What Does Tokyo Electron Do? The Coater-Track Model

Tokyo Electron sells the coater/developer tracks that sit beside every scanner, plus etch, deposition, clean and probers: it gets paid when a tool or a Field Solutions contract is recognised, while the die still waits on an ASML slot.

SEMICONDUCTORS · Data as at 30 Jul 2026 · Next update: after Tokyo Electron Q2 FY2027 results (company calendar, 30 Oct 2026)

Affects: Semiconductors · Taiwan · Korea · Japan · US

It is the track-and-chamber company on the semiconductors layer: the CLEAN TRACK that coats and develops the wafer around EUV, then the etch, deposition, clean and wafer-prober tools that follow. Logic and foundry still take the heavier new-equipment mix this quarter. Taiwan is the rising invoice; China is still the largest geo. See who makes wafer fab equipment and foundry and advanced packaging.

In short: Tokyo Electron coats, processes and services. The April–June 2026 quarter printed ¥732.3 billion of net sales — a record — of which SPE new equipment was ¥531.3 billion and Field Solutions ¥188.0 billion. Non-memory was 57% of new equipment. The layer’s stuck step is still the scanner slot.

Education only, not advice to buy or sell any security. Figures are from company filings and company releases dated as shown. Outlook is labelled outlook.

Numbers that matter

Sources: Tokyo Electron Summary of Consolidated Financial Results for the first quarter ended 30 Jun 2026, 30 Jul 2026; Q1 FY2027 presentation, 30 Jul 2026; transcript, 30 Jul 2026. Status chips are Second Order’s read. Application mix is of SPE new equipment, not of total sales. First-half figures are guidance. The print for that guide is due 30 Oct 2026.

Our arithmetic. Net sales were +33.3% year over year (732,388 ÷ 549,586 = 1.333) and +2.9% sequentially against the presentation’s ¥711.8 billion March quarter (732.3 ÷ 711.8 = 1.029). Gross margin 46.8% (342,712 ÷ 732,388 = 0.468). Operating margin 28.9% (211,407 ÷ 732,388 = 0.289). Net margin 22.4% (164,341 ÷ 732,388 = 0.224). SPE new equipment was 72.5% of the quarter (531.3 ÷ 732.3 = 0.725); Field Solutions 25.7% (188.0 ÷ 732.3 = 0.257). Those two disclosed SPE buckets sum to ¥719.3 billion of the ¥732.3 billion print; TEL reports a single SPE segment. Field Solutions was +15.5% sequentially (188.0 ÷ 162.8 = 1.155). China 30.4% of total is about ¥223 billion; Taiwan 25.4% about ¥186 billion. DRAM at 32% of SPE new equipment is about ¥170 billion (0.32 × 531.3); non-memory 57% about ¥303 billion. Do not apply the 57/32/11 split to the ¥732.3 billion total — the mix is new equipment only. Implied July–September sales against the ¥1,620 billion first-half guide is ¥887.7 billion (1,620.0 − 732.3). That is outlook.

What does Tokyo Electron actually do?

Three steps, in order.

Coat. Process. Service.

1. It coats and develops beside the scanner. It does not print the pattern. Lithography prints. Tokyo Electron’s CLEAN TRACK sits next to the scanner: coat photoresist, develop after exposure. ASML prints. Applied, Lam and Tokyo Electron build and carve. KLA checks. The 30 July presentation names a newly released CLEAN TRACK LITHIUS Pro DICE as a FY2027 coater/developer growth driver. That line is outlook, not this quarter’s ¥531.3 billion.
2. It ships a counted invoice, not a counted scanner plan. April–June SPE new equipment was ¥531.3 billion. Non-memory (logic, foundry, others) was 57%, DRAM 32%, non-volatile memory 11%. NAND new-equipment sales rose 32% sequentially, the CFO said on the 30 July call. Do not read ¥531.3 billion as a shipped accelerator.
3. It services the installed fleet. Field Solutions printed ¥188.0 billion, 25.7% of the quarter: parts and services ¥135.7 billion, used equipment and modifications ¥52.3 billion. That invoice does not wait for a new tool to ship. Utilisation at the customer fab is the clock.

Tokyo Electron also sells into advanced packaging. The 30 July presentation’s FY2027 outlook is advanced-packaging sales growth of 70%+ year over year, with wafer probers expected to significantly exceed ¥100 billion for the full year. That is outlook, and a neighbour clock on CoWoS. It is not the April–June ¥531.3 billion.

How does Tokyo Electron make money?

