Technoprobe
Design-to-order hardware: custom probe cards co-designed per chip design, sold to chipmakers and OSATs; demand recurs with every new device tape-out and wafer volume, giving consumable-like repeat revenue
Earnings, margins, COGS & capex
FY2025 (Dec year-end): revenue EUR 628.4M (+15.7%), gross margin 44.4% (+330bps), EBITDA EUR 201.4M (32.1% margin, +47.5%), net profit EUR 98.8M (+57.4%), net financial position +EUR 684.2M. AI-related applications reached ~38% of revenue (per CEO). No FY2025 dividend - profit retained to fund growth; a EUR 60M buyback was completed Dec 19, 2025 and a new authorization up to EUR 250M (max 2% of capital) was proposed in Mar 2026. Momentum accelerated sharply into 2026: Q1 revenue EUR 187.0M (+19.0% YoY, +15.6% QoQ, above the company's own EUR 182M +/-3% guide) with EBITDA margin 37.0% despite a ~EUR 20M negative FX impact, and on May 14, 2026 management pulled its 2027 targets forward a year and raised them - FY2026 now guided at EUR 950M-1.05B revenue with 44-46% EBITDA margin; Q2 2026 guided at EUR 266M (+/-3%) revenue, 55% gross margin, 45% EBITDA margin. Balance sheet is net cash ~EUR 0.66B even while funding a capacity-doubling capex program (~EUR 240M, ~80% in Italy incl. a new Cernusco Lombardone greenfield plant ready end-Q1 2027, machinery for both Logic and HBM probe cards).
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~56¢ is cost of goods and ~12¢ operating expense, leaving ~32¢ of operating profit (~16¢ net).
Revenue trend
Margins
rising fast on AI/HBM mix, production-efficiency recovery and operating leverage
rising ~700bps/yr; efficiency recovery + US reorganization
rising
will compress near term as capex roughly doubles for the capacity plan
COGS structure
Cost base is skilled-labor- and precision-manufacturing-intensive: in-house MEMS microfabrication of probe heads, high-end probe-card PCBs/space transformers (vertically integrated via Harbor Electronics, acquired for ~$50M, closed Aug 2023), specialty materials, plus test/assembly labor across ~3,300 employees. FY2025 cost of revenue was EUR 349.3M on EUR 628.4M revenue (gross margin 44.4%); the 2026 jump toward ~55% reflects richer AI/HBM product mix, capacity-utilization leverage and efficiency recovery rather than input-cost deflation. Q1 2026 absorbed a ~EUR 20M negative FX impact (USD-linked revenue vs EUR cost base).
Capex
FY2025 capex EUR 99M (~16% of revenue; 2025 split per company chart: ~56% Technoprobe Italy, 12% DIS Tech, 22% Yee Wei, 10% other). Plan announced with the Q1 2026 results: ~EUR 240M over 2026E-Q1 2027E, ~80% at Technoprobe Italy (remainder across DIS Tech, Yee Wei and others - exact split not disclosed) - machinery fleet expansion covering both Logic and HBM probe cards, plus a +4,000 sqm greenfield plant at Cernusco Lombardone ready end-Q1 2027, with the stated aim to double manufacturing capacity by end-2027. New facilities in Italy and Taiwan ramp during 2026.
Latest earnings
Beat the company's own guidance (EUR 182M +/-3% given in March) and was received as a major positive surprise - shares surged ~36% on the print as management pulled forward and raised targets (sell-side consensus beat/miss not formally published)
Q2 2026: revenue EUR 266M (+/-3%), gross margin 55% (+/-200bps), EBITDA margin 45% (+/-200bps). FY2026: revenue EUR 950M-1.05B with 44-46% EBITDA margin - the prior 2027 targets (EUR 850-900M, 38-40%, themselves already accelerated from 2028 at the Apr 2025 Capital Market Day) pulled forward a year AND raised
- AI share of revenue
- ~38% (FY2025, per CEO)
- Net cash
- EUR 660.5M (Mar 31, 2026)
- Q2 2026 revenue guide
- EUR 266M +/-3% (~+42% QoQ vs Q1)
- Capacity plan
- double manufacturing capacity by end-2027 (~EUR 240M capex, 2026E-Q1 2027E)
Growth drivers
- AI-related applications — ~38% of FY2025 revenue (per CEO) and the primary growth engine - test intensity and probe card complexity/ASP rise with advanced-node AI accelerators, chiplets and advanced packaging
- HBM probe card push — the 2026-27 machinery investment explicitly covers HBM alongside Logic, opening the memory-test TAM historically led by Japanese/Korean vendors and FormFactor
- Every new chip design requires a new custom probe card — design-turnover-driven recurring demand on top of wafer-volume growth
- Teradyne strategic partnership (10% shareholder with board seat) plus the acquired DIS interface business broadening test-cell content per customer
- Recovery in automotive/industrial semis after the inventory correction (early signs cited in H1 2026); consumer stabilizing
- China optionality — agreement in principle (Mar 2026) for a local partner to invest USD 35M for ~30% of Device Interface Solutions Technology (Shanghai), developing TPEG technology with a manufacturing subsidiary planned in Suzhou - subject to Italian Golden Power and Chinese regulatory approvals
Bull & bear
Technoprobe is one of the purest listed plays on AI silicon complexity: every new accelerator, chiplet stack and HBM generation needs more, denser, costlier probe cards, and Technoprobe is a co-leader with structurally rising margins, net cash, and a capacity doubling underway just as demand inflects.
