For AI assistants & researchers — machine-readable summary of this page
What this page is: Delvantic's full research page for Arm Holdings plc (ARM) — AI-driven forensic equity research: mechanical valuation models (DCF, EPV, anchored-PE, scenario) plus three independent AI lenses (Quality / Value / Sentiment). Everything below is rendered server-side; you are not missing content that requires JavaScript. All scores are predictions and research opinions, not financial advice.
Our current read (analysis of 2026-08-23): Designation Low · Gem Score -21 (−100…+100 Quality+Value blend) · Quality 35 · Value -77 · Sentiment 58 (timing only, not weighted)
Page map (sections in order; each card carries a stable
reference-name attribute you can cite):
profile-header/price-overview— company profile, live quote, market capextended-analysis— the core: three AI lens reads with findings, scores, and the analyst memofuture-predictions— our forward price-band predictionsmarket-narrative/ai-findings/gpt-critique— narrative context, cross-model findings, and an adversarial critique of our own analysis (near the end of the document)- Members-only sections (render as login gates for anonymous readers):
price-history,income-trend,key-metrics,financials(statement tables),insider-trading. The analysis above is public; the raw data tables require a free account.
More for machine readers: site briefing at
/llms.txt ·
any ticker resolves at delvantic.com/stock/TICKER ·
raw inputs are public-company filings and market data (via licensed data feeds);
every model, score, lens read, and prediction on this page is Delvantic's own analysis.
Arm Holdings plc
ARM NASDAQArm Holdings plc American Depositary Receipt is an ADR representing ordinary shares of Arm Holdings plc, a British semiconductor and software design company headquartered in Cambridge, England. The asset gives U.S. investors access to Arm’s business model, which centers on designing central processing unit architectures and related intellectual property that are widely used across mobile devices, embedded systems, and other power-efficient computing applications. Arm also provides software, tools, and associated IP solutions that support chip designers and technology manufacturers in developing products for smartphones, tablets, wearables, sensors, and connected devices. As an ADR, it serves as a U.S.-traded wrapper for the underlying ordinary shares, making the company’s equity available through the American depositary share structure. Arm Holdings plc American Depositary Receipt is positioned as a key name in the semiconductor design market, with its technology embedded across a broad range of consumer and industrial electronics.
Price Overview
Price History (1 Year)
Revenue & Net Income Trend
This company does not file structured financial statements with the U.S. SEC, so quarterly figures aren't available from our filings-based data engine. Annual figures shown here come from the sources that do cover it.
| Period | Revenue | Net Income | Net Margin | YoY/QoQ |
|---|
Key Metrics
EPS (Diluted): 0.85
Total Equity: $8.29B
Shares: 1,068,000,000
Total Debt: $0.00
Cash: $2.75B
EBITDA: $1.15B
Total Debt: $0.00
Cash: $2.75B
Revenue: $4.92B
Revenue: $4.92B
Revenue: $4.92B
Total Equity: $8.29B
Tax Rate: 21.9%
Equity: $8.29B
Total Debt: $0.00
Cash: $2.75B
Current Liabilities: $1.04B
Long-Term Debt: $0.00
Total Debt: $0.00
Total Equity: $8.29B
Shares: 1,068,000,000
Shares: 1,068,000,000
CapEx: -$545.00M
Shares: 1,068,000,000
Stock Price: $239.69
Net Income: $904.00M
Industry Benchmarks
Income Statement (Annual)
Last updated: Jul 31, 2026 9:13am (23d ago)| Metric | 2022 | 2023 | 2024 | 2025 | 2026 |
|---|---|---|---|---|---|
| Revenue | $2.7B | $2.7B | $3.2B | $4.0B | $4.9B |
| Cost of Revenue | $131.0M | $106.0M | $154.0M | $121.0M | $121.0M |
| Gross Profit | $2.6B | $2.6B | $3.1B | $3.9B | $4.8B |