It gets paid when a system is recognised, and again when the installed fleet is serviced. Three things decide which cash prints, and when:

• Recognition — a tool in advances received is not the quarter’s sales until TEL’s accounting says it is. Advances received were ¥294.3 billion at 30 June 2026, up from ¥270.5 billion at 31 March 2026 (tanshin). Inventories were ¥751.5 billion.
• Mix — logic/foundry versus DRAM versus NAND; new systems versus Field Solutions; Taiwan versus China versus Korea.
• Process intensity — more EUV layers, tighter overlay, HBM stacks and hybrid-bonded packages mean more coater, etch, deposition and clean steps per wafer. That is how TEL’s invoice can rise even when wafer starts do not.

On 30 July 2026 the tanshin printed ¥732,388 million of net sales: gross profit ¥342,712 million, operating income ¥211,407 million, net income attributable to owners of parent ¥164,341 million. Free cash flow was ¥80.5 billion (presentation: operating ¥118.7 billion minus investing ¥38.1 billion). Full-year FY2026 net sales were ¥2,443.5 billion, operating income ¥624.9 billion (same 30 July deck, actuals). TEL reports a single SPE segment.

Who gets paid. Tokyo Electron, on the recognised tool and on Field Solutions. Who pays. The tanshin does not name a customer above 10%. The geo table is the proxy: China ¥222.9 billion, Taiwan ¥186.1 billion, Korea ¥152.4 billion this quarter. TSMC, Samsung, Intel Foundry, SK hynix and Micron sit on the same wafer-buyer list. Money sticks at the recognised CLEAN TRACK, etch or deposition tool and the Field Solutions contract, not at a foundry capex headline. See the Taiwan capex-split Pulse.

Guidance for the first half ending September 2026 was revised to ¥1,620 billion of net sales, operating income ¥458 billion, operating margin 28.3%. SPE new equipment for the half was revised to ¥1,240 billion, of which the second quarter is anticipated at ¥708.7 billion (+33% sequentially) — outlook. Implied Q2 sales versus June is +21.2% (887.7 ÷ 732.3 = 1.212). That is outlook. The print is the 30 October call.

What limits it?

Scanner slots first. Track-and-chamber conversion second. China mix third. Neighbour clocks fourth.

1. A shipped CLEAN TRACK does not print a leading-edge die by itself. Every advanced logic wafer and most new DRAM still need an ASML scanner. On 14 October 2026 ASML said planned supply for 2027 and 2028 will not fully meet demand. See the Wire. TEL can raise a product invoice while the wafer still waits on a Low-NA slot. High-NA adds overlay and coat/develop steps; it does not replace the scanner bind. Do not read ¥531.3 billion as a shipped accelerator.
2. The track still has to convert. Process intensity is TEL’s own clock: more EUV layers, tighter specs, HBM stacks. A 57% non-memory new-equipment mix can look leading-edge while TSMC CoWoS still queues the package. See the Taiwan CoWoS Pulse. DRAM at 32% of new equipment, NAND at 11%, can stay “strong” while cubes wait on SK hynix allocation and Micron cleanroom timing. A CLEAN TRACK in a bay is not a tested stack.
3. China is a mix, not an EUV escape. The presentation puts China at 30.4% of total revenue, Taiwan 25.4%, Korea 20.8%, Japan 8.8%, North America 8.1%. Full-year FY2027 China is expected in the low 30s% — outlook. That print is process tools and Field Solutions. It is not a scanner substitute, and it moves with export licences.
4. A coated wafer still needs a stack and a slot. HBM and CoWoS sit after the coat-and-etch step. The presentation’s FY2027 coater/developer 50%+ year-over-year, etch 25%+, and advanced packaging 70%+ are outlook. They do not move the scanner bind. Calendar-2026 WFE of $150 billion+ is outlook as of July 2026, not this tanshin. The Besi Wire is the same neighbour from the bonder’s side: HBM still uses bumps.

The honest uncertainty is conversion: a ¥1,620 billion first-half guide can still miss if installs slip or licences stall.

Who else does this?

Roles, not a league table.

What separates them is who can ship a qualified EUV box — that is ASML — versus who gets paid on the hundreds of other steps. Tokyo Electron sits on the coat-and-chamber seat. They overlap on etch and deposition. They do not overlap on production EUV.

What would change this view?

The read on the model weakens if:

• The 30 Oct 2026 print comes in well below the implied ¥887.7 billion second-quarter sales against the ¥1,620 billion first-half guide
• SPE new equipment rolling over, or Field Solutions falling well below ~26% of the quarter without a matching systems offset
• China geo jumping well above the low-30s% full-year outlook, or export licensing cutting the 30.4%
• Advances received climbing while new-equipment recognition stalls
• ASML 2027–2028 supply remaining short while TEL’s product invoice is sold as if wafers were unconstrained
• CoWoS or HBM remaining so tight that coated wafers queue anyway — the neighbour clock stealing the bind

It stays intact while SPE new equipment stays ~73% of the invoice, Field Solutions keeps about a quarter of the quarter, non-memory holds the new-equipment majority, China stays a process-tool mix rather than an EUV story, and the scanner slot remains the layer’s stuck step.

