Overview of ASML: Competing with Moore’s Law (Business Breakdowns — REPLAY)
This episode is a deep-dive conversation with Tom Walsh (Baillie Gifford) about ASML — the Dutch company that manufactures the world’s most advanced photolithography machines (notably EUV). The discussion covers ASML’s unlikely rise from a troubled Philips spin‑out to a near-monopoly supplier of leading‑edge lithography, explains photolithography and extreme ultraviolet (EUV) in plain language, lays out the company’s business model and economics, and highlights key competitive advantages, risks, and investor takeaways.
Company history & trajectory
- Founded as a Philips spin‑out in 1984 — initially the “10th of 10” players, weak balance sheet and doubtful prospects.
- Survived early cycles, improved equipment and execution; by mid‑1990s became one of the top three lithography vendors (with Nikon and Canon).
- Became industry leader (overtook Nikon) around 2002 and continued aggressive R&D and “moonshot” investments.
- Delivered the first high‑volume production EUV systems in 2019 — the pivotal technology that enables today’s most advanced chips.
- Key people: long‑time CTO Martin van den Brink (instrumental engineering leader since the spin‑out) and CEO Peter Wennink (joined 1999, CEO since 2013).
What photolithography and EUV are (simple explanation)
- Photolithography: like projecting an image through a mask onto a coated silicon wafer; the “image” is the circuit pattern. Conceptually simple, but physically extremely hard because feature sizes are now a few nanometers (a handful of atoms).
- Two primary technical elements: the light source and the optics (lenses/mirrors). Over decades the industry moved from visible light → deep ultraviolet (DUV, 193 nm) → extreme ultraviolet (EUV, ~13.5 nm).
- EUV essentials:
- Uses a novel light source (laser hitting tiny tin droplets to create plasma) and reflective optics (mirrors) inside vacuum.
- Producing EUV light requires firing at micron-scale tin droplets thousands of times per second, producing plasma hotter than the Sun’s surface — a hugely difficult engineering challenge.
- EUV enables much finer patterns (smaller transistors) and simplifies some multi‑patterning tricks used with older wavelengths.
Business model & economics
- Scale and pricing:
- Despite large revenue, unit volumes are low: only a few hundred machines sold per year (transcript cites ~345 units in one year; leading‑edge EUV units are in the low double‑digits annually).
- Top‑end EUV systems cost >€100–150 million each.
- ASML’s machines account for roughly 20–25% of wafer fab equipment spend (higher in leading‑edge logic fabs).
- Revenue mix:
- New machine sales ≈ 75% of revenue; services, upgrades and spares ≈ 25%.
- Customers typically pay deposits; machines are long‑lived (stat: ~90% of machines sold in the past 30 years still operate somewhere).
- Profitability and cash flow:
- 2022: ~€21bn revenue and ~€6.5bn operating profit (host notes later growth to ~€32bn by 2025).
- Gross margins ~50% (improved from ~30% in the 2000s); operating margins ≈30%.
- R&D spend ~15–16% of revenue; strong free cash flow conversion; uses cash for dividends, buybacks and heavy CapEx to expand capacity.
- Manufacturing model:
- ASML calls itself an “architect and integrator.” ~80% of COGS are supplied components from an ecosystem of highly specialized vendors; ASML assembles, integrates and finalizes complex machines on‑site at customer fabs.
- This modular approach enables incremental upgrades and de‑risked development.
Competitive advantages & moats
- Technological lead in EUV is extremely wide:
- Decades of development and >€10bn+ invested in EUV R&D.
- Critical components (e.g., the EUV light source provider Cymer, precision mirrors from ZEISS) are both specialized and hard to replicate; ASML acquired Cymer and holds strong supplier relationships.
- The next generation (high‑NA EUV) further increases complexity and barriers.
- Ecosystem & customer collaboration:
- Deep, cooperative relationships with the few customers who can afford and exploit leading‑edge tools (TSMC, Samsung, Intel, etc.). Customers co‑invested in ASML historically.
- Pricing philosophy is collaborative — ASML aims to split the incremental economic benefit roughly 50/50 with customers rather than exploit monopoly pricing, preserving long‑term cooperation.
