互联网金融背景下“软信息”对于违约行为的影响——基于Logit模型

发布时间：2018-10-08 15:34

【摘要】：互联网金融背景下,网络平台积累了越来越多的用户行为信息,而且这些信息与借款人提供的传统信息之间存在较大差异。如何有效地利用这些信息来降低借款人与投资人之间信息的不对称,对于提高风险控制水平、预判违约行为发生的可能性具有重要意义。借助拍拍贷平台提供的海量数据,从"软信息"的角度出发,探索其与借款人违约行为的关系。实证分析结果表明:当借款人在申请借款时修改信息的数目高于12项时,违约的机会概率会提高;当借款人在贷款成交之前修改内容的天数越多、修改时距离成交的日期越近时,违约的机会概率相应会越高。而当借款人最近3个月在社交网络存在去欧洲或者美国旅游的信息,或者填写借款信息的完整程度较高时,相应的违约机会概率则较低。
[Abstract]:Under the background of Internet finance, more and more user behavior information has been accumulated on the network platform, and there is a great difference between these information and the traditional information provided by the borrower. How to use these information effectively to reduce the asymmetry of information between borrower and investor is of great significance to improve the risk control level and predict the possibility of default. With the help of the massive data provided by PPDAI platform, this paper explores the relationship between the soft information and the borrower's breach of contract from the point of view of "soft information". The results of empirical analysis show that the probability of default increases when the borrower modifies more than 12 items of information when applying for the loan, and the more days the borrower modifies the content before the loan transaction. The closer the date of the transaction, the higher the chance of default. When borrowers have been traveling to Europe or the United States on social networks for the last three months, or filling in more complete information on borrowing, the probability of default is lower.
【作者单位】：北京工商大学经济学院;北京工商大学中国食品安全研究中心;
【基金】：国家社科基金重大项目:粮食安全目标下市场起决定作用的粮食价格形成机制研究(14ZDA034) 北京市社科基金项目:北京文化创意产业发展效应研究(15JGA031);北京市社科基金项目:北京市出租车市场管制下的租金耗散与管制改革(14JGB029) 2016年研究生科研能力提升计划项目资助
【分类号】：F832.4;F724.6

【相似文献】