
Teradyne
Capital-equipment vendor: sells test platforms (Semiconductor Test, Product Test) plus recurring consumables/service and instruments; cobots/AMRs via Robotics. Revenue is project-/order-driven and cyclical, levered to chipmakers' and OSATs' test capacity capex.
- 2026-08-04This market capitalisation previously read ~$67-73B (as of Jun 24 2026). Restated to $57.19B on this refresh, roughly 18% higher.
Sources — 13 figures with citations
- Q2 FY26 revenue, gross profit, operating income and EPSfiled2026-06-28Net revenues $1,328.990M vs $1,282.494M (Q1'26) and $651.797M (Q2'25); cost of revenues $534.372M; gross profit $794.618M (59.8%); operating expenses $356.808M (S&A $192.520M, E&D $156.284M, acquired-intangible amortization $4.972M, restructuring and other $3.032M); income from operations $437.810M (32.9%); pre-tax $443.619M; tax $66.788M; equity in net earnings of affiliate -$1.946M; net income attributable to Teradyne $374.533M; GAAP diluted EPS $2.38, basic $2.39; diluted shares 157.693M; dividend declared $0.13. H1: revenue $2,611.484M, income from operations $910.813M, net income $773.442M, diluted EPS $4.91sec.gov — 8-K Item 2.02 Exhibit 99.1 (release dated 2026-07-28, filing accepted 2026-07-29), Condensed Consolidated Statements of Operations.
- Q2 FY26 GAAP-to-non-GAAP reconciliation and sequential margin comparisonfiled2026-06-28Q2'26: gross profit non-GAAP $794.6M (59.8%); income from operations GAAP $437.8M (32.9%), non-GAAP $448.3M (33.7%) after adding back $5.0M intangible amortization, $3.0M restructuring, $2.5M ERP expenses. Q1'26: gross profit 60.9%, income from operations GAAP 36.9%, non-GAAP $480.4M (37.5%). Q2'25: gross profit 57.2%, GAAP operating margin 13.9%, non-GAAP $98.2M (15.1%). Non-GAAP net income attributable to Teradyne $389.0M, non-GAAP diluted EPS $2.47 vs $0.57 in Q2'25sec.gov — GAAP to Non-GAAP Reconciliation table, which prints Q2'26, Q1'26 and Q2'25 side by side — this is the source for the 60.9% to 59.8% gross-margin and 37.5% to 33.7% non-GAAP operating-margin sequential declines.
- Q2 FY26 revenue by product line and geographyfiled2026-06-28Semiconductor Test: System-on-a-Chip $842.971M, Memory $212.333M, IST $66.521M (total $1,121.825M); Robotics $99.917M; Product Test $107.248M; total $1,328.990M. Q2'25 comparatives: SOC $396.588M, Memory $60.943M, IST $34.347M (Semi Test $491.878M), Robotics $74.866M, Product Test $85.053M, total $651.797M. Geography Q2'26: Asia Pacific $1,158.866M, Americas $109.579M, EMEA $60.545M. Q2'25: Asia Pacific $506.658M, Americas $89.265M, EMEA $55.874M. Timing: point-in-time $1,219.450M, over-time $109.540Msec.gov — Form 10-Q filed 2026-07-31, Note E: REVENUE, Disaggregation of Revenue. Growth rates derived: SOC +112.6%, Memory +248.4%, IST +93.7%, Semi Test +128.1%, Robotics +33.5%, Product Test +26.1%. Asia Pacific share 1,158.866/1,328.990 = 87.2% vs 506.658/651.797 = 77.7%.
- Q2 FY26 cash flow and capexfiled2026-06-28Net cash provided by operating activities $469.138M (Q2'25 $182.088M); purchases of property, plant and equipment $90.706M (Q2'25 $50.408M); acquisitions of businesses net of cash acquired $165.611M. H1: OCF $734.262M, capex $155.439Msec.gov — Condensed Consolidated Statements of Cash Flows in the 8-K exhibit. Free cash flow is NOT a Teradyne-reported measure; $469.138M - $90.706M = $378.432M = 28.5% of revenue is derived.
