ARMArm Holdings plc
Industry AI semis Updated 24 Sep 2026
The idea
~100x+ FY28 earnings; 74% of FY26 revenue growth came from related parties (SoftBank and Arm China) while external revenue grew 7%. Upside needs the AGI CPU and compute-subsystem royalties to inflect.
Events to watch
- Royalty growth vs the guided 13%
- related-party share of revenue
- compute-subsystem adoption
- SoftBank's 86.4% as an overhang
How we got the target
Dossier: base 60x FY28 model EPS $2.81. Only the bull case ($288) comes near today's price. Time frame: FY2028 (Mar 2028).
Target: from the research notes. Prices: market data get_latest_stock_price, close of 2026-09-22 (fetched 2026-09-23)
What would prove it wrong
6 warning signsRelated-party revenue stays above ~30% of the total, or external growth falls back below 10%.
Details
The ratio is disclosed on the face of every quarterly income statement, so this is checkable in one line every three months. Q1 FY27: related parties $388m of $1,289m, 30.1%; external +24%.
Royalty growth undershoots the 13% management guided for Q2 FY27.
Details
Royalty is the annuity; licence revenue is lumpy and, in FY26, mostly related-party.
The Arm China relationship changes.
Details
Arm has no equity, no board seat of its own and no independent verification of the numbers Arm China reports to it. Any interruption removes substantially all of the 18% of revenue that comes from the PRC.
GAAP operating income keeps falling while revenue grows.
Details
Q1 FY27 operating income was $91m against $114m a year earlier on 22% more revenue, because stock compensation was $343m in the quarter — 3.8x operating income. Two more quarters of that and the non-GAAP framing is fiction.
Related-party unbilled contract assets keep outrunning related-party revenue (added 2026-09-24 from the earnings-quality check).
Details
$22m → $152m → $646m at FY26 year end, $577m at 2026-06-30, on the balance-sheet parenthetical of every 6-K. Above ~$650m again, or above half of a year's related-party revenue, means the growth is being booked faster than the parent pays for it.
SoftBank sells down, or does not.
Details
An 86.4% holder is both the overhang and the support; a registered secondary would be the first genuine price discovery this stock has had.
How far along
7 of 10 steps done · 3 done in a group review
Next step
decide whether Arm's FY26 growth is a business result or a parent's decision. Related-party revenue — SoftBank Group and Arm China — rose $676m, 82%, and supplied 74% of the $913m of total revenue growth, while revenue from external customers grew 7%. The auditor's sole critical audit matter is revenue recognition on long-term contracts "including related parties". A controlling shareholder at 86.4% can time a licence. Q1 FY27 partly answers it — external revenue reaccelerated to +24% — and the next two quarterly splits settle it. The 2026-09-24 backfill sharpens it: $494m of FY26's $676m related-party revenue growth was still unbilled at year end (related-party contract assets $152m → $646m), and management declined to discuss the step-up on three calls running. Watch the parenthetical on each 6-K balance sheet: billed and collected, the worry fades; still growing, it is the thesis question.
Needs your decision
S6 product test / expert call, S7 conviction. open Open (routine): the number of shares withheld on net settlement (to price the ~$807m of net settlement, a buyback in all but name), and whether the $646m of related-party contract assets is billed and collected — the next two 6-Ks answer the second.
Failed checks
none. S4/S7 are not closed by design (cluster stages, carried forward). S8 closed closed reliable 2026-09-24. ⚠️ Watch, not a formal §3.1 thread: SoftBank / related-party revenue was asked about in three consecutive calls (FY26-Q1, Q2, Q3) and drew one scored dodge and two explicit refusals — the same line that supplied 74% of FY26 growth and whose unbilled balance rose to $646m. ⚠️ Earnings quality 3 flags (S3 EQ:).
Who runs it
The Front OfficeStep by step
- S0How it does as AI gets cheapDone in a group review
carry-forward (cluster: AI-capex complex — cluster pass, 2026-09-20)
cluster verdict is.
⚠️ SPLIT; ARM sits on the "seller of the shortage" side that passes. Owner checks dropped 2026-09-21 (README §Stage 0 override).
- S1Where the idea came fromDone
watchlist (owner, Apple Stocks, 2026-09-20); no Tier 0 promotion.
- S2Quick deal-breaker checkDone
FY26 rev $4,920m +23%, GM 97.5%, op margin 18.3%, OCF $1,524m, no borrowings of any kind on the balance sheet, $3,058m cash at Q1 FY27. Survives easily.
⚠️ but 74% of FY26 revenue growth came from related parties, and R&D is 56% of revenue — see S3. · capital allocation: neutral (FY22–FY26, two pre-IPO years). No buyback (20-F Item 16E "None"), no dividend by policy, no debt, no cash acquisition until DreamBig (~$265m, closed 2026-07-01), no goodwill impairment. SBC ~$1bn a year offset only by net settlement ($807m withholding tax FY24–26); diluted shares 1,025m → 1,068m (+4.2%). Capex 92 → 219 → 545. ROIC (GAAP, ex-cash) 31% FY23 → 17.6% FY26; incremental ~6.8% (NOPAT +195 on IC +2,849, FY23→FY26) — about half the added capital is receivables and unbilled contract assets. Ampere stake: $246m fair-value loss, exited to SoftBank. Bonus paid on revenue + non-GAAP operating income. dossier-2026-09-24.md.
- S3Reading the company's reportsDone
(gate-minimum)
FY26 20-F Item 3.D risk factors + Item 5 MD&A + the auditor's critical audit matter, Q1 FY27 6-K financials and MD&A, EDGAR form census. No proxy exists (FPI)
Full notes
Transcripts: all eight archived calls (FY25-Q2 → FY27-Q1) read Q&A-first on 2026-09-24 (backfill), which closes the 2026-09-21 zero-transcripts gap. · EQ: 3 flags — related-party unbilled contract assets $22m → $152m → $646m (43% of a year's related-party revenue. $494m of FY26's $676m related-party growth unbilled at year end). SBC above FCF every year (FCF after SBC −130 / −721 / −170); capitalised internal-use software +101% vs +23% revenue. Watch: Q1 FY27 non-GAAP add-backs +48% on +22% revenue. Annual bridge otherwise clean. dossier-2026-09-24.md §Earnings quality.
- S4Rivals and suppliersDone in a group review
carry-forward (cluster: AI-capex complex — cluster pass, 2026-09-20)
- S5Who owns the stockDone
SoftBank at 86.4%, 13F register, FINRA short interest, Form 4 cadence, no shelf on file.
⚠️ 3spread returned zero 13F filers for the ADS CUSIP, so the register came from the Equibles fallback, option lines included.
