ASML · one lens
Hamilton Helmer on ASML
ASML clears Helmer's five-part Cornered Resource test cleanly — except the one a Chinese lab is now trying to fail it on.
Helmer’s Cornered Resource carries five tests for telling a real Power from a coveted asset that just happens to sit inside a good business: idiosyncratic, non-arbitraged, ongoing, non-replicable, sufficient (7 Powers). ASML is close to a clean pass on four of them. It is the only company that builds extreme ultraviolet lithography tools, the method required to print the smallest, densest chips — every advanced processor from TSMC, Samsung, or Intel passes through one of its machines first. The position isn’t rented at a price that captured the rents; it’s built from ASML’s own patent stack, its optics supply relationship, and decades of Dutch state co-investment, the way Helmer’s Post-it example turns on owned patents rather than a purchased ingredient. And it keeps compounding rather than sitting on past work: ASML raised its 2026 sales forecast for the second time this year in July, to €43B–€45B, on results that beat estimates rather than merely matched them (texxr.com/1172967).
The fifth test, non-replicable, is the one actually being contested in the record right now. Reuters reported in December 2025 that a team of former ASML engineers, having reverse-engineered the company’s own machines, had built a working EUV prototype in China earlier that year (texxr.com/893813). As of the most recent reporting, that remains a lab claim, not a shipped, qualified production tool. A prototype doesn’t clear Helmer’s bar; a shipped tool at volume would — which makes this the one test with a genuinely open verdict.
That the record’s most sustained pressure on ASML runs through legislators and cabinet officials rather than a competing supplier is itself consistent with a Cornered Resource holding — by April 2026, outside coverage had started using the word “chokepoint” as a plain descriptor rather than a hedge (texxr.com/1167786). When a Barrier can’t be out-engineered, the contest moves to policy.
The seven Powers, one by one
| Power | Status | Why |
|---|---|---|
| Scale Economies | absent | not argued in the record |
| Network Economies | absent | not argued in the record |
| Counter Positioning | absent | not argued in the record |
| Switching Costs | absent | not argued in the record |
| Branding | absent | not argued in the record |
| Cornered Resource | contested | clears 4 of Helmer's 5 sub-tests cleanly; non-replicability is open — a reverse-engineered China EUV prototype remains a lab claim, not a shipped production tool |
| Process Power | absent | not argued in the record |
An editorial application of Hamilton Helmer's published framework — not their actual view, affiliation, or advice.
Where this lens comes from
Hamilton Helmer's framework is set out at the lens page, drawn from 7 Powers. This page applies it to ASML and nothing else — the company's full record is in the dossier.