I 摘要 2019年，世界金融市场增势放缓，全球信贷业务走势分化，国内信贷业务有 所收紧。一方面，人民银行印发《金融科技(FinTech)发展规划(2019-2021年)》继 续鼓励地方政府与金融机构深化和发展大数据技术，提升信贷市场的内生动力。另 一方面，多地政府陆续开展关于高利息、暴力催收、侵犯个人隐私等社会乱象的整 治工作。以数据驱动为发展思路的信贷业务新业态正在逐渐成为金融企业转型的 必要路径。如何在信贷业务的各个环节中合理布局大数据技术，成为了传统金融企 业转型的首要难题。当具备了一定的大数据技术，如何科学设计、合理优化原有的 业务模式成为了传统金融企业业务突破的关键瓶颈。 M金融机构成立于2015年，是一家由中国银保监会批准成立的消费金融公司。 自成立之初，M金融机构即确立了“数据驱动”的发展战略，并持续进行大数据技 术的研发投入，长期进行大数据信贷业务的模式探索。截至2019年，M金融机构 通过实践积累了诸多的案例，积累了诸多成功运营的大数据信贷产品的经验，总结 出值得借鉴的大数据信贷业务模式设计思路。首先，本文在梳理了全球宏观金融环 境趋势变化、国内外金融大数据行业发展现状的基础上，明晰了大数据技术和信贷 业务的相关理论。接下来，本文通过对M金融机构的公司背景、战略规划、业务 发展规模、产品矩阵的部署进行解读。其次，本文对信贷业务中的营销环节、信贷 生命周期进行拆解，分析大数据技术在风险识别、风险评价、风险管理的应用。 医疗整形美容消费分期贷款业务（简称“医美分期”）是本文进行分析和论证 的核心案例。本文深度复盘了M金融机构探索“医美分期”业务模式的整个过程， 比较了不同技术手段下“医美分期”业务贷款风险的特征与防范手段。本文对“医 美分期”业务的商业模式进行了剖析，对业务开展的各个流程中风险防范的具体手 段进行了归纳和梳理。最后，本文结合M金融机构2015年至2019年间的财务数 据、交易数据、业务数据，对M金融机构进行业务规模评价、客群分析、财务报 表分析，对其业务产生的信贷资产进行贷款质量分析。从而论证了大数据技术的部 署为M金融机构的经营成果带来正向帮助。作者提出，金融企业布局大数据技术 应遵循划分环节、小范围试错、逐步替换的思路，并充分考虑风险与收益的平衡性。 设计大数据信贷业务模式应严格区分数据源和业务场景，以数据为决策的驱动力、 模式与流程的制衡力来识别风险、评价风险、管理风险。本文为经营信贷业务的传 统金融企业转型数据驱动型企业提供参照物和建议。 关键词：信贷业务，大数据技术，业务模式，风险收益，医美分期 ABSTRACT II ABSTRACT In 2019, the world financial market grows slowly; global credit business emerges divergent trends, and domestic credit business has going through constriction. On the one hand, for increasing inner impetus of credit market People’s Bank of China issued Fin Tech Development Plan (2019-2021), which continuously encourages local governments and financial organizations to push and deepen the development of big data technology. On the other hand, many governments, one after the other, carried out a series of rectification work on high interest rates, violent collection, invasion of privacy and other social chaos. Data-driven credit business has gradually become the necessary path of traditional financial enterprises’ transformations. How to allocate big data technology in every link of credit business reasonably is the first challenge of transformation; while after application of big data technology, how to scientifically design and reasonably optimize original business models becomes the key bottleneck of business breakthrough. Financial Institution M established in 2015, which is a consumer finance company ratified by China Banking and Insurance Regulatory Commission. At very beginning of establishment, financial institution M set “Data-driven” as its development strategy, and it continuously put effort in big data technology R&D, exploring data-driven credit business models for a long time. Generally speaking, this article sorted related theories of big data technology and credit business based on trends and changes of macroscopic financial environment , besides domestic and international circumstance of financial big data industry. Specifically, this article introduced background information, strategic planning, development scale, and product line deployment of financial institution M, clearly demonstrated universality as well as uniqueness of cases used. Furthermore, this article dissected marketing section, credit life cycle (pre, mid, and post) of credit business, analyzing and evaluating benefits of big data technology applications on risk identification, risk assessment, and risk management. Instalment business of medical cosmetology （IBMC） is the core case for analysis and demonstration in this paper. Firstly, this paper comprehensively replied the whole procedure about how Financial Institution M explored business model of medical cosmetology, as well as the characteristics and risk-prevision measures of IBMC under different technical means. Secondly, this paper analyzed the business model of IBMC, ABSTRACT III and summarized specific measures for risk-prevention in every stage of business conduction. Last but not the least, based on financial data、transaction data、business data from 2015 to 2019, this paper conducted business scale evaluation、costumer analysis、 financial analysis, and loan quality analysis which produced by its credit assets of Financial Institution M. With all above it is demonstrated that the deployment of big data technology brought positive influence to business performance of Financial Institution M. The author raised that except give full consideration to the balance between risks and benefits, the layout of big data technology should follow the ideas of dividing links, trial and error on a small scale, and gradual replacement. Designing data-driven credit business models should strictly distinguish data sources and business scenarios; identify, evaluate, and manage risks based on data-driven decisions, models, as well as checks and balances of procedure. This article provides references and suggestions for traditional financial enterprises engaged in credit business that wants to transform into data-driven enterprises. Key words: Credit Operations, Big Data, business models, Risk Management 目录 IV 目录 第一章 绪论 ..................................................................................................................... 1 1.1 研究的背景 ......................................................................................................... 1 1.2 国内外研究现状 ................................................................................................. 2 1.2.1 国内大数据赋能的信贷业务发展现状 ................................................... 2 1.2.2 国外大数据赋能的信贷业务发展现状 ................................................... 4 1.3 研究的目的和意义 ............................................................................................. 5 1.4 论文可能的创新点与局限 ................................................................................. 5 1.4 总结 ..................................................................................................................... 6 第二章 金融大数据概述与信贷业务相关理论 ............................................................. 7 2.1 金融大数据的定义 ............................................................................................. 7 2.2 大数据技术发展的简史 ..................................................................................... 7 2.3 大数据赋能的贷业务模式分类 ......................................................................... 9 2.3.1 按主体划分 ............................................................................................... 9 2.3.2 按产品业务划分 ..................................................................................... 10 2.4 大数据信贷业务相关理论 ............................................................................... 11 2.4.1 互联网金融产品相关理论 ..................................................................... 11 2.4.2 金融大数据技术相关理论 ..................................................................... 12 2.4.3 信贷业务相关理论 ................................................................................. 14 2.5 总结 ....................................................