Goldman Sachs-style institutional equity research methodology for single-stock work: narrative → fundamentals → SOTP valuation → fragility audit → analyst note.
更像机构研究方法论仓库，而不是工具仓库：叙事 → 基本面 → SOTP 估值 → 脆弱性审计 → analyst note。
Positioning note
This repository is inspired by institutional sell-side research structure and Goldman Sachs-style deliverable discipline. It is not affiliated with, endorsed by, or published by Goldman Sachs.
Quant Stock Analysis Valuation is a bilingual methodology repo for people who want single-stock work to read like institutional equity research, not like a loose memo, a retail blog post, or a spreadsheet with a target multiple pasted on top.
It is designed for workflows such as:
Analyze [ticker]Give me an institutional view on [company]Build a target price framework for [company]Re-underwrite the narrative for [stock]
The ambition here is not “yet another stock template.”
The ambition is to make the research chain coherent from first principle to final conclusion.
This repository should be read as an institutional research methodology repository.
The intended keywords are:
institutional equity researchequity research methodologySOTP valuation modelsingle stock valuation frameworkGoldman Sachs style researchnarrative to fundamentalsfragility audit
A serious research process should connect five layers into one chain:
market narrative → fundamental structure → SOTP / hybrid valuation → fragility audit → analyst conclusion
If any one of these layers is missing, the conclusion is weaker than it looks.
A lot of public stock-analysis repos stop too early.
Some stop at:
- data collection,
- memo writing,
- target multiple templates, or
- thematic commentary.
Institutional work requires more. It requires a process that can explain:
- what the market thinks the company is,
- what the company economically is,
- what part of the story is actually new,
- how that story should enter valuation,
- what fragility could invalidate the rerating, and
- what target price / rating conclusion follows from that chain.
A rerating story has no analytical value unless it can be mapped back to:
- revenue pools,
- profit-pool quality,
- segment economics,
- business mix, or
- capital allocation logic.
A company with mixed-quality businesses should not be forced into one blunt multiple. This framework prefers:
- segment SOTP when multiple businesses deserve different treatments,
- hybrid PE / PS when one segment is profit mature and another is still scaling,
- simple core PE only when additional complexity would be false precision.
This methodology treats fragility as a valuation input. That includes:
- geographic concentration,
- channel concentration,
- litigation / IP exposure,
- policy dependence,
- supply-chain bottlenecks,
- inventory / delivery / warranty risk.
The question is not whether risk exists.
The question is whether the risk deserves disclosure only, a multiple haircut, or a scenario discount.
The objective is a readable conclusion with:
- company framing,
- variant view,
- valuation derivation,
- target price,
- upside / downside,
- thesis risks,
- evidence references.
In this repository, “Goldman Sachs-style” means:
- debate framing before conclusion,
- valuation as a structured bridge rather than a slogan,
- segment logic instead of theme-chasing,
- explicit assumptions,
- explicit downside / variant view,
- institutional presentation discipline.
It does not mean imitation of any firm's proprietary model. It means adopting the standard of clarity and rigor expected from institutional sell-side work.
This framework emphasizes Sum-of-the-Parts (SOTP) when a company is really several economic engines under one ticker.
Instead of asking only:
- “What PE should the stock trade on?”
it asks:
- What are the real business segments?
- Which segments are already profit-bearing?
- Which segments are still in revenue scale-up?
- Which segments deserve PE vs PS?
- Which narratives are real and which are still optionality?
- What is the implied equity value when those pieces are valued appropriately?
That is the core difference between a superficial valuation memo and an institutional underwriting framework.
This repository includes a reusable structure for:
- market snapshot
- business decomposition
- narrative classification
- fragility audit
- valuation routing
- final analyst-style document generation
It is a methodology + template + lightweight script repository. It is not a fully autonomous research engine.
git clone https://github.com/oierkid-crypto/quant-stock-analysis-valuation.git
cd quant-stock-analysis-valuationpython3 scripts/quote_snapshot.py --ticker 300124.SZ
python3 scripts/quote_snapshot.py --ticker 06656.HKpython3 scripts/scaffold_input.py --ticker 300124.SZ --output work/inovance.jsonUse the build command to turn structured input into the final analyst note.
The repository is organized around four layers:
- methodology — the research logic and valuation discipline
- template — the final analyst-style structure
- scripts — lightweight helpers for quote snapshot, scaffolding, report build, and document rendering
- examples — a sample structured input for onboarding
This repository does not replace actual research judgment. You still need to:
- read source materials,
- decide what the real narrative is,
- judge whether it is already in the numbers,
- assign valuation logic thoughtfully.
The point of the repository is to make that work institutional, repeatable, and readable.
