Onto Innovation
Instruments that measure and inspect chips mid-manufacture without touching them, so a chip factory learns a wafer is bad before paying for the next step.
The scores
Growth3/4 ↑
Revenue grew 1.8% in FY2025 with Semilab inside it, then 22.1% in the first half of 2026. Real, but under two years and partly acquired.
Business quality2/4 ↓
Qualification-led positions at 190 customers argue for 3, but systems revenue fell in the boom year and margins are half the cohort's.
Financial position3/4 ↓
Zero-coupon notes maturing outside the horizon and strong operating cash, but $1.5 billion of new debt replaced a debt-free balance sheet.
Execution record2/4 →
Non-GAAP guidance is generally beaten, but GAAP earnings per share missed its guided range once, and restructuring rose in each of four years.
Per share capture3/4 →
Share count up 1.4% and acquisitions paid in cash, with a capped call well above the conversion price. Buybacks do not quite outrun compensation.
Potential2/4 ↑
The upside is a bounded cycle: contracts yet to convert and somebody else's capital plans, with the largest agreements conditional and unfiled.
Risk profile3/4 →
Three customers are 49% of revenue and China export controls bind, but the concentration is disclosed, falling, and no single event resolves the case.
Valuation2/4 →
The price assumes the margin convergence completes. On trailing operating profit it is twice the cohort; on the guided quarter annualised, near it.
Evidence quality3/4 →
The restructuring table, the reconciliations and the cohort are all filed primary sources, but the backlog is a quote and the latest call is unread.
The Verdict
We rate Onto neutral. It sells tools that inspect chips mid-manufacture, so a factory spots a bad wafer early. Sales are booming, but the business is mixed: it keeps far less of each sales dollar as profit than its named rivals, partly because it has booked restructuring charges four years running. The price looks cheap on sales and dear on profit, so it is full, not cheap: it assumes official profit rises to meet management's adjusted figure. The view hinges on reported margins reaching the promised range. We think the promise has missed once already and must show in the numbers.
The long caseAI spending is building new chip factories, which install inspection tools first; Onto rides only a small slice of that, and its reported margins must still rise.
The evidence under this - 8 claims
- Annualising the third-quarter GAAP operating income guide of $81.4-89.7 million gives roughly 38 times enterprise value - below the cohort's trailing median of 45. If the guided quarter is the run rate the premium is already gone.
- The pack's enterprise value nets only cash and equivalents against the notes; on the balance sheet's $1,882.2 million of cash and marketable securities the enterprise value is about $629 million lower, which cuts the multiple by roughly five percent.
- The company guided first-quarter 2026 GAAP diluted EPS to $0.74-$0.84 and delivered $0.67, below its own range, while the non-GAAP figure beat its range - so the GAAP half of this guidance has missed once in the last two quarters it was given.
- Restructuring has been guided as ending, and has instead risen in each of the last three fiscal years.
The company
A chip passes through hundreds of steps, and a flaw made early is often found only at the final test, after every later step has been paid for. Onto's instruments shine light at a wafer and work backwards from what returns to measure layers too thin to see, or photograph the whole wafer to spot defects.
It also sells machines that print patterns onto chip packages, and software that turns the readings into decisions. Its newest inspection tool sees details below 200 nanometres, against about 800 in the past.
The evidence under this - 3 claims
- The customer is not named and the win is not quantified; management said G5 was under 10% and perhaps 5% of advanced packaging growth in 2026.
- Camtek and KLA sell competing advanced-packaging inspection and neither has disclosed losing this position.
How it makes money
Onto is paid mainly when a chipmaker buys a tool, and chipmakers buy when they build new production lines, so orders follow their building plans. Three unnamed customers took 49% of 2025 sales.
The evidence under this - 2 claims
Where the money comes from, stream by stream
| Stream | Share of sales (first half of 2026) | How it earns it |
|---|---|---|
| Systems and software | 85.2% | Metrology, inspection and packaging-lithography tools plus yield software, recognised mostly when a tool ships or is accepted. |
| Parts | 8.7% | Spare parts for the installed base of tools, sold as customers run and upgrade them. |
| Services | 6.1% | Service contracts and field support on installed tools, recognised over the contract term. |
Who runs it
Michael Plisinski has run Onto since it was formed by a 2019 merger; Brian Roberts became finance chief in 2025. Sales have nearly doubled under Plisinski, but cleanup charges have risen four years running. Insiders own about 0.5% of the company. The bonus rides on sales and on profit before those charges, and half the shares on how the stock does against other chip companies.
