Lasertec: the quiet monopoly behind every EUV chip
A neutral, evidence-first reading of Lasertec — the Japanese maker whose tools verify the photomasks for every leading-edge chip, holding a near-monopoly so profitable, and so concentrated, that it became both Japan's most-traded stock and a short-seller's biggest target.
44 sourcesAs of 7 June 202610 analysis sections
In the year ended June 2025 Lasertec turned over a record ¥251.5B (up 17.8%) and earned a 48.8% operating margin — economics far above typical hardware-maker margins[11][22]. It does it from a single, narrow franchise: it is effectively the only company in the world selling tools that inspect extreme-ultraviolet (EUV) photomasks at the actual 13.5nm wavelength chips are printed with[41].
Every advanced chip starts from a photomask — the master stencil ASML's EUV scanners project onto silicon. A defect on that mask repeats on every wafer, so chipmakers like TSMC, Samsung and Intel must inspect masks with light of the same 13.5nm wavelength. Lasertec's ACTIS (patterned masks) and ABICS (mask blanks) tools are the only commercial way to do that[17][18]. That position produced a near-sevenfold revenue rise and a stock that rose roughly 1,500–1,800% over five years[4]. The open questions are not whether the franchise is real — it plainly is — but how durable the monopoly is[42], whether record FY2025 marks a cyclical peak[23], and what to make of a 2024 fraud allegation that a special committee dismissed but markets never fully shook off[43]. The evidence cuts both ways. This study lays out both cases; the verdict is yours.
The decisive questions
Each links to the section that lays out the evidence on both sides.
Net sales, ¥B, fiscal years ending June. FY2026 is the company's revised guidance (~¥220B), not a reported figure — the first down year after the record run.
Lasertec net sales, FY2021–FY2026E (¥B)
⚖️
What reasonable people disagree about
Whether a monopoly with no legal protection is durable as KLA and customers push for a second source[42]; whether FY2025's record is a peak or a pause, given a ~60% order drop against a 2nm/high-NA EUV tailwind[22][24]; whether the 2024 short-seller report exposed real weaknesses or was an overreach a committee rightly dismissed[29][43]; and whether a once-parabolic “story stock” is fairly valued[38]. Informed observers genuinely disagree on each; the sections below lay out the strongest evidence on both sides, and the dated signals that would change the call.
How to read this
Ten sections, each built the same way: a neutral synthesis, a two-sided case-for / case-against ledger, sourced data and charts, and dated facts. Start with the question that interests you, or read in order from the Overview.
🔍
Independent research artifact, not affiliated with or endorsed by Lasertec. Financial figures are from Lasertec's FY2025 results and FY2026 guidance via reputable press; market-size and customer-mix figures are third-party estimates and are labeled as such. A meaningful share of sources are Japanese-language. See Methodology & Limits.
Overview & Timeline
From near-death to indispensable
A 1960s optics lab that nearly didn't survive the financial crisis, then bet everything on inspecting EUV photomasks — and became one of Japan's most valuable, most volatile companies.
Founded 1960TSE: 6920 · Prime / Nikkei 225HQ Yokohama, Japan
Lasertec is a 1960-founded optical-inspection maker that almost failed after 2008, when its display-inspection business collapsed into a ¥6.5B loss[3]. Its pivot to EUV mask inspection turned it into a near-monopoly, lifting revenue roughly sevenfold and the share price ~1,500–1,800% in five years[4] — a rise so steep it made the stock both a retail favourite and a short-seller magnet[38].
What Lasertec actually makes
Lasertec designs inspection and measurement systems built on optical and laser technology. Its founding credo, in its own words, was to “launch a product totally new to the world every year”[2] — and its history is a string of world-firsts: the first LSI photomask inspection system (1976), the first color laser scanning microscope (1985)[1]. Today the engine is semiconductor mask inspection — the ACTIS, ABICS and MATRICS lines — with smaller businesses in flat-panel inspection and laser microscopy[13].
The revival
The pivot was born of necessity. After the 2008 crash wiped out demand for LCD-panel inspection, Lasertec's FY2009 collapsed to roughly ¥9.2B in sales and a ¥6.5B loss, with its market value falling below net assets[3]. Rather than defend a commoditized business, it pushed into the hardest problem in advanced chipmaking: verifying the photomasks used in EUV lithography. By June 2023 that bet had carried its market cap past ¥2 trillion — larger than far bigger industrial names — on fewer than 700 employees[3].
Timeline
Year
Milestone
1960
Founded in Yokohama as Tokyo ITV Laboratory, initially building X-ray cameras for medical equipment.
1976
Launches the world's first LSI photomask inspection system — the start of the semiconductor franchise.
1985–86
Introduces the world's first color laser scanning microscope; renamed Lasertec Corporation (1986).
2009
Post-Lehman, the display-inspection business collapses: ~¥9.2B sales, ~¥6.5B loss, market cap below net assets.
2017
Releases ABICS E120 — the first actinic EUV mask-blank inspection system.
2019
Releases ACTIS A150 — the world's first actinic EUV patterned-mask inspection system.
2022
Moves to the TSE Prime market; becomes the most actively traded stock on the Tokyo exchange.
2023
Added to the Nikkei 225; market cap surpasses ¥2 trillion with under 700 employees.
Jun 2024
Scorpion Capital publishes a 334-page report alleging fraud; shares fall for eight straight sessions.
Special investigation committee reports no accounting irregularities found.
FY2025
Record year — ¥251.5B sales, ¥84.7B net income — but orders fall ~60%; FY2026 guided to decline.
Leadership
On July 1, 2024 — weeks after the short-seller report — long-time president Osamu Okabayashi moved to Chairman, Tetsuya Sendoda became CEO, and Hisashi Yokokawa became CFO; the change was resolved at the April 30, 2024 board meeting, before the report[5]. Supporters note the timing was pre-planned; skeptics argue a CFO change landing amid a fraud allegation is, at minimum, awkward optics[39].
