Cadence: 22% Growth, but the Profit Has to Reach the Share

The 60-second read

Five things to know

1. Sales growth outpaced operating profit. Revenue rose 22%, while GAAP operating income increased less than 1% and GAAP EPS remained $1.05.
2. Recurring demand is strong but not a pure organic measure. Recurring revenue grew 18%, with acquired scope and product mix requiring separate analysis.
3. The guidance raise did not raise cash guidance. FY2026 revenue midpoint increased $102M and adjusted EPS midpoint increased $0.30; OCF remains approximately $2.0B.
4. Repurchases did not shrink diluted ownership. Q3 buybacks used about $400M, while diluted weighted-average shares rose 0.54% year over year.
5. The valuation methods disagree. Earnings-based routes give about $304 per share; an owner-cash route gives about $209, supporting a rounded $270 anchor with High uncertainty.

Three scenarios, one probability-weighted number

Explicit analyst probabilities; all forward values are assumptions.

Signposts: what would change our mind

February 2027 is an analytical review window for Q4/FY2026; exact future release date is not verified.

The tape — growth needs to reach the share

Our call is AVOID for fresh capital at the $322.17 reference price, with three-out-of-five conviction and High uncertainty. This is a valuation judgment about the price of the business, not a prediction that the company will stop growing. Our three explicitly modeled routes average $272.33, rounded to a $270 fair-value anchor. The method range is approximately $209–$304. Its width is meaningful: a high adjusted earnings multiple and a cash model that recognizes the economic cost of compensation give very different answers.

Cadence reported strong revenue and recurring-demand growth, but the third quarter shows how much work remains between a promising product cycle and shareholder economics. Revenue rose 22%, while GAAP operating income rose less than 1% and diluted GAAP earnings per share remained $1.05. Acquisition accounting explains part of that gap. Stock compensation, restructuring, product mix and a larger cost base also matter. The quarter merits a conversion analysis rather than another celebration of a headline beat.

The reference price is the completed October 26 close from FMP. The release date is verified as October 26; its exact clock time is not established here. We therefore do not label that close, or its 1.79% daily increase, as a reaction to the earnings release. The SEC acceptance time and the webcast appointment are not substituted for the release clock. No independently verified post-release trading quote is used in this valuation.

The retrieved 52-week range is $262.75–$416.69. That range shows substantial changes in expectations, not a statistically valid valuation band. Our $216 entry checkpoint is a 20% discount to the $270 anchor and applies only with the operating thesis intact. A lower price caused by permanent damage to design demand would require a new valuation. Conversely, better cash conversion and stronger evidence that AI tools raise revenue per customer could justify a higher anchor before that price is reached.

The print — sales rise faster than profit

Revenue was $1.633 billion, compared with $1.339 billion a year earlier, a calculated 21.99% increase. Product and maintenance contributed $1.487 billion and services contributed $146.0 million. GAAP operating income was $428.7 million versus $425.4 million, an increase of only 0.78%. The incremental operating profit on the additional $294.5 million of revenue was approximately 1.13%. That ratio is a diagnostic of this particular comparison, not a forecast of the margin on every future sale.

GAAP operating margin fell from 31.8% to 26.2%, or roughly 560 basis points. Non-GAAP operating margin also declined, from 47.6% to 44.1%, so acquisition amortization is not the entire explanation. Adjusted EPS increased from $1.93 to $2.11, while GAAP EPS was unchanged at $1.05. The two measures answer different questions and both belong in the analysis. Quoting only adjusted EPS would conceal the size of the expense bridge that an owner still needs to understand.

The retained provider earnings record shows consensus revenue of $1.613 billion and EPS of $2.05. Actual results exceeded those snapshots by approximately 1.26% and 2.93%, respectively. The EPS beat uses the non-GAAP actual supplied by the provider. It is not a comparison of GAAP EPS with an adjusted estimate. Consensus data can update after a release, so this packet preserves the retrieved record rather than claiming a uniquely definitive pre-announcement consensus.

Management raised its full-year revenue midpoint from $6.300 billion to $6.402 billion, an increase of $102 million, and its adjusted EPS midpoint from $8.10 to $8.40. The new revenue midpoint implies approximately 20.86% growth over the rounded $5.297 billion fiscal 2025 base. Full-year operating-cash-flow guidance remains approximately $2.0 billion and capital expenditures remain approximately $240 million. An improved earnings outlook without a higher cash forecast is one reason to keep the conversion question central.

