一、评价依据与理论来源
This algorithm does not presuppose a single standard of a 'good financial institution'; instead, it synthesizes modern financial theory, regulatory frameworks and international evaluation practice and maps them onto the model's four dimensions:
本算法不预设单一的"好金融机构"标准,而是综合现代金融理论、监管框架与国际评价实践,将其映射到模型的四个维度:
- 金融功能观(Merton & Bodie,1995):金融体系的本质在于其履行的资源配置、风险分散、支付清算等经济功能,机构形态只是载体——为"贡献维度(服务实体经济)"提供理论根基。
- 巴塞尔协议(Basel III):以资本充足率、杠杆率、流动性覆盖率等指标刻画银行经营的稳健性与质量——为"质量维度"提供监管依据。
- 金融稳定理事会(FSB)系统重要性金融机构评估:综合规模、复杂度、关联度、可替代性等指标——为"规模之外多维评价"提供国际参照。
- G20 普惠金融全球合作伙伴(GPFI)指标体系:以金融服务的可得性、使用情况与质量衡量金融体系的包容性——为"贡献维度(普惠金融)"提供方法依据。
- 金融中介理论(Gurley & Shaw)与信息不对称理论(Stiglitz):金融机构通过规模与网络降低交易成本、缓解信息不对称——为"规模维度"提供经济学依据。
- 金融科技(FinTech)与数字金融研究:科技投入与数字化转型重塑金融机构的效率、风控与服务边界——为"创新维度"提供依据。
二、模型的数学结构
The composite index adopts a 'range-standardization + weighted summation' model. Each indicator is first mapped onto the [0,1] interval by min-max normalization; large-span scale indicators such as revenue, total assets, market value and real-economy funding deployment are first log-transformed and then normalized, to avoid head institutions monopolizing the scores:
综合指数采用极差标准化 + 加权求和模型。各指标先经 min-max 归一化映射到 [0,1] 区间;营收、总资产、市值、服务实体投放等大跨度规模类指标先取对数再归一化,避免头部机构垄断分值:
指标′ = (指标 − 指标min) / (指标max − 指标min)
规模 S = 0.5·log₁₀(营收)′ + 0.5·log₁₀(总资产)′
质量 Q = 0.5·净利润′ + 0.5·log₁₀(市值)′
创新 I = 金融科技投入′
贡献 C = 0.5·log₁₀(服务实体投放)′ + 0.5·普惠金融′
综合指数 = 100 × (0.30·S + 0.30·Q + 0.15·I + 0.25·C)
The four weights always sum to 1.0. Scale and Quality each account for 0.30, forming the basic foundation of 'volume and substance'; Contribution accounts for 0.25, reflecting CPPPS's tilt toward 'serving the real economy' (a dimension generally absent from international rankings); and Innovation accounts for 0.15, measuring future-oriented fintech capability.
四项权重之和恒为 1.0。规模与质量各占 0.30,构成"体量与成色"的基本盘;贡献占 0.25,体现 CPPPS 对"服务实体经济"的倾斜(国际榜单普遍缺失该维度);创新占 0.15,衡量面向未来的金融科技能力。
三、指标与权重依据
| 维度 | 指标 | 处理 | 维度权重 | 依据 |
| 规模 | 营业收入 | 取对数 | 0.30 | 金融中介规模经济;Gurley & Shaw |
| 总资产 | 取对数 | 金融中介理论;资产规模 |
| 质量 | 净利润 | 线性 | 0.30 | 盈利能力;Basel III 稳健性 |
| 市值 | 取对数 | 市场定价与股东价值 |
| 创新 | 金融科技投入 | 线性 | 0.15 | 金融科技/数字金融研究 |
| 贡献 | 服务实体经济投放 | 取对数 | 0.25 | 金融功能观;资源配置功能 |
| 普惠金融 | 线性 | G20 GPFI 普惠金融指标 |
The logic of the weight design: Scale and Quality are treated equally (0.30 each), avoiding both 'scale-only thinking' and 'hollow bloat' divorced from operational volume; Contribution (0.25) is weighted higher than Innovation (0.15), reflecting CPPPS's value orientation that finance should serve the real economy—the 'contribution' of finance is first and foremost serving the real economy, and only secondarily its own technological evolution.
权重设计逻辑:规模与质量并重(各 0.30),既避免"唯规模论",也避免脱离经营体量的"虚胖";贡献(0.25)权重高于创新(0.15),体现 CPPPS 对金融服务实体经济的价值取向——金融的"贡献"首先是服务实体,其次才是自身的技术进化。
四、数据来源与局限性
Operating revenue, total assets, net profit, market value, fintech investment, real-economy funding deployment and financial-inclusion scores all adopt public-knowledge approximations: mainly referencing the annual reports of banks, insurers and securities firms, regulatory disclosures of the People's Bank of China and the National Financial Regulatory Administration, and industry statistics of the China Banking Association, the Insurance Association of China and the Securities Association of China. All approximations are uniformly marked 'subject to official figures'. Market value is approximated in the combined A+H caliber.
营业收入、总资产、净利润、市值、金融科技投入、服务实体投放与普惠金融评分均采用公开常识近似值:主要参照各银行、保险、券商年度报告,中国人民银行、国家金融监督管理总局监管披露,以及中国银行业协会、保险行业协会、证券业协会的行业统计。所有近似值一律标注"以官方为准"。市值按 A+H 合计口径近似。
This model is structurally transparent and reproducible, but has limitations: (1) banks, insurers and securities firms differ greatly in business structure, and cross-industry comparability is constrained by caliber approximations; (2) financial inclusion is a composite approximate score with some subjectivity; (3) risk-control and compliance indicators such as capital adequacy and non-performing-loan ratio, and the full ESG dimension, are not included; (4) institutions differ in their disclosure calibers for fintech investment. The ranking is for research reference only and does not constitute any investment advice.
本模型结构透明、可复现,但存在局限:①银行、保险、券商三类机构业务结构差异大,跨行业可比性受制于口径近似;②普惠金融为综合近似评分,含一定主观性;③未纳入资本充足率、不良率等风控与合规指标,也未纳入 ESG 完整维度;④金融科技投入各机构披露口径不一。排名仅供研究参考,不构成任何投资建议。
五、参考文献
- Merton, R. C., & Bodie, Z.(1995):"A Conceptual Framework for Analyzing the Financial Environment",载《The Global Financial System: A Functional Perspective》,Harvard Business School Press。
- 巴塞尔银行监管委员会(BCBS):《Basel III: A Global Regulatory Framework for More Resilient Banks and Banking Systems》。
- 金融稳定理事会(FSB):《Global Systemically Important Financial Institutions: Assessment Methodology》。
- G20 普惠金融全球合作伙伴(GPFI):《G20 Financial Inclusion Indicators》。
- Gurley, J., & Shaw, E.(1960):《Money in a Theory of Finance》,Brookings Institution。
- Stiglitz, J., & Weiss, A.(1981):"Credit Rationing in Markets with Imperfect Information",American Economic Review。
- 中国人民银行:《金融机构贷款投向统计报告》(普惠金融/绿色金融口径)。
- 国家金融监督管理总局:银行业、保险业主要监管指标及统计数据。
- 中国银行业协会、中国保险行业协会、中国证券业协会:年度行业报告。
- 相关金融机构年度报告及公开财务披露。