“Lasertec's inspection equipment is indispensable for chipmakers to use EUV efficiently.”
original · ja“レーザーテックの検査装置は不可欠だ。”
Nikkei Business · on the EUV pivot · 2024 · source
What the history shows in its favour
+A genuine turnaround: from a ¥6.5B FY2009 loss to a near-monopoly worth over ¥2 trillion[3].
+A repeatable innovation culture — a long line of “world-first” systems[1][2].
+Lean and focused: outsized value created by fewer than ~700 people[3].
What the history flags as caution
−The same display business once thought solid collapsed almost overnight — concentration cuts both ways[3].
−A ~1,500–1,800% run made it a “story stock,” inviting volatility and short interest[4][38].
−A leadership reshuffle arriving with a fraud allegation raised governance questions[39].
Market & Value Chain
A tiny market at the choke point of advanced chips
Mask inspection is a small slice of semiconductor equipment by dollars — but a step no leading-edge chip can skip. Lasertec sits at the narrowest, most defect-sensitive point of the EUV supply chain.
Every advanced chip needs a defect-free photomask, and the only way to truly verify one is to inspect it with the same 13.5nm light the scanner uses — “actinic” inspection[6]. That makes Lasertec's niche mission-critical but small: the entire EUV mask-blank market was only ~$194M in 2024, growing to ~$551M by 2032[7]. A choke point, not a vast market.
Why “actinic” matters
EUV lithography prints chips with light at a 13.5nm wavelength — far shorter than the older deep- ultraviolet (DUV) light. A photomask is the master stencil; a single defect on it repeats on every wafer it prints. The catch: a defect that is invisible to ordinary optical inspection can still “print” under 13.5nm light. Only inspecting at 13.5nm — actinic inspection — sees the mask the way the scanner does, catching printable defects others miss[6][9]. As nodes shrink below 7nm, that capability moves from nice-to-have to mandatory[8].
Where Lasertec sits
The leading-edge mask supply chain, from light source to fab. Lasertec's tools verify the mask between writing and high-volume manufacturing.
EUV scanner
ASML
Projects 13.5nm EUV light through a mask onto the wafer. ~$200–300M per tool.
Mask blank
AGC · Hoya
Defect-free multilayer substrate. AGC + Hoya ≈ 93% of the market.
Mask writing
NuFlare et al.
E-beam writers pattern the circuit onto the blank.
Mask inspection
Lasertec · KLA
Verify the mask before it prints billions of times. Lasertec owns actinic.
Wafer fab
TSMC · Samsung · Intel
Run the qualified mask in high-volume manufacturing.
A concentrated supply chain
Lasertec's fate is tied to a handful of upstream and downstream players. The mask blanks it inspects come almost entirely from two Japanese firms — AGC and Hoya, together ~93% of production-grade EUV blanks[7] — and the masks ultimately run in the fabs of TSMC, Samsung and Intel. That concentration is a double-edged sword: it makes Lasertec's tools indispensable, but it leaves demand hostage to a few customers' capex decisions[8].
🔬
The bottleneck framing
Fewer than a dozen actinic inspection tools exist worldwide; analysts call qualified mask inspection a “silent bottleneck” in EUV — small in dollars, large in consequence[9].
Why the market position is attractive
+A mandatory step with no real substitute at 13.5nm — actinic inspection is becoming unavoidable below 7nm[6][8].
+Sits at the most defect-sensitive choke point of the EUV chain, where tool value is high[9].
+Upstream concentration (AGC/Hoya) and EUV adoption both pull demand toward Lasertec[7].
Why the market is also limiting
−The addressable market is small in absolute dollars (~$194M mask-blank market in 2024)[7].
−Demand depends entirely on EUV adoption pace — a niche, not a broad equipment market[7].
−A chain this concentrated transmits any single customer's pause straight to Lasertec[8].
Business Model
Fabless optics, monopoly margins
Lasertec designs the optics and algorithms, outsources the metal-bending, and sells a handful of multi-million-dollar tools a year to the world's most advanced chipmakers — at margins most hardware makers never see.
48.8% operating margin (FY2025)~92% sales overseas~70% of staff are engineers
Lasertec runs a fabless model — it owns the optics and software and outsources precision parts, doing final assembly in Japan[10]. The result is a 48.8% operating margin in FY2025, far above the machinery-industry norm[11] — but the revenue comes almost entirely from abroad (~92%), concentrated in a few chipmaking regions[12].
How it makes money
Lasertec sells capital equipment: a small number of high-value inspection systems (ACTIS, ABICS, MATRICS) each priced in the millions to tens of millions of dollars, with long 1–2 year lead times. Because it is fabless — concentrating on R&D and algorithms and outsourcing production to vetted suppliers — fixed costs stay low and gross margins stay high[10]. Roughly 70% of its workforce are engineers[10]. A growing service and maintenance stream rides on the expanding installed base, adding more recurring, higher-visibility revenue over time[13].
📈
Why the margins are so high
With one commercial actinic supplier and customers who cannot run leading-edge fabs without qualified masks, Lasertec has real pricing power — the 48.8% FY2025 operating margin is the financial fingerprint of a near-monopoly[11].
Where the revenue comes from
FY2025 net sales by region (by customer location). About 92% is overseas; the US, Taiwan and Korea — home to the leading-edge fabs and chip designers — dominate.
Lasertec FY2025 net sales by region (%)
United States — 27%
Taiwan — 26%
South Korea — 21%
Other Asia — 9%
Japan — 8%
Europe — 8%
The concentration trade-off
The flip side of those margins is dependence. With ~92% of sales overseas and the bulk in just three regions — the United States, Taiwan and South Korea[12] — Lasertec carries currency, customer and geopolitical exposure that a more diversified vendor would not. Demand also arrives in lumps: a single large order or cancellation can swing a year (as FY2025's order decline showed). High margins and high concentration are two sides of the same coin.
Strengths of the model
+Fabless structure keeps fixed costs low and capital efficiency high[10].