- Net cash position at 2026-06-28derived2026-06-28$517.1M net cash: cash and equivalents $349.538M + current marketable securities $5.291M + long-term marketable securities $162.274M = $517.103M, against zero debt (short-term debt $0 vs $200.000M at 2025-12-31; no long-term debt line). Total liabilities $1,452.304M, total Teradyne shareholders' equity $3,437.162M, noncontrolling interests $36.151Msec.gov — Component dollars are filed in the Condensed Consolidated Balance Sheets; the $517.103M sum and the zero-debt conclusion are derived by inspection of the liability section, which shows no debt line other than the now-zero short-term debt.
- MultiLane Test Products (MLTP) acquisitionfiled2026-04-08On 2026-04-08 Teradyne and HTP Holding SAL ('MultiLane') formed MultiLane Test Products Holding LLP, in which Teradyne holds a controlling 75% interest (25% noncontrolling), for a total purchase price of $157.8M subject to post-closing adjustments. Noncontrolling-interest fair value $35.8M. Purchase-price allocation: goodwill $131.6M (not tax-deductible), acquired intangibles $46.7M with a 4.7-year weighted average life, net tangible assets $15.3M. Purpose: serve AI-datacenter equipment demand by accelerating test solutions for critical high-speed data connectionssec.gov — Form 10-Q Note D: ACQUISITIONS. Goodwill share of price = 131.6/157.8 = 83.4% (derived).
- Goodwill, accumulated impairment and intangiblesfiled2026-06-28Balance at 2026-06-28: gross goodwill $1,426.383M less accumulated impairment losses $762.566M = net goodwill $663.817M (from $521.019M at 2025-12-31). H1 2026 acquisition goodwill additions $153.939M, including $131.634M from MLTP; foreign-currency adjustment -$11.141M. Net amortizable intangibles include developed technology $83.180M, customer relationships $11.329M, tradenames and trademarks $7.401M. No intangible impairment recorded in H1 2026sec.gov — Form 10-Q goodwill rollforward and intangible-assets tables. The $762.566M of accumulated impairment exceeds the $663.817M of net goodwill currently carried.
- Q3 FY2026 guidancefiled2026-07-28Revenue $1,200M-$1,300M; GAAP net income attributable to Teradyne $1.79-$2.09 per diluted share; non-GAAP $1.85-$2.15 per diluted share. Non-GAAP guidance excludes acquired-intangible amortization and amortization on the Technoprobe investment plus related tax impacts. No full-year guidance providedsec.gov — Guidance paragraph in the release. Sequential math (derived): revenue midpoint $1,250M vs $1,328.990M = -5.9%; non-GAAP EPS midpoint $2.00 vs $2.47 = -19.0%.
- Beat versus company guidance and management commentaryfiled2026-07-28Release states the quarter was the 'second consecutive quarter of record revenue, exceeding the high end of our Q2 Guidance', with 'Revenue up 104% and earnings up over 300% from Q2'25' and 'Record Memory revenue driven by continued strength in DRAM and a resurgence in NAND final test'. CEO Greg Smith: 'rapid increase in wafer fab equipment investment sets the stage for continued growth in 2027 and beyond'sec.gov — Headline bullets and management-commentary paragraph. The beat-versus-own-guidance claim is the company's own statement; the specific prior Q2 guidance range is not restated numerically in this release.
- FY2025 revenue (base for TTM math)filed2025-12-31$3,190.024Mdata.sec.gov — SEC XBRL company-concept API, FY2025 annual fact from the FY2025 Form 10-K (2025-01-01 to 2025-12-31).
- TTM revenue through 2026-06-28derived2026-06-28$4,464.0Msec.gov — FY2025 $3,190.024M - H1'25 $1,337.477M + H1'26 $2,611.484M = $4,464.031M. H1 figures from this exhibit; FY2025 from the 10-K XBRL fact above.
- Share price and market capitalisationmarket2026-08-03$365.80 closing price on 2026-08-03 (-$1.89 / -0.51% that day); market cap $57.19B; 156.34M shares outstandingstockanalysis.com — Regular-session close, not an intraday high. Cross-check: $365.80 x 156.34M = $57.19B, matching the quoted market cap. Enterprise value = $57.19B - $0.517B net cash = ~$56.67B, or ~12.7x TTM revenue; $365.80 / ($2.47 x 4) = ~37.0x annualised Q2 non-GAAP EPS (both derived). Retrieved 2026-08-04.