The evidence under this - 3 claims
- The operating-income goal is struck after adding back $63.4 million of restructuring and $39.4 million of amortisation, so the charges this thesis turns on (C06) cost the bonus nothing.
- FY2025 revenue of $1,005.3 million fell short of the $1,100.0 million target, so the plan did pay below target when the business missed.
Why it wins, and who could beat it
Why customers come back
A tool is approved for one named step at one named customer, and its recipes and library of known defects are built against that customer's process over months. Replacing it means re-approving the step, not swapping a box. KLA, the industry giant, names Onto as a competitor in its own annual report, and Onto holds 423 patents.
The weak spot showed in 2025: tool sales fell as fewer inspection units shipped for AI chip packaging, a market that was growing.
The evidence under this - 4 claims
The threat
KLA sells into the same factories and earned $12.2B from this kind of equipment last year, at a 41.7% operating margin. Camtek and Nova sell to the same chipmakers, and Camtek gets about half its sales from AI, including the packaging market where Onto's inspection sales fell.
The evidence under this - 3 claims
- Camtek's revenue is a fifth of Onto's, so it can grow off a smaller base without Onto losing absolute share.
- Onto's own advanced packaging revenue reached a record in the second quarter of 2026, so whatever happened in FY2025 has reversed at least once.
The rivals: KLAC, CAMT, NVMI, PDFS, and 1 more that is private, in-house or listed abroad.
Who else does this, rival by rival
| Ticker | Rival | How far along | Priced against it |
|---|---|---|---|
| KLAC | KLA CorporationPriced against it: Yes | On sale | Yes |
| CAMT | Camtek Ltd.Priced against it: Yes | On sale | Yes |
| NVMI | Nova Ltd.Priced against it: Yes | On sale | Yes |
| GigaVis Co. Ltd.Priced against it: Listed abroad | On sale | Listed abroad | |
| PDFS | PDF SolutionsPriced against it: One segment only | On sale | One segment only |
How big it could get
How big this could get
Chipmakers are forecast to spend $143.9B on factory equipment in 2026, up 23.1%. Measuring and inspection is about a tenth of that: four listed suppliers sell at least $14.6B a year of it, and KLA alone is $12.2B. Nobody publishes the exact size, so that is a floor, not the market.
The evidence under this - 3 claims
- SEMI's own December 2025 forecast had wafer fab equipment growing 9.0% in 2026; the mid-year revision to 23.1% is a large move in seven months and the series is a forecast, not a measurement.
- A forecast published by the industry's trade association is made by an organisation whose members benefit from the spending it projects.
- Three of the suppliers KLA names - Applied Materials, Hitachi High-Tech and Lasertec - publish no process-control line, so the true market is larger than this sum by an unknown amount.
- The periods do not align: KLA's year ends in June, the others in December or January, and the wafer fab equipment figure is a calendar-2026 forecast.
Who has signed
The backlog passed $1B in June, a figure that appears in a quote from the chief executive but not in the accounts. The largest single deal is a purchase agreement worth over $240M with an unnamed maker of high-bandwidth memory chips, supporting its expansion through 2027.
The evidence under this - 2 claims
- Onto tags no remaining performance obligation in its XBRL and discloses no backlog table, so the figure cannot be checked against anything filed.
- A semiconductor equipment backlog is an order book customers can move or cancel; Onto's own risk factors describe fluctuating customer capital spending.
- The counterparty is not named, the agreement is not filed as an exhibit, and the value is described as an estimate rather than a committed minimum.
- A volume purchase agreement in semiconductor equipment typically sets pricing against expected volumes rather than obliging the buyer to take them.