Product mix — recurring is not the same as organic

Core EDA accounted for 69% of third-quarter revenue, Semiconductor IP for 15% and System Design and Analysis for 16%, using the rounded issuer mix table. Management reported growth of 20%, 24% and 30%, respectively. These are product categories within one reportable segment; standalone operating profits for the categories are not supplied. We do not apply the consolidated margin to each category and then present those manufactured amounts as reported segment earnings.

Recurring revenue grew 18% year over year and represented 79% of the quarter. Revenue recognized over time accounted for 74%, with other recurring revenue contributing another 5%. The remaining 21% was up-front revenue. This recurring mix is a valuable stabilizer, but it is not synonymous with a subscription-only business. The company includes specified services, royalties, maintenance and hardware-related arrangements in its recurring definition. Hardware shipment timing and license arrangements still affect period comparisons.

The trailing-twelve-month recurring share was 78%, compared with 80% a year earlier. A lower mix percentage does not mean recurring dollars declined; current recurring dollars grew strongly while other categories also expanded. Conversely, recurring growth is not an organic-growth disclosure. Cadence acquired Hexagon’s design and engineering business in February, and the current earnings release does not provide a complete Q3 acquired-revenue bridge. The packet therefore avoids labeling either 22% total growth or 18% recurring growth organic.

Backlog reached $8.2 billion, while revenue expected in the next twelve months from remaining performance obligations was $4.2 billion. The latter is about 51% of the backlog, not an instruction to recognize all $8.2 billion next year. Backlog is evidence of contracted demand whose timing depends on contract terms and performance. The most useful follow-up is whether the near-term obligation pool converts as expected while new bookings replace it, rather than assuming a record stock of contracts guarantees an immediate acceleration in cash collections.

Profit bridge — the adjustment is larger than GAAP income

The income statement records $173.9 million of product and maintenance costs and $61.8 million of services costs. Together they are $235.6 million, leaving $1.398 billion of GAAP gross profit, or approximately 85.6% of revenue. The provider’s current income-statement mapping omitted product costs from cost of revenue and consequently overstated gross profit. We use the primary statements throughout; a clean-looking data feed is not a substitute for checking the actual row definitions.

Research and development was $537.5 million, marketing and sales $242.5 million, and general and administrative expense $113.0 million. The statement separately shows $33.8 million of acquired-intangible amortization and $42.2 million of restructuring. Total costs and expenses were $1.205 billion, leaving $428.7 million operating income. Interest expense of $30.4 million and other income of $10.0 million produce $408.3 million pretax income; tax expense of $118.6 million leaves $289.7 million net income.

The non-GAAP net-income bridge adds $144.8 million stock compensation, $79.9 million total acquired-intangible amortization, $25.0 million acquisition and integration costs and $42.2 million restructuring, with smaller deferred-compensation, investment and tax effects. Adjusted net income is $582.2 million. The $292.4 million net adjustment is slightly larger than GAAP net income itself. Total amortization exceeds the separately displayed expense line because some acquired-technology amortization is classified in costs of revenue; those amounts must not be added twice.

Amortization is noncash in the current quarter, but the acquired assets required cash or shares when purchased. Stock compensation is also noncash at grant expense recognition, yet it either dilutes owners or requires cash repurchases to offset issuance. Restructuring may improve future efficiency, but it is still an economic cost whose cash timing deserves monitoring. The appropriate response is to understand and compare the bridge, not to declare all adjustments worthless or to treat all excluded costs as permanently irrelevant.

Earnings quality — dilution survives the buyback

The quality assessment separates cash generation from per-share ownership. Third-quarter operating cash flow was $423.7 million and capital expenditures were $61.8 million, producing $361.9 million free cash flow. That is a solid positive result and exceeds GAAP net income. However, the quarter also included $144.8 million of stock compensation, while cash repurchases were approximately $400.0 million. It is possible to generate cash, return substantial cash and still fail to reduce the diluted share denominator.

Diluted weighted-average shares increased from 273.798 million to 275.278 million, or about 0.54% year over year. Hexagon consideration included approximately 3.2 million issued shares, and employee equity programs create additional movement. Weighted-average shares are not a point-in-time share count and will not exactly reconcile to cash repurchases during a single quarter. The important conclusion is narrower: repurchase spending alone is insufficient evidence that each remaining share owns a larger fraction of the business.

Subtracting stock-compensation expense from free cash flow gives an economic diagnostic of approximately $217.2 million for the quarter and $821.3 million year to date. These are our computed proxies, not issuer non-GAAP measures or GAAP cash-flow lines. They are deliberately conservative indicators of the cash required if compensation were replaced economically, not claims that the expense must be paid again dollar for dollar. A valuation must avoid subtracting the same compensation cost twice through both cash flows and an inconsistent share forecast.

