Quality compounder research asks whether a company can reinvest at returns above its cost of capital for long enough to create material value. The disagreement worth testing is usually about how long those returns last compared with what the price requires, not whether the business is good.
Separate existing returns from returns on the next investment
High reported returns can reflect a small accounting capital base, acquisitions, expensed intangible investment or a temporary profit peak. Fix a consistent definition of invested capital and after-tax operating profit before comparing periods.
Then ask what the next unit of reinvestment earns. A mature franchise with few places to reinvest is different from one with a long runway. Customer economics, retention, pricing, distribution and competition help test how long the returns persist.
Price the duration of the advantage
The case has to separate durable economics from expectations that already pay for them. Model a period of excess returns and test both faster erosion and longer persistence. History is context; it does not establish that this firm's advantage will hold.
In an Amten story this shows up twice. Inside the numbers, the business is read per $100 of sales: in the illustrative NVIDIA case $62 of operating profit against $18 for the median chipmaker, and margins that wide invite competition. Then the hook and the reality check say how long the price needs that to last, and how often companies have managed it.
Define the evidence that could erode the case
Watch returns on new investment, reinvestment quality, competitive pricing and customer economics. Set a threshold and a date before any deterioration appears.
Do not treat every margin gain as evidence of an advantage. Mix shifts, accounting changes and underinvestment can lift near-term results while weakening future economics. In Amten, the readings that argue against duration (such as customers building their own products) sit beside the ones that support it, and following the story shows when one moves.
Common questions
Is a high return on capital enough?
No. Check how capital is defined, how durable the returns are and what new investment earns, then compare that with what the price requires.
What is a compounder variant?
A supported disagreement about the rate, scale or duration of profitable reinvestment compared with the baseline reflected in the price.