Where the growth is coming from, business by business
Where the growth is coming from
| Business | Share of revenue | Growth: last quarter | Last half-year | Last year |
|---|---|---|---|---|
| Systems And Software Revenue | 86% | +37.1% | +21.4% | +6.4% |
| Parts Revenue | 8% | +44.7% | +45.3% | +32.4% |
| Service Revenue | 6% | +5.9% | +5.7% | +17.4% |
| Whole company | 100% | +35.3% | +22.1% | +9.1% |
The filed split through 2026-06-30. These rates are not carried forward: the reading below says why.
How we read it: The score believes neither the 23% the mix implies nor the 8% trailing: the half-year rate includes Semilab, bought November 2025, and laps a weak 2025 base; excluding Semilab, first-half growth was about 13%.
The results
The numbers
Sales grew 8.1% over the twelve months to Jun 2026, faster than the 1.8% in the year to Jan 2026.
Operating margin is 12%, down from 14% in the year to Dec 2023.
Running costs take 38¢ of each sales dollar, down from 74¢ in the year to Jan 2022.
Each dollar put into the business earns 5.8% a year after tax, down from 14% in the year to Dec 2022.
It turned 23% of sales into free cash, $252M over the last twelve months.
Yearly sales growth has ranged from −19% to +42% over the last 5 years, with sales falling in 1 of them.
The share count barely moved in 4.5 years, and pay in shares costs 2.8% of sales.
All the numbers
Filed figures from the company's 10-K and 10-Q, read 2026-10-04. Fiscal years end in January.
| Jan 2022 | Dec 2022 | Dec 2023 | Dec 2024 | Jan 2026 | 12 mo to Jun 2026 | |
|---|---|---|---|---|---|---|
| Size and growth | ||||||
| Sales | $789M | $1.01B | $816M | $987M | $1.01B | $1.12B |
| Sales growth | +42% | +27% | −19% | +21% | +1.8% | +8.1% |
| Profit | ||||||
| Gross margin | 54% | 54% | 52% | 52% | 50% | 50% |
| Operating margin | −20% | 24% | 14% | 19% | 13% | 12% |
| Net margin | 18% | 22% | 15% | 20% | 14% | 12% |
| What growth costs, per sales dollar | ||||||
| Research and development | 12% | 11% | 13% | 12% | 13% | 13% |
| Sales and marketing | 7.3% | 6.5% | 7.6% | 7.7% | 7.0% | 7.1% |
| General and admin | 8.6% | 6.9% | 9.8% | 8.1% | 11% | 11% |
| Pay in shares | 2.5% | 2.4% | 3.1% | 2.9% | 2.7% | 2.8% |
| Capital spending | 1.5% | 1.8% | 2.8% | 3.2% | 2.8% | 1.2% |
| Cash | ||||||
| Cash from operations | $175M | $137M | $172M | $246M | $328M | $266M |
| Free cash flow | $163M | $118M | $149M | $214M | $300M | $252M |
| Balance sheet | ||||||
| Cash and short-term investments | $511M | $548M | $698M | $852M | $640M | $1.88B |
| Returns and shares | ||||||
| Return on capital | - | 14% | 6.4% | 9.3% | 5.5% | 5.8% |
| Shares (diluted) | 49.7M | 49.8M | 49.3M | 49.7M | 49.3M | 49.7M |
What growth costs
Growth has not yet made Onto more profitable: sales have risen strongly, but spending on research, selling and running the company rose just as fast, partly because of the business it bought. Cleanup charges and the cost of past deals sit on top.
The evidence under this - 2 claims
Semilab USA, bought in November 2025, brought its own costs with it (C10, C36), so part of the FY2025 and trailing rise is the acquired business rather than organic spending.
The last report (6 Aug)
Quarter ended June 2026, filed figures.
The second quarter of 2026 was a record: sales of $343.1M, up 35.3% on a year earlier, and operating profit of $63.6M. By the company's adjusted measure it looked better still.
The gap between those two measures is the story. Restructuring charges rose from $10.6M in 2023 to $63.4M in 2025, with $30.7M more in the first half of 2026, and about two-thirds of them land in the cost of the tools themselves. Management now promises reported operating margin of 21.4% to 22.4% in the September quarter.
