Teradyne (TER) · Deep analysis

Report previewTER.USReport date Last updated
← Back to reports
Report contents

At a glance

  • Business model: Teradyne sells semiconductor test systems, about 80% of revenue, with roughly 10% each from robotics and product test. Orders are discrete and tied to device programs, so revenue is cyclical capital equipment. Rising test intensity, from 4% of chip capital spending in 2023 to 8% in early 2026, supports growth, while a short list of buyers constrains it.
  • Financial quality: Earnings convert to cash, with operating cash flow at 1.22 to 1.29 times net income in FY2023 through FY2025 and free cash flow at 0.82 to 0.96 times. The June 2026 balance sheet shows no long-term debt and liabilities at 29% of assets. First-half 2026 operating cash flow fell to 0.95 times net income during the ramp.
01

Business model

Business model: one test business carries the AI case

Teradyne sells semiconductor test systems, a business that produced about 80% of revenue in the mix CEO Greg Smith described at the Goldman Sachs Communacopia + Technology Conference on September 9, 2026. Robotics and product test contribute roughly 10% each. The buyers are chip designers and manufacturers, plus a growing set of cloud companies commissioning their own silicon. Test tools are purchased in discrete orders tied to a device program and a capacity plan, so revenue arrives with each order rather than through a subscription, and management described no recurring revenue stream of consequence. That makes this a cyclical capital-equipment manufacturer, currently in an expansion phase. Investing.com's coverage, citing InvestingPro data, put trailing-twelve-month revenue at $4.5 billion, up 58% from a year earlier, with gross margin at 59%, inside the 59% to 61% range Smith said the company expects to hold.

Test intensity carries as much of the case as share gains do

Test equipment's share of total semiconductor capital spending rose from 4% in 2023 to 8% in the first five months of 2026, Smith said. Multi-die packaging multiplies the quality checks a finished device requires, and smaller nodes (1.4 nanometers against 3 nanometers in his example) pack more transistors onto each wafer, so each wafer needs more test time. Management projects industry capital spending of $250 billion to $300 billion in 2028 to 2030, up from about $120 billion in 2025. Those are management's estimates, not observed results, and they do much of the work in the growth case. If test intensity keeps climbing, Teradyne grows with chip investment even without taking share. If it stalls near 8%, the company is back to riding the equipment cycle, which is why revenue in this business has historically swung hard. Memory and system-level test were both listed among 2026 share gains, so the demand base is wider than AI compute alone.

Second source today, volume share in 2027

The near-term revenue question is whether qualification work turns into production volume. At the lead merchant compute account, Teradyne qualified as a second vendor in mid-2025, released its first part in early 2026 and is running a small fleet of testers; more device qualifications are expected in the second half of 2026, with meaningful volume share in 2027 and a target of about 30% of the compute portion of that account over three to five years. In custom AI silicon, the company went from qualification to majority share at one hyperscaler between 2023 and 2026, is a second source at another, and expects volume there in 2027. Management sizes each hyperscaler program at hundreds of millions of dollars to $1 billion of annual tester demand. Networking and AI ASIC test both grew from 2025 to 2026, with management pointing to more upside in AI ASICs, where accelerators carry higher test intensity. Second-source status is the hinge: it caps near-term revenue at whatever volume the customer allocates, while the multiyear qualification cycle that makes entry slow also makes Teradyne expensive to remove once it is designed in.

Twenty percent of revenue sits in businesses that have not scaled

Robotics and product test each contribute about a tenth of revenue, and management ties both to the same AI build-out. Product test reaches parts, modules and trays rather than only wafers and packaged chips, while physical AI (robots and autonomous systems) could be, in Smith's phrasing, one of the most important growth drivers of the next few years. Neither has delivered that at scale. What protects the economics meanwhile is operating leverage: Smith said operating expense growth is being held below half of revenue growth and that gross margin stays in the 59% to 61% range, which keeps incremental revenue valuable while the company invests in merchant compute, custom silicon and silicon photonics. The flip side of selling several product lines to the same handful of hyperscalers is that those buyers know their own leverage and have alternative suppliers for every tool they buy. Management also conceded that growth will not be smooth and that competition remains intense.

What would overturn this read

Two things carry the judgment above: converting second-source positions into volume, and doing it without paying in price. Tester shipments for the compute account staying small through 2027, test's share of chip capital spending flattening near 8%, gross margin falling below the low end of the stated range as volume scales, or a hyperscaler program moving back toward a competing platform would each break part of the case. Disclosure limits what can be tracked. Teradyne does not publish per-account share, a customer concentration split or a robotics order backlog, and the conference remarks give no geographic revenue breakdown. Conversion will have to be read from segment revenue, tester shipments and the cloud buyers' own capital spending plans.