- Consensus beat magnitudemarket2026-07-28Non-GAAP EPS $2.47 vs ~$2.05 consensus (~20% surprise); revenue $1.329B vs ~$1.22B consensus (~9% beat); Q3 revenue guidance characterised by third-party analysis as above expectationsfuturumgroup.com — Third-party sell-side consensus and interpretation, not company- or SEC-reported. Exact consensus is approximate; the direction and rough magnitude are corroborated across multiple outlets.
The thesis on this name
State of the Memory Supercycle
Semiconductor-test leader with a fresh HBM memory-test product cycle — the Magnum 7H tester supports HBM3/3E and next-gen HBM4/4E up to 4.5Gbps, with volume shipments ramping at the largest HBM makers (Teradyne IR, 2026). Semi-test revenue rose 23% sequentially in Q3'25 on AI memory/accelerator demand. Memory test intensity rises with HBM stack complexity and at-speed HBM4 requirements, giving Teradyne a specific HBM-attach growth lever on top of its broader SoC-test franchise. More diversified than the pure probe/inspection names, so a steadier expression of the memory test-spend wave.
State of the Memory Supercycle
Memory + SoC test toll-booth — HBM and DRAM test demand accelerating beyond expectations, Q1 FY26 rev $1.28B +87% YoY; test intensity rises with HBM stack complexity.
State of the Memory Supercycle
HBM/DRAM test intensity rising; Q1 +87% YoY. Diluted by non-memory SoC/robotics, so satellite-sized.
Earnings, margins, COGS & capex
Teradyne delivered a record Q1 FY26 — $1.282B revenue, +87% YoY, with Semiconductor Test crossing $1.0B for the first time ($1,111M) and memory test at $203M on HBM/DRAM strength and the Magnum 7 ramp (fact). ~70% of revenue is AI-related, up from ~60% in Q4 FY25 (fact). The catch: management guided Q2 FY26 down sequentially ($1.15-1.25B) and said FY26 is first-half-weighted (55-60% of revenue in H1) with limited H2 visibility — so the run-rate is front-loaded, not a new permanent base (fact). Balance sheet is pristine: zero debt, net cash, ~16% FCF margin.
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~40¢ is cost of goods and ~27¢ operating expense, leaving ~33¢ of operating profit (~31¢ net).
Revenue trend
Margins
up (vs 57.2% Q4 FY25, 58.3% FY25)
up sharply (vs 29.0% Q4 FY25, 22.3% FY25) on volume leverage
up (vs 17.4% FY25)
stable/up (~14% FY25)
COGS structure
COGS is dominated by purchased hardware — precision instruments, mixed-signal/digital ASICs, FPGAs, power and RF components, plus mechanical handlers/probes — assembled into test systems; Teradyne is more design/integration than fab, so material and component cost is the bulk of COGS. Gross margin swings with product mix (memory/HBM and high-end SoC testers carry richer margin), volume leverage on fixed manufacturing, and pricing power vs the small ATE customer set. The 60.9% Q1 FY26 gross margin reflects favorable mix + scale.
Capex
Light capex for the sector: $64.7M in Q1 FY26 (~5% of revenue), $224M FY2025 (~7%). Funds test-and-measurement lab/engineering capacity, internal IT, and capacity for instrument/board production — Teradyne outsources most manufacturing, so it is far less capex-intensive than a fab or even lithography peer.
Latest earnings
Large beat. Revenue $1.329B exceeded the high end of the company's own Q2 guidance range (stated in the release) and topped consensus of roughly $1.22B by ~9%. Non-GAAP EPS $2.47 versus roughly $2.05 consensus, about a 20% surprise (third-party consensus, not a filed figure). Third-party coverage also read the Q3 guide as above expectations, even though it steps down sequentially from Q2.
Q3 FY2026: revenue $1,200M-$1,300M; GAAP diluted EPS $1.79-$2.09; non-GAAP diluted EPS $1.85-$2.15. Non-GAAP guidance excludes acquired-intangible amortization and amortization on the Technoprobe investment plus related tax effects. Note the midpoints imply revenue DOWN ~6.0% QoQ from $1,328.990M and non-GAAP EPS DOWN ~19% QoQ from $2.47 — Teradyne guides one quarter at a time and issued no full-year number. Management framed the medium term positively: 'rapid increase in wafer fab equipment investment sets the stage for continued growth in 2027 and beyond.'