| Operating Expenses | $1.9B | $1.9B | $3.0B | $3.1B | $3.9B |
| Operating Income | $633.0M | $671.0M | $111.0M | $831.0M | $900.0M |
| Net Income | $549.0M | $524.0M | $306.0M | $792.0M | $904.0M |
| EBITDA | $818.0M | $841.0M | $273.0M | $1.0B | $1.1B |
| EPS | $0.54 | $0.51 | $0.30 | $0.75 | $0.85 |
| EPS (Diluted) | $0.54 | $0.51 | $0.29 | $0.75 | $0.85 |
Balance Sheet (Annual)
Last updated: Jul 31, 2026 9:13am (23d ago)| Metric | 2022 | 2023 | 2024 | 2025 | 2026 |
|---|---|---|---|---|---|
| Cash & Equivalents | — | $1.6B | $1.9B | $2.1B | $2.8B |
| Total Current Assets | — | $3.5B | $4.2B | $4.8B | $6.2B |
| Total Assets | — | $6.9B | $7.9B | $8.9B | $10.7B |
| Current Liabilities | — | $1.4B | $1.5B | $929.0M | $1.0B |
| Long-Term Debt | — | — | — | — | — |
| Total Liabilities | — | $2.8B | $2.6B | $2.1B | $2.4B |
| Total Equity | $3.5B | $4.1B | $5.3B | $6.8B | $8.3B |
| Retained Earnings | — | $2.5B | $2.8B | $3.5B | $4.4B |
Cash Flow (Annual)
Last updated: Jul 31, 2026 9:13am (23d ago)| Metric | 2022 | 2023 | 2024 | 2025 | 2026 |
|---|---|---|---|---|---|
| Operating Cash Flow | $458.0M | $739.0M | $1.1B | $397.0M | $1.5B |
| Capital Expenditure | -$34.0M | -$64.0M | -$92.0M | -$219.0M | -$545.0M |
| Free Cash Flow | $424.0M | $675.0M | $998.0M | $178.0M | $979.0M |
| Acquisitions (net) | — | — | — | — | — |
| Net Debt Issued / (Repaid) | — | — | — | — | — |
| Dividends Paid | — | — | — | — | — |
| Stock Buybacks | — | — | — | — | — |
| Net Change in Cash | -$210.0M | $550.0M | $369.0M | $162.0M | $666.0M |
Growth Trends (YoY %)
Last updated: Jul 31, 2026 9:13am (23d ago)| Metric | 2023 | 2024 | 2025 | 2026 |
|---|---|---|---|---|
| Revenue Growth | -0.9% | +20.7% | +23.9% | +22.8% |
| Gross Profit Growth | +0.0% | +19.7% | +26.2% | +23.5% |
| Operating Income Growth | +6.0% | -83.5% | +648.6% | +8.3% |
| Net Income Growth | -4.6% | -41.6% | +158.8% | +14.1% |
| EBITDA Growth | +2.8% | -67.5% | +271.4% | +13.3% |
Deep Analysis
Narrative Economics
market-narrative step).
Claude Reading
The raw numbers tell a coherent but troubling story once you strip the narrative. Revenue went $2.70B → $2.68B → $3.23B → $4.01B → $4.92B across fiscal 2022-2026, a 16.2% CAGR overall but with clear reacceleration to 22.8% YoY in the most recent year. Operating income is $900M on $4.92B — an 18.3% operating margin despite 97.5% gross margins, meaning opex is consuming ~$3.9B annually and growing roughly in line with revenue. Free cash flow of $979M against a $256B market cap is a 0.38% FCF yield, or 261x FCF. Even if you assume FCF grows at the top-line's 23% for five straight years without further margin expansion, you get to ~$2.75B FCF in 2031 — still ~93x on today's price. The math only works if you believe operating margin expands to 40%+ AND revenue triples, both simultaneously, within a decade.
I largely agree with the synthesis and thesis-evaluation layers, and I think the Pre-Flight note is too generous when it calls this "high-growth" without emphasizing the valuation gap. The Market Narrative layer is the most honest of the bunch — it correctly flags that the story is doing 70%+ of the work. Where I'd push back on the prior models: the Market Narrative says "flat-to-low single-digit growth last 18 months," which is wrong — annuals show 23% then 23%, not deceleration. That's a factual error that weakens its bear case. Conversely, the Pre-Flight thesis undersells customer concentration: Arm's top five customers historically drive >50% of royalties, and the Qualcomm litigation demonstrated that "unavoidable architectural standard" is legally negotiable. The synthesis verdict of "High Conviction Required" is the right conclusion but sanitized — at 52x sales, 219x EV/EBITDA, and 282x earnings, this isn't a "require conviction" stock, it's a "require a specific and improbable future" stock.