- S6What workers and customers sayDone
carry-forward (cluster: AI-capex complex — cluster pass 2: S6 for 41 names, S0/S4/S7 for three, 2026-09-21)
closed there for all 41 members. Product test / expert call open OPEN (user)
- S7The case for it, and pricesDone in a group review
carry-forward (cluster: AI-capex complex — cluster pass, 2026-09-20)
conviction open OPEN (user) · valuation (bear/base/bull; moved from S8 2026-09-23): ~132x model FY27 and ~98x model FY28 non-GAAP EPS at $275.61 (2026-09-18); 325x reported FY26 GAAP EPS. Guidance evidence is two rows only, so the model leans on trajectory, not on a management guide.
- S8Do the bosses keep their word?Done
**14 rows: 4 beat, 6 met, 3 missed, 1 withdrawn
reliable; current team (Haas CEO, Child CFO) made 14 of 14, nobody departed.** Every written quarterly revenue and EPS guide met or beaten, at or above midpoint The misses are the spoken royalty colour (FY26 "mid-20%" → 20.5% H2 FY26 +10–15% q/q → Q4 −9%) and the DreamBig close (a quarter late); withdrawn: the Armv9 royalty share the CEO said he could "guarantee" would be well above 25%, then stopped being published No full-year guide since May 2025.
⚠️ Haas is also CEO of SoftBank Group International since 2026-04-20 (part-time). Dodges: 2.
- S9What to watch nextDone
trigger written. Tier 0 EDGAR sweep covers the feed (cik 1973239)
Sources and notes
Sources 7 documents
- Full notes, 24 Sep 2026
- Full notes, 21 Sep 2026
- AI-capex complex — cluster pass, 2026-09-20
- AI-capex complex — cluster pass 2: S6 for 41 names, S0/S4/S7 for three, 2026-09-21
- EDGAR filings CIK 1973239
| Short name | Document | Where it came from | Read on |
|---|---|---|---|
| 20F-FY26 | Arm Holdings plc 20-F for FY ended 2026-03-31, filed 2026-05-26 | https://www.sec.gov/Archives/edgar/data/1973239/000197323926000097/arm-20260331.htm | 2026-09-21 |
| 6K-Q1FY27 | Arm Holdings plc 6-K, condensed financials for the quarter ended 2026-06-30, filed 2026-07-29 | https://www.sec.gov/Archives/edgar/data/1973239/000197323926000114/arm-20260630.htm | 2026-09-21 |
| 13F-Q2-26 | Institutional holders, 13F report date 2026-06-30 (option lines included) | Equibles GetTopHolders ARM | 2026-09-21 |
| SI-AUG26 | FINRA short interest, settlement 2026-08-31 | Equibles GetShortInterest ARM | 2026-09-21 |
| 6K-2026-04-21 | Arm Holdings plc 6-K, Rene Haas appointed CEO of SoftBank Group International, filed 2026-04-21 | https://www.sec.gov/Archives/edgar/data/1973239/000197323926000059/arm-20260421.htm | 2026-09-24 |
| SL-Q1FY27 | Arm Holdings plc shareholder letter Q1 FY27, Ex-99.2 to the 6-K filed 2026-07-29 | https://www.sec.gov/Archives/edgar/data/1973239/000197323926000113/exhibit992fye27q130-junx26.htm | 2026-09-24 |
| EDGAR-INDEX | Arm EDGAR filing index, forms filed since 2023-01-01 (CIK 1973239) | https://data.sec.gov/submissions/CIK0001973239.json | 2026-09-21 |
Links to other companies 16 links
| From | Link | To | Date | Where it says so |
|---|---|---|---|---|
| SoftBank Group Corp. | governance structure | ARM | 2026-05-21 | 20F-FY26 — Item 3.D, Risks Relating to Our Status as a Controlled Company and Foreign Private Issuer — controlled-company risk factor |
| SoftBank Group Corp. | governance structure | Shareholder Governance Agreement between Arm and SoftBank Group, entered at the 2023 IPO | 2026-03-31 | 20F-FY26 — Item 7.B, Related Party Transactions — Shareholder Governance Agreement, the three-bullet process |
| SoftBank Group Corp. | customer of | ARM | 2026-03-31 | 20F-FY26 — Item 5.A, results of operations — 'Revenu e from external customers increased $237 million, or 7%,...' |
| Arm Technology (China) Co. Limited ("Arm China") | channel concentration | ARM | 2026-03-31 | 20F-FY26 — Item 3.D, 'We utilize our commercial relationship with Arm China to access the PRC market for IP revenue...' |
| Arm Technology (China) Co. Limited ("Arm China") | key person risk | Intellectual property license agreement (IPLA) between Arm and Arm China | 2026-03-31 | 20F-FY26 — Item 3.D, 'Neither we nor SoftBank Group control the operations of Arm China, which operates independently of us.' |
| Allen Wu | litigation watch | Arm Technology (China) Co. Limited ("Arm China") | 2026-03-31 | 20F-FY26 — Item 3.D, Arm China independence risk factor — the April 2022 litigation paragraph |
| QUALCOMM Incorporated | partner and competitor | ARM | 2026-03-31 | 20F-FY26 — Item 3.D, litigation risk factor |
| SoftBank Group Corp. | channel concentration | ARM | 2026-03-31 | 20F-FY26 — Item 3.D, 'A significant portion of our total revenue comes from a limited number of customers...' |
| Deloitte (Arm's independent registered public accounting firm) | auditor of | ARM | 2026-05-26 | 20F-FY26 — Report of Independent Registered Public Accounting Firm — Critical Audit Matter, 'Revenue Recognition for Material Long-Term Revenue Contracts' |
| Arm Technology (China) Co. Limited ("Arm China") | supplier dependency | ARM | 2026-06-30 | 6K-Q1FY27 — Condensed consolidated balance sheet, line items 'Accounts receivable, net (including receivables from related parties of $ 281 and $ 270...)' and 'Contract assets (including contract assets from related parties of $ 577 and $ 646...)' |
| Invesco Ltd. | holds | ARM | 2026-06-30 | 13F-Q2-26 — row 1, 'Invesco Ltd. | Common | 12,816,588' |
| Susquehanna International Group, LLP | holds | ARM | 2026-06-30 | 13F-Q2-26 — rows 2 and 3, 'SUSQUEHANNA INTERNATIONAL GROUP, LLP | Put | 9,099,400' and '| Call | 8,445,200' |
| The Nasdaq Stock Market LLC | regulator of | ARM | 2026-09-21 | EDGAR-INDEX — filings.recent form census: 20-F ×3, 6-K ×34, SC 13G ×1, no F-3 or F-3ASR |
| Rene Haas | officer of | ARM | 2026-03-31 | 20F-FY26 — Item 6.A, Directors and senior management table, as of March 31, 2026 |
| Jason Child | officer of | ARM | 2026-03-31 | 20F-FY26 — Item 6.A, Directors and senior management table, as of March 31, 2026 |
| Rene Haas | officer of | SoftBank Group Corp. | 2026-04-21 | 6K-2026-04-21 — body of the 6-K, 'Rene Haas Appointment to SoftBank Group International' |
History 3 entries
- 2026-09-21 — first pass, slot
names-ai-capex-2, model-drafted unattended by the cloud research routine. S1, S2, S3 (gate-minimum), S5 and S9 closed; S0/S4/S7 carried forward from the 2026-09-20 cluster file and S6 from the 2026-09-21 one; S8 modelled and left open. Evidence and access limits in dossier-2026-09-21.md. Nothing here is human-verified. - 2026-09-23 — stage redefinition, owner decision (WATCHLIST_PLAN.md §25): S8 is now Management credibility and is not yet run for this name; the old S8 valuation figures moved to the
valuation:line under S7, and any proposed buy price or position size was removed (the pipeline no longer drafts either). No research was redone. S2 capital allocation and S3 earnings quality are also pending, in the same backfill slot. - 2026-09-24 — backfill slot
backfill-2026-09-23-2, model-drafted in a session (subagent), nothing human-verified. Added S2 capital allocation (verdict neutral), S3 earnings quality (3 flags) after reading all eight archived calls Q&A-first, and S8 management credibility (14 rows, reliable; ✅). Dodge log started (2 dodges). One kill criterion added (related-party contract assets); S7 valuation line unchanged (no haircut — S8 is not over-promises). Sources: 20-Fs FY24–FY26, ten shareholder letters, 6-Ks, XBRL, one market data call, no Equibles call. Evidence in dossier-2026-09-24.md.