- Semiconductor Test revenue
- $1,121.8M, +128.1% YoY — 84.4% of total company revenue (from 75.5% in Q2'25)
- Memory test revenue
- $212.333M record, +248.4% YoY from $60.943M — the fastest-growing line, driven by DRAM/HBM and a NAND final-test resurgence
- SOC test revenue
- $842.971M, +112.6% YoY — still the largest line at 63.4% of total revenue
- Robotics revenue
- $99.917M, +33.5% YoY — 7.5% of total, the slowest-growing segment
- Geographic concentration
- Asia Pacific $1,158.866M = 87.2% of Q2 revenue, up from 77.7% in Q2'25; Americas 8.2%; EMEA 4.6%
- Sequential margin give-back
- Non-GAAP operating margin 37.5% (Q1'26) to 33.7% (Q2'26) — opex rose $48.9M QoQ to $356.808M while gross margin fell 60.9% to 59.8%
- MLTP acquisition
- 2026-04-08: 75%-controlling JV with HTP Holding SAL (MultiLane) forming MultiLane Test Products Holding LLP, $157.8M purchase price, targeting AI-datacenter high-speed-interconnect test
- MLTP purchase-price allocation
- $131.6M goodwill (83% of price, non-deductible), $46.7M intangibles (4.7-yr weighted life), $15.3M net tangible assets; $35.8M noncontrolling interest
- Balance sheet
- Zero debt (the $200.0M short-term debt at 2025-12-31 was repaid); $517.1M cash + marketable securities; accounts receivable $1,109.711M vs $786.913M at year-end
- Goodwill history
- $663.817M net goodwill against $762.566M of cumulative accumulated impairment losses — Teradyne has written off more acquisition goodwill than it currently carries
- Technoprobe stake
- $514.957M equity-method investment; equity in net earnings of affiliate was a $1.946M drag in Q2 FY26 (a $5.927M drag in Q2 FY25)
- Effective tax rate
- 15.06% GAAP in Q2 FY26 ($66.788M on $443.619M pre-tax)
Growth drivers
- HBM/DRAM memory test — rising test intensity per stack as HBM3E→HBM4 stack height and complexity grow; memory revenue $203M in Q1 FY26 with Magnum 7 ramp (fact)
- AI SoC/compute test — hyperscaler custom accelerators (ASICs) and GPUs drive high-end SoC tester demand; Semi Test $1.111B in Q1 FY26 (fact)
- Networking/co-packaged optics test for AI clusters (fact, named on call)
- Rising test intensity broadly — advanced nodes, chiplets/advanced packaging and tighter quality requirements raise tester time per device (estimate)
- Share gains in compute/DRAM/HBM — TER targets ATE share moving from ~30% toward 35-38% (estimate, sell-side)
- Recurring service + consumables as the installed base of testers grows (estimate)
Reported financials — SEC EDGAR
Audited GAAP figures pulled from SEC filings · latest filing 2026-02-19. The audited primary-source spine — not financial advice.
Revenue — annual (GAAP)
Margins & balance sheet — FY’25
Bull & bear
Teradyne is being paid for its position on the AI test toll booth exactly as designed: revenue doubled YoY, Semiconductor Test grew +128% with Memory up +248%, the company beat the high end of its own guidance, and it does it with zero debt, $517M of net cash and a 28.5% free-cash-flow margin — with management pointing at rising wafer-fab-equipment investment as the 2027 setup.
- The magnitude is hard to argue with: +103.9% YoY revenue growth on a $1.3B quarterly base, and a second consecutive record quarter that exceeded the high end of the company's own guidance rather than merely a sell-side estimate.
- Memory is the new engine: $212.333M, +248.4% YoY from $60.943M, on DRAM/HBM strength plus a NAND final-test resurgence. Memory test had been the cyclical laggard; its inflection broadens the story beyond SOC and adds a second AI-levered growth vector.