+Near-monopoly pricing power → a 48.8% FY2025 operating margin[11].
+A growing installed base feeds recurring, higher-margin service revenue[13].
Fragilities of the model
−~92% of sales overseas concentrates currency and geopolitical risk[12].
−Revenue is lumpy — a handful of multi-year orders drive each year[10].
−Outsourced production depends on scarce specialty components (EUV sources, precision optics)[10].
Competitive Landscape
Alone in actinic, shoulder-to-shoulder with KLA
In its core niche Lasertec has no commercial rival. One step out — into the broader mask-inspection market — it shares the field with KLA, a far larger company whose customers would welcome a second actinic source.
Sole actinic supplierKLA + Lasertec ≈ 65–70% of mask inspectionTSMC + Intel + Samsung ≈ 77% of revenue
Lasertec is the only commercial supplier of actinic patterned-mask inspection[41], but the wider mask-inspection market is shared with KLA, whose Teron tools dominate die-to-database pattern inspection; together they hold an estimated 65–70% of mask-inspection revenue[14]. The competitive pressure is less about today's rivals than about buyer power and a possible second source.
The KLA question
KLA is the broad leader in semiconductor process control and reticle inspection, with optical and e-beam (die-to-database) tools used in mask shops worldwide[14]. What it does not sell, today, is a commercial actinic patterned-mask inspection tool — the exact niche Lasertec owns. The debate is whether that gap holds: bulls point out that by 2027 over 80% of high-volume masks are expected to need actinic inspection, a wave Lasertec is positioned to ride[15]; bears note KLA has publicly researched actinic inspection and customers want competition (covered in The Actinic Moat).
Five Forces — mask inspection
Click a force for the rated pressure and the evidence behind it. The standout is not rivalry but buyer power — Lasertec's customers are few, sophisticated, and decisive.
EUV mask inspection
Competitive rivalry — Medium. In its core niche — actinic (13.5nm) EUV patterned-mask inspection — Lasertec is effectively the only commercial supplier, so direct rivalry is low. But the broader EUV mask-inspection market is shared with KLA, whose Teron e-beam/optical reticle tools dominate die-to-database pattern inspection; the two together hold an estimated ~65–70% of mask-inspection revenue. Rivalry is real one step out from the actinic niche.
Customer power is the real pressure
Lasertec's buyers are among the most powerful in technology: per Scorpion Capital's analysis, TSMC, Intel and Samsung alone are roughly 77% of revenue[16]. They co-develop tools, qualify them over years, and can delay or cancel orders — which is exactly what helped push the ACTIS line to negative bookings in FY2025. A monopoly on the supply side meets an oligopsony on the demand side; the balance of power is more even than “monopoly” suggests.
Lasertec's competitive strengths
+Sole commercial actinic supplier — no direct rival in its core niche[41].
+A secular tailwind: >80% of high-volume masks expected to use actinic inspection by 2027[15].
+Years of customer qualification make switching costly and slow[14].
Where it is exposed
−KLA is far larger and dominates adjacent die-to-database pattern inspection[14].
−Extreme buyer power: three customers ≈ 77% of revenue[16].
−Customers' desire for a second source is a standing invitation to a rival[14].
The Actinic Moat
A monopoly built on physics — and on no one else trying yet
Lasertec's edge is structural: only it sells tools that inspect EUV masks at the exposure wavelength. But the moat rests on technical difficulty and time, not patents or law — and the field is empty by choice, not by rule.
The moat rests on a measurable lead: ACTIS (2019) was the world's first actinic EUV patterned-mask inspection system and ABICS (2017) the first for mask blanks[17][18], and Lasertec keeps extending the lead with faster tools (ACTIS A200HiT, 3× speed) and high-NA-ready systems[19]. The vulnerability: the lead is not legally protected, and KLA has researched the same actinic approach[42].
What makes the moat deep
Three things compound into a real barrier. First, physics: building a tool that inspects with a 13.5nm EUV light source requires an EUV source and extreme-precision optics that few organizations can assemble. Second, a decade of first-mover learning: ABICS shipped in 2017 and ACTIS A150 in 2019, and the installed base has been refined ever since[17][18]. Third, customer qualification: chipmakers spend years validating an inspection tool before trusting it in production, so even a capable challenger faces a long, slow path to adoption.
A continuous product cadence
Lasertec defends the lead by moving the target. The ABICS E320 (late 2024) tightened defect thresholds for high-NA EUV mask blanks[20], and the ACTIS A200HiT (October 2025) delivers triple the inspection speed of the A150 for incoming and periodic fab inspection while keeping sensitivity[19]. Each generation raises the bar a would-be rival must clear.
Positioning
Breadth of portfolio vs. depth of actinic-EUV / leading-edge lock-in. Hover a point for the basis. Lasertec is the deepest, narrowest specialist; the giants are broad but not in commercial actinic patterned-mask inspection.
Hover a point to see the basis for its placement.
⚠️
The moat's soft spot
A monopoly that exists because rivals haven't built the product is different from one protected by patents or scale. KLA's published work on actinic patterned-mask inspection shows the approach is not Lasertec's alone — and customers have every incentive to fund a second source[42].
Why the moat should hold
+First-mover since 2017–2019 with a continually refreshed product line (A200HiT, E320)[17][19].
+Enormous technical barrier — EUV source + optics + years of customer qualification[18].
+Demand tailwind: actinic expected on >80% of high-volume masks by 2027[15].
Why it could erode
−No legal protection; KLA has publicly researched actinic patterned-mask inspection[42].
−Powerful customers actively want a second source to break the monopoly[14].
−A single-product moat means one technical leapfrog could reset the field[42].
Financials & Orders
A record year, and an order cliff
FY2025 was the best year in Lasertec's history by every income measure. In the same twelve months, orders fell roughly 60% — and management guided FY2026 to the first profit decline of the EUV era.