- Long product life, field upgrades and service economics create recurring revenue and high customer lock‑in.
Key numbers & vivid details
- Market cap (approx): ~€250bn (as of the episode day).
- ASML share of wafer fab equipment ≈ 20–25%.
- Top two customers ≈ 60% of revenue; major regional concentration: Taiwan (~40%), South Korea (~30%), China (~~15% cited).
- Target capacity expansion: producing up to ~90 EUV machines/year (vs. ~40 sold/year recently) and hundreds of DUV units; target ramp for high‑NA units around 2027–2028.
- Example physical scale: EUV tool ~size of a double‑decker bus; shipping parts can require three jumbo jets.
Risks & failure modes
- Supply‑chain bottlenecks: ASML depends on a small number of highly specialized suppliers; if suppliers cannot scale or innovate, ASML may need to acquire/stake in them (has happened with Cymer/ZEISS).
- Disruptive technology/alternate manufacturing routes: if an alternative (e.g., new transistor architectures, 3D stacking, or another lithography approach) becomes more cost‑effective, demand for certain lithography spend can decline (example: transition in NAND to 3D stacking reduced lithography share).
- Cyclicality: semiconductor capex cycles cause order variability; while ASML’s position has dampened cyclicality recently, demand still fluctuates (customer capex cuts matter).
- Geopolitics & export control: concentration of customers in Taiwan, South Korea and China exposes ASML to export restrictions, political pressure and regional instability; controls on selling EUV to China are material to future growth.
- Long, capital‑intensive R&D and heavy upfront investment: missteps in next‑gen (e.g., high‑NA) would be costly.
Notable quotes / insights
- “Make things smaller, make them sharper and make them faster.” — succinct summary of ASML’s mission.
- “ASML’s competition is not really another company, but Moore’s Law.” — framing ASML as an enabler of continued scaling.
- The EUV source: “shooting a laser at droplets of tin… turn it into plasma, which is 40 times hotter than the surface of the sun” — illustrates the extreme engineering involved.
Investor takeaways & monitoring checklist
- Structural bull case:
- Continued secular demand for more and smaller transistors (AI, datacenter, advanced mobile, HPC) supports long‑term lithography spend.
- ASML’s unique position in EUV and its planned capacity ramp should translate to revenue and margin expansion if demand materializes.
- What to watch (milestones and data points):
- Order intake and backlog for EUV and DUV machines (timing, cancellations, ASPs).
- Progress and timelines for high‑NA EUV qualification and customer adoption.
- Supplier health and investments (Zeiss optics, Cymer light sources, other key vendors).
- Customer capex guidance from TSMC, Samsung, Intel, Micron (their roadmaps drive future ASML demand).
- Geopolitical developments and export control rules impacting shipments to China.
- Margin trajectory (gross and operating margins) as production scales.
- Balance potential: ASML is a high‑quality, cash‑generative franchise with outsized technical moats, but it’s exposed to cyclicality, geopolitics and execution risk on capacity/supply chain.
Lessons beyond ASML
- Look beyond short cycles to long‑term structural value when a company controls a critical industry bottleneck.
- Strong supplier/customer ecosystems can be both a moat and a vulnerability — managing partnerships is as important as owning assets.
- Incremental modular innovation and relentless engineering can extend a company’s relevance across multiple technology generations.
For someone who wants the core story without listening to the full episode: ASML transformed from an underdog spin‑out into the indispensable, highly technical supplier of the lithography tools required for the world’s most advanced chips. Its EUV lead is a rare, multi‑decade, multi‑billion‑euro technological moat, but success depends on supply‑chain execution, customer demand cycles, continued R&D, and geopolitics.
![ASML: Competing with Moore’s Law - [Business Breakdowns, REPLAY]](https://megaphone.imgix.net/podcasts/8e22228a-1900-11f1-9c46-c71411cc72f0/image/089860611643f35b87f4d7906b431e38.jpg?ixlib=rails-4.3.1&max-w=3000&max-h=3000&fit=crop&auto=format,compress)