摘要
【目的/意义】P2P网贷从兴盛到衰亡的案例提示我们,前瞻性的金融监管非常重要。对此,金融情报学可望发挥“耳目、尖兵、参谋”作用。本文探索为P2P网贷构建风险情报体系,为互联网金融全面风险管理体系和监管框架奠定基础。【方法/过程】在构建风险情报测度、框架和模型等理论体系的基础上,本文侧重实证研究,构建覆盖上千家P2P网贷平台的PoC Demo,调集约5万CPU核进行大规模金融风险情报挖掘和金融超算,研制相对和绝对风险测度等情报产品,并作微观、中观、宏观分析研究。【结果/结论】研发实时监测预警全国百十万家互金主体的全国风险地图和可视化风险曲线等情报产品,具备可行性和有效性。我们呼吁构建这样由“大数据、大模型、大计算”情报范式驱动的国产安全、自主可控的金融大数据情报分析与全国金融情报超算“天眼”系统。【创新/局限】通过实证验证所提出的内嵌前瞻性,有一定科学性、系统性的互金风险情报测度和分析框架。局限:数据不易获得。
【Purpose/significance】The case of P2P Online Lending's rise and fall reminds us that forward-looking financial supervision is very important.In this regard,Financial Intelligence Studies is expected to play the role of"ears and eyes,spearhead and staff".This paper explores the construction of risk intelligence system for P2P online lending,and lays the foundation for the comprehensive risk management system and regulatory framework of Internet Finance.【Method/process】On the basis of theoretical foundations such as risk intelligence measurement,framework and model,this paper focuses on empirical research.It builds PoC Demo covering thousands of P2P platforms,mobilizes around 50000 CPU cores for large-scale financial risk intelligence mining and financial supercomputing,develops intelligence products such as relative and absolute risk measurement,and conducts micro,meso and macro analysis and research.【Result/conclusion】It is feasible and effective to develop national risk maps,visual risk curves and other intelligence products for real-time monitoring and early-warning of hundreds of thousands of Internet Finance entities nationwide.We call for building such a domestic safe,autonomous and controllable Financial Big Data Intelligence Analysis and National Financial Intelligence Supercomputing"TianYan"System driven by the Intelligence Studies Paradigm of"Big Data,Big Model and Big Computing".【Innovation/limitation】Used empirical research to verify the proposed Internet Financial Risk Intelligence Measurement and Analysis Framework,which is forward-looking embedded,and scientific and systematic.The limitation of this paper is that data is not easy to obtain.
作者
丁晓蔚
赵笑宇
丁毅杰
王献东
DING Xiao-wei;ZHAO Xiao-yu;DING Yi-jie;WANG Xian-dong(School of Information Management,Nanjing University,Nanjing 210023,China;Nanjing University Inclusive&Rural FinTech Research Center,Nanjing 210023,China)
出处
《情报科学》
CSSCI
北大核心
2022年第12期42-55,95,共15页
Information Science
基金
国家社会科学基金一般项目“基于大数据的情报分析在互联网金融风险防控场景中的应用研究”(19BTQ070)。
关键词
金融大数据
互联网金融
P2P网贷
情报超算
国家金融安全
financial big data
Internet finance
P2P online lending
intelligence supercomputing
national financial security