The nine-check scorecard also flags the adjustment gap, margin decline, cash coverage of repurchases, acquisition comparability, collection timing, inventory build and guidance quality. Reported DSO was 66 days versus 55 a year earlier. Inventory was $448.8 million versus $303.5 million at year-end, alongside a larger hardware and acquired-business footprint. Those changes are monitoring signals, not proof of revenue manipulation. The new revenue and EPS guidance is constructive, while unchanged cash guidance keeps the conversion test open.

Cash and capital — acquisitions change the funding picture

Year-to-date operating cash flow of $1.414 billion less $163.3 million capital expenditures leaves $1.251 billion free cash flow. Cash used in business combinations was $2.143 billion, and repurchases used $800.0 million. These investments and returns were financed from a combination of cash generation, starting liquidity, revolving-credit movements and other equity-related cash flows. They explain why a company reporting positive free cash flow can still finish the period with substantially less cash.

Cash and equivalents fell from $3.001 billion at year-end to $1.452 billion. Debt principal was $2.650 billion, including a $150 million revolving-credit balance. The carrying value was $2.633 billion after debt-accounting adjustments. Our valuation uses principal debt less cash, or $1.198 billion net debt, so its capital claim is explicit. The provider’s current balance-sheet mapping omitted the revolver from total debt and treated long-term liabilities as total liabilities; primary-source totals replace those mappings.

Hexagon’s design and engineering business closed on February 23. The prior interim filing records gross purchase consideration of $3.101 billion, including approximately $2.2 billion cash and $902.2 million of shares, before the effect of cash acquired. The transaction expands system-level design capabilities and may strengthen cross-selling, but it also adds integration work and acquired-intangible amortization. Goodwill plus acquired intangibles now represent approximately 55% of total assets. That is not an impairment prediction; it increases the importance of testing acquired returns.

Management expects approximately $2.0 billion operating cash flow and $240 million capex for 2026, implying approximately $1.760 billion free cash flow under the simple issuer definition. Against year-to-date actuals, the arithmetic implies roughly $508.8 million fourth-quarter free cash flow if both rounded annual guides are achieved. This is a derived remainder, not a separately issued Q4 FCF forecast. Management expects approximately $400 million Q4 repurchases and more than half of annual free cash flow to be used for repurchases. The new $2.5 billion authorization is capacity, not an obligation or a completed cash return.

Valuation — price the compensation and the growth

The earnings route assumes $9.50 normalized EPS and a thirty-two-times multiple, producing $304.00 per share. This EPS is close to the retrieved 2027 consensus of $9.552, while the multiple still grants Cadence a substantial premium for its embedded design workflows, recurring contracts and potential AI productivity gains. It is a deliberately favorable earnings-based case, not a low-multiple liquidation estimate. Adjusted EPS excludes material costs, so the multiple cannot be compared casually with a peer whose earnings include those costs.

The enterprise-value route assumes $3.4 billion normalized EBITDA at twenty-five times, subtracts $1.198 billion principal-based net debt, then divides by 275.278 million diluted weighted-average shares. It produces $304.43 per share. This is a simplified capital bridge that does not separately capitalize leases. We hold the share denominator constant instead of assuming future repurchases automatically create accretion. The normalized EBITDA input and the multiple are analyst assumptions, not company guidance or a disguised provider estimate.

The cash route uses five annual enterprise owner-cash assumptions of $1.95 billion, $2.30 billion, $2.70 billion, $3.10 billion and $3.50 billion. At an 8.5% discount rate and 3.5% terminal growth, enterprise value is approximately $58.612 billion; subtracting the same net debt and dividing by the same shares gives $208.57. These cash flows include an assumed economic replacement cost for stock compensation. They are not the issuer’s levered free cash flow relabeled as enterprise cash. The first year already requires improvement over the current cash-after-compensation diagnostic, so this route is not a no-growth stress case.

The equal-weight mean is $272.33, rounded to $270. Equal weighting is a disclosed judgment, not a claim that the three methods are independent estimates with equal statistical precision. At $322.17, thirty-two times requires approximately $10.07 sustainable EPS, while twenty-five times enterprise EBITDA implies $3.595 billion. A perpetual 8.5% discount and 3.5% growth equation requires approximately $4.342 billion enterprise cash flow. That last figure is a steady-state illustration, not a next-year forecast. The wide method dispersion and terminal-value dependence explain High uncertainty and the required margin of safety.

