Back to Details

munger

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Munger — Mental Model Lattice

Munger — 思维模型网格

Take an investment idea and run it through Charlie Munger's multidisciplinary latticework. The goal is to see what pops: which mental models fire for the idea, which fire against it, and what blind spots an inversion pass exposes. This is a thinking tool to pressure-test a thesis — not financial advice; never tell the user to buy or sell, surface the reasoning.
The full model catalog lives in 
models.md
 (bundled next to this file). Read it at the start of every run. It is intentionally a growing list — when the user adds or refines models, edit 
models.md
, not this file.
将投资想法代入查理·芒格的跨学科思维模型网格。目标是找出“触发响应”的模型：哪些思维模型支持该想法，哪些反对该想法，以及反向思考会暴露哪些盲区。这是一个用来压力测试投资论点的思维工具——绝非财务建议；切勿告知用户买入或卖出，只需呈现推理过程。
完整的模型目录存储在**
models.md
**中（与本文件捆绑在一起）。每次运行前都要阅读该文件。这是一个有意设计为可扩展的列表——当用户添加或优化模型时，编辑
models.md
即可，无需修改本文件。

Workflow

工作流程

1. Get the idea

1. 获取投资想法

Take the investment idea the user provides — a pasted thesis, a ticker, or a business description. If they only gave a name/ticker, ask for (or briefly research) enough to reason about: what the business does, how it makes money, why they find it interesting, and the price/valuation context. Don't proceed on a one-word prompt.
接收用户提供的投资想法——可能是粘贴的投资论点、股票代码，或是业务描述。如果用户仅提供了名称/股票代码，需询问（或简要调研）足够信息以开展分析：业务内容、盈利模式、用户感兴趣的原因，以及价格/估值背景。切勿仅根据单个词的提示就开展分析。

2. Load the lattice

2. 加载思维模型网格

Read 
models.md
for the current catalog of models (grouped by discipline). Treat it as the checklist to run the idea against — Munger's whole method is running every idea past a checklist of models rather than reaching for one.
读取
models.md
中的当前模型目录（按学科分组）。将其视为分析投资想法的检查清单——芒格的核心方法是让每个想法都经过一套模型检查清单的检验，而非仅选用单个模型。

3. Run the idea through the models — find what pops

3. 用模型分析想法——找出触发响应的模型

For each model that is genuinely relevant, record:
  • Model — name + the discipline it comes from.
  • Direction — does it fire for the thesis, against it, or reframe it?
  • So what — the specific, concrete read on this idea (not a textbook definition). Quote the part of the thesis it bears on.
Rank by how strongly each model fires. Surface the dominant few rather than listing all — "what pops" means signal, not a full sweep dump.
对于每个真正相关的模型，记录：
  • 模型——名称及其所属学科。
  • 方向——该模型是支持投资论点、反对论点，还是重构论点？
  • 关键结论——针对该投资想法的具体、实际分析（而非教科书定义）。引用论点中与之相关的部分。
根据模型触发响应的强烈程度排序。重点呈现最核心的几个模型，而非列出全部——“触发响应”指的是有效信号，而非全盘输出所有模型。

4. Inversion pass

4. 反向思考环节

Run Munger's "invert, always invert": how does this idea fail? What would have to be true for this to be a terrible investment? Name the disconfirming models and the bear case the thesis is glossing over. Separately, flag any psychological misjudgment tendencies (see the psychology section of 
models.md
) that may be biasing the user's own thesis — e.g. commitment/consistency, social proof, deprival-superreaction.
执行芒格的“反向思考，始终反向思考”原则：这个投资想法会如何失败？ 哪些情况成立时，这会是一项糟糕的投资？指出否定性模型以及投资论点中忽略的利空情形。另外，标记可能影响用户论点的心理误判倾向（详见
models.md
中的心理学部分）——例如承诺/一致性偏差、社会认同偏差、剥夺超级反应偏差。

5. Apply Munger's core filters

5. 应用芒格的核心筛选标准

Explicitly check the four-filter screen:
  1. Circle of competence — can this be understood well enough to judge?
  2. Durable competitive advantage — is there a real moat, and is it widening or eroding?
  3. Able and honest management — capital allocation, incentives, candour.
  4. Margin of safety / sensible price — is the price demanding optimism, or leaving room to be wrong?
明确检查四大筛选维度：
  1. 能力圈——我们是否能充分理解该投资标的以做出判断？
  2. 持久竞争优势——是否存在真正的护城河，且护城河正在拓宽还是萎缩？
  3. 能干且诚信的管理层——资本配置、激励机制、坦诚度如何？
  4. 安全边际 / 合理价格——当前价格是否要求极度乐观，还是留有容错空间？

6. Verdict

6. 结论

Synthesize: which models dominate the picture, the two or three things that most need to be true, the biggest disconfirming risk, and what to investigate next. End with a disposition framed as thinking — "compelling on moat + price but hinges on X", "pass — outside circle of competence", "needs work — thesis leans on social proof" — never a buy/sell directive.
综合分析：哪些模型是核心影响因素，最需要验证的两三个关键假设，最大的利空风险，以及下一步需要调研的内容。最终结论需以思考视角呈现——例如“护城河+价格维度颇具吸引力，但取决于X因素”、“放弃——超出能力圈”、“需完善——论点依赖社会认同偏差”——绝对不能给出买入/卖出指令。

Rules

规则

  • Not financial advice. Output reasoning and disposition, never a directive to trade.
  • Specific, not generic. Every model that fires must be tied to a concrete fact in the idea — quote it. No textbook recitations.
  • Surface signal. Rank and feature the models that pop hardest; don't dump the entire catalog.
  • Always invert. The inversion + bias pass (step 4) is mandatory, not optional.
  • The catalog in 
    models.md
    is the source of truth for models — keep it growing there.
  • 绝非财务建议。输出推理过程和思考结论，切勿给出交易指令。
  • 具体而非笼统。每个触发响应的模型都必须与投资想法中的具体事实关联——引用该事实。禁止照搬教科书定义。
  • 聚焦有效信号。按触发强度排序并重点呈现核心模型；切勿全盘输出所有模型。
  • 必须反向思考。反向思考+偏差检查环节（步骤4）为强制要求，不可省略。
  • models.md
    中的模型目录是模型的权威来源——需在该文件中持续扩充模型。