Full notes, 24 Sep 2026 22 min read
ARM — Arm Holdings plc — dossier 2026-09-24 (backfill)
Backfill slot backfill-2026-09-23-2. Model-drafted, not human-verified. This file holds only the three checks added on 2026-09-23 (WATCHLIST_PLAN.md §25): S2 capital allocation, S3 earnings quality (with the transcript read that feeds it) and S8 management credibility. Nothing else was redone; the other stages stand as written in dossier-2026-09-21.md.
Figures are $ millions unless stated; Arm's fiscal year ends 31 March (FY26 = year to 2026-03-31). Quotes from calls were located by exact substring search in the archived market data transcript bytes (ledger/research/_transcripts/ARM/FY<y>-Q<q>) and the speaker is the Name: label the quote sits under; the call date is the locator (earnings-day precision, owner decision 2026-09-23). Quotes from filings were located in a text extraction of the EDGAR document at the URL below, which collapses whitespace; the words are the filing's. Anything from background knowledge is marked [background]. Roles come from EDGAR (20-F Item 6.A, the 6-K of 2026-03-24, the letters' signature blocks), never from the transcript. Transcript quirk: the FY2026-Q3 archive file renders every spoken "Arm" as "Arm Holdings plc American Depositary Shares"; quotes from that call avoid the word.
Sources
| Key | Document | Locator |
|---|---|---|
20F-FY26 | Form 20-F, FY ended 2026-03-31, filed 2026-05-26 (acc 0001973239-26-000097) | https://www.sec.gov/Archives/edgar/data/1973239/000197323926000097/arm-20260331.htm |
20F-FY25 | Form 20-F, FY ended 2025-03-31, filed 2025-05-28 (acc 0001973239-25-000016) | https://www.sec.gov/Archives/edgar/data/1973239/000197323925000016/arm-20250331.htm |
20F-FY24 | Form 20-F, FY ended 2024-03-31, filed 2024-05-29 (acc 0001973239-24-000012) — cash flows FY22–FY24 | https://www.sec.gov/Archives/edgar/data/1973239/000197323924000012/arm-20240331.htm |
SL-Q4FY24 … SL-Q1FY27 | Shareholder letters, Ex-99.2 to each results 6-K — the "Guidance and Results" table is the written guide; the GAAP to non-GAAP reconciliation is the bridge | e.g. 000197323924000007/exhibit992fye24q431-marx24.htm (2024-05-08), 000197323924000017/…q130-junx24.htm (2024-07-31), 000197323924000036/…q230-sepx24.htm (2024-11-06), 000197323925000006/…q331-decx24.htm (2025-02-05), 000197323925000010/…q431-marx25.htm (2025-05-07), 000197323925000023/…q130-junx25.htm (2025-07-30), 000197323925000042/…q230-sepx25.htm (2025-11-05), 000197323926000005/…q331-decx25.htm (2026-02-04), 000197323926000062/…q431-marx26.htm (2026-05-06), 000197323926000113/…q130-junx26.htm (2026-07-29), all under https://www.sec.gov/Archives/edgar/data/1973239/ |
6K-Q2FY26, 6K-Q3FY26, 6K-Q1FY27 | 6-K condensed financials for the quarters ended 2025-09-30, 2025-12-31, 2026-06-30 | 000197323925000043/arm-20250930.htm, 000197323926000006/arm-20251231.htm, 000197323926000114/arm-20260630.htm |
6K-2026-03-24 | 6-K, "Arm Everywhere" investor conference — CFO reaffirms Q4 FY26 guide (acc 0001973239-26-000052) | 000197323926000052/arm-20260324.htm |
6K-2026-04-21 | 6-K, Rene Haas appointed CEO of SoftBank Group International (acc 0001973239-26-000059) | 000197323926000059/arm-20260421.htm |
PR-2024-11-07 | 6-K Ex-99.1, Chief People Officer succession (acc 0001973239-24-000038) | 000197323924000038/exhibit991fye25q27-novx24p.htm |
XBRL | SEC company facts, CIK 0001973239 (us-gaap tags named inline) | https://data.sec.gov/api/xbrl/companyfacts/CIK0001973239.json |
CALL-FY25Q2 … CALL-FY27Q1 | Earnings calls 2024-11-06, 2025-02-05, 2025-05-07, 2025-07-30, 2025-11-05, 2026-02-04, 2026-05-06, 2026-07-29 | ledger/research/_transcripts/ARM/FY2025-Q2 … FY2027-Q1 (market data) |
MKT-PR | market data get_profitability_ratios NASDAQ:ARM annual — ROIC 8.6% FY26 on its own definition (cross-check only) | connector, 2026-09-24 |
Connectors: no Equibles call was made (EDGAR answered every question the budget was for — the written guide history is in the shareholder letters, the non-GAAP bridge in their reconciliations, buybacks in 20-F Item 16E). One market data call. Alpha Vantage, Bigdata.com and FMP not called.
S2 — Capital allocation
Five years are available because the FY24 20-F carries the FY22 and FY23 cash-flow statements from before the September 2023 IPO; FY22–FY23 are therefore pre-IPO years under SoftBank's sole ownership, and read as such.