- SOC did not stall to make room: $842.971M, +112.6% YoY. Both major Semiconductor Test lines more than doubled in the same quarter, which is a market-expansion signal rather than share-shift or a mix trick.
- Operating leverage YoY is enormous: gross margin 57.2% to 59.8% and GAAP operating margin 13.9% to 32.9%, so incremental revenue dropped through at a very high rate. GAAP EPS went $0.49 to $2.38.
- Cash generation is real and unlevered: Q2 operating cash flow $469.138M and ~$378.4M of free cash flow (28.5% margin) with zero debt, having repaid the $200.0M of short-term borrowings outstanding at year-end. $517.1M of net cash plus a $515.0M Technoprobe stake sit outside the operating business.
- The MLTP joint venture is aimed at the right adjacency: a 75%-controlled vehicle with MultiLane for $157.8M, explicitly to build test solutions for AI-datacenter high-speed data connections. Interconnect test is where content growth is migrating as rack-scale AI systems proliferate.
- Duopoly structure and installed-base economics mean the demand does not need to be won twice: ATE is a two-player global market, and testers pull recurring service and consumable-adjacent revenue ($109.5M of Q2 revenue was recognised over time).
- Management's forward framing is about capacity, not inventory: 'rapid increase in wafer fab equipment investment sets the stage for continued growth in 2027 and beyond' — a WFE-cycle statement, which historically leads ATE demand rather than coinciding with it.
The sequential picture contradicts the YoY headline: Q3 guidance implies revenue down ~6% and non-GAAP EPS down ~19% from Q2, gross margin already fell 60.9% to 59.8% and non-GAAP operating margin 37.5% to 33.7% as opex jumped $48.9M in one quarter, and 87.2% of revenue now comes from Asia Pacific — all capitalised at roughly 37x annualised non-GAAP earnings on what is structurally a cyclical equipment business.
- Q3 guidance steps down on both lines: revenue midpoint $1,250M versus $1,328.990M delivered (-6.0% QoQ) and non-GAAP EPS midpoint $2.00 versus $2.47 (-19.0% QoQ). Teradyne gives no annual guidance, so there is no company-sourced visibility past September.
- Margins peaked in Q1, not Q2: gross margin 60.9% to 59.8% and non-GAAP operating margin 37.5% to 33.7% — a ~380bps sequential contraction on 3.6% MORE revenue. Memory test carries lower gross margin than SOC, so the fastest-growing line is the margin-dilutive one.
- Operating expenses grew far faster than revenue sequentially: $307.947M to $356.808M (+15.9%) against +3.6% revenue. S&A rose $25.8M and E&D $20.7M, part of it MLTP consolidation from 2026-04-08. If revenue steps down as guided while this cost base persists, decremental margins will be steep.
- Geographic concentration worsened sharply: Asia Pacific is 87.2% of Q2 revenue (from 77.7%), with SOC alone at $814.023M of $842.971M sold into Asia Pacific. The company's own risk factors name US export controls and China tariffs — a single regulatory action hits nearly nine-tenths of revenue.
- The MLTP deal is 83% goodwill: $131.6M of the $157.8M purchase price, non-deductible, plus $46.7M of intangibles amortising over only 4.7 years. Against a track record of $762.566M of cumulative accumulated goodwill impairment — Teradyne has already written off more acquisition goodwill than the $663.817M it currently carries.
- Robotics remains the unfixed problem: $99.917M (7.5% of revenue) growing 33.5% — respectable but far slower than the test business, and it sits inside a goodwill base that has absorbed large historical impairments. It dilutes the 'pure AI test toll booth' framing.
- The Technoprobe equity stake is a slow drag: carrying value fell to $514.957M from $537.098M, and equity in net earnings of affiliate was a $1.946M deduction in Q2 FY26. A ~$515M asset producing negative income is dead capital until it turns.
- Working capital is inflating with the boom: accounts receivable $1,109.711M versus $786.913M at 2025-12-31 (+41%) and prepayments $468.174M. That is the classic late-cycle equipment pattern — shipments ahead of collections.
- Valuation prices continuation, not cyclicality: $365.80 against annualised Q2 non-GAAP EPS of $9.88 is ~37x, and ~12.7x TTM revenue on an enterprise value of ~$56.7B. Even after fading from a $437.92 52-week-high close, the multiple assumes the ATE cycle does not roll.