Practical takeaways

• Tokyo Electron’s model is Coat → Process → Service: CLEAN TRACK around the scanner, a recognised SPE invoice, then Field Solutions on the fleet
• April–June 2026 quarter: ¥732.3bn sales (record), ¥531.3bn SPE new equipment (72.5%), ¥188.0bn Field Solutions (25.7%), 46.8% gross margin, ¥211.4bn operating income, ¥164.3bn net income
• Mix: non-memory 57% / DRAM 32% / NAND 11% of SPE new equipment. Geo: China 30.4%, Taiwan 25.4%, Korea 20.8%. No named >10% customer in this tanshin
• Layer stuck step: ASML scanner slots for 2027–2028. TEL’s clock: coat-and-etch intensity and the ¥1,620bn first-half guide, due 30 Oct 2026
• A coated wafer still waits on HBM and CoWoS

FAQ

What does Tokyo Electron do?
Tokyo Electron designs, builds and services semiconductor production equipment — coater/developer tracks that sit beside the scanner, plus etch, deposition, cleaning systems and wafer probers — then charges again through Field Solutions to keep those tools running. It does not sell production EUV scanners. SPE new-equipment sales in the April–June 2026 quarter were ¥531.3 billion.

What does Tokyo Electron do in simple terms?
It makes the machines a chip factory uses to coat photoresist beside the scanner, then etch, deposit and clean the layers, and it charges again to service those machines.

Is Tokyo Electron profitable?
Yes on the dated print. The April–June 2026 quarter’s net income was ¥164.3 billion on ¥732.3 billion of net sales, a 22.4% net margin, with a 46.8% gross margin and a 28.9% operating margin (tanshin, 30 Jul 2026, margins ours). That is not a forecast.

How is Tokyo Electron different from ASML and Applied Materials?
ASML sells the production EUV scanner. Applied sells the widest process-chamber catalogue. Tokyo Electron sells the coater/developer that sits beside the scanner, plus etch, deposition, clean and probers. Field Solutions was 25.7% of TEL’s April–June quarter. See What Does ASML Do? and What Does Applied Materials Do?.

What share of Tokyo Electron’s sales is new equipment versus Field Solutions?
On the 30 July 2026 presentation, SPE new equipment was ¥531.3 billion, 72.5% of the April–June quarter, and Field Solutions was ¥188.0 billion, 25.7%. Application mix (non-memory 57%, DRAM 32%, NAND 11%) is of new equipment, not a second income statement.

Should I buy Tokyo Electron shares?
We do not make buy or sell calls, and nothing here is a recommendation. What we can tell you is what the company does, how it gets paid, and which figures would change the picture. The decision is yours.

Keep reading on the map

• The layer: Semiconductors
• The mechanism: Who makes wafer fab equipment · Foundry and advanced packaging
• The scanner: EUV, explained · High-NA EUV · What Does ASML Do?
• The foundry: What Does TSMC Do?
• The neighbours: CoWoS · HBM · Applied Materials · Lam Research · KLA
• The Wire: ASML 2027–2028 supply short · Besi: HBM still uses bumps
• Where it happens: Taiwan · TSMC CoWoS · TSMC capex split · Korea HBM · Micron Idaho

The members cut on the long clock is already live: Where does the stuck step move?.

How we check this

Figures come from company filings and company releases, each dated below. Application mix (non-memory 57%, DRAM 32%, NAND 11%) is the 30 July presentation; yen splits from those percentages applied to SPE new equipment are our arithmetic. China 30.4%, Taiwan 25.4% and Korea 20.8% are the presentation’s geo table. First-half ¥1,620 billion, Q2 SPE new equipment ¥708.7 billion, and the FY2027 growth rates are outlook. TEL reports a single SPE segment. Where we work something out ourselves, the arithmetic is shown in full.

Last reviewed: 28 Oct 2026 · Next review: after Tokyo Electron Q2 FY2027 results (company calendar, 30 Oct 2026) print sales against the ¥1,620 billion first-half guide

Spot an error? Tell us; we will correct it here. Editorial standards →

Sources

1. Tokyo Electron, Summary of Consolidated Financial Results for the first quarter ended 30 Jun 2026 (Japanese GAAP tanshin), 30 Jul 2026
2. Tokyo Electron, Q1 FY2027 presentation, 30 Jul 2026
3. Tokyo Electron, Q1 FY2027 transcript, 30 Jul 2026
4. Tokyo Electron, FY2027 Q1 Financial Announcement (newsroom), 30 Jul 2026
5. Tokyo Electron, FY2027 Q2 earnings release (scheduled), 30 Oct 2026, 16:00 JST
6. ASML, via Wire: 2027–2028 supply short, 14 Oct 2026

All external links rel="nofollow noopener".