What it did to the thesisIt confirmed the business is growing fast, but it sharpened the case against paying up: restructuring charges keep eating profit year after year. The view stays Neutral until reported profit actually reaches the band management promised for the September quarter.
The quarter in detail
Sales growth from the previous quarter, filed figures.
Earnings per share $1.93 against $1.69 expected; $1.21 as filed, the gap being what the company adds back.
The evidence under this - 4 claims
- The company guided first-quarter 2026 GAAP diluted EPS to $0.74-$0.84 and delivered $0.67, below its own range, while the non-GAAP figure beat its range - so the GAAP half of this guidance has missed once in the last two quarters it was given.
- Restructuring has been guided as ending, and has instead risen in each of the last three fiscal years.
What management did with the money
It borrowed $1.5B through notes that pay no interest and turn into shares at $381.80, then spent about $720M on 27% of Rigaku, a company it does not control. It also bought back $205M of stock, yet the share count still rose, to 49.7 million from 49.0 million.
The evidence under this - 3 claims
- Management framed the stake as self-funding: foregone interest income offset by licensing revenue plus about $7 million of Rigaku dividends.
- The stake buys access to X-ray metrology that Onto's own chief executive placed several device generations out, so the operating return is deferred rather than absent.
The price, and where we disagree
In short. Full. Cheap on sales, dear on profit. The price pays for Onto keeping far more profit from each sale, a recovery management promises but its official figures do not yet show.
The shares gained 162.9% over the past year on the strength of the boom. The test of what that assumes comes on 5 November, with the September-quarter results and the first promised step up in reported margin.
The same kind of promise has missed once already. In the March quarter, reported earnings per share fell short of guidance while the adjusted figure beat it.
The evidence under this - 2 claims
- The company guided first-quarter 2026 GAAP diluted EPS to $0.74-$0.84 and delivered $0.67, below its own range, while the non-GAAP figure beat its range - so the GAAP half of this guidance has missed once in the last two quarters it was given.
- Restructuring has been guided as ending, and has instead risen in each of the last three fiscal years.
How we read the multiple
On sales, Onto is the cheapest of the rivals its own annual report names. On operating profit, the profit from the business itself, it is the dearest: about 97 times a year's profit, against about 45 for rivals.
It keeps so little of each sales dollar: it earns 3.4% on the money invested in it, against 7.4% over its past six years and 10.0% for rivals.
So the price bets that its official profit rises to meet management's higher 'adjusted' one.
Against its own history. Far above its own normal on every ratio, struck on a profit line held down by charges it takes every year.
The evidence under this - 3 claims
- Annualising the third-quarter GAAP operating income guide of $81.4-89.7 million gives roughly 38 times enterprise value - below the cohort's trailing median of 45. If the guided quarter is the run rate the premium is already gone.
- The pack's enterprise value nets only cash and equivalents against the notes; on the balance sheet's $1,882.2 million of cash and marketable securities the enterprise value is about $629 million lower, which cuts the multiple by roughly five percent.
Priced against its rivals
The competitors its own annual report names, each confirmed against the rival's filing. Measured on company value for each $1 of operating profit, cheapest first:
| Ticker | Company | company value for each $1 of operating profit |
|---|---|---|
| KLAC | KLA | 43.6× |
| NVMI | Nova | 45.0× |
| CAMT | Camtek | 59.9× |
| ONTO | Onto Innovation | 96.9× |
The middle of the other 3 is 45.0. Onto Innovation Inc. is dearest of 4.
Rivals' figures come from a data provider, so read the spread of the group rather than any single pair.
The gap is Onto's own rather than the group's: it sits at the cheap end on sales and alone at the dear end on operating profit.
Private or foreign rivals that cannot be priced: Ushio, Inc., Canon, Inc., GigaVis Co. Ltd.
Where we disagree with the market
A record quarter, a first $1B backlog and AI-driven demand make Onto a clear winner in chip inspection.
The gap is narrowing, but a tenth of sales still goes on charges the company has taken every year for four years.
The adjusted number is the one everyone quotes. The difference appears only by adding up four years of restructuring from the notes to the accounts, and recent purchases add a yearly write-off of what was bought on top.
We would be wrong if reported operating margin lands in the promised band this quarter and restructuring drops below $10M.