Quant Stock Analysis Valuation 是一个双语的机构研究方法论仓库。它不是散装 memo，不是零售风格的 stock blog，也不是套一个 target multiple 就结束的估值模板。
它适用于这类工作流：
帮我分析 [公司 / ticker]给我一个机构视角的结论帮我重做这家公司目标价框架重新 underwrite 这只股票的叙事
这个仓库想做的，不是“又一个股票模板”。
它想做的是：把单票研究从第一性原理到最终结论，真正连成一条完整链路。
这个仓库应该被理解成一个机构研究方法论仓库，而不是一个展示工具或文档渲染仓库。
它的关键词应该是：
institutional equity researchequity research methodologySOTP valuation modelsingle stock valuation frameworkGoldman Sachs style researchnarrative to fundamentalsfragility audit机构研究方法论分部估值单票估值框架
一套像样的研究流程，应该把五层逻辑真正打通：
市场叙事 → 基本面结构 → SOTP / 混合估值 → 脆弱性审计 → analyst 结论
这五层里缺任何一层，最后的结论都会看起来比它实际上更强。
很多公开的股票分析 repo，停得太早。
有的停在：
- 数据抓取，
- 研究 memo，
- target multiple 模板，或
- 主题式评论。
但机构研究需要更多。
它需要一套流程，能够回答：
- 市场以为这家公司是什么，
- 这家公司在经济意义上到底是什么，
- 什么部分才是真正的新叙事，
- 这个叙事应该如何进入估值，
- 什么脆弱性会让这次重估失败，
- 沿着这条链路，最后应该落到什么目标价 / 评级。
任何重估故事，如果不能映射回：
- 收入池，
- 利润池质量，
- 分部经济结构，
- 业务组合，或
- 资本配置逻辑，
它的分析价值就非常有限。
一家公司如果本质上由多个质量不同的业务组成，就不应该被粗暴地塞进一个总 PE。
这套框架优先使用：
- 分部 SOTP：当多个业务应当被区别对待时
- PE / PS 混合估值：当一部分业务已成熟盈利，另一部分仍处于放量阶段时
- 核心业务 PE：只有在继续复杂化会变成假精确时才使用
这套方法把脆弱性直接视为估值输入项。包括：
- 地域集中度，
- 渠道集中度，
- 诉讼 / IP 暴露，
- 政策依赖，
- 供应链瓶颈，
- 库存 / 交付 / 质保风险。
关键不是“有没有风险”。
关键是这些风险到底只需要披露，还是应该对应multiple haircut 或 scenario discount。
最终目标不是工程中间件，而是一份可读的研究结论，里面应该有：
- 公司定位，
- variant view，
- 估值推导，
- 目标价，
- upside / downside，
- 核心风险，
- 证据引用。
在这个仓库里，所谓 Goldman Sachs-style，指的是一种机构 sell-side 纪律：
- 先定义 debate framing，再给结论
- 估值是结构化推导，不是口号
- 先讲分部逻辑，不追逐主题热词
- 假设必须显式写出来
- downside / variant view 必须清楚
- 呈现方式要有机构研究的节制感
它不代表模仿任何机构的 proprietary model。
它代表的是：用机构卖方研究应有的清晰度和严谨度来组织研究。
这套框架强调 SOTP（Sum-of-the-Parts），因为很多公司在一个 ticker 下面，其实装着几个不同的经济引擎。
所以问题不只是：
- “这只股票应该给多少 PE？”
而是：
- 真实的业务分部是什么？
- 哪些分部已经是利润池？
- 哪些分部还在收入放量？
- 哪些分部该用 PE，哪些该用 PS？
- 哪些叙事已经成立，哪些还只是 optionality？
- 当这些部分被合理估值后，隐含股权价值到底是多少？
这就是表面估值 memo 和机构式 underwrite 框架的区别。
这个仓库提供的是一套可复用结构，用于承载：
- 市场快照
- 业务拆解
- 叙事分类
- 脆弱性审计
- 估值路由
- 最终 analyst-style 文档生成
它本质上是一个 methodology + template + lightweight script 仓库。
它不是一个全自动研究引擎。
git clone https://github.com/oierkid-crypto/quant-stock-analysis-valuation.git
cd quant-stock-analysis-valuationpython3 scripts/quote_snapshot.py --ticker 300124.SZ
python3 scripts/quote_snapshot.py --ticker 06656.HKpython3 scripts/scaffold_input.py --ticker 300124.SZ --output work/inovance.json使用 build 命令把结构化输入转成最终 analyst-style 研究文档。
这个仓库实际分成四层：
- methodology —— 研究逻辑和估值纪律
- template —— 最终 analyst-style 结构
- scripts —— 行情抓取、脚手架生成、报告生成、文档渲染等轻量辅助脚本
- examples —— onboarding 用的样例输入
这个仓库并不能替代真实的研究判断。
你仍然需要：
- 阅读源材料
- 判断真正的新叙事是什么
- 判断它是否已经进入数字
- 谨慎分配估值逻辑
这个仓库的作用，是把这些工作变得更机构化、可复用、可读。
MIT