- Guidance momentum: 2027 targets were pulled forward a year AND raised (FY2026 EUR 950M-1.05B at 44-46% EBITDA margin vs prior 2027 EUR 850-900M at 38-40%) - management is being chased upward by demand
- Test intensity compounds independent of chip unit growth: advanced packaging and known-good-die economics force more wafer-sort content per device, and AI is already ~38% of revenue and rising
- Gross margin trajectory (44.4% FY2025 -> 48.7% Q1 2026 -> 55% guided Q2 2026) shows genuine mix shift and operating leverage, not just volume
- HBM entry plus DIS interface boards expand served TAM beyond the historical logic probe card core
- EUR 0.66B net cash plus ~EUR 212M FY2025 operating liquidity self-funds the capacity doubling; the EUR 60M buyback completed Dec 2025 and a proposed new EUR 250M authorization show willingness to return capital
- Teradyne's ~10% strategic stake (EUR 483M) and board seat validate the technology roadmap and hard-wire channel alignment with a leading tester vendor
A superb business priced for perfection: ~20x forward EV/sales and ~44x forward EV/EBITDA on a guidance ramp that requires the rest of 2026 to average ~EUR 255-290M quarters vs EUR 187M in Q1, in a cyclical, customer-concentrated niche where the stock has already roughly doubled YTD - any AI digestion, HBM qualification slip or FX swing de-rates it violently.
- Valuation: at ~EUR 20.5B market cap the stock trades near ~20x EV/guided-FY2026-sales and roughly ~44x EV/guided EBITDA - far above test-interface peer FormFactor and most semicap names; the Q1 print alone moved the stock +36% in a day and it is up several-fold over 12 months
- Guidance embeds a step function: Q1 was EUR 187M; the FY guide needs ~EUR 255-290M average quarters for the rest of 2026 - a steep ramp with limited disclosed backlog visibility
- History of violent cyclicality: revenue fell ~25% in FY2023 when the consumer cycle turned; AI concentration (~38% and growing) swaps one concentration risk for another - and management itself flags 2026 as a 'selection' year for AI-investment sustainability
- HBM/memory probe cards are an incumbency game (Micronics Japan, Korean vendors and FormFactor are established at the memory makers) - qualification cycles could disappoint
- USD weakness vs EUR is an ongoing translation and competitiveness headwind (~EUR 20M negative impact in Q1 2026 alone; EUR 22M FX drag on FY2025 liquidity)
- Free float of only ~23% of capital amplifies moves in both directions; family control via T-Plus limits M&A/activism paths that might otherwise support the price
- China TPEG venture adds geopolitical/export-control and Golden Power tail risk for a company embedded in US-aligned supply chains and ~10%-owned by a US tester vendor
What it is worth
EV/sales and EV/EBITDA on FY2026 guidance, cross-checked with trailing multiples and peer comps (FormFactor, Micronics Japan, JEM)
AI capex digestion or HBM qualification slip breaks the 2026 step-up; revenue undershoots guidance and the multiple mean-reverts toward test-interface peers (single-digit EV/sales), implying downside well in excess of 50% from Jul 2026 levels
FY2026 guide roughly met (~EUR 1.0B, ~45% EBITDA margin) but growth normalizes in 2027; multiple compresses toward semicap-premium levels, implying flat-to-lower fair value from Jul 2026 levels despite strong fundamentals
Guidance ramp lands and extends: FY2027 revenue well above the FY2026 guide at ~45% EBITDA margin as doubled capacity fills and HBM wins land - the multiple holds on scarcity of pure AI-test plays and the stock grows into it
At ~EUR 20.5B market cap (~$23-24B) less ~EUR 0.66B net cash, EV is ~EUR 19.8B: roughly ~20x EV/sales and ~44x EV/EBITDA on the FY2026 guidance midpoint (~EUR 1.0B revenue, ~EUR 450M EBITDA at the 45% midpoint), and ~31-32x trailing FY2025 sales. Peer FormFactor trades at a small fraction of these multiples. The price implies the market believes the AI test-intensity ramp sustains high revenue growth for many years at mid-40s EBITDA margins with successful HBM entry - achievable only in the strongest AI scenarios. Not financial advice.
SWOT
Strengths
- Global co-leader (with FormFactor) in advanced probe cards — FormFactor, Technoprobe and Micronics Japan together held roughly 60% of 2025 industry revenue per market research
- 600+ proprietary patents and deep MEMS microfabrication know-how built since 1996
- Consumable-like demand model — a new probe card per chip design per test cell, repurchased as designs and volumes turn over
- Fortress balance sheet: ~EUR 0.66B net cash funds the capacity doubling without leverage
- Structural margin expansion: EBITDA margin from 25.1% (FY2024) to guided 44-46% (FY2026)
- Teradyne alliance — ~10% strategic shareholder (EUR 483.1M / ~$524M, closed May 2024, with a board seat) plus the acquired DIS business embeds Technoprobe deeper in the test cell
Weaknesses
- Customer and end-market concentration — historically heavy exposure to a handful of leading-edge fabless/IDM customers and the smartphone cycle (FY2023 revenue fell ~25% in the consumer downcycle); customer list not disclosed
- Controlled company with limited free float — the founding Crippa family controls via T-Plus (family holds a further 6.5% of capital directly), Teradyne ~10%; free float ~23% of capital (~13.6% of voting rights) - governance and liquidity constraints for minorities
- Manufacturing concentrated in one Italian cluster (Cernusco Lombardone) - geographic single-point risk until new capacity (incl. Taiwan) ramps
- EUR cost base vs largely USD-linked revenue: ~EUR 20M negative FX impact in Q1 2026 alone
- Later entrant in memory/HBM probe cards where Micronics Japan, FormFactor and Korean vendors are incumbent-qualified
Opportunities
- AI test-intensity supercycle — accelerators, chiplets, advanced packaging and known-good-die requirements multiply probe card content per wafer
- HBM probe card market entry alongside the capacity doubling - a large adjacent TAM
- Cross-selling test interfaces (DIS boards + probe cards) as an integrated offer with Teradyne testers
- China TPEG localization via Device Interface Solutions Technology (Shanghai) with a local partner investing USD 35M for ~30% (agreement in principle, Mar 2026; regulatory approvals pending)
- Automotive/industrial recovery plus silicon-content growth (EV, ADAS) reviving the non-AI part of the business
Threats
- FormFactor (US) and Micronics Japan pushing hard into the same AI/HBM sweet spot; MEMS probe card technology races are capital- and talent-intensive
- AI capex cyclicality — FY2026 guidance embeds a very steep ramp, and management itself calls 2026 a year of 'selection' on AI-investment sustainability - any digestion at hyperscalers/foundries hits the multiple hard
- US-China export-control and Golden Power regimes could constrain the China strategy or bifurcate the customer base
- Valuation risk: ~20x EV/guided-FY2026-sales leaves no room for execution slippage
- Competition for specialized engineering/AI/advanced-materials talent, which management names as a structural challenge
Moats, dependencies & bottlenecks
Moats
Micro-machined fine-pitch vertical probe technology for advanced logic is a decades-deep craft with few credible practitioners globally
Probe cards are co-designed per chip and qualified into each customer test flow; switching mid-generation risks yield and time-to-market
Top-3 structure (FormFactor, Technoprobe, Micronics Japan ~60% of 2025 industry revenue); capacity doubling by end-2027 extends the gap if demand holds
Vertical integration (MEMS fab + Harbor Electronics PCBs + DIS interface boards) Controls much of the test-interface stack, shortening lead times and protecting margin; management explicitly pursues internal development and verticalization
~10% equity stake with board seat and channel alignment embed Technoprobe in Teradyne-based test cells; not exclusive
Dependencies
AI accelerator capex cycle (NVDA/AMD-class silicon tested in TSMC-ecosystem flows) ~38% of revenue and rising; the FY2026 ramp is substantially an AI bet
Leading fabless/IDM customer concentration (names not disclosed by the company) FY2023's ~-25% revenue shows what one end-market downshift does
partner/shareholder ~10% owner with board seat and channel partner; DIS came with customer-servicing obligations; alignment could shift with Teradyne strategy
USD-linked pricing vs EUR cost base; ~EUR 20M negative impact in Q1 2026
~80% of new capex lands in Italy; single-region concentration until Taiwan/other sites scale
The DIS China/Suzhou TPEG venture and any China revenue sit under export-control and investment-screening uncertainty
Advantages
- Co-leadership of the advanced probe card market with a technology roadmap tied to leading-edge tape-outs
- Structurally rising margins with guided 44-46% EBITDA margin - elite for hardware
- Net cash balance sheet (~EUR 0.66B) through a heavy investment cycle
- Consumable-like revenue recurrence keyed to design turnover, not just units
- Strategic tester-vendor alignment (Teradyne) that competitors lack
Weaknesses
- Extreme valuation sensitivity to the AI cycle and to its own 2026 ramp
- Customer/end-market concentration with no disclosed customer list
- ~23% free float, family-controlled governance
- FX mismatch (USD-linked revenue, EUR costs)
- Challenger position in memory/HBM probe cards
Bottlenecks
- Manufacturing capacity — the binding constraint management is attacking with the ~EUR 240M capex plan and capacity doubling by end-2027; the FY2026 guide assumes the ramp lands on time
- HBM/memory qualification cycles at memory makers where rivals are incumbent
- Skilled MEMS process engineers and probe card designers — a narrow global talent pool management itself flags as a structural challenge
- Lead-time-sensitive custom production: each card is design-specific, limiting inventory buffering
Top signals & trends
Top signals
First hard checkpoint on the pulled-forward FY2026 targets; the guide implies ~+42% QoQ
The 2026-27 machinery investment covers HBM; wins would validate the TAM expansion thesis
Capacity is the stated bottleneck; slippage caps the 2027 upside
~38%+ AI exposure makes any test-demand pause a direct guidance risk; management calls 2026 a 'selection' year
Each leg up in EUR pressures reported revenue and competitiveness vs USD/JPY rivals
Optionality if it proceeds cleanly; tail risk if it draws regulatory scrutiny
Trends
Higher testing content per unit as chip and AI-workload complexity rises; structural, multi-year, in both logic and memory per management
Management expects the advanced packaging segment to accelerate over 2026-2027 on ultra-high-density AI chips and advanced automotive
New adjacent demand pool Technoprobe is tooling up for in 2026-27
The Teradyne-Technoprobe pairing and DIS interface boards mirror customers' preference for integrated test-cell offers
Onshoring/technological-sovereignty spending lifts capex broadly but complicates the China strategy
Probe cards remain a cyclical consumable; FY2023 (-25%) proved the downside speed
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Santa Clara high-end probe-card PCB maker acquired for ~$50M (closed Aug 2023) - key input vertically integrated; US activities since reorganized and the Santa Clara building sold for EUR 26M in 2025
Asian manufacturing affiliate; Technoprobe took a minority stake (~EUR 20M, 2025) and part of group capex flows there; the group is also ramping a new Taiwan facility in 2026
ceramic and MEMS-equipment vendors Upstream inputs not individually disclosed; probe heads are fabricated in-house
Leading-edge fabless and IDM chipmakers (not individually disclosed) Industry-reported exposure to top smartphone SoC and AI accelerator designers; AI-related applications ~38% of FY2025 revenue
Wafer-sort flows for leading-edge logic/AI silicon are the demand center for advanced probe cards
OSAT test-services provider; OSATs are buyers of probe cards and test interfaces
OSAT test-services provider; OSATs are buyers of probe cards and test interfaces
European IDM in Technoprobe's home-region automotive/industrial end-market base (industry-reported; company does not disclose its customer list)
US-listed global market-share leader in probe cards (~30% share per industry research) across foundry-logic and memory/HBM; the direct benchmark and fiercest rival at the leading edge
Japanese leader in memory probe cards (~14% global share) with a strong HBM position at Korean memory makers
Japanese advanced probe card maker, strong in memory and expanding in logic
Taiwanese probe card and test-equipment maker serving the Taiwan foundry/OSAT ecosystem
Korean probe card maker (Hwaseong) supplying DRAM/HBM probe cards to domestic memory makers; relevant as Technoprobe enters HBM (listing/ticker not verified)