Study on Credit Card Risk of Commercial Banks in Prefectural-Level Banks Based on Logistic Model

• 作者: Yuwen Deng
• 来源期刊: 经济管理学刊:中英文版
• 学科: 经济
• 关键词: CREDIT Risk Logistic MODEL SCORING MODEL Prefecture-Level Bank
• 年，卷（期）: 2018,（1）
• 页码范围: 69-80
• 页数: 12页
• 分类号: F

摘要

Substantial income has been brought to the bank since credit card business entering the Chinese market,giving rise to a growing number of credit card issuers and more frequent transactions,which also improves convenience of cardholders.However,the booming market of credit card caused a series of credit risk.Credit risk in commercial banks and medium-sized banks in prefectural-level city is an operational risk that cannot be ignored,if not properly handled;it will exacerbate risk control pressure.Credit card risk mainly concerns default risk,and factors causing breach varied.Credit card risk can exist in the whole process,including the customers?application for credit card,card insuring,and transaction settlement.Finding an effective way to identify a variety of credit card risk,and developing a complete and efficient monitoring system to reduce the risk of credit loss is essential for large commercial banks.In terms of smaller-scale banks in prefecture-level city,a credit scoring system to evaluate the customer's credit ability is particularly important.Dataset in this paper mainly comes from a prefecture-level city bank,and the information is anonymous and authentic.This paper starts with the more than 700 customer data of a prefecture-level city bank and comprehensively considers the status quo of credit card development in China's commercial banks and successful domestic and foreign credit risk management experiences,followed with the causes and characteristics of credit card risks,solutions,and proposals,systematically expounding credit card business risk management.This article adopts the Logistic model and the credit scoring model.Through the screening and analysis of dozens of customer's characteristic variables and the use of various commands of statistical software,a prediction model of customer default probability will be constructed.At the same time,a scoring model was introduced to set the threshold for issuing cards in a quantifiable manner to help banks predict the possibility of customer default be