Uses of cash (20F-FY24, 20F-FY26 consolidated statements of cash flows; 6K-Q1FY27)
| FY22 | FY23 | FY24 | FY25 | FY26 | Q1 FY27 | |
|---|---|---|---|---|---|---|
| Operating cash flow | 458 | 739 | 1,090 | 397 | 1,524 | 902 |
| Capex — property and equipment | (34) | (64) | (92) | (219) | (545) | (197) |
| Intangibles bought + intangible obligations paid | (78) | (69) | (91) | (79) | (97) | (40) |
| Acquisitions (cash, net) | — | — | — | — | — | — (DreamBig ~$265m closed 2026-07-01, i.e. Q2 FY27) |
| Equity investments bought | (8) | (15) | (32) | (57) | (11) | (74) |
| Divestiture / investment exits | — | — | — | — | +127 business, +205 equity investments | +1 |
| Buybacks | none | none | none | none | none (20-F Item 16E: "None.") | none |
| Dividends | (43)* | none | none | none | none | none |
| Debt raised / (repaid) | +50 short-term | — | — | — | — | — |
| Withholding tax paid on vested shares (net settlement) | — | — | (158) | (120) | (529) | (278) |
Stock-based compensation (ShareBasedCompensation) | 26 | 79 | 1,037 | 820 | 1,052 | 343 |
Diluted weighted shares, m (WeightedAverageNumberOfDilutedSharesOutstanding) | 1,025 | 1,028 | 1,044 | 1,063 | 1,068 | — |
Cash + short-term investments, year end (XBRL) | — | 2,215 | 2,923 | 2,825 | 3,601 | 3,888 |
\* FY22's $43m is the cash transferred with the distribution of Treasure Data and IoTP to SoftBank before the IPO, not a dividend to public holders. No borrowings of any kind since FY22.
Reading it:
- Arm returns nothing to shareholders in cash, by policy. "We intend to retain any earnings for use in our business and do not currently intend to pay dividends on our ordinary shares or ADSs." (
20F-FY26, Dividend Policy.) Item 16E, issuer purchases: none. There is therefore no buyback average price to set against today's — the check does not apply. - The only share-count defence is net settlement. Arm paid $807m over FY24–FY26 (and $278m more in Q1 FY27) of withholding tax on vesting awards instead of issuing the shares to cover it. That is economically a repurchase at the vesting price, but the number of shares withheld is not tagged in XBRL and was not found in the 20-F, so no average price can be computed — said rather than estimated. Diluted shares still rose 1,025m → 1,068m (+4.2%) over four years; with SBC at ~$1bn a year and no buyback, nothing offsets it beyond net settlement.
- Capex has turned on. Property and equipment spend went 92 → 219 → 545 (FY24→FY26) and 197 in Q1 FY27 alone, the year Arm moved into selling its own silicon (the AGI CPU). Net PP&E 185 (FY23) → 772 (FY26) → 923 (2026-06-30) (
PropertyPlantAndEquipmentNet). - Acquisitions: none paid in cash in the five years. DreamBig (~$265m, all cash) closed on 2026-07-01, a quarter after the first expected close date (S8 row 13).
- What the investments were worth. No goodwill impairment in any year (
GoodwillImpairmentLoss0 for FY23–FY26; goodwill flat at ~$1,620m since FY23). The one visible investment outcome is Ampere: a loss "primarily due to a $246.4 million fair value loss with respect to our investment in Ampere Computing Holdings LLC" in FY25 (20F-FY25, Item 5), followed by an exit when SoftBank bought Ampere — "the Company, as a minority shareholder of Ampere, received $143.4 million in gross proceeds for its equity interest in Ampere and $39.3 million from the settlement of the convertible promissory note with Ampere" (20F-FY26, Note on equity investments). The exit price was set in a transaction by the controlling shareholder; it is a related-party capital outcome and it was a loss.
Returns
Invested capital = shareholders' equity − cash − short-term investments (Arm has no debt); NOPAT = operating income × (1 − 15%), 15% being the non-GAAP effective tax rate the letters report. Invested capital so defined still contains ~$1.6bn of goodwill, the equity-investment book and — the point — receivables and contract assets.
| FY23 | FY24 | FY25 | FY26 | |
|---|---|---|---|---|
Equity (StockholdersEquity) | 4,051 | 5,295 | 6,839 | 8,286 |
| Invested capital | 1,836 | 2,372 | 4,014 | 4,685 |
| GAAP operating income | 671 | 111 | 831 | 900 |
| NOPAT, GAAP basis | 570 | 94 | 706 | 765 |
| ROIC, GAAP basis (on average IC; FY23 on year-end) | 31% | 4.5% | 22.1% | 17.6% |
| Non-GAAP operating income (SBC excluded) | — | 1,408 | 1,871 | 2,115 |
| ROIC, non-GAAP basis | — | 56.9% | 49.8% | 41.3% |
- Incremental return, GAAP basis, FY23 → FY26: NOPAT +195 (570 → 765) on invested capital +2,849 (1,836 → 4,685) = ~6.8%. FY25 → FY26 alone: +59 on +671 = ~8.8%.
- Incremental return on the flattering basis, non-GAAP FY24 → FY26: +601 on +2,313 = ~26%. The two readings differ by exactly the ~$1bn a year of stock compensation.
- market data reports 8.6% ROIC for FY26 on its own (cash-inclusive) definition (
MKT-PR) — the same direction. - Where the extra capital went: receivables plus contract assets 1,269 (FY23) → 2,597 (FY26), +1,328, and net PP&E +587. About half of the incremental capital Arm employs is money owed to it and not yet billed, and the fastest-growing part of that is owed by related parties (S3 check 4).
Bonus paid on: the CEO's annual bonus is scored on revenue and non-GAAP operating income — "The award will be subject to the Company's attainment of goals relating to Revenue (50% weighting) and the Non-GAAP Operating Income (50% weighting)." (20F-FY26, Item 6.B) — i.e. on an operating profit measure that excludes the ~21% of revenue paid in stock. FY26 revenue came in at $4,920m against a $4,996m internal bonus target (92% outcome), and the Remuneration Committee "exercised discretion" on one strategic measure. No buyback exists, so the EPS-plus-buyback flag does not arise. A market-capitalisation VCP award ($1.0tn / $1.5tn / $2.0tn by 2029–2031) was added for FY27.
Capital-allocation verdict: neutral. Nothing was destroyed — no debt, no goodwill impairment, no buybacks at inflated prices, no large acquisitions — and the balance sheet compounds cash. But the cash return to shareholders is nil by policy, dilution is only partly offset by net settlement, the one visible investment (Ampere) ended in a loss realised through the parent, and the incremental GAAP return on the capital added since FY23 is single-digit (~7%) because that capital is mostly unbilled receivables and new PP&E. Not a destroyer; not a compounder on the evidence yet.
S3 — Earnings quality
Read: the GAAP to non-GAAP reconciliations in SL-Q4FY24, SL-Q4FY26 (which carries FY25 and FY26) and SL-Q1FY27; the FY26 and FY24 20-F cash-flow statements, balance-sheet parentheticals and the internal-use software note; XBRL company facts.
