Initiated 2026-09-21 · Price $263.43 (as of 2026-09-18, stockanalysis.com quote) · Mkt cap $12.93bn · Information technology / semiconductor process control · Model: verified

Rating: Sell — Conviction: Medium

An excellent business at a price that needs the AI-packaging capex cycle to run further and longer than any case I can honestly build. Revenue is inflecting hard — Q2 2026 up 35% year over year, backlog above $1bn for the first time — but the market capitalizes a non-GAAP stream that excludes ~$80mm a year of recurring merger and restructuring charges, and looks past how slowly growth converts to cash: working capital absorbs about 40 cents of every incremental revenue dollar. Discounted at a 5% risk-free rate, even a bull case with 25% multi-year growth and mid-30s EBITDA margins lands ~30% below the price.

Model value range vs price
Bear $41.77Base $104.13Bull $185.44Price $263.43

Business overview

Onto sells inspection, metrology and lithography systems plus process-control software to chipmakers; it was formed by the 2019 Rudolph–Nanometrics merger. Demand splits into Advanced Nodes, Specialty Devices and Advanced Packaging, and software/services. In Q2 2026 Advanced Nodes revenue grew 50% sequentially to a record, and Specialty Devices and Advanced Packaging hit historic highs on 2.5D logic, HBM and silicon photonics. Three variables drive earnings: process-control intensity at advanced nodes and in packaging, systems-versus-software mix, and the working-capital cost of scaling a long-lead-time systems business. In August 2026 Onto paid ~$720mm for 27% of Japan's Rigaku Holdings from a $1.3bn+ 0.00% convertible; the stake is fair-valued, not consolidated.

Bull case

  1. AI packaging and HBM intensity — each layer of 2.5D/3D stacking adds metrology steps, so revenue outgrows wafer starts. Plays out if accelerator-driven HBM and packaging capacity additions run through 2027; backlog above $1bn already reaches into 2027. Model: rev_growth
  2. Advanced Nodes share gain — the 50% sequential jump was broad-based across logic and memory, which reads as design-win conversion, not one customer's pull-in. Plays out if Q3/Q4 hold the run-rate. Model: rev_growth
  3. Leverage as charges roll off — management targets non-GAAP operating margin of at least 33% by year-end against 30.0% in Q2, and the $4.3mm/quarter of merger expense is transitory. Plays out if restructuring decays with it. Model: ebitda_margin
  4. Rigaku broadens the served market — X-ray addresses the exotic materials and buried 3D structures optical metrology struggles with. A years 3–5 effect. Model: rev_growth

Bear case

  1. Working capital eats the growth — receivables plus inventory less payables have run 45.7%–61.8% of revenue since FY2021, $577mm in June. FY2026 consumes ~$140mm of cash in working capital alone, so free cash flow lags the margin story in exactly the years the multiple assumes cash. Model: nwc_pct_rev
  2. The price capitalizes non-GAAP earnings — the 25.6x forward P/E rests on EPS adding back $19.6mm a quarter of merger and restructuring charges. Onto creditably does not exclude share-based compensation ($17.0mm in H1), so the add-backs are operating charges. On GAAP-inclusive EBITDA the stock is at 40.5x trailing. Model: ebitda_margin
  3. Air pocket after a pull-forward — semicap revenue mean-reverts: FY2022's $1,005mm fell 19% to $816mm in FY2023. A 42% year is usually followed by digestion; bear assumes one down year in FY2027. Model: rev_growth
  4. Duration risk at a 5% risk-free rate — beta 1.58 at 11x sales, 10Y at 5.00%, 16 of 18 FOMC participants expecting another hike. Why S3 and S12 dominate the stress test. Model: wacc, exit_ev_ebitda

Valuation & balance sheet

Metric (definition) Current Own history (range or 5y avg) Peers Source, as-of
Forward P/E (consensus non-GAAP EPS, NTM) 25.6x n/a (unverified) KLA 33.8x · Camtek 35.6x stockanalysis statistics / GuruFocus, 2026-09-21
EV/EBITDA (TTM, EV $12.52bn) 40.5x n/a (unverified) KLA ~42.9x · Nova ~36.6x · Camtek ~30.4x stockanalysis statistics / GuruFocus, 2026-09-21
EV/Sales (TTM revenue $1.12bn) 11.2x n/a (unverified) n/a (unverified) stockanalysis statistics, 2026-09-21
EBITDA margin (GAAP op. income + D&A / revenue) 19.2% FY2025 · 23.2% H1 2026 19.2%–30.0%, 5y avg 25.0% n/a (unverified) SEC EDGAR XBRL + Q2 2026 release, 2026-09-21
FCF yield (TTM FCF $252.5mm / mkt cap) 2.0% n/a (unverified) n/a (unverified) stockanalysis statistics, 2026-09-21

Model-implied value range (from model-summary.json; generic module, DCF-Gordon and DCF-exit-multiple, midpoints): Bear $41.77 · Base $104.13 · Bull $185.44 per share, i.e. implied returns of −84.1% / −60.5% / −29.6% vs $263.43. These ranges show how the bull and bear drivers translate into value; they are not price targets. The price sits above the bull case: $263 needs the bull revenue and margin path and a 9% WACC and a 22x exit together.