FY2025 (ended June 2025) set records: net sales ¥251.5B (+17.8%), operating income ¥122.8B (+51.0%), net income ¥84.7B (+43.3%)[22]. But orders received collapsed ~60% to ¥105.2B, and FY2026 is guided down — the first profit fall after the record streak[22][23].
Net income rose every year — until now
Net income, ¥B, fiscal years ending June. FY2026 is the company's guidance (~¥60B), not a reported figure.
Lasertec net income, FY2021–FY2026E (¥B)
Across five years revenue compounded from ¥70.2B (FY2021) to ¥251.5B (FY2025), and net income from ¥19.3B to ¥84.7B[21]. The FY2025 operating margin reached 48.8% — the high-water mark of the monopoly economics[22]. Then the order book turned.
The order cliff
Orders received vs. sales, FY2024–FY2025, ¥B. Record shipments drew down a backlog booked in prior years even as new orders fell sharply.
Lasertec orders, sales and backlog (¥B)
Net sales FY25
¥251.5B
Orders FY24
¥272.8B
Orders FY25
¥105.2B
Backlog FY25
¥315.9B
Orders received fell from ¥272.8B to ¥105.2B, and the backlog shrank from ¥462.2B to ¥315.9B[22]. The ACTIS line alone booked negative on a single cancellation and a lost order — a vivid illustration of how lumpy this business is[37]. Record revenue and a collapsing order book in the same year is the central tension of the Lasertec story.
Two derived ratios size that tension (our arithmetic over the disclosed figures; both illustrative). FY2025 book-to-bill — orders received ÷ net sales — was ¥105.2B ÷ ¥251.5B ≈ 0.42[22]: for every ¥1 of equipment shipped, only about ¥0.42 of new work was booked. And the ¥315.9B year-end backlog covers the revised ~¥220B FY2026 sales outlook for only ≈ 1.4 years (315.9 ÷ 220)[22][24]. The implication is checkable: for the backlog not to shrink again, FY2026 orders must roughly double from ¥105.2B toward the ~¥220B shipment pace — the single number to watch at the August 2026 full-year results.
FY2026: decline, then maybe recovery
For FY2026 (ending June 2026) Lasertec guided to lower results — initially ¥200B sales / ¥85B operating income / ¥60B net income, its first guided decline after the record run[23]. By October 2025 it had raised the sales outlook to ~¥220B as order enquiries improved, and management expects bookings to recover in calendar 2026 on the 2nm/GAA and A16/A14 ramp and AI demand[24]. Through H1 FY2026, sales were ~¥128.3B (roughly flat) with net income ~¥45.7B (+5.6%)[25] — a softer year, but not a collapse.
📊
Bull and bear read the same numbers
Bulls see a backlog-driven record, a raised outlook and a 2026 recovery setup[24]. Bears see a 60% order drop and a guided profit decline as the cycle's first crack[23]. Both are reading the same disclosures — but the raised outlook and a roughly flat first half[25] give the pause reading the better of it, provided bookings actually recover in calendar 2026 as management expects[24].
+Outlook raised to ~¥220B for FY2026; management sees orders recovering in 2026[24][25].
+Five years of compounding revenue and profit through FY2025[21].
Financial cautions
−Orders fell ~60% and the backlog shrank sharply in FY2025[22].
−FY2026 guided to the first profit decline of the EUV era[23].
−The ACTIS line booked negative on a single cancellation — demand is lumpy and cyclical[37].
The Short-Seller Attack
'The largest outright fraud in the world' — and the rebuttal
In June 2024 Scorpion Capital aimed one of the most aggressive short reports in recent memory at Lasertec. The company denied it; a special committee found no fraud. On the weight of the evidence the episode reads as an overreach — one that nonetheless left legitimate disclosure questions standing.
334-page report · Jun 2024Committee: no irregularitiesHBS case study
On June 5, 2024 Scorpion Capital published a 334-page report calling Lasertec “the largest outright fraud in the world”[26]. Lasertec “clearly denied” improper accounting[28], and an Aug 2024 special committee found no irregularities[29] — yet the stock fell for eight straight sessions and the questions never fully cleared[34].
What Scorpion alleged
The report — released in both English and Japanese — made overlapping technical, operational and accounting claims[27]:
Accounting: that Lasertec books equipment as “work-in-process” rather than finished goods to avoid write-downs, inflating inventory, revenue and earnings[27].
Product: that the ACTIS tool has a “fatally flawed” EUV light source, is inferior to KLA's and kept only for R&D, and that the next-gen ACTIS A300 was a “hoax”[26].
Operations: that investigators found the Yokohama innovation park “deserted,” and that customers like TSMC had stopped buying its EUV machines[26][16].
Governance: a 2018 auditor switch, which short-sellers associate with concealment[27].
Lasertec's response
Lasertec rejected the allegations on June 6, 2024, saying it “clearly denies” improper accounting[28]. Its substantive defence on the inventory point: its tools are customized and formally accepted at customer sites, so until acceptance they are correctly held as work-in-process under standard accounting; the swelling work-in-process reflects 1–2 year lead times and order growth, not concealment[28]. It separately said inventory was valued properly and ACTIS orders were strong[32].
“Until acceptance, items are not treated as products but are recognized as work-in-process under generally accepted accounting standards.”
original · ja“検収時点までは製品とはみなされず、一般に公正妥当と認められる会計基準に従い仕掛品と認識している。”
Lasertec · rebuttal, via Nikkei · 6 Jun 2024 · source
The independent investigation
Lasertec convened a special investigation committee — two outside directors plus one external expert — that worked from June 18 to August 5, 2024, verifying the existence of work-in-process inventory through stocktaking and interviews with logistics staff and management[30]. On August 5, 2024 it concluded there were no work-in-process items casting doubt on asset value and that no fraud was found[29]. The stock rebounded on the finding.
