一、会计师事务所的多维定义与评价依据
This algorithm does not equate accounting firms with a "revenue ranking"; instead, drawing on classic theories of auditing and market regulation, it maps them onto four dimensions:
本算法不把会计师事务所等同于"营收排行榜"，而是综合审计学与市场监管的经典理论，将其映射到四个维度：
- 看门人理论（Watts & Zimmerman，1983）：审计是缓解委托代理冲突的独立第三方鉴证机制，注册会计师以独立、客观、公正的执业，保障资本市场信息披露的公信力——这是评价事务所"质量"的根本依据。
- 《中华人民共和国注册会计师法》：确立会计师事务所与注册会计师从事审计鉴证业务的法定地位与执业规范——为"专业能力"提供法律依据。
- 中注协综合评价框架：以业务收入、人才规模、执业质量等多维评价事务所——本算法借鉴其"多维"思路，但采用自有指标与权重，不摘抄其综合排名名次。
- 审计质量的不可直接观测性（DeAngelo，1981）：审计质量难以直接量化，故以"人均创收"（专业定价能力与高端客户结构）与"执业质量指数"作为代理变量。
- 数字化审计变革：大数据、人工智能、云审计正在重构审计方法——数字化审计能力是衡量事务所创新与未来竞争力的关键维度。
- 贡献维度（CPPPS 特色）：会计师事务所的贡献不仅在于创收，更在于审计了多少上市公司、服务了多少实体经济主体——这是国际榜单普遍缺失的本土视角。
二、模型的数学结构
This index adopts a four-dimensional weighted-sum model. Each indicator is first min-max normalized into the [0,1] interval (revenue and the number of listed-company clients are log-transformed first to curb top-tier monopoly), then synthesized into four dimensions by secondary weights, and finally weighted by primary weights and multiplied by 100:
本指数采用四维加权求和模型。每个指标先经 min-max 归一化映射到 [0,1] 区间（营收、上市公司客户数先取对数以抑制头部垄断），再按二级权重合成四个维度，最终按一级权重加权并乘以 100：
Here x′ denotes the min-max normalized value of indicator x: x′ = (x − xmin) / (xmax − xmin). Normalization removes scale differences, and the weights sum to 1, ensuring that the index falls within a stable 0~100 range.
其中 x′ 表示指标 x 的 min-max 归一化值：x′ = (x − xmin) / (xmax − xmin)。归一化消除量纲差异，权重之和为 1，保证指数落在 0~100 的稳定区间。
三、指标与权重依据
| 指标 | 方向 | 权重 | 所属维度 | 依据 |
|---|
| 业务收入（对数） | 正向 | 规模 0.60 | 规模 | 经营体量与市场份额；对数抑制头部垄断 |
| 注册会计师人数 | 正向 | 规模 0.40 | 规模 | 专业人才储备，执业能力根基 |
| 人均创收 | 正向 | 质量 0.60 | 质量 | 专业定价能力与高端客户结构（审计质量代理） |
| 执业质量指数 | 正向 | 质量 0.40 | 质量 | 质量控制与专业胜任能力（公开常识近似） |
| 数字化审计能力 | 正向 | 创新 1.00 | 创新 | 大数据/人工智能/云审计应用水平 |
| 上市公司审计客户数（对数） | 正向 | 贡献 0.60 | 贡献 | 服务资本市场广度（看门人职能的直接体现） |
| 本土服务实体经济系数 | 正向 | 贡献 0.40 | 贡献 | 覆盖中小微、民营与地方实体经济的广度 |
Primary-weight design: Scale and Quality are given equal weight (0.30 each), reflecting a "volume × professionalism" dual-engine drive; Contribution is 0.25, highlighting service to the capital market and the real economy; Innovation is 0.15, reflecting the long-term value of digital auditing. Revenue and listed-company client counts are log-transformed to prevent leading firms from monopolizing scores by sheer volume.