Wall Street — dated targets are not a fresh response

The retrieved FMP target consensus is $403.27, with a $300–$450 range and $420 median. Its ratings collection contains 26 Buy, four Hold and one Sell observations. These are provider records; they are not a verified count of independent analysts who all updated their views after this quarter. The aggregate sits above our valuation, but its timing and underlying earnings definitions matter as much as the headline difference.

The latest specific target in the saved feed is Deutsche Bank’s October 7 increase to $420 from $400. FBN Securities initiated at $375 on September 11. Both precede the October 26 earnings release. We therefore use them as the dated Street backdrop and make no claim that either is a response to Q3 results. The full proprietary reports were not supplied, so we do not invent bank-specific arguments or assume the cited target changes reflect the precise cash and compensation treatment used here.

The annual estimate snapshot gives approximately $6.323 billion revenue and $8.150 EPS for 2026, below management’s new midpoint of $6.402 billion and $8.40. That is consistent with a provider snapshot that may not yet incorporate the print; it is not evidence that analysts have already rejected the guidance. For 2027, the snapshot shows $7.180 billion revenue and $9.552 EPS, with fifteen EPS contributors. For 2028, it shows $8.024 billion revenue and $10.989 EPS, with eight EPS contributors. Less distant coverage and narrower uncertainty should not be assumed from the presence of precise decimals.

Our interpretation differs primarily in the price paid for owner economics, not in denying demand for design software. The bull case can plausibly support a premium valuation if recurring contracts, new AI capabilities and system-design expansion deliver sustained earnings and cash growth. The current reference already pays a large multiple of management’s adjusted EPS guide, however, and the cash yield is much lower than that multiple might suggest. A future Street upgrade becomes more informative when it specifies the earnings basis, share count and compensation assumptions.

Management — measure outcomes beyond activity

Management describes strong demand for AI-driven designs and a widening opportunity from agentic design tools. Maestro is a portfolio of specialized AI models, and the release reports initial workflow results of up to 30% higher quality and five times greater token efficiency than general-purpose models. Those are issuer-described workflow results, not independently verified companywide financial gains. They should not be multiplied across all revenue or translated directly into margin improvement without evidence of adoption, pricing and deployment economics.

The company reports more than one hundred active AI Super Agent customer engagements, sixteen new digital and signoff full-flow wins and another record hardware quarter. These are useful signs of engagement and product reach. An engagement is not necessarily a completed sale, and a design win may monetize over time. The next useful disclosures would connect deployment to paid usage, contract renewals, cross-selling or demonstrable reductions in support and delivery cost. The packet treats activity metrics as leading indicators rather than substituting them for recognized profit.

Capital allocation deserves a parallel scorecard. The Hexagon transaction expands the addressable workflow, but its success must eventually appear in organic demand, retained customers, integration progress and cash returns relative to the purchase price. The current release does not separately quantify all acquired Q3 revenue, and category-level profitability is not disclosed. That uncertainty prevents a clean acquisition return calculation today. It is better to state the missing bridge than to fabricate an apparently precise return on invested capital from incomplete data.

The board’s additional $2.5 billion authorization signals flexibility and an intention to return capital. The actual Q3 program bought approximately 1.305 million shares at a computed average $306.51, below the current reference. That short-term price comparison does not establish whether repurchases created value; the relevant comparison is intrinsic value and the alternative uses of cash. Our scorecard gives credit for raised revenue and EPS guidance, retains a question on unchanged cash guidance, and requires the diluted denominator to confirm that repurchases benefit continuing owners.

Risks — premium expectations need durable economics

The first risk is conversion. Demand can remain strong while product mix, acquired costs, engineering investment and compensation prevent revenue growth from reaching owners. Both GAAP and non-GAAP operating margins declined in the current comparison. If that persists after integration and restructuring costs normalize, a premium earnings multiple becomes more difficult to defend. The observable warning would be repeated margin misses or unchanged cash guidance despite further revenue increases.

The second risk is acquisition execution. Goodwill and intangibles constitute a large share of assets, and Hexagon increased both the operating scope and the share base. Cross-selling may create genuine value, but integrating engineering products, customers and employees is not automatic. Acquired revenue can make total growth appear stronger while obscuring weaker internal demand. An impairment charge would be a later accounting signal; the earlier economic signs would be weak bookings, customer losses, integration delays or returns below the cost of the capital deployed.

The third risk is export and compliance exposure. China accounted for 16% of current-quarter revenue. Cadence’s July 2025 DOJ and BIS settlements resolved historical export-violation matters and involved $140.6 million aggregate net penalties and forfeitures, while ongoing audit and compliance obligations remain. The prior interim filing also discusses evolving restrictions and a dated suspension of an affiliate rule. We do not assume that a historical policy schedule proves current legal status; the live rules and any subsequent company updates must be checked before a specific regulatory catalyst is traded.