Questions worth following

  • Can Teradyne convert its compute second-source win into volume share?

    Management expects about 30% of the lead merchant compute account's tester volume within three to five years, up from a small fleet in 2026. Whether device qualifications convert into production share decides how much of the 80% test business grows with AI spending rather than with the wider equipment cycle.

  • Will hyperscaler custom silicon keep expanding Teradyne's test demand?

    Custom AI silicon across two cloud hyperscalers and one automotive and robotics customer is where management sees the largest near-term upside, with each program worth hundreds of millions to $1 billion of annual tester demand. New program awards and buyer budgets decide whether current revenue growth extends or reverts to the chip cycle.

  • Can robotics become a second growth engine for Teradyne?

    Robotics is roughly a tenth of revenue, and management calls physical AI a potentially major driver over the next few years. If the segment scales, it diversifies a business whose results hinge on a handful of chip and cloud customers; if it does not, the AI test cycle remains the whole story.

02

Financial quality

Financial quality: strong cash conversion, cyclical earnings power

Verdict: Strong, with the profit level set by the test cycle. Teradyne's earnings turn into cash, the balance sheet carries almost no debt, and capex runs at 5–7% of revenue. What no ratio resolves is that the earnings base moves with semiconductor test demand: revenue fell from $3.70 billion in FY2021 to $2.68 billion in FY2023, and net margin fell with it, from 27.4% to 16.8%, before the AI-driven recovery pushed first-half 2026 net margin to 29.9%.

MetricFY2021FY2022FY2023FY2024FY2025H1 2026Assessment
Revenue (USD bn)3.703.162.682.823.192.61Adequate, cyclical
Gross margin (%)59.659.257.458.558.660.5Strong
Operating margin (%)32.426.318.720.921.634.9Strong
Net margin (%)27.422.716.819.217.229.9Strong
ROE (%)42.528.518.020.319.625.0Strong
Liabilities/assets (%)32.730.027.624.033.129.4Strong
Current ratio (x)3.203.033.282.91n/dn/dStrong
Operating cash flow (USD bn)1.100.580.580.670.670.74Strong
Free cash flow (USD bn)0.970.420.430.470.450.58Strong
Operating cash flow/net income (x)1.080.811.291.241.220.95Strong
Free cash flow/net income (x)0.950.580.960.870.820.74Adequate
Capex/revenue (%)3.55.15.67.16.96.1Adequate
Inventory turns (x)2.871.582.022.40n/dn/dAdequate
Receivable turns (x)7.066.425.866.31n/dn/dAdequate

*H1 2026 flows cover six months and are not annualized; balance-sheet ratios use June 30, 2026 balances. n/d = not disclosed.

A cyclical earnings base, not a fragile one

ROE slid from 42.5% in FY2021 to 18.0% in FY2023 and 19.6% in FY2025. Equity barely moved over that stretch, from $2.56 billion to $2.80 billion, so the decline came from earnings rather than a change in capital structure. That is the opposite of the leverage-flattered returns a high ROE sometimes hides. Liabilities are 24–33% of assets, interest expense was $10 million in FY2025 against $670 million of operating cash flow, and no long-term debt is reported for these periods; short-term borrowings reached $220 million at the end of 2025 and were back to $20 million by March 2026. The liability side is mostly working capital and customer accruals, so debt-to-assets overstates financial risk.

Operating leverage cuts both ways, and the swing is wide: operating margin went from 32.4% in FY2021 to 18.7% in FY2023 and back to 34.9% in the first half of 2026. Teradyne reported third-quarter 2026 revenue of $1.31 billion, up 70% from a year earlier, with GAAP net income of $342.4 million and GAAP EPS of $2.17. Semiconductor Test supplied $1,072 million of that, or 82% of revenue. Management guided the fourth quarter to $1.25–$1.35 billion of revenue and GAAP EPS of $1.88–2.18, which implies the surge levels off rather than continues.

