also spend extra cost, it was not subordinate to profit.Credit review for credit card-holder profile is the first line of defense against bad debt, not only can apply to filter poor credit condition, to prevent cases of deliberated fraud, as well as to discover the healthy consumer to increase profit. Given to the most suitable credit limit it can make good use of the credit reviews in order to bring the credit card business profitable.The credit card business is a financial product that depends highly on human and system analysis. Specifically, credit card personal credit score is a comprehensive assessment, making use of data, statistical quantification techniques to build up credit scoring models. In addition, it would be on behalf of an individual to fulfill the possibility repayment obligations in the future to a certain score.After investigation, we find that gender, age, annual income, educational attainment, marital status, titles, conduct industry do has the significance influence on the difference between abnormal with the default cards and credit card normal piece. In this article, we use regression model to conduct and analysis a Taiwan’s credit card-issuing institution in2011’s data, including gender as female students year age more long, the higher the annual income, educational process level the higher, marital status as married, the higher the titles identity, industry not as a soldier, Catholic personnel, go to ways for banks VIP by the cardholder default risk would be the lowest in all segments.