The fourth risk is valuation duration. A large portion of the cash model’s value comes from earnings beyond the explicit five-year horizon. Higher discount rates, lower long-run growth or less efficient compensation replacement can reduce value without a near-term revenue collapse. The bull case instead requires sustained recurring demand, better margin conversion, successful system-design integration and stronger cash per share. We assign 25% to a $200 bear case, 50% to the $270 base and 25% to a $418 bull case. The $289.50 weighted value is below the reference price; these probabilities are transparent assumptions, not measured frequencies.

Checkpoints — give the next report a test

The next earnings review should start with the margin bridge rather than the headline revenue surprise. Management’s fourth-quarter outlook is $1.695–$1.725 billion revenue, 30%–31% GAAP operating margin, 44%–45% non-GAAP margin and $2.19–$2.25 adjusted EPS. Meeting those ranges would be useful progress, especially the recovery in GAAP margin from Q3. A beat driven by a favorable tax item or share denominator would not carry the same information as better operating conversion.

The full-year cash test is separately measurable. The approximately $2.0 billion operating-cash-flow guide and $240 million capital plan imply a stronger fourth-quarter contribution than Q3. Receivables, inventory and deferred-revenue timing need to be read alongside that result. A temporary working-capital movement can explain a quarter, but persistent divergence between revenue, adjusted earnings and cash should affect the valuation. Our thresholds are research checkpoints intended to trigger investigation, not automatic trading signals.

Backlog and recurring revenue remain central to the demand thesis. We would look for recurring growth above fifteen percent and clarity on the $4.2 billion of next-twelve-month revenue expected from remaining obligations. That threshold is our analytical hurdle, not company guidance. Stable or improving customer commitments combined with stronger cash collections would reduce uncertainty. A headline backlog record without information about timing, acquired scope or cancellation conditions would provide less incremental evidence.

The ownership test completes the framework. Cadence expects Q4 repurchases of approximately $400 million, but we will inspect diluted and period-end shares separately. A stable or falling denominator would support the claim that returns benefit continuing holders; rising shares despite heavy spending would require explanation from acquisitions or employee issuance. The future report date has not been independently confirmed, so February 2027 is a review window rather than a claimed scheduled release. Until new evidence arrives, the current price offers insufficient cushion against the $270 anchor.

Source notes — definitions stay attached to the numbers

The source of record is Cadence’s October 26 Form 8-K, accession 0000813672-26-000116, including Exhibit 99.01 and the Exhibit 99.02 CFO commentary. The covered quarter ended September 30, 2026. The filing’s earliest event date is October 23 because it also reports the board’s repurchase authorization; that date is not used as the earnings release date. The current Q3 10-Q had not appeared in the source set when research was prepared, so the earnings release and commentary establish current figures and the prior Q2 filing supplies detailed acquisition context.

The source archive retains nine earnings releases from Q3 2024 through Q3 2026, with their original tables, primary URLs, hashes and retrieval times. Quarterly income statements, below-the-line items, GAAP-to-non-GAAP bridges, balance sheets and category/geography tables remain available in the dataset. Quarterly cash flow is computed from issuer year-to-date statements where necessary; the oldest quarter uses its CFO commentary. Three annual reports supply cash history, repurchases, compensation, corporate structure and employee counts. Missing category profits are explicitly unavailable, not estimated as reported facts.

All financial statement tables use U.S. dollars. Primary release statements are generally in thousands, while the packet converts major amounts to millions or billions and keeps per-share values in ordinary dollars. Cost-of-revenue and balance-sheet errors in the current provider mapping are corrected from SEC statements. Principal and carrying debt are both shown, with principal used consistently in the valuation. The historical 2024 provider repurchase field differs from the issuer’s program table; the packet uses the issuer’s $550.0 million program figure instead of silently accepting the larger provider field.

Market prices, target records, ratings and consensus are dated FMP snapshots. Target news is not treated as a substitute for proprietary analyst reports. Release-date daily returns are broad windows and do not establish a post-earnings reaction. The valuation, scenario probabilities and decision thresholds are analyst assumptions computed in the retained Python work, while management guidance is labeled separately. Historical free cash flow, the cash-after-compensation diagnostic and enterprise valuation cash flows are different definitions. Keeping those distinctions visible is necessary to assess the conclusion and its limitations.

Current SEC earnings release · CFO commentary · FY2025 annual report