“There were no work-in-process items casting doubt on asset value; no irregularities were recognized.”
original · ja“資産性に疑義を生じさせる仕掛品はなかったとして「不正は認められなかった」と結論づけた。”
Special investigation committee, via Nikkei · report · 5 Aug 2024 · source
How to weigh it
Independent observers landed in the middle. A detailed Japanese analysis judged some Scorpion points worth scrutiny (the inventory build, the thorough fieldwork) but found the core thesis weak: a “product nobody can use” is hard to square with surging sales, and the report offered no plausible motive — no controlling shareholder stood to benefit from inflation[31]. One analyst summed the report up as “some of the accusations are disgraceful, some not”[43]. The episode is now taught as a Harvard Business School case on short-seller attacks and how a “story stock” defends itself[33].
⚖️
The honest read
A committee found no fraud, and surging shipments are hard to fake — but a self-convened committee is not an outside regulator, and the market took eight sessions to stop falling[29][34]. The weight of evidence favours Lasertec; certainty does not.
Why the fraud thesis looks weak
+An independent committee found no accounting irregularities after stocktaking and interviews[29][30].
+The work-in-process treatment fits long lead times and acceptance-based revenue recognition[28].
+No plausible motive and surging real sales undercut a “useless product” claim[31].
Why some questions linger
−The committee was company-convened, not an independent regulator's probe[30].
−The stock fell for eight straight sessions; the market didn't treat it as fully settled[34].
−Analysts judged at least some accusations as legitimate questions, not pure noise[43].
Peer Comparison
The smallest giant in the room
Against the wafer-fab-equipment majors, Lasertec is tiny — a fraction of even KLA, its nearest rival. But on profitability and niche dominance it stands apart, which is exactly why it trades like something bigger.
By revenue Lasertec (~$1.7B) is a rounding error next to ASML, Applied Materials and Tokyo Electron, and roughly one-seventh of KLA, its closest mask-inspection rival[35]. What sets it apart is focus and margin: a single-niche monopoly earning a 48.8% operating margin[11].
Scale: a specialist among giants
Most-recent fiscal-year revenue, US$B (approximate; fiscal years and FX differ). ASML, Applied and Tokyo Electron are shown for scale context — they are broad-line equipment makers, not direct mask-inspection rivals.
Revenue vs. equipment peers (US$B, most-recent FY)
ASML
$35B
Applied Materials
$28.4B
Tokyo Electron
$16B
KLA
$12.2B
Lasertec
$1.7B
Margin: where the monopoly shows
Operating margin, %. Lasertec's FY2025 figure is reported; the others are approximate most-recent-FY figures shown for scale, not exact comparisons.
Operating margin, % (approximate, most-recent FY)
Lasertec
48.8%
KLA
40%
ASML
33%
Applied Materials
29%
How the players relate
These companies are interdependent rather than head-to-head. ASML makes the EUV scanners; AGC and Hoya make the mask blanks; mask writers pattern them; and Lasertec and KLA inspect them[36]. TSMC is a confirmed Lasertec equipment buyer[36]. The only true overlap is Lasertec vs. KLA in mask inspection — and even there they specialize: KLA in die-to-database pattern inspection, Lasertec in actinic.
Company
Role in the chain
Relationship to Lasertec
KLA
Process control / reticle inspection
Main rival; far larger; no commercial actinic tool yet
ASML
EUV lithography scanners
Upstream; Lasertec's tools serve ASML's masks
AGC · Hoya
EUV mask blanks (~93%)
Customers for ABICS blank inspection
Applied Materials · Tokyo Electron
Broad WFE (deposition, etch, etc.)
Scale context, not direct rivals
🔎
Why size isn't the whole story
Lasertec is small but un-substitutable in its niche — a different risk/reward than a broad-line peer. Investors have rewarded the focus with a premium valuation, which is also its biggest vulnerability (see Risks)[38].
What the comparison flatters
+Best-in-class operating margin (48.8%), above the broad-line giants[11].
+Sole supplier in a niche none of the giants commercially serve[36].
−~$1.7B revenue — a fraction of KLA (~$12B) and the WFE majors[35].
−One product line vs. peers' diversified portfolios — far less shock-absorbing[35].
−A premium valuation leaves little margin for a cyclical stumble[38].
Risks & Skeptics
What could go wrong
The bull case rests on a durable monopoly riding an AI-driven EUV super-cycle. The bear case is that a tiny, single-product, customer-concentrated 'story stock' is uniquely exposed to a cyclical air-pocket and a second source.
Four risks dominate: customer concentration (TSMC/Intel/Samsung ~77%)[16], cyclicality (a ~60% FY2025 order drop on a single cancellation)[37], valuation (a story stock that swung from ~¥13,700 to ~¥46,700 in a year)[38], and a lingering governance overhang from the 2024 short report and the close-timed CFO change[39].
Concentration and cyclicality
The same focus that drives Lasertec's margins makes it fragile. With three customers at ~77% of revenue[16] and a single product family, one buyer's capex pause moves the whole company — as FY2025 showed, when the ACTIS line booked negative on one cancellation and a lost order and total orders fell ~60%[37]. Semiconductor equipment has always been cyclical; a monopoly does not exempt Lasertec from the cycle, it just concentrates the swings.
Share of Lasertec revenue: top-3 customers vs. everyone else (%)
TSMC + Intel + Samsung
77%
All other customers
23%
Top three named customers are ~77% of revenue[16]; the ~23% remainder is the arithmetic complement (100% − 77%), shown illustratively. The split quantifies the concentration the bear case turns on.
Valuation and the “story stock” problem
Lasertec's ~1,500–1,800% run made it a retail favourite and Japan's most-traded stock — and a magnet for shorts. Over one 52-week window the shares ranged from roughly ¥13,715 to ¥46,720[38], a swing that implies investors are pricing a wide range of outcomes. A premium multiple on a cyclical, single-product business leaves little room for a stumble.
Governance overhang
The 2024 Scorpion report (see The Short-Seller Attack) was dismissed by a special committee, but the optics of a CEO and CFO change landing weeks later emboldened short-sellers and left a reputational residue[39]. The change was board-approved before the report — but perception lags facts.