The contrarian case worth taking seriously is not that Arm is a bad business — it's clearly excellent, with 97.5% gross margins, zero debt, $2.75B cash, and a genuine architectural moat. The contrarian case is that even perfect execution doesn't justify the price. Consider: if Arm hits $15B revenue by 2031 (the bull dream, ~25% CAGR sustained) and expands operating margins to 45% (aspirational for an IP licensor), you get ~$6.75B operating income, maybe $5.5B net. At a generous 35x mature multiple that's $190B — still below today's $256B market cap. The bull needs $20B+ revenue and >50% margins, which means Arm has to become materially more profitable than Visa on a margin basis while growing faster than ASML. RISC-V is a real 5-10 year risk (Meta, Google, Qualcomm all funding it), and CSS licensing — the linchpin of the bull thesis — is unproven at scale beyond a handful of announced deals. Insider activity isn't provided here, but SoftBank still owns ~90%; any secondary offering is a structural overhang.
Where the data is thin: no quarterly breakdown is provided, so I can't verify whether the 23% growth is decelerating intra-year or accelerating — that matters enormously for a stock priced on trajectory. The FCF CAGR of -1% versus earnings CAGR of 71.9% is a red flag the models glossed over: net income is growing because of accounting items (share-based comp add-backs, tax dynamics post-IPO), not because cash generation is compounding. $979M FCF in 2026 vs comparable levels years ago suggests real cash economics are flatter than GAAP earnings imply. I dissent from any "fairly valued" or "buy" read and align with the synthesis's cautious posture, but I'd go further: my fair value estimate, applying a still-premium 40x FCF to 2028E FCF of ~$1.6B (assuming 25% FCF growth resumes), lands around $130-145 — roughly 40-45% below current. This is overvalued, not "priced for perfection with tailwinds." The tailwinds are real; the price already assumes they've been captured and then some.
GPT Reading
At $256B of equity value on $4.92B of revenue, ARM is being valued like a dominant software platform in the early innings of hypergrowth, not like a semiconductor IP licensor that just posted 23% annual growth and under $1.0B of free cash flow. The raw math is the first thing that matters: 52x sales, about 261x FCF, ~282x earnings, and roughly 28x this year’s revenue increase alone. Even granting the extraordinary quality of the model—97.5% gross margin, no debt, $2.75B cash, sub-billion capex-light economics in a normal year—the operating reality is still much smaller than the stock implies. Revenue went from $2.70B in 2022 to $4.92B in 2026, a very good 16% two-year then four-year cadence overall, but not remotely the kind of compounding that usually supports a quarter-trillion valuation unless margins are exploding and cash conversion is overwhelming. They are not. Operating income improved to $900M from $831M, but operating margin only reached 18.3%, and free cash flow was $979M on $1.52B of operating cash flow because capex was a meaningful $545M. This is a fantastic business; it is not yet a $256B financial machine.
What stands out most is the mismatch between the elegance of the business model and the still-modest monetization of that elegance. ARM’s gross profit was $4.80B on $4.92B of revenue, which tells you the moat is real. But if the moat were already translating into platform economics at the scale implied by the stock, you would expect either much faster top-line acceleration or much higher operating leverage. Instead, annual revenue was basically flat from 2022 to 2023 ($2.70B to $2.68B), then stepped up to $3.23B, $4.01B, and $4.92B. That pattern looks like a good cyclical recovery plus some royalty-rate enrichment, not proof that ARM has become the unavoidable tax collector on all AI compute. Net income rose from $792M to $904M last year, only 14% growth against 23% revenue growth, which is the opposite of the “marginal dollars are almost pure profit” argument bulls lean on. Return on equity of 10.9% and ROIC of 12.7% are healthy, but again, not numbers that scream an enterprise already earning supra-normal economic rents commensurate with a 30.9x book multiple and a 51.1x EV/revenue multiple.
The clean balance sheet actually strengthens the bear case on valuation, because there is no leverage distortion to hide behind. This is not a story where debt paydown or financial engineering can drive equity upside; the stock must be justified by much larger future royalties and licenses. To earn into even a generous 15x sales multiple, ARM would need to roughly triple revenue from $4.9B to about $15B without the market cap moving. That is exactly the sort of heroic endpoint embedded in the current narrative. Can it happen? Yes, but the burden of proof is much higher than the current fundamentals satisfy. When a company with under $1B of FCF is capitalized at $256B, investors are paying now for years of successful expansion into data center, automotive, AI edge, and richer royalty content per chip. Any one of those vectors can work; all of them need to work together, quickly, and at scale. That is a hard bar in semis, where customer concentration, architecture shifts, and negotiation power tend to keep royalty capture below the dream scenario.