- The non-GAAP bridge (operating income, add-backs by year):
| FY24 | FY25 | FY26 | Q1 FY26 | Q1 FY27 | |
|---|---|---|---|---|---|
| GAAP operating income | 111 | 831 | 900 | 114 | 91 |
| SBC (equity settled) | 1,037 | 820 | 1,052 | 241* | 343 |
| Employer taxes on SBC, net of R&D incentive | 189 | 213 | 150 | — | 90 |
| Acquisition-related | 4 | — | 2 | — | — |
| Restructuring / disposal / IPO / other operating | 67 (IPO readiness 42, one-time benefit 19, other 6) | 7 | 11 | — | 7 |
| Total add-backs | 1,297 | 1,040 | 1,215 | 298 | 440 |
| as % of revenue | 40.1% | 26.0% | 24.7% | 28.3% | 34.1% |
| Non-GAAP operating income | 1,408 | 1,871 | 2,115 | 412 | 531 |
\* SBC from the Q1 FY26 cash-flow statement; the Q1 FY26 total is the letter's non-GAAP minus GAAP operating income. Sources: SL-Q4FY24 FY24 reconciliation; SL-Q4FY26 FY26 and FY25 reconciliations; SL-Q1FY27 Q1 FY27 reconciliation; SL-Q1FY26 summary bullets.
On the annual trend the bridge is clean: FY24 is inflated by IPO-year awards, and add-backs grew 17% FY25 → FY26 against 23% revenue growth. No "one-off" charge recurs at size — the small disposal/restructuring/other lines are $7–11m a year after the IPO year. The employer-tax add-back is recurring by nature and labelled as such. But the latest quarter breaks the trend: Q1 FY27 add-backs were +48% year on year on +22% revenue, SBC alone +42%, and GAAP operating income fell to $91m from $114m while non-GAAP rose 29%. One quarter; a watch, not yet a flag.
- Stock-based compensation vs revenue and FCF. SBC was 32.1% / 20.5% / 21.4% of revenue in FY24 / FY25 / FY26 and 26.6% in Q1 FY27. Against the company's own non-GAAP free cash flow (OCF − PP&E − intangibles − intangible obligations: 907 / 99 / 882,
SL-Q4FY26FCF reconciliation and the same arithmetic on20F-FY24), SBC exceeded FCF in all three years — 114%, 828%, 119% — so FCF after SBC was −130 / −721 / −170. ⚠️ Flag (the test is FCF after SBC below half of FCF; here it is below zero). - Cash conversion and accruals. OCF / net income 3.56× (FY24), 0.50× (FY25), 1.69× (FY26); 1.50× cumulatively. Accruals ((NI − OCF) / average total assets): −10.6%, +4.7%, −6.3%. FY25's weak conversion is explained in the letter — "Full-year results were negatively impacted by unfavorable working capital movements in receivables and the timing of payroll tax liabilities related to the IPO share vesting event" (
SL-Q4FY25). Not a flag on its own, with one caveat: OCF includes the ~$1bn SBC add-back, and FY26 OCF excluding SBC was $472m against $904m of net income (0.52×). Q1 FY27 OCF of $902m was helped by receivables collections and a +$285m move in other liabilities including payables to related parties (6K-Q1FY27cash-flow statement) — timing, by the company's own account: "benefiting from favorable timing of receivables collections and tax payments" (SL-Q1FY27). - Working capital — receivables and unbilled contract assets (the licensor's inventory).
| Year end | Revenue | AR, net | Contract assets (curr. + non-curr.) | AR days | CA days | Contract assets from related parties | Related-party revenue |
|---|---|---|---|---|---|---|---|
| FY23 | 2,679 | 999 | 270 | 136 | 37 | — | — |
| FY24 | 3,233 | 781 | 576 | 88 | 65 | 22 | 724 |
| FY25 | 4,007 | 1,107 | 988 | 101 | 90 | 152 | 823 |
| FY26 | 4,920 | 1,300 | 1,297 | 96 | 97 | 646 | 1,499 |
| Q1 FY27 (×4) | 5,156 | 1,119 | 1,247 | 79 | 88 | 577 | 1,552 |
Sources: AccountsReceivableNetCurrent, ContractWithCustomerAssetNetCurrent / …Noncurrent (XBRL); related-party parentheticals "contract assets from related parties of $646 and $152 as of March 31, 2026 and 2025, respectively" (20F-FY26 balance sheet) and "contract assets from related parties of $152 and $22 as of March 31, 2025 and 2024, respectively" (20F-FY25 balance sheet); Q1 FY27 from 6K-Q1FY27. ⚠️ Flag. Contract-asset days rose two periods running (65 → 90 → 97) while revenue grew, and the whole of the rise is related-party: unbilled amounts owed by SoftBank Group and Arm China went $22m → $152m → $646m — from 3% to 18% to 43% of a year's related-party revenue — while contract assets owed by external customers fell from ~$836m to ~$651m in FY26. Of the $676m by which related-party revenue grew in FY26, $494m was added to unbilled related-party contract assets in the same year. That is the S3 finding of 2026-09-21 (74% of FY26 growth came from related parties) seen from the balance sheet: most of the related-party revenue booked in FY26 had not been billed at the year end. Q1 FY27 reversed a little ($646m → $577m).
- Costs moved onto the balance sheet. "The Company capitalized $173.6 million, $86.2 million and $68.8 million, and amortized $92.3 million, $79.8 million and $73.7 million of internal-use software in the fiscal years ended March 31, 2026, 2025 and 2024, respectively." (
20F-FY26, intangible assets note.) ⚠️ Flag: capitalisation doubled (+101%) in FY26 against +23% revenue, and now runs $81m a year ahead of amortisation — ~9% of FY26 GAAP operating income. Product R&D is still expensed ("has not historically capitalized software development costs for software to be sold"). - Sector equivalents: n/a (not an insurer, bank or REIT).
EQ: 3 flags — related-party unbilled contract assets $22m → $152m → $646m (43% of a year's related-party revenue; $494m of FY26's $676m related-party growth unbilled at year end); SBC above FCF every year (FCF after SBC −130 / −721 / −170); capitalised internal-use software +101% vs +23% revenue. Watch: Q1 FY27 non-GAAP add-backs +48% on +22% revenue. Annual non-GAAP bridge otherwise clean.
Transcripts read (§3.1)
All eight archived calls, FY2025-Q2 (2024-11-06) to FY2027-Q1 (2026-07-29), read in §3.1 order: safe harbour skipped, CFO section skimmed for guidance, Q&A read in full. Speakers on the company side are always Rene Haas (CEO) and Jason Child (CFO) — in the FY2026-Q1 file labelled "Rene Anthony Andrada Haas" and "Jason E. Child". Roles: 20F-FY25 and 20F-FY26 Item 6.A ("Rene Haas | 63 | Chief Executive Officer and Director", "Jason Child | 57 | Executive Vice President and Chief Financial Officer"), and 6K-2026-03-24 ("the Company’s Chief Financial Officer, Jason Child, reaffirmed the Company’s fourth quarter 2026 guidance"). IR host varies (Jeff Kvaal, Ian Thornton, Jessica Vall).