Geopolitics — a double edge
US and Dutch export controls lock China out of leading-edge EUV, so Lasertec has little direct China leading-edge business to lose — and Chinese inspection-tool makers are kept away from the leading edge too, insulating the actinic franchise[40]. The flip side: it ties Lasertec's fate even more tightly to the capex of a few non-Chinese fabs, and any tightening that slows TSMC/Samsung/Intel investment would hit it directly.
SWOT
Strengths
•Near-monopoly in actinic EUV mask inspection: ACTIS (patterned masks) and ABICS (mask blanks) are the only commercial actinic tools, inspecting at the 13.5nm exposure wavelength (s17, s18, s41).
•Exceptional economics: FY2025 operating margin 48.8% on a fabless model with ~70% engineers and low fixed costs (s10, s11).
•Record FY2025 — ¥251.5B sales (+17.8%), ¥84.7B net income (+43.3%) — and a fresh product cadence (ACTIS A200HiT, ABICS E320) extending the franchise into high-NA EUV (s19, s20, s22).
Weaknesses
•Tiny and undiversified: ~$1.7B revenue against KLA's ~$12B, with fate tied almost entirely to EUV mask inspection and a few buyers (s35, s16).
•Extreme customer + geographic concentration: TSMC/Intel/Samsung ~77% of revenue; ~92% of sales overseas (s16, s12).
•Lumpy, cyclical bookings: FY2025 orders fell ~60% and the ACTIS line booked negative on a single cancellation/lost order (s22, s37).
Opportunities
•The 2nm/GAA and A16/A14 ramp plus high-NA EUV lift demand for actinic inspection — expected on >80% of high-volume masks by 2027 (s15, s24).
•Export controls lock China out of leading-edge EUV, insulating Lasertec's actinic franchise from Chinese inspection-tool entrants at the leading edge (s40).
•A growing installed base feeds higher-margin recurring service/maintenance revenue (s13).
Threats
•KLA (which has researched actinic inspection) or a state-backed entrant breaking the monopoly, especially as customers seek a second source (s43, s14).
•Capex cyclicality and customer pauses: a single buyer's delay swings results, as the FY2025 order cliff showed (s37, s23).
•Reputational/governance overhang from the 2024 Scorpion short report and the closely-timed CEO/CFO change, plus story-stock valuation volatility (s26, s39, s38).
Source ids (e.g. s11, s16) refer to the Sources page; each SWOT item is backed in the section prose above.
🚩
The bear's strongest point
Strip away the monopoly narrative and you have a ~$1.7B, single-product company with three customers and a cyclical order book trading at a premium — the kind of profile that punishes a single bad year[37][38].
Why the risks may be manageable
+Export controls insulate the actinic franchise from Chinese rivals at the leading edge[40].
+The fraud allegation was investigated and dismissed; surging shipments are hard to fake[16].
+A 2nm/high-NA tailwind underpins a 2026 order recovery the company expects.
Why the risks may bite
−Three customers ≈ 77% of revenue — extreme concentration[16].
−Orders fell ~60% on a single cancellation; the cycle can turn fast[37].
−A volatile premium valuation plus a governance overhang amplify any miss[38][39].
The weighing
On the durability of the actinic mask-inspection monopoly: the evidence leans toward a real but legally unprotected moat (medium confidence). The controlling evidence is that Lasertec remains the only supplier of actinic (13.5nm) mask inspection while running a 48.8% operating margin[22], which outweighs the second-source threat because no rival has yet shipped a qualified competing tool at production volume; the moat is technical lead-time, not patent. The strongest surviving counter-argument: the position rests on three customers — TSMC, Intel and Samsung — that together make up ~77% of revenue[16], any one of which could fund and qualify a KLA alternative. What would flip this reading: a credibly qualified actinic inspection tool from KLA winning a named fab's purchase order before FY2027 close; or Lasertec's operating margin compressing below 40% for two consecutive fiscal years as pricing power erodes. Pre-mortem: if this looks wrong in two years, the most likely reason is that the bull underweighted how fast a single ~$12B rival[16] can close a one-product gap — or, on the other side, that the bear underestimated how sticky actinic qualification is once a fab has validated it.
On whether record profit masks a turning cycle: the evidence leans toward a genuine downturn already underway (high confidence). The controlling evidence is that FY2026 is guided to the first profit decline, with revenue stepping back to roughly ¥220B[23] after FY2025's record ¥251.5B (+17.8%)[11], which outweighs the headline record because guidance reflects an order book that already fell ~60% on a single cancellation[37]. The strongest surviving counter-argument: orders this concentrated are lumpy, so one cancellation can overstate a trend that a single 2nm or high-NA award could reverse. What would flip this reading: book-to-bill climbing back above 1.0 for two straight quarters through FY2026; or management reporting a return to revenue growth above the prior record peak by FY2027 results. Pre-mortem: if this looks wrong in two years, the most likely reason is that the bear extrapolated a lumpy order book into a structural decline — or, on the other side, that the bull treated a record print as a floor rather than a peak.
On whether the fraud allegation was ever substantiated: the evidence leans toward not substantiated (medium confidence). The controlling evidence is that the special committee convened June–August 2024 found no irregularities[29], which outweighs the 334-page Scorpion report[26] because the latter remains an unverified short thesis rather than a finding by an independent reviewer with access to the books. The strongest surviving counter-argument: the market did not treat the matter as closed — the stock fell eight straight sessions around the episode[34], signalling residual doubt a committee review did not dispel. What would flip this reading: a regulator or auditor opening a formal investigation or issuing a restatement before FY2027; or a second independent report surfacing specific accounting discrepancies that the special committee did not address. Pre-mortem: if this looks wrong in two years, the most likely reason is that the bull trusted a company-appointed committee too readily — or, on the other side, that the bear mistook a sharp share-price reaction for proof of the underlying claim.