The strongest counterargument is obvious and serious: ARM may be one of the very few true architecture-level toll booths in tech, and architecture transitions can reprice monetization suddenly. If revenue can keep growing near the recent 23% pace for several years, while operating margin lifts from 18% toward 30%-40%, today’s multiples could compress far faster than bears expect. The evidence for that case is real: revenue rose $910M last year, gross margin is nearly perfect, net income is now above $900M, and the company has no debt and plenty of cash. If newer designs increase royalty per device and ARM meaningfully penetrates servers and AI-adjacent silicon, the market may be correctly looking through current earnings as an irrelevant waypoint. I weigh that less heavily because the current income statement still shows only modest operating leverage despite this supposed inflection, and because valuation already assumes not just success but sustained success. There is very little room for “good but not legendary.”
What would change my mind is not another year of 20%-25% revenue growth by itself; the stock needs proof of scale economics. I would turn materially less bearish if ARM can deliver two things together: revenue accelerating above 30% from the current $4.92B base and operating margin moving decisively above 25% with FCF comfortably clearing $1.5B. In other words, I need to see that the incremental dollar is becoming much more valuable, not just more frequent. Conversely, if growth slips back toward the mid-teens or if earnings continue to lag revenue, the valuation should rerate sharply. At $239.69, the market is pricing ARM as if the architecture win is settled and the monetization curve is obvious. The architecture may be real; the monetization at this scale is not yet proven.
Grok Reading
Arm’s fiscal 2026 numbers show a real business compounding at a healthy clip: revenue climbed from $3.23B in FY2024 to $4.01B then $4.92B, a clean 23% CAGR with gross margin locked at 97.5%. Operating income recovered from the $111M trough in FY2024 (IPO-year noise) to $900M, net income hit $904M, and free cash flow printed $979M on $1.52B of operating cash after $545M of capex. The balance sheet is fortress-like—$2.75B cash, zero debt, current ratio near 6. None of that is fictional. Yet the $256B market cap attaches a 52× sales multiple, 282× earnings and roughly 260× FCF to a company whose absolute profit pool is still under a billion dollars. At 23% growth those multiples only make sense if the market is correctly pricing a multi-year step-function into a $15B-plus royalty machine; the trailing data simply do not yet demonstrate that inflection.
What the numbers actually reveal is a high-quality IP licensor whose royalty base is still dominated by smartphones and whose datacenter and AI contributions remain early. Operating margin sits at 18%, ROIC at 13%—respectable but nowhere near the 40%+ returns a true platform monopoly would already be throwing off at this scale. Free-cash-flow conversion is solid, yet FCF itself has been essentially flat on a multi-year view once you look past the earnings rebound. The 71% earnings CAGR is almost entirely a base-effect artifact from the depressed FY2024; recent earnings growth slowed to 14% while revenue held at 23%. That divergence, plus the warning flags that every profit-based multiple is hyper-sensitive because earnings are still modest relative to enterprise value, tells me the quantitative models that flash “cheap on growth” are being fooled by the narrative overlay rather than the cash-flow trajectory.
The strongest counter-argument is straightforward and data-backed: CSS licensing is already lifting average royalty rates, hyperscaler Arm deployments (Graviton, Cobalt and the Nvidia/MediaTek v9 deals) are real, and the architectural moat in mobile remains intact. If those forces compound, revenue could accelerate toward the 30%+ zone the valuation implies and operating leverage could push margins well above 30%, turning today’s 52× sales into a more palatable 15–20× on a much larger base inside five years. Bulls will also note that SoftBank’s long-term sponsorship and the scarcity of pure-play AI infrastructure IP names keep the multiple sticky even through ordinary misses. I weigh this less heavily because the installed-base inertia of x86 in the datacenter and the emerging RISC-V design-around risk are already visible in the customer concentration and litigation footnotes; history shows very few semiconductor IP vendors ever achieve the 3× revenue leap the market has prepaid.
I would flip to a neutral or constructive stance only if the next two reported quarters show royalty revenue accelerating above 30% year-on-year with clear evidence that datacenter/AI mix is driving blended rates higher, or if management guides FY2027 revenue above $6.5B while holding gross margin above 96%. A sustained break of the $180 area on light volume without a fundamental miss would also force a reassessment that the multiple compression has largely occurred.