Guidance changes (step 2)
- Full-year guidance stopped in May 2025. FY25 was guided in writing from May 2024; for FY26 the CFO said "As a result, we do not consider it prudent to issue full year guidance." (
CALL-FY25Q4, Jason Child, citing tariff visibility). It has not come back: FY27 has quarterly guides and spoken colour only. - The royalty-growth colour has been trimmed every year.
- FY25: guided "high teens" at the Q2 call; the analyst asking said it had been lowered — "I think you have changed this year's contribution to, I think, high teens growth from low 20s and then mid-20s" (
CALL-FY25Q2, Vivek Arya). The earlier guides are in calls the archive does not hold, so that path is the analyst's, not verified here. - FY26: "we would expect royalties to grow probably somewhere in the mid-20% growth rate" (
CALL-FY25Q3, Jason Child) → no FY26 guide (May 2025) → "full year royalties, I would expect to be in that north of you know, 20% range" (CALL-FY26Q3) → 20.5% actual. - FY27: "So I expect that the royalty growth is going to be in the roughly 20-ish percent for the year." (
CALL-FY26Q4, 2026-05-06) → 84 days later "Right now, if I had to guess, that's probably somewhere closer to the high teens right now." (CALL-FY27Q1, Jason Child), and Q2 FY27 royalty guided at ~13%. - SoftBank licence and design services stepped from ~$126m to ~$178m (Q2 FY26) to $200m a quarter and is guided flat: "We expect the quarterly run rate for the rest of the year to be around $200 million." (
CALL-FY27Q1, Jason Child.) - Arm AGI CPU (own silicon): $1bn of revenue across FY27–FY28, first production revenue (~$90m) in Q4 FY27 — "However, we are maintaining our outlook of $1 billion while we pursue supply chain capacity, and we still expect the first revenues from production chip sales to land in the fourth quarter of this fiscal year." (
CALL-FY26Q4, Jason Child). Demand stated as >$2bn; an update promised at the Q3 FY27 results. Gross margin guided high-30s/low-40s first generation. Not yet checkable. - Opex: quarterly non-GAAP opex has come in at or just under the guide every quarter (e.g. Q1 FY27 $733m vs ~$760m); FY27 guided to grow mid-single digits a quarter.
What analysts asked about (step 3), by topic
- Royalty growth and smartphones — every call (Arya, Quatrochi, Leopold, Shi, Naji, Pajjuri): memory-price-driven handset weakness, v9/CSS rate uplift, the path of royalty growth.
- SoftBank and related-party revenue — FY26-Q1 (Rakesh: the SoftBank licence expansion and Stargate), FY26-Q2 (Schneider: the related-party step-up; Arya: how long the SoftBank run-rate lasts), FY26-Q3 (Rakesh: a SoftBank custom ASIC; Leopold: would SoftBank sell Arm shares).
- Own silicon / chiplets / AGI CPU — from FY25-Q3 (Gardiner, Simpson) through every FY26–FY27 call: timing, supply, margin, channel conflict with licensees (Naji, FY26-Q4).
- Data-centre share and CPU TAM — Quatrochi, Lipacis, Arya, O'Malley, Sankar, Simpson.
- Qualcomm litigation — FY25-Q2 (Gardiner), FY25-Q3 (Shi): answered; forecasts assume Arm loses.
- Licensing / ACV / RPO mechanics — Sur, Seymore, Gardiner, Schulze-Melander.
Dodges (step 4)
- FY2025-Q2 (2024-11-06) · the cut to the FY25 royalty-growth guide · Vivek Arya (BofA). Asked "And then the overall royalty growth, what is the need to take it down?" Rene Haas answered with the v9 generation-on-generation pricing mechanism and CSS rates and closed "I think maybe we answered the extent of that question." The reason for lowering the guide was not given. A dodge on a topic management itself guided. Locator:
CALL-FY25Q2. - FY2026-Q2 (2025-11-05) · the step-up in related-party revenue · Jim Schneider (Goldman Sachs). Asked "I noticed in your disclosures that you saw a material step-up in related party revenue." and "Can you maybe give us any kind of color you can on the nature of that relationship". Rene Haas answered with the size of the Stargate opportunity for the SoftBank family ("So without getting into too many of the specifics"; "it provides huge opportunity for Arm to provide solutions into that space") — nothing on the revenue, its terms, its billing or its duration. Locator:
CALL-FY26Q2.
Logged, not scored (explicit refusals, not evasions), all on the same SoftBank / related-party topic: FY26-Q1, Rene Haas to Vijay Rakesh on the SoftBank licence — "there's some details you asked for that I can't really make a comment on"; FY26-Q2, Jason Child to Vivek Arya on how long the SoftBank run-rate persists — "we're not at liberty to say yet"; FY26-Q3, Jason Child to Vijay Rakesh on a SoftBank custom chip — "Nothing we can say specific about any products that you're you're asking about." Also refused, other topics: chiplet/ASIC plans (FY25-Q3 Gardiner, FY26-Q1 O'Connor, FY26-Q2 Seymore), FY28 outlook ("In terms of '28, we haven't thrown anything out there yet.", FY26-Q3).
Thread: the related-party/SoftBank line was asked about in three consecutive quarters (FY26-Q1, Q2, Q3) and got one scored dodge and two explicit refusals — the same line that supplied 74% of FY26 growth and $494m of new unbilled contract assets (S3 check 4). Under §3.1 it is one dodge, not a two-quarter thread; it is carried to the scorecard's ⚠️ line anyway because of what it sits next to.
S8 — Management credibility
Arm guides quarterly revenue, non-GAAP opex and non-GAAP EPS in writing (the letters' "Guidance and Results" table, which also prints the result against the guide), gave a written full-year guide for FY25 only, and adds spoken colour on royalty and licence growth. Fourteen rows over FY25-Q2 → FY27-Q1, each scored against the first guide. Written outlooks are attributed to the letters' signatories — Rene Haas, Chief Executive Officer | Jason Child, Chief Financial Officer (signature block of every letter read) — spoken ones to the speaker label.