Methodology & Limits
How this was built — and where it's uncertain
An independent, neutral compilation. Every load-bearing claim traces to a source fetched during the research run. Where the evidence supports a lean, the study states one — and shows the work behind it.
As of 7 June 202644 sourcesIndependent
🔍
Independence
This is an independent research artifact. It is not affiliated with, sponsored by, or endorsed by Lasertec Corporation or any competitor, and it is not investment advice — no rating, price target, or recommendation to buy or sell any security. No relationship, no compensation, no access beyond public sources.
How the research was done
We ran fan-out web search and fetched the underlying pages, in both English and Japanese: Lasertec's own IR and product pages[6][17], Japanese financial press and analysts (Nikkei, Rakuten Securities/Toushiru, kabutan)[22][28], English trade and financial press (Fortune, Bloomberg, Semiconductor Engineering)[4][9], third-party market data for industry sizing[7], and the primary documents of the 2024 short-seller episode and the company's response[26][29]. Roughly a third of the sources are Japanese-language, where the domestic debate — especially around the short report — is richest. Every URL cited was opened during the run; the bibliography is machine-checked for dead links.
Frameworks used
Pyramid Principle — answer-first synthesis (the balanced state of the debate), then evidence.
Porter's Five Forces — industry structure of EUV mask inspection.
2×2 positioning — portfolio breadth vs. actinic-EUV / leading-edge lock-in.
Peer comparables — revenue and operating margin vs. KLA and the WFE majors.
SWOT — applied even-handedly, with weaknesses and threats given equal weight.
Disclosed vs. estimated
Lasertec's revenue, operating income, net income, orders and backlog are disclosed figures from its FY2025 results and FY2026 guidance, reported via Japanese press[22][23]. Treat as estimates: the EUV mask-blank market size and the ~93% AGC/Hoya share (third-party estimates)[7]; the ~77% customer-concentration figure (from Scorpion's analysis, not a company disclosure)[16]; peer revenues and operating margins (most-recent fiscal year, different period-ends, approximate FX at ~¥150/$); and the FY2026 outlook (company guidance, not a result). Valuation and share-price ranges move daily.
Neutrality
Each section presents both supporting and countervailing evidence and a two-sided ledger. The source base is tagged by stance — supporting, critical and neutral in roughly balanced shares — to keep the compilation balanced rather than advocating. The 2024 fraud allegation is presented with the short-seller's claims, the company's rebuttal, and the independent committee's finding — and then weighed: this study's read is that the committee's finding and the absence of a plausible motive outweigh the unresolved optics. Where we interpret, we say so and show the basis.
🗓️
This is a point-in-time artifact dated 7 June 2026. Lasertec's next quarterly results (its fiscal year ends June), the pace of the 2nm/high-NA EUV ramp, any move by KLA into actinic inspection, and shifts in export policy can each change the picture quickly. Re-check the primary sources before relying on any figure.
Bibliography
Sources
Every cited source was fetched or read during the research run. Tiers: 1 = primary/official (Lasertec releases, IR, the HBS case, KLA), 2 = reputable press/research (Nikkei, Rakuten/Toushiru, Fortune, Bloomberg, Semiconductor Engineering), 3 = tertiary (market-data sites, community/analyst write-ups).
Lasertec joined the TSE Prime market (April 2022), was added to the Nikkei 225 (2023), and became the most actively traded stock on the Tokyo Stock Exchange; shares rose roughly 1,500–1,800% over five years.
Lasertec is a Japanese optical-inspection maker (TSE: 6920) whose actinic EUV mask-inspection tools are mission-critical to advanced chipmaking, giving it a near-monopoly in actinic patterned-mask and mask-blank inspection.
Independent analysts read the Scorpion episode as mixed — some accusations are baseless, but others raise legitimate questions — rather than a clean exoneration, underscoring that reasonable observers still disagree about Lasertec.
Lasertec was founded in 1960 as Tokyo ITV Laboratory in Yokohama; developed the world's first LSI photomask inspection system (1976) and color laser scanning microscope (1985); renamed Lasertec in 1986.
After the 2008 crisis, Lasertec's display-inspection business collapsed (FY2009: ~¥9.2B sales, ~¥6.5B loss, market cap below net assets); it then pivoted to EUV mask inspection and grew revenue roughly sevenfold, surpassing ¥2 trillion market cap by June 2023 with under 700 employees.
Leadership change effective July 1, 2024: Osamu Okabayashi moved from President & CEO to Chairman; Tetsuya Sendoda became CEO; Hisashi Yokokawa became CFO. Resolved at the April 30, 2024 board meeting.
Lasertec became a heavily-traded 'story stock': shares ranged from roughly ¥13,715 to ~¥46,720 over a 52-week window — extreme volatility that makes valuation a live risk, and it has been a favourite of both Japanese retail and short-sellers.
EUV lithography uses 13.5nm light; actinic inspection (inspecting at the same 13.5nm wavelength) detects printable defects that DUV/optical tools miss, and is becoming essential as nodes scale below 7nm.
The EUV mask blanks market is highly concentrated — AGC and Hoya together hold ~93% of production-grade EUV mask blanks; the market was ~US$194M in 2024, projected to ~US$551M by 2032 (~15% CAGR).
EUV mask blank/photomask makers (Hoya, AGC) invest heavily in actinic inspection systems to identify phase/amplitude defects; mask blank business is heating up as EUV scales.
Lasertec's tools verify masks used with ASML's EUV scanners (which cost ~$200–300M each); actinic inspection at 13.5nm finds more defects than optical techniques — a quiet bottleneck in the EUV supply chain.
Lasertec runs a fabless model — it concentrates on R&D and algorithms, outsources precision parts to vetted suppliers, and does final assembly/calibration in Japan. ~70% of staff are engineers.
Lasertec posted FY2025 operating margin of 48.8% (operating income ¥122.8B on ¥251.5B sales) — far above the Japanese machinery-industry average; analysts attribute the pricing power to its near-monopoly position.
