A reviewable investment thesis states its purpose, a sourced market baseline, a material variant, independent evidence, valuation assumptions, the strongest alternative explanation, catalysts and what would end the view. Date it, and judge outcomes against the original claim rather than a rewritten one.
Check the argument before the conclusion
Complete it in your own words and leave open questions visible. The goal is an argument another person could rebuild, including the parts they would reject.
- Purpose: why this company, over which horizon, and what decision the research informs.
- Business: how cash is earned, which drivers matter, and what kind of company it is.
- Sources: dates, primary evidence, dependencies and gaps.
- Baseline: what the price implies and what published sources say.
- Variant: your alternative assumption on one material driver, and its effect on value.
- Challenge: the strongest case that the market is right.
- Resolution: catalysts, what would end the view, and the evidence needed to check them.
- Review: a dated record and a routine for updating it.
How a story maps onto the checklist
An Amten story does much of the gathering, in a fixed order. The judgment stays with you.
| Checklist item | Where it lives in a story |
|---|---|
| Baseline | The hook: what the price needs, with its working |
| Business | Inside the numbers: of every $100 of sales and assets |
| Sources | The evidence sidebar on every beat |
| Variant | The missing chapter: hidden cast, early signals, contradictions |
| Challenge | Competing readings, each with what would confirm and refute it |
| Resolution | Another ending, and research branches with dates to watch |
| Review | Following and versions: "How this story changed" |
Keep process and outcome apart
After new evidence, compare the changed claim with the dated view. Ask whether the operating prediction was right, whether the market recognized it, and whether your confidence was calibrated. Do not rewrite the original reasoning to fit the outcome.
Amten's versions help here because they are immutable: version 1 still says what it said. A precise model and a well-written story can both hide weak inputs, so check source independence and sensitivity before trusting either. "No variant" is better than an unsupported story.
Common questions
What is the difference between a company summary and a thesis?
A summary describes the business. A thesis explains a material disagreement with a measured baseline, its effect on value, and how it could be tested.
Should open questions stay in the thesis?
Yes. Visible gaps make the next evidence requests and the review more reliable.