普惠金融中关系评分的信息含量研究——来自中小商业银行线上快贷平台的证据
DOI:
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:


The information content of relationship scores under inclusive finance:Evidence from an online lending platform affiliated with a rural commercial bank
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    有效识别长尾群体的信用风险,是商业银行推动普惠金融发展的关键问题.本研究基于湖南省某商业银行线上快贷平台数据,评估关系评分的信息含量及对长尾群体信用风险评估的影响机制.实证结果表明,关系信息提高长尾群体信用评估精度,其作用来源于关系信息产生的约束机制.同时,关系信息可弥补传统信用数据缺乏,识别借款人风险偏好,提高违约判断信息含量.研究结论揭示银行关系信息在“普惠金融+数字金融”背景下的影响机制与作用方式,为提高我国金融高质量发展提供现实支撑.

    Abstract:

    Identifying the credit risk of borrowers residing in lower-tier cities and rural areas remains a critical challenge for commercial banks to improve inclusive finance. Based on the data of an online lending platform affiliated with a commercial bank in Hunan Province, this study evaluates the information content of relationship score and its impact on the credit risk assessment. Empirical results demonstrate that relational information significantly enhances the accuracy of credit risk assessments for borrowers located in lower-tier cities and rural areas, and this effect operates primarily through a constraint mechanism. At the same time, relational information enhances the information content of credit risk assessment by mitigating gaps in traditional credit data and revealing borrowers risk preferences. This study reveals the influence mechanism and operational pathways of bank relational information in the context of “inclusive finance+digital finance”, which provides support for advancing high-quality development in Chinese financial sector.

    参考文献
    相似文献
    引证文献
引用本文

陈收,刁安琪,黄宇漩,朱琦,杨晓临.普惠金融中关系评分的信息含量研究——来自中小商业银行线上快贷平台的证据[J].管理科学学报,2026,(5):176~190

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2026-05-29
  • 出版日期:
您是第位访问者
管理科学学报 ® 2026 版权所有
通讯地址:天津市南开区卫津路92号天津大学第25教学楼A座908室 邮编:300072
联系电话/传真:022-27403197 电子信箱:jmsc@tju.edu.cn