商业银行信用卡动态额度管理研究

发布时间：2018-06-01 06:46

本文选题：动态额度 + 行为评分　；参考：《中国海洋大学》2013年硕士论文

【摘要】：自1985年我国发行第一张信用卡“中银卡”以来,信用卡发卡规模不断扩大,授信额度急速增长,银行收益逐步显现,同时银行将面临着巨大的信用卡风险。近些年信用卡业务更是高速发展,通过中国人民银行公布的信用卡相关风险指标数据可知,信用卡风险已经开始凸显,然而银行并没有有效的风险防范措施和手段,跟不上业务发展速度。目前我国信用卡行业的风险控制主要集中在发卡阶段,这是由我国当前以信用卡营销为主,过多关注事前信用卡额度审批的信用额度情况决定的,而在银行发卡以后,持卡人的信用消费情况、还款情况以及资信状况的后续跟踪调查却很少关注。其实银行的风险控制任务并没有完成,信用卡的账户不能放任自流、自生自灭,恰恰相反,对信用卡账户进行持续性的、科学性的、全方位的管理,将有利于银行提高持卡者的额度使用率、保持和客户沟通交流、增加客户忠诚度、防止客户的流失等,从而控制信用卡风险,增加银行的盈利性。而本文正是从信用卡管理核心阶段—动态额度管理入手,在分析国内外信用卡动态额度管理的基础上,考虑我国银行业信用卡发展状况以及经济环境变化等因素下构建了一套更加合理的信用卡动态额度管理指标体系。以期达到,在微观层面,给予客户更高的信用额度,以促进信贷资产的增长和收益的提高、提高信用卡的市场竞争力、巩固客户的忠诚度,降低信用额度过低而导致的客户流失；同时,合理的降低信用卡的额度,可以减少潜在的坏账准备,控制风险,并可以降低信用卡里的“睡眠卡”,减少过度授信所带来的资本消耗越。在宏观层面,可以促进银行信用卡业务的发展、降低银行风险、提高银行业的竞争力,从而对整个银行业的发展起到促进作用。在对指标体系进行评价时,本文通过对各种方法的分析,最终选取BP神经网络模型。并在获得银行内部相关数据的基础上,运用BP神经网络模型对指标体系进行了可行性分析,达到了良好的效果。在我国商业银行的信用卡业务开始从量变到质变转化过程开始时,本文提出对信用卡动态额度进行更加有效的管理,通过选取持卡者基本信息、消费行为信息、银行内部状况以及宏观经济变量四方面指标构建了信用卡动态额度管理指标体系,相比于当前信用卡动态额度管理中仅以行为评分模型中客户消费行为指标进行评价,指标体系更加的全面性,评价更具有科学性、合理性。并且在模型评价时,不同于以分值划分客户等级,而是采用BP神经网络方法对信用卡动态额度管理指标体系进行评价,其高度并行计算能力、自学能力和容错能力使评价更科学更准确。
[Abstract]:Since China issued the first credit card "Bank of China" in 1985, the scale of issuing credit card has been continuously expanded, the credit line has increased rapidly, the bank income has gradually appeared, and the bank will be faced with huge credit card risks. In recent years, credit card business has been developing at a high speed. According to the credit card risk indicators published by the people's Bank of China, credit card risks have begun to highlight, but the banks do not have effective risk prevention measures and means. We can't keep up with the speed of business development. At present, the risk control of the credit card industry in our country is mainly focused on the credit card issuing stage, which is determined by the credit limit approved in advance by the credit card marketing in our country, and after the bank has issued the card, Follow-up investigations into cardholders' credit consumption, repayment and creditworthiness have paid little attention. In fact, the risk control task of banks has not been completed. Credit card accounts cannot be left alone. On the contrary, credit card accounts are managed continuously, scientifically, and comprehensively. It will help banks to improve the quota utilization rate of card holders, maintain communication with customers, increase customer loyalty, prevent customer loss, and thus control credit card risks and increase the profitability of banks. This paper starts from the core stage of credit card management-dynamic quota management, on the basis of analyzing the domestic and foreign credit card dynamic quota management. A more reasonable index system of credit card dynamic quota management is constructed under the consideration of the development of credit card and the change of economic environment in the banking industry of our country. In order to achieve, at the micro level, to give customers a higher credit line, in order to promote the growth of credit assets and earnings, improve the market competitiveness of credit cards, consolidate customer loyalty, Reduce customer losses caused by low credit lines; at the same time, reduce credit card quotas reasonably, reduce potential bad debt preparation, control risk, and reduce "sleep cards" in credit cards, Reduce the capital consumption caused by excessive credit. At the macro level, it can promote the development of bank credit card business, reduce the bank risk, improve the competitiveness of the banking industry, thus promoting the development of the whole banking industry. In the evaluation of the index system, through the analysis of various methods, the BP neural network model is finally selected. On the basis of obtaining the internal data of the bank, the feasibility of the index system is analyzed by using BP neural network model, and good results are achieved. When the credit card business of commercial banks in our country begins to change from quantity to quality, this paper puts forward a more effective management of the dynamic credit card quota, through selecting the basic information of the cardholder and the information of consumption behavior. The index system of credit card dynamic quota management is constructed from four aspects of bank internal condition and macroeconomic variables. Compared with the current credit card dynamic quota management, the index of customer consumption behavior is evaluated only in the behavior score model. The index system is more comprehensive and the evaluation is more scientific and reasonable. And in the evaluation of the model, it is different from dividing the customer grade by the score, but using the BP neural network method to evaluate the index system of credit card dynamic quota management, and its high parallel computing ability. Self-learning and fault tolerance make evaluation more scientific and accurate.
【学位授予单位】：中国海洋大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.2;TP18

【参考文献】