Cash has followed earnings, with one soft half

FY2023 through FY2025 were the clean stretch: operating cash flow ran 1.22–1.29 times net income and free cash flow 0.82–0.96 times net income, while capex absorbed about a third of operating cash flow. In the first half of 2026, operating cash flow of $740 million was 0.95 times net income of $780 million and free cash flow was 0.74 times net income, the first half below 1.0 since FY2022, and it arrived while revenue nearly doubled. Building working capital ahead of a ramp is the benign reading; receivables or inventory growing faster than sales would mean the reported profit is less cash-backed than in the prior three years. FY2022 is the precedent to keep in view: revenue rose that year, but operating cash flow was only 0.81 times net income as inventory reached $820 million at year-end.

The March-to-June 2025 quarters deserve the same attention. Inventory halved, from $710 million to $350 million, in the quarter that gross margin slipped 3.4 points to 57.2%. Shipment of previously built stock implies the roughly 60% gross margin of 2026 is a volume and mix story; write-downs would mean part of the improvement rests on a lower base. The reported figures do not separate the two.

What capital return and deal making say about reinvestment

Cumulative net income of roughly $3.3 billion across FY2021–FY2025 added only $240 million to equity, while cumulative financing outflows reached $3.2 billion. Most of the cash went back to shareholders, which is defensible when capex is light and ROE sits near 20%, but it leaves growth dependent on the existing asset base or on acquisitions. Goodwill rose from $520 million at the end of 2025 to $660 million at June 2026, and non-GAAP results exclude acquired intangible amortization, which is why third-quarter 2026 non-GAAP EPS of $2.29 exceeded GAAP EPS of $2.17. A 5% gap is modest. The question is whether the acquired revenue carries margin with it. Cash fell from $1.12 billion at the end of FY2021 to $290 million at the end of FY2025 and stood at $350 million in June 2026, thinner than it once was.

PeriodRevenueNet incomeGross margin (%)Total assetsTotal liabilitiesOCFFCFAssessmentTrend
FY2025 Q10.690.1060.63.710.910.160.10AdequateCyclical trough
FY2025 Q20.650.0857.23.760.930.180.13WeakTrough
FY2025 Q30.770.1258.43.961.250.050.00AdequateEarly recovery
FY2025 Q41.080.2657.44.171.380.280.22StrongInflecting
FY2026 Q11.280.4060.94.431.290.270.20ExceptionalAccelerating
FY2026 Q21.330.3860.24.931.450.470.38StrongCash catching up
FY2026 Q31.310.34n/dn/dn/dn/dn/dStrongPlateau at high level

All figures in USD bn. Fiscal quarters match calendar quarters. FY2026 Q3 comes from the company's October 27, 2026 release, which gave revenue, GAAP net income attributable to Teradyne and EPS but no balance-sheet or cash-flow detail. Fixed assets are reported annually, not quarterly.

Three things worth checking
  • Cash conversion during the ramp. Operating cash flow at 0.95 times net income while revenue doubles is the most informative number here. Two quarters above 1.0 would support the earnings; two more near 0.9 with receivables outgrowing revenue would not.
  • Goodwill and acquired intangibles. $140 million of goodwill added in one quarter needs a matching lift in operating profit. Segment margins are where that shows, and a rising goodwill balance against a flat consolidated margin would say the deals bought revenue.
  • Concentration and the turn. Semiconductor Test was 82% of third-quarter 2026 revenue, with memory and hard-disk-drive test at records. Robotics, at $96 million and profitable in the quarter, is the only meaningful piece with a different demand driver.

Questions worth following

  • Can Teradyne convert record AI test demand into cash?

    The company's profit is accelerating faster than its cash generation for the first time since 2022. Whether receivables and inventory absorb the gain determines if the 2026 earnings are as durable as the last three years suggest.

  • Can Teradyne hold peak-level margins through the next test cycle?

    Net margin has ranged from 16.8% to 29.9% over five years on roughly stable gross margin, so the level depends on volume against a fixed cost base. Customer capital spending decides whether the 2026 margin is a new baseline or a peak.

  • Are Teradyne's acquired businesses earning their cost of capital?

    Goodwill and intangibles are growing faster than revenue, and non-GAAP reporting excludes the amortization that follows. If the acquired revenue does not carry margin, capital is being converted into balance-sheet assets rather than earnings power.

Generated by DeepVal AI. Data and models may contain errors or be outdated. For research only; not personalized investment advice.

Go beyond the preview

Continue reading the full report

  1. Key risksWhat could break the investment case
  2. Valuation and margin of safetyWhat the current price assumes, and how much room for error it leaves
  3. SummaryOur overall judgement and what to watch next

Free to sign up · 20 research points · unlocking this report uses 10

20 points cover 2 full reports. No card needed.

Research reflects its generation date. Preview the first two modules.

More on Teradyne Inc