一级权重设计：规模与质量并重（各 0.30），体现"体量×专业"双轮驱动；贡献 0.25 突出资本市场与实体经济服务；创新 0.15 反映数字化审计的长期价值。营收、上市客户取对数，避免头部大所凭绝对体量垄断评分。
四、贡献维度之辨（CPPPS 特色）
Internationally prevalent accounting-firm rankings mostly sort by business revenue alone, causing the "Big Four", with their high unit prices and multinational major clients, to monopolize the top positions year after year, while domestic firms that are deeply rooted locally and audit the largest number of listed companies are underestimated. The CPPPS Contribution dimension deliberately corrects this bias: it looks not only at "how much was earned" but also at "how many listed companies were audited and how many real-economy entities were served".
国际通行的会计师事务所榜单多以业务收入单一排序，导致单价高、服务跨国大客户的"四大"常年垄断头部，而深耕本土、审计最多上市公司的内资本土大所被低估。CPPPS 贡献维度刻意纠偏：不仅看"赚了多少"，更看"审计了多少上市公司、服务了多少实体经济主体"。
Leading domestic firms (BDO China, Pan-China, and others) audit hundreds of A-share listed companies and cover a large number of small, medium and micro enterprises, private enterprises, and local state-owned enterprises, serving as the "capillaries" of the real economy; the "Big Four" audit a smaller number of A-share listed companies but handle large individual engagements with a high degree of internationalization, mainly serving major central SOEs, financial institutions, and multinational enterprises. The two groups have distinct divisions of labor and contributions. Under the multi-dimensional model, leading domestic firms that excel in Scale and Contribution can stand shoulder-to-shoulder with the "Big Four", forming a complementary pattern in which "the Big Four lead on quality and domestic firms lead on contribution".
本土头部大所（立信、天健等）审计数百家 A 股上市公司，覆盖大量中小微企业、民营企业与地方国企，是服务实体经济的"毛细血管"；"四大"审计的 A 股上市公司数量较少，但单体规模大、国际化程度高，主要服务大型央企、金融与跨国企业。两者各有分工、各有贡献。多维模型下，规模与贡献突出的本土头部所得以与"四大"并驾齐驱，形成"四大领质量、本土领贡献"的互补格局。
五、数据来源与局限性
Business revenue, the number of certified public accountants, and the number of listed-company audit clients use approximate values from public common knowledge (mainly the 2023 reporting period), marked "subject to official sources"; per-capita revenue is computed in-house as "business revenue ÷ number of CPAs"; the digital-audit capability index, practice quality index, and domestic real-economy service coefficient are approximate scores based on public knowledge. The data refers to CICPA industry statistics, each firm's annual disclosures and transparency reports, and public regulatory information from the CSRC and the Ministry of Finance. This model does not include factors that are difficult to quantify uniformly, such as practice-quality inspection penalty records, audit-failure litigation, and industry cycles, and the approximate values contain errors; the ranking is for research and reference only.
业务收入、注册会计师人数、上市公司审计客户数采用公开常识近似值（2023 年度口径为主），标注"以官方为准"；人均创收由"业务收入 ÷ 注册会计师人数"自行计算；数字化审计能力指数、执业质量指数、本土服务实体经济系数为公开知识近似评分。数据参考中注协行业统计数据、各所年度信息披露与透明度报告、中国证监会及财政部公开监管信息。本模型不包含执业质量检查处罚记录、审计失败诉讼、行业周期等难以统一量化的因素，且近似值存在误差，排名仅供研究与参考。
六、参考文献
- Watts, R. & Zimmerman, J.（1983）："Agency Problems, Auditing, and the Theory of the Firm"——审计的委托代理理论根基。
- DeAngelo, L.（1981）："Auditor Size and Audit Quality"，Journal of Accounting and Economics——审计质量的代理变量讨论。
- 《中华人民共和国注册会计师法》——会计师事务所与注册会计师执业的法律依据。
- 财政部：《会计师事务所执业许可和监督管理办法》《会计师事务所一体化管理办法》——执业质量与一体化管理规范。
- 中国注册会计师协会：《会计师事务所综合评价办法》及各年度行业统计数据——本榜仅采用其底层公开数据，不摘抄其综合排名名次。
- 中国证监会：《证券服务机构从事证券服务业务备案管理规定》——证券期货审计业务备案与监管。
- 《中国注册会计师审计准则》——审计执业质量标准。
- 各会计师事务所年度信息披露与透明度报告（公开资料，以官方为准）。