| # | Promise (metric, period, figure) | Made — where, who | Path | Result — where | Score |
|---|---|---|---|---|---|
| 1 | Q2 FY25 revenue $780–830m; non-GAAP EPS $0.23–0.27 | SL-Q1FY25 2024-07-31, written (Haas/Child) | — | $844m; $0.30 — SL-Q2FY25 | beat |
| 2 | Q3 FY25 revenue $920–970m; EPS $0.32–0.36 | SL-Q2FY25 / CALL-FY25Q2 2024-11-06, Jason Child (CFO) | — | $983m; $0.39 — SL-Q3FY25 | beat |
| 3 | Q4 FY25 revenue $1,175–1,275m; EPS $0.48–0.56 | SL-Q3FY25 / CALL-FY25Q3 2025-02-05, Jason Child | — | $1,241m; $0.55 — SL-Q4FY25 | met |
| 4 | FY25 revenue $3.8–4.1bn; EPS $1.45–1.65; opex ~$2.05bn | SL-Q4FY24 2024-05-08, written (first guide) | reiterated ×2, narrowed to $3.94–4.04bn / $1.56–1.64 (2025-02-05) | $4.01bn; $1.63; $2.05bn — SL-Q4FY25 | met |
| 5 | Q1 FY26 revenue $1.0–1.1bn; EPS $0.30–0.38; royalty growth 25–30% | SL-Q4FY25 / CALL-FY25Q4 2025-05-07, Jason Child — "We expect to start the year with strong 25% to 30% royalty growth." | — | $1,053m; $0.35; royalty $585m, +25.3% (CFO: "I would say we were in the lower end of that range.") — SL-Q1FY26, CALL-FY26Q1 | met |
| 6 | Royalty +10–15% sequentially in each of Q3 and Q4 FY26 | CALL-FY25Q4 2025-05-07, Jason Child — "I expect to see somewhere in the 10-ish -- 10% to 15% kind of sequential growth in each of the last two quarters in the back half of the year" | Q4 re-guided to "low teens" year on year (2026-02-04) | Q3 $737m, +18.9% q/q; Q4 $671m, −9.0% q/q — SL-Q3FY26, SL-Q4FY26 | missed (Q4) |
| 7 | FY26 royalty growth "mid-20%" | CALL-FY25Q3 2025-02-05, Jason Child | no FY26 guide (May 2025) → full-year royalties guided to north of 20% (2026-02-04) | $2,613m vs $2,168m, +20.5% — quarterly royalty lines in SL-Q1FY26…SL-Q4FY26, SL-Q4FY25 | missed (by ~4–5 points) |
| 8 | Q2 FY26 revenue $1.01–1.11bn; EPS $0.29–0.37; royalty and licence ~flat q/q | SL-Q1FY26 / CALL-FY26Q1 2025-07-30, Jason E. Child | — | $1,135m; $0.39; royalty +6% q/q, licence +10% q/q — SL-Q2FY26 | beat |
| 9 | Q3 FY26 revenue $1.225bn ± $50m; EPS $0.41 ± $0.04; royalty "just over 20%" | SL-Q2FY26 / CALL-FY26Q2 2025-11-05, Jason Child | — | $1.242bn; $0.43; royalty +27% — SL-Q3FY26 | met (royalty beat) |
| 10 | Q4 FY26 revenue $1.47bn ± $50m; EPS $0.58 ± $0.04; royalty low teens, licence high teens | SL-Q3FY26 / CALL-FY26Q3 2026-02-04, Jason Child | reaffirmed by the CFO 2026-03-24 (6K-2026-03-24) | $1.490bn; $0.60; royalty +11%, licence +29% — SL-Q4FY26 | met |
| 11 | Q1 FY27 revenue $1.26bn ± $50m; EPS $0.40 ± $0.04; royalty and licence ~+20% | SL-Q4FY26 / CALL-FY26Q4 2026-05-06, Jason Child | — | $1.289bn; $0.45 (top of range $0.44); royalty +22%, licence +23% — SL-Q1FY27 | beat |
| 12 | Armv9 share of royalties "considerably higher" than 25% a year on | CALL-FY25Q2 2024-11-06, Rene Haas — "I can guarantee you, it will not be 25% a year from now. It will be considerably higher than that." | "north of 30%" (2025-05-07), then the metric was dropped from quarterly reporting — "one of the things we mentioned was we're not going to give the v9 adoption rate every quarter going forward." (CALL-FY25Q4, Jason Child) — and moved to annual (CALL-FY26Q1) | no year-end figure published: the FY26 letter and 20-F carry no Armv9 share | withdrawn |
| 13 | DreamBig acquisition (~$265m cash) to close by end of Q4 FY26 | 6K-Q2FY26 2025-11-05, company (written): "The transaction is expected to close by the end of the fourth quarter of the fiscal year ending March 31, 2026" | → end of Q1 FY27 (6K-Q3FY26) → end of Q2 FY27 (20F-FY26) | "On July 1, 2026, the Company completed the acquisition of DreamBig" — 6K-Q1FY27 | missed (one quarter late; each slip disclosed, regulatory approvals the stated condition) |
| 14 | SoftBank licence and design services ~$200m a quarter | CALL-FY26Q3 2026-02-04, Jason Child — "I would expect that $200 going forward is the right run rate going forward." | restated "around $200 million" (2026-07-29) | Q4 FY26 $200m, Q1 FY27 $193m — CALL-FY26Q4, CALL-FY27Q1 | met |
Not scored (period open or first guide not in the archive): FY27 royalty growth (~20% → "closer to the high teens", 2026-05-06 → 2026-07-29 — a revision in progress, and the royalty colour coming down again); FY25 royalty "high teens" (first guide not in the archive; +20.3% actual); AGI CPU ~$90m in Q4 FY27 and $1bn across FY27–FY28; the $15bn AGI CPU / $25bn total revenue / $9 EPS targets for FY31; the long-term revenue-growth target of 20% ("This midpoint represents about a 24% year-on-year growth, which is ahead of our long-term target of 20%.", CALL-FY25Q3), exceeded in FY25 (+24%) and FY26 (+23%).
Tally: 14 rows — 4 beat, 6 met, 3 missed, 1 withdrawn. Every written quarterly revenue guide in the window landed at or above its midpoint and none below the range; non-GAAP EPS likewise. The misses are all in the spoken medium-term royalty colour (rows 6, 7) and a deal-close date (row 13); the withdrawal is a metric the CEO said he could guarantee and the company then stopped publishing (row 12).
Are the promisers still there? Yes — every row was made by Rene Haas (CEO) or Jason Child (CFO), or by the company in writing under their joint signature, and both still hold those roles: 20F-FY26 Item 6.A (as of 2026-03-31), 6K-2026-03-24 (Child as CFO), and the signature block of SL-Q1FY27 (2026-07-29). EDGAR full-text search of Arm's 6-Ks since 2024 (efts.sec.gov; queries on resignation, CFO appointment, and the CEO stepping down) returned no officer departure; the one executive change in the window is the Chief People Officer succession of 2024-11-07 (PR-2024-11-07), not a promiser. Nothing is departed. One flag of a different kind: since 2026-04-20 Rene Haas is also CEO of SoftBank Group International, the controlling shareholder's operating platform — "Mr. Haas’s role at SBGI is expected to be limited and part-time and is not intended to detract from his responsibilities at the Company." (6K-2026-04-21). The man who makes Arm's promises now also runs part of its largest related-party counterparty's portfolio.