Lasertec sells almost entirely overseas (~92% of FY2025 sales): US ~27%, Taiwan ~26%, South Korea ~21%, Other Asia ~9%, Japan ~8%, Europe ~8% (by customer location).
Lasertec's revenue mix is anchored in EUV-related semiconductor inspection (ACTIS, ABICS, MATRICS lines), with a growing service/maintenance stream as the installed base expands.
KLA is the dominant force in reticle/mask pattern inspection (Teron platforms, die-to-database e-beam/optical); Lasertec and KLA together hold an estimated ~65–70% of total EUV mask-inspection revenue, but Lasertec is alone in commercial actinic patterned-mask inspection.
By 2027 over 80% of high-volume EUV masks are expected to undergo actinic inspection — the segment Lasertec currently owns; actinic technology detects defects optical tools cannot.
ACTIS A150 (released 2019) is the world's first actinic EUV patterned mask inspection system; it inspects at the 13.5nm exposure wavelength and can inspect through pellicles.
Lasertec released the ABICS E120 actinic EUV mask blank inspection/review system in 2017 — the first to use EUV light to selectively detect printable defects on mask blanks.
Lasertec keeps extending the actinic franchise: the ACTIS A200HiT (Oct 31, 2025) triples A150 inspection speed for incoming/periodic fab inspection while maintaining sensitivity.
The ABICS E320 (introduced late 2024) improves defect-detection accuracy and resolution to support tighter thresholds required by high-NA EUV lithography.
Lasertec's actinic lead is not legally protected: KLA has publicly researched actinic patterned-mask inspection, and customers' desire for a second source is a standing pressure on the monopoly.
Lasertec's revenue and net income rose every year through FY2025: rev ¥70.2B/NI ¥19.3B (FY21), ¥90.4B/¥24.9B (FY22), ¥152.8B/¥46.2B (FY23), ¥213.5B/¥59.1B (FY24), ¥251.5B/¥84.7B (FY25).
FY2025 (ended June 2025) set records — net sales ¥251.5B (+17.8%), operating income ¥122.8B (+51.0%), net income ¥84.7B (+43.3%) — but orders received collapsed ~60% to ¥105.2B from ¥272.8B, and the backlog fell to ¥315.9B from ¥462.2B.
For FY2026 (ending June 2026) Lasertec guided to lower results — its first profit decline after years of records — initially net sales ¥200B / operating income ¥85B / net income ¥60B.
Lasertec expects orders to recover in calendar 2026 on the ramp of 2nm/GAA and A16/A14 nodes and AI demand; in Oct 2025 it raised its FY2026 outlook and the stock rose.
H1 FY2026 (Jul–Dec 2025) net sales were ~¥128.3B, roughly flat year-over-year, with net income ~¥45.7B (+5.6%); the full-year sales forecast was revised up to ¥220B.
On June 5, 2024 short-seller Scorpion Capital published a 334-page report calling Lasertec 'the largest outright fraud in the world,' alleging a flawed EUV light source, that the ACTIS A300 was a 'hoax,' a 'deserted' Yokohama innovation park, and inventory/revenue manipulation.
Scorpion's report (in English and Japanese) alleged Lasertec books equipment as work-in-process to avoid write-downs, that its tools are inferior to KLA's and kept only for R&D, and flagged a 2018 auditor switch; it claimed TSMC had stopped buying its EUV machines.
Lasertec rejected the allegations on June 6, 2024, saying it 'clearly denies' improper accounting, that equipment is customized and signed off at customer sites and so is held as work-in-process until acceptance, and that rising work-in-process reflects long 1–2 year lead times and order growth.
Lasertec convened a special investigation committee (two outside directors + one external expert) that worked June 18–Aug 5, 2024, verifying work-in-process existence via stocktaking and interviews; on Aug 5, 2024 it concluded no accounting irregularities were found.
Details of the special-committee process: composed of two outside directors and one external expert, running June 18–Aug 5, 2024, centering on the existence of work-in-process inventory; reported no fraud.
A balanced Japanese analysis: some Scorpion points (inventory build warrants scrutiny, thorough fieldwork) have merit, but the 'unusable product' claim is hard to square with strong sales growth, and the report offers no plausible motive (no controlling shareholder benefits).
The episode is studied as a Harvard Business School case ('Short Attack: Lasertec Faces the Sting'), reflecting the broader debate over short-seller reports, governance, and how a 'story stock' responds to attack.
Despite the 'clear denial,' the stock fell for eight straight sessions after the report as buying interest stayed weak — illustrating how hard such attacks are to fully rebut in the market.
Among process-control/inspection peers, KLA is by far the largest (process control / inspection leader, FY2025 revenue ~$12.2B); Lasertec is a focused specialist (~$1.7B equivalent FY2025) that owns the actinic EUV niche.
Lasertec's tools sit downstream of ASML's EUV scanners: ASML makes the EUV lithography machines, the mask blanks come from AGC/Hoya, masks are written (e.g. by NuFlare), and Lasertec/KLA inspect them — a tightly interdependent value chain. TSMC is a confirmed equipment buyer.
Lasertec's results swing with a handful of leading-edge buyers and the EUV capex cycle: the ~60% FY2025 order drop (ACTIS bookings turned negative on a cancellation/lost order) shows how concentrated and lumpy demand is.
The July 2024 CEO and CFO handover, landing weeks after the short report, emboldened short-sellers and raised governance questions about the timing of leadership change.
China is locked out of leading-edge EUV by US/Dutch export controls, so Lasertec's actinic franchise has little direct China leading-edge market — concentrating its fate in TSMC/Samsung/Intel/SK hynix capex.
Cross-checked at build time by an automated link checker. A few sources — notably some Nikkei and Bloomberg articles and the Smartkarma analyst note — sit behind paywalls or bot-walls; their headlines and the cited figures were read during the run, and equivalent figures are corroborated by open sources where possible. See Methodology & Limits.