Rigaku is excluded from every case. The stake is deducted from cash at its ~$720mm cost and never added back, because the generic module has no line for a fair-value holding. That understates all three cases by ~$14.50 per share at cost, and does not change the direction of the call.

Balance sheet: net leverage 1.60x FY0 EBITDA (0.8x on base FY2026E); coverage not meaningful — the only borrowing is a 0.00%-coupon convertible, modeled EBITDA/net interest 240x in Y1. Liquidity $1,162mm of cash and securities pro forma for Rigaku against $1,471.7mm of notes, the whole stack maturing at once on 2031-06-01. No agency rating retrieved this session.

Model note: built and LibreOffice-verified (3,188 formula cells matched, 0 failing error checks). One warning check fails: market.cash is unverified, derived as the 2026-06-30 balance of $1,882.2mm less the ~$720mm Rigaku payment. No assumptions lack a basis; no scenario CHECKs remain. Two judgment inputs are weaker than the rest and say so in their basis: capex at 3% of revenue, and the 15% cash tax rate inferred from Q2. Cost-of-debt is charged on net debt only, ignoring ~$45mm a year of interest income on gross cash. Generator note (2026-09-24): in this module share_change, capital_return_pct and cost_of_debt move EPS and the net-cash path only, and peer_pe feeds a cross-check only; none changes the value range, so bull/bear tags no longer cite them. Scenario deltas were rebuilt after multiple_pct shocks were extended to the Gordon method (previously they moved only the exit-multiple half), and EPS now uses average diluted shares.

Scenario stress test

Reasoned from the bull/bear drivers above. The model column comes from the scenario overlays (Base case + shock).

Scenario Effect Mechanism Magnitude Model Δ value vs Base ($/sh)
S1 Fast equity crash − Beta 1.58 at 11x sales: multiple compresses on de-risking, backlog untouched High −$24.91
S2 Slow bear / recession − Customers defer deliveries; backlog stretches instead of shipping, underutilization hits margin High −$37.32
S3 Rapid rate shock − Long-duration equity de-rates on the yield spike; no floating debt, so no interest-cost channel High −$26.40
S4 Slow rate grind − Same channel, slower and shallower, across the full horizon Med −$17.20
S5 Soft-landing cuts + Cheaper money plus intact fab investment is the best regime for a high-multiple equipment name High +$21.50
S6 Recession-driven cuts − Capex cuts dominate the discount-rate relief High −$21.45
S7a Credit liquidity shock − Mostly a multiple event: no maturity before 2031, no revolver to roll Low −$13.99
S7b Slow default cycle − Large chipmakers are investment grade, but levered OSAT and specialty buyers slow orders Low −$7.13
S8 Stagflation − Input and labor inflation squeeze margin faster than system pricing resets; discount rate rises Med −$20.18
S9a Dollar spike − Asia-concentrated revenue translates lower; the yen-denominated Rigaku stake marks down Low −$4.66
S9b Dollar slide + Mirror image: translation tailwind plus a yen-asset gain Low +$3.96
S10 Melt-up + Rewards exactly this profile: high beta, AI-levered, narrow leadership High +$27.13
S11 Energy supply shock − Freight and energy input costs on globally shipped systems; no direct demand channel Low −$4.60
S12 Mega-cap/AI derating − The re-rating rests on AI packaging and HBM intensity; a leadership unwind takes multiple and orders both High −$35.64

Magnitude labels re-ranked 2026-09-25 to match the model's dollar deltas (skill 3a-v-b item 7; the 2026-09-24 generator fix enlarged multiple-shock rows): S5 Med→High.

Currently active/on watch per the playbook: S3 partially active (10Y 5.00%, price leg met, pace legs short); S8, S10, S11 on watch, S11 escalated after the 2026-09-19/20 Houthi strikes on Riyadh and Yanbu.

Model value change vs Base, by scenario
S2 Slow bear / recession−$37.32S12 Mega-cap/AI derating−$35.64S3 Rapid rate shock−$26.40S1 Fast equity crash−$24.91S6 Recession-driven cuts−$21.45S8 Stagflation−$20.18S4 Slow rate grind−$17.20S7a Credit liquidity shock−$13.99S7b Slow default cycle−$7.13S9a Dollar spike−$4.66S11 Energy supply shock−$4.60S9b Dollar slide+$3.96S5 Soft-landing cuts+$21.50S10 Melt-up+$27.13

What would change the call

Upgrades if: the bull operating profile shows up — revenue above a $1.6bn annualized run-rate, GAAP-inclusive EBITDA margin above 30%, working capital below 38% of revenue — while the price de-rates into the $185–235 bull range; or a sustained fall in the 10Y resets the discount rate the whole group is capitalized at. Downgrades if: conviction goes to High rather than the rating changing — backlog below $1bn, restructuring above $15mm a quarter into FY2027, or the GAAP-to-non-GAAP gap widening on new charges.

Watch items

Sources