Dodge log rolled up: 2 scored dodges in eight calls — the FY25 royalty-guide cut (FY25-Q2, a topic management itself guided — the strongest signal type) and the related-party revenue step-up (FY26-Q2). Plus three explicit refusals on the SoftBank/related-party topic across FY26-Q1 to Q3. No topic was dodged (as opposed to refused) in two quarters.
Verdict: ✅ reliable. The written quarterly guide is met or beaten every time, usually above the midpoint — on that alone Arm would read conservative. It does not get the stronger word because the spoken royalty-growth colour is trimmed year after year (FY26 "mid-20%" → 20.5%; FY27 "roughly 20-ish" → "high teens" within one quarter), the one dated metric the CEO said he could guarantee was withdrawn from disclosure, the first scored dodge sits on the royalty guide itself, and full-year guidance stopped in May 2025. Current team made all 14 rows. S7's base case can take the quarterly written guide as given; it should not take the medium-term royalty colour at face value — the valuation table in the scorecard already leans on trajectory rather than a management guide, so no haircut is applied here.
Full notes, 21 Sep 2026 11 min read
ARM — dossier, 2026-09-21
Slot names-ai-capex-2. Model-drafted unattended by the cloud research routine; nothing here is human-verified. Sources and locators are in relationships.yaml; figures below carry theirs inline. Anything from background knowledge is marked [background].
Price reference: $275.61 close, 2026-09-18 (market data get_latest_stock_price NASDAQ:ARM; volume 6,946,800 shares, about $1.9bn traded). Shares outstanding 1,064,055,252 at 2026-03-31 (20-F cover, dei EntityCommonStockSharesOutstanding) → roughly $293bn market capitalisation, of which ~13.6% is public float — about 145m shares, ~$40bn.
S1 — Source
watchlist (owner, Apple Stocks, 2026-09-20), cluster ai_semis, coverage sec_fpi, cik 1973239. No Tier 0 promotion. As a foreign private issuer Arm files a 20-F and 6-Ks — no 10-Q, no proxy.
S2 — Kill test
XBRL company facts from the SEC filings (data.sec.gov/api/xbrl/companyfacts, CIK 0001973239); fiscal years end 31 March.
| FY (ends) | Revenue $m | Gross profit $m | GM % | R&D $m | Op income $m | Op margin % | Net income $m | OCF $m | Diluted WA shares m |
|---|---|---|---|---|---|---|---|---|---|
| FY22 (2022-03-31) | 2,703 | 2,572 | 95.2 | 995 | 633 | 23.4 | 549 | 458 | 1,025 |
| FY23 (2023-03-31) | 2,679 | 2,573 | 96.0 | 1,133 | 671 | 25.0 | 524 | 739 | 1,028 |
| FY24 (2024-03-31) | 3,233 | 3,079 | 95.2 | 1,979 | 111 | 3.4 | 306 | 1,090 | 1,044 |
| FY25 (2025-03-31) | 4,007 | 3,886 | 97.0 | 2,071 | 831 | 20.7 | 792 | 397 | 1,063 |
| FY26 (2026-03-31) | 4,920 | 4,799 | 97.5 | 2,776 | 900 | 18.3 | 904 | 1,524 | 1,068 |
Quarterly (6-K XBRL):
| Quarter ended | Revenue $m | Op income $m | Net income $m | Stock comp $m |
|---|---|---|---|---|
| 2025-06-30 (Q1 FY26) | 1,053 | 114 | 130 | 241 |
| 2025-09-30 (Q2 FY26) | 1,135 | 163 | 238 | — |
| 2025-12-31 (Q3 FY26) | 1,242 | 185 | 223 | — |
| 2026-06-30 (Q1 FY27) | 1,289 | 91 | 270 | 343 |
Survives, comfortably — the balance sheet is not the question here. Arm carries no borrowings at all: the XBRL facts contain no debt tag of any kind, only lease liabilities. Cash was $3,058m at 2026-06-30 and equity $8,630m, both rising every quarter in the series. Operating cash flow was $1,524m in FY26.
Three things the kill test does surface:
- Dilution is mild by semiconductor standards — 1,025m diluted shares (FY22) → 1,068m (FY26), +4.2% in four years. That is the one capital-structure line that flatters Arm against its peers.
- R&D is 56% of revenue ($2,776m on $4,920m) and rising; it was 37% in FY22. Operating margin has been sacrificed to it, deliberately.
- GAAP operating income went backwards in Q1 FY27 — $91m against $114m a year earlier on 22% more revenue — because stock compensation was $343m in the quarter, 3.8x operating income and up 42% year on year. Net income of $270m in the same quarter is below the operating line only because interest and tax ran the other way; it is not operating earnings.
- Analyst coverage count: not obtained — see "Evidence that was thin".
S3 — Filings, gate-minimum
Read: FY26 20-F (Item 3.D summary of risk factors and the individual factors on concentration, the PRC, Arm China and controlled-company status; Item 5 results of operations; the auditor's report and its critical audit matter; Item 7.B related-party transactions), the Q1 FY27 6-K (condensed financials, balance sheet related-party lines, Item 5-equivalent MD&A), and the EDGAR form census. No transcript was read. No proxy exists — FPIs do not file one, and the skill's gate-minimum table says to skip it.
The three biggest risks management itself discloses
- Customer concentration that includes the controlling shareholder. "our top five customers (including Arm China and SoftBank Group) collectively accounted for approximately 57%, 56% and 54% of our total revenue for the fiscal years ended March 31, 2026, 2025 and 2024, respectively" (20-F, Item 3.D). Two of the five top customers are the parent and an entity the parent owns. Separately, Qualcomm — a litigation adversary — "accounted for 9% of our total revenue for the fiscal year ended March 31, 2026".
- The PRC reaches Arm only through an entity Arm does not control. "Substantially all of our PRC-related revenue is earned through the intellectual property license agreement ("IPLA") with Arm China" (20-F, Item 3.D); PRC revenue was 18% of FY26 total, 19% in FY25, 22% in FY24. Arm transferred its entire equity interest in Arm China to a SoftBank subsidiary on 2022-03-28, has no board seat of its own, and "Arm China's payments due to us are determined based on the financial information that Arm China provides to us". Its former CEO has been litigating against its governance in PRC courts since April 2022.
- Customers can design around the product they license. "Customers may decide to license our architecture and develop their own processors based on our ISA, rather than utilize our predeveloped products through an implementation license" (20-F, Item 3.D) — and Arm is now competing with those same customers: "some of our customers may face direct competition from us in silicon production products, such as with the Arm AGI CPU."
The finding this run exists to record
Arm reports revenue in two lines on the face of the income statement — external customers, and related parties (SoftBank Group and Arm China):
| FY | External $m | % | Related parties $m | % | Total $m |
|---|---|---|---|---|---|
| FY24 | 2,509 | 78 | 724 | 22 | 3,233 |
| FY25 | 3,184 | 79 | 823 | 21 | 4,007 |
| FY