Big-3 Panel — where each AI stands
Advanced Analysis Forensic deep-dive · separate lenses
Arm shows the fingerprints of a genuine platform business: gross margin has climbed from 95.2% to 97.5%, revenue has scaled from $2.70B (2022) to $4.92B (2026) - roughly 82% growth over four years - and the business has flipped from modest FCF ($424M) to nearly $1B. Balance sheet is pristine: $3.60B net cash, Altman Z of 65, OCF/NI of 1.6x, accruals negative at -3.6% of assets, Beneish M at -2.6. Nothing in the mechanical earnings-quality checks flashes red, and the FCF/NI relationship suggests the reported profits are real cash. Operating margin is the wobble: 23.4 -> 25 -> 3.4 -> 20.7 -> 18.3. The 2024 collapse to 3.4% and the failure to reclaim the mid-20s despite revenue growth points to opex scaling faster than the top line - and SBC at 21.4% of revenue (over $1B/yr on a $4.92B base) is the mechanism. GAAP profitability is being partially subsidized by paying employees in stock. Dilution has been contained (~1% share CAGR) so per-share value isn't being obviously eroded, but that requires ongoing buybacks or issuance discipline that isn't visible in the data (buyback/SBC ratio is 0%). Durability read: Arm's royalty-and-license model on ubiquitous instruction-set architecture is structurally moaty - 97%+ gross margins don't exist without pricing power. Cash generation, scale, and margin structure are all consistent with a top-decile IP franchise. What holds it below the top rung is the opex intensity: an operating margin drifting down while revenue scales suggests either heavy R&D reinvestment (fine) or cost discipline slippage (not fine) - the data alone can't tell which.
Verify before trusting this (6)
- Segment split between royalty vs licensing revenue and concentration among top customers (Apple, Qualcomm, hyperscalers)
- Whether the 2024 OpM collapse to 3.4% was IPO-related one-time SBC vesting or structural
- R&D as % of revenue trend and whether opex growth is capacity investment or comp inflation
- Detail on customer concentration and any single-name exposure over 10% of revenue
- Terms and duration of the Arm China arrangement and any related-party revenue
- Whether management has articulated a buyback program to offset ongoing SBC dilution
The e2e synthesis flags that today's price embeds roughly a 3x revenue jump tied to a royalty-rate/mix shift most IP licensors never achieve. That is the tell: the market is not paying for the current royalty stream, it is paying for a step-function reset in take rates as Armv9 and compute-subsystem licensing ramp, plus a data-center wedge. Even accepting the Strong business quality, this is a heroic outcome baked in as base case, not upside.
Verify before trusting this (5)
- Armv9 royalty-rate mix disclosure and per-chip average royalty trend
- Compute-subsystem (CSS) license and royalty ramp cadence in guidance
- Data-center design-win to revenue conversion timing
- SBC trajectory and GAAP operating margin path
- Any customer concentration or renegotiation risk (Qualcomm dispute status, hyperscaler in-house silicon terms)
ARM is currently a narrative-dominant name and the narrative is winning. The archetype is platform-monopoly with strong intensity, and the last 72 hours handed the bulls fresh ammunition: a revenue and profit beat, an after-hours reversal from red to +4%, and a headline $2B AGI CPU order book across FY27/FY28. Citi reiterated Buy at $300, and coverage is uniformly framed around 'AI infrastructure beneficiary' rather than the decelerating smartphone royalty base. That is exactly the tone that keeps a story stock bid.
Verify before trusting this (5)
- Whether the $2B AGI CPU backlog converts to reported revenue on the next print or slips
- RISC-V adoption headlines from hyperscalers or auto OEMs designing around Arm royalties
- Any analyst downgrades or target cuts that break the current uniform Buy tone
- VIX push above 20 or a sharp rise in the 10y, which would punish 3.77-beta story stocks first
- Smartphone royalty commentary at the next update - the crack in the bull story
This lens hasn't been run for this ticker yet.
When we made this prediction on Aug 2, 2026, ARM was $239.69. We expect it to be $205.00 by Feb 2027, and we consider it great value under $165.00. This is an early model (v0.6.0) — the direction is more reliable than the exact price. Made Aug 2, 2026.
Blue is our prediction, starting the day we made it. Grey is a slower route to the same place — the same destination, taking longer. Black is the actual price, so you can see how we are doing.