What analysts say, and the evidence
The evidence under this - 3 claims
- The company guided first-quarter 2026 GAAP diluted EPS to $0.74-$0.84 and delivered $0.67, below its own range, while the non-GAAP figure beat its range - so the GAAP half of this guidance has missed once in the last two quarters it was given.
- Restructuring has been guided as ending, and has instead risen in each of the last three fiscal years.
What could go wrong, and how we would know
The risks
- Charges that never end. Restructuring has been booked four years running. A fourth year carried into 2027 with no end date would make it part of the business.
- Ground lost in packaging. Tool sales fell in 2025 on fewer AI packaging inspection units, while Camtek, a rival, gets half its sales from AI.
- A $720M minority bet. Onto owns 27% of Rigaku, valued at market prices and outside its control; a write-down would hit earnings.
- Few, large customers. Three customers took 49% of 2025 sales, and the biggest new deal is with one memory maker.
The evidence under this - 6 claims
- Camtek's revenue is a fifth of Onto's, so it can grow off a smaller base without Onto losing absolute share.
- Onto's own advanced packaging revenue reached a record in the second quarter of 2026, so whatever happened in FY2025 has reversed at least once.
- Management framed the stake as self-funding: foregone interest income offset by licensing revenue plus about $7 million of Rigaku dividends.
- The stake buys access to X-ray metrology that Onto's own chief executive placed several device generations out, so the operating return is deferred rather than absent.
- The counterparty is not named, the agreement is not filed as an exhibit, and the value is described as an estimate rather than a committed minimum.
- A volume purchase agreement in semiconductor equipment typically sets pricing against expected volumes rather than obliging the buyer to take them.
What we still do not know
- Which product lines earn the money. Onto reports as one business, with no split by line.
- When the restructuring ends. Management has not dated it.
- What Rigaku will pay back. Management would not size the licensing income it expects from the stake.
The evidence under this - 3 claims
- Management framed the stake as self-funding: foregone interest income offset by licensing revenue plus about $7 million of Rigaku dividends.
- The stake buys access to X-ray metrology that Onto's own chief executive placed several device generations out, so the operating return is deferred rather than absent.
What would change our view
- Neutral→AttractiveThe margin comes back
It needs all of:
- ○Reported operating margin rises above 18% through the December 2026 quarter, with 2026 restructuring charges below $60M.checked at the next results, due by 5 Nov
- ○The return on money invested climbs from 3.4% toward its own six-year normal of 7.4%.checked at the next results, due by 5 Nov
- Neutral→UnattractiveThe margin does not, or the base narrows
It needs any of:
- ○Tool and software sales fall from a year earlier in any 2027 quarter, or the top three customers exceed 55% of 2026 sales.checked at the next results, due by 5 Nov
- ○Onto writes down its Rigaku stake, or pays for another minority stake.watched as it happens
We would also be wrong if
- Restructuring charges exceed $60M again in 2026, or reported operating margin stays below 18% through the December quarter.
What is coming
- 5 November 2026 Q3 2026 results: Whether GAAP operating margin reaches the guided 21.4%-22.4% and restructuring falls to $15-17 million, the reading the thesis pre-registered.
Every dated event ahead
- 5 November 2026 Q3 2026 results: Whether GAAP operating margin reaches the guided 21.4%-22.4% and restructuring falls to $15-17 million, the reading the thesis pre-registered.
- 17 December 2026 Analyst event at the New York Stock Exchange: Whether management dates the end of the restructuring programme and sets a GAAP margin target, and how it means to earn a return on Rigaku and Semilab.
- February 2027 Q4 and full-year 2026 results and the FY2026 10-K: Whether a full year of Semilab and the memory upturn lifts revenue with costs per sales dollar falling, and whether a fifth year of restructuring is guided.
Since we wrote this
- 5 OctThesis reviewThe long case now sits under the verdictAfter the five-name pilot, the owner asked for the long case on every thesis.
- 4 OctThesis reviewOwner's checklist: revenue streams, cost of growth, management, catalystsThe owner's checklist found the method silent on who runs the company, what growth costs, where the money comes from and what is dated ahead; this pass fills those four blocks from the proxy and filings.