What it is worth
Forward P/E on non-GAAP EPS, cross-checked vs the ATE duopoly (Advantest) and against a cyclical-peak haircut
~$300-345 (where it traded post-print drawdown) if H2 visibility disappoints / AI-capex digestion compresses the multiple (fact for the ~$345 level; estimate for the path)
~$400-440 (recent trading range
52-wk-high area ~$438 close) — fair if FY26 lands near sell-side ~$6B and H2 re-books (estimate)
~$525 (BofA target raised Jun 23 2026, on share-gain + ~$6B revenue path) (fact)
Quality, net-cash duopoly asset trading at a premium AI multiple (~40-50x fwd; trailing optics >100x) — the debate is peak-earnings risk vs durable test-intensity growth, not business quality.
SWOT
Strengths
- Duopoly position with Advantest — together >80% of global ATE; near-impossible for a new entrant to win qualified test slots (fact)
- Magnum-family memory testers are an industry reference for HBM test; Magnum EPIC/Magnum 7 well-positioned for HBM stack-complexity growth (fact)
- Fortress balance sheet — $0 debt, net cash, ~16% FCF margin; funds buybacks/dividend through the cycle (fact)
- High incremental margins — non-GAAP operating margin jumped to 37.5% in Q1 FY26 on volume, showing strong operating leverage (fact)
- Diversified test franchise across SoC, memory, networking, plus Product Test and Robotics optionality
Weaknesses
- Deeply cyclical, lumpy, order-driven revenue — Q2 FY26 guided down sequentially and H2 visibility is thin (fact)
- Customer concentration — a handful of leading-edge logic/memory makers and OSATs drive most demand (estimate)
- Robotics (UR + MiR) has underperformed — revenue declined YoY through 2025, prompting a ~10% headcount cut and reorg; a drag/distraction vs the test core (fact)
- Valuation leaves no margin for error — forward P/E in the ~40-50x range, so any 'peak demand' hint resets the stock (fact)
- Memory test share contested — Advantest holds a large slice of certain GPU/HBM-related niches (estimate)
Opportunities
- HBM4 / 16-Hi stacks raise test time and tester count per wafer — secular test-intensity tailwind into 2026-27 (fact/estimate)
- Custom AI ASIC proliferation at hyperscalers expands the high-end SoC TAM beyond merchant GPUs (estimate)
- ATE share gains in compute/DRAM toward 35-38% on a $12-14B market = path to ~$6B revenue (estimate, sell-side)
- Advanced packaging / chiplet test and system-level test as new attach points (estimate)
- Capital return — net-cash balance sheet supports continued buybacks at scale
Threats
- AI capex digestion / inventory correction would hit test orders fast given the lumpiness (estimate)
- Advantest taking memory/HBM share in key sockets (estimate)
- China exposure / export-control escalation could curb a meaningful slice of ATE demand (estimate)
- Customer in-house or alternative test approaches reducing external tester intensity (low-probability, estimate)
- Multiple compression — if AI growth normalizes, a ~40-50x forward multiple is the primary downside risk (fact)
Moats, dependencies & bottlenecks
Moats
TER + Advantest >80% of global ATE; testers are qualified into a customer's production flow and rarely swapped — switching means re-qualifying every device (fact)
Magnum EPIC/Magnum 7 are reference HBM testers, but Advantest contests memory share — leadership is real but not uncontested (fact/estimate)
Large installed tester base drives recurring service, spares and software revenue and anchors the next-gen sale (estimate)
Test-engineering depth and breadth across SoC/memory/networking is hard to replicate at sub-scale; rising test complexity favors incumbents (estimate)
Dependencies
GPU/ASIC makers, memory makers) ~70% of revenue is AI-related; an AI-capex digestion or HBM inventory correction would hit test orders quickly (fact)
SK Hynix, Samsung, Micron, Nvidia/ASIC programs, ASE/Amkor) Concentrated set of buyers; one program's timing moves a quarter (fact/estimate)
FPGAs, power/RF parts, handlers) COGS is mostly purchased hardware; component availability/pricing flows straight to gross margin (estimate)
Share is split socket-by-socket; losing a key HBM socket caps the best-growth niche (estimate)
