# Output style and language

Use for a longer research answer or a public example.

Lead with a concrete judgment: what to study first, why, and the key unresolved question. For theme scans, explain layer priorities before company priorities. Use one compact comparison table when helpful, followed by the strongest evidence, counterargument and next check.

Link each material claim to its source close to the claim. For long PDFs, identify the page or section. Keep disclosed facts and interpretation distinguishable without turning every paragraph into a disclaimer.

## Chinese investment language

Use natural Chinese and explain technical terms where they affect the decision:

- 产业链卡点: the particular constraint, not a generic “core supplier” label.
- 商业进展: distinguish development, sampling, qualification, delivery and revenue.
- 研究优先级: which question or company deserves the next research effort.
- 反方理由: the best competing explanation, not a generic list of every possible risk.
- 失效条件: an observable fact that would change the judgment.

For example, in a hypothetical company challenge:

> 我会继续研究这家公司。公开材料已经支持相关产品开始交付，但还没有单列收入。接下来最有价值的是确认这部分业务有多大，以及客户能否更换供应商。只看公司整体利润增长，还不足以判断这个新业务的贡献。

This is a wording example, not a current company claim. Do not copy an old example's sector or company ranking into a new research task.

## Reports and ordinary conversation

A structured memo can use [the template](../assets/thesis-template.md). An ordinary question should receive a direct answer at the requested depth. Translate the template labels when answering in another language.

Keep technical detail proportional to its investment relevance. Explain how a capability, constraint or competing route affects orders, margins, cash flow or valuation. Place the research date and material access or valuation limitations where the reader can see them.