- 6 SepThesis writtenNeutral, medium convictionCheapest of its named rivals on sales and dearest on profit, because charges running at about a tenth of sales keep the reported margin down. Management says they will fall.
Also checked 5 times with nothing new to act on.
Sources 15
Sources - 15
Every claim above cites one of these, or states the reasoning that derived it. The quote is frozen at the date it was read: a link can rot, and an agent fetch returns a reading of a page rather than the page. Backs counts the whole claim ledger, so it can name a claim this page does not surface.
- S01primarycompany filing2026-02-24Item 1 business and named competitors, customer concentration table, the three-year restructuring table split between cost of goods sold and operating expenses, revenue by source, the Semilab purchase price allocation, single-segment statementbacksC01C02C03C05C06C07C10C15C16C26C31C32C33C37C39C43
- S02primarycompany filing2026-08-06Note 15 restructuring split, Note 2 on the Rigaku share purchase agreement, legal matters, share repurchase detail, the segment table, the new headquarters leasebacksC02C06C07C09C20C22C34C36C38C39
- S03primarycompany filing2026-08-06the full GAAP-to-non-GAAP reconciliation for the quarter, half and prior-year comparatives, the balance sheet, the third-quarter outlook with its own reconciliation, and the backlog statementbacksC04C05C08C09C12C17C19C21C22
- S04primarycompany filing2026-02-19FY2025 and Q4 2025 GAAP and non-GAAP results, the first-quarter 2026 guidance that Q1 later missed on GAAP, and the volume purchase agreement with an HBM manufacturerbacksC03C18C23
- S05primarycompany filing2026-05-21conversion price, capped call terms and cost, the concurrent share repurchase price, and the stated intention to fund the Rigaku acquisitionbacksC19
- S06primarycompany filing2026-08-10the closing, the yen and dollar consideration, the fair value option accounting and the statement that Rigaku will not be consolidatedbacksC20C21
- S07primarytranscript2026-05-05First quarter 2026 earnings call transcriptmanagement's answers on the Dragonfly G5 head-to-head win, the panel opportunity and the Rigaku economics; nine analyst questions, which are a separate origin from the answers
We are seeing things now below 200 nanometers, where historically, 800 nanometers might have been about the limit.
backsC30 - S08primarycompany filing2026-08-06the largest process-control supplier's own segment revenue, its GAAP operating margin, and its Competition section naming Onto Innovation - evidence about Onto from outside OntobacksC05C12C24
- S09primarycompany filing2026-02-17a competitor's own description of the same product sold to the same buyers, confirming cohort membershipbacksC25
- S10primarycompany filing2026-03-19a competitor's own statement that roughly half its revenue serves AI applications including HBM and advanced packaging - the market Onto's inspection volume fell inbacksC25C26
- S11primaryindustry data2026-07-14the rate behind machine.current, measured by the industry's own trade association rather than by any supplier in it
the Wafer Fab Equipment (WFE) segment...is projected to increase 23.1% to $143.9 billion in 2026.
backsC27 - S12primaryindustry data2026-07-14the memory half of the driver, which is where Onto's HBM-linked advanced packaging demand sits
DRAM equipment sales are projected to rise 39.0% to $38.8 billion in 2026
backsC27 - S13primaryvendor data2026-09-06Research pack, frozen at content/ONTO/research-pack.jsonprice, returns, the analyst split, the volatility bucket, insider and short-interest filings, and the competitors:filed cohort multiplesbacksC05C12C13C14C23C29
- S14primarycompany filing2026-04-06beneficial ownership as of March 24, 2026, the executive officers and their start dates, the FY2025 cash bonus goals and payouts, the PSU design on relative TSR against the SOXbacksC40C41C42
- S15primarycompany release2026-08-10the date of the company's analyst event; the Business Wire original refuses this repository's client, so the release was read through press_fetch.py from a syndicated copy
Onto Innovation Inc. (NYSE: ONTO) (“Onto Innovation,” “Onto,” or the “Company”) today announced that it will host a Company-sponsored Analyst Event on Thursday, December 17, 2026, at the New York Stock Exchange in New York City.
backsC44