ATE demand has meaningful China exposure; export-control escalation could curtail a slice of the market (estimate)
Advantages
- Toll-booth economics — paid per chip tested, with test intensity rising faster than unit volume as HBM/chiplets get more complex (fact/estimate)
- Duopoly pricing power and qualification lock-in (>80% share with Advantest) (fact)
- Capital-light model — ~5-7% capex intensity, outsourced manufacturing, ~16% FCF margin (fact)
- Net-cash, zero-debt balance sheet that funds buybacks and weathers the down-cycle (fact)
- Operating leverage proven in Q1 FY26 — incremental volume converted to 37.5% operating margin (fact)
Weaknesses
- Cyclical, order-driven revenue with low forward visibility — guided down sequentially for Q2 FY26 (fact)
- Premium valuation (~40-50x forward) magnifies any demand-peak signal into a sharp drawdown (fact)
- Memory/HBM test share contested by Advantest (estimate)
- Robotics segment declining YoY and restructured (~10% headcount cut) — a persistent drag (fact)
- High customer concentration in a handful of leading-edge chip/memory makers (estimate)
- China/export-control exposure adds a regulatory tail risk (estimate)
Bottlenecks
- Order timing / program lumpiness — a single large AI test program slipping a quarter swings reported revenue (fact)
- Second-half FY26 visibility — H1-weighted 55-60% means H2 must be re-booked; the bottleneck is forward order flow, not capacity (fact)
- Memory-test share vs Advantest in the hottest HBM/GPU sockets (estimate)
- Component/instrument supply for high-end testers if AI demand spikes faster than the supply chain (estimate)
- Robotics turnaround — sub-scale segment that consumes management attention without moving the needle (fact)
Top signals & trends
Top signals
Direct read on HBM/DRAM test-intensity thesis; sustained growth validates the toll-booth (fact)
The single biggest swing factor — H2 must be re-booked; bookings will confirm peak vs pause (fact)
Mix shift toward AI test is the growth engine; further rise is confirmatory (fact)
Watch for TER losing key HBM4 sockets — caps the best-growth niche (estimate)
At ~40-50x forward, the multiple is the dominant downside lever if growth normalizes (fact)
Small, but a return to growth would remove an overhang; continued decline keeps the 'sell it' question open (fact)
Trends
More test time and tester count per stack — core test-intensity tailwind (fact/estimate)
Broadens high-end SoC test TAM beyond merchant GPUs (estimate)
New test-attach points as devices disaggregate (estimate)
Same AI wave that drives demand can correct sharply, given order lumpiness (estimate)
Regulatory tightening could remove a slice of addressable demand (estimate)
Robotics end-markets weak — drove TER's UR/MiR reorg and headcount cut (fact)
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.
AMD/Xilinx, Lattice, analog/power/RF suppliers) Supply the instruments, ASICs, FPGAs and power/RF parts that make up most of tester COGS
Teradyne outsources most system manufacturing — keeps it capital-light
Probe cards/interface hardware used alongside testers in the test cell
Handlers/contactors that integrate with test systems (also a competitor in places)
Leading-edge foundry; tests advanced-node logic/AI SoCs — drives high-end SoC tester demand
GPU/AI accelerator demand cascades to SoC + HBM test capacity
Korea-listed; leading HBM/DRAM maker, a primary memory-test buyer (HBM4 ramp 2026)
US memory maker ramping HBM4 — direct HBM/DRAM test customer
Korea-listed; memory + foundry, both SoC and memory test demand
Outsourced assembly & test houses run large tester fleets — major ATE buyers
The other half of the ATE duopoly; ~31% global ATE share (2024) and a strong/contested position in memory/HBM and GPU SoC test — TER's primary rival (fact)
Test handlers, contactors, and back-end test/inspection; overlaps in test-cell hardware and consumables, more handler-focused (fact)
Probe cards and test-and-measurement for wafer-level test, including HBM/DRAM — adjacent/complementary but competes in the test ecosystem (fact)
Wafer-level burn-in/test; niche but exposed to the same AI/memory test-intensity theme (fact)
Taiwan-listed; competes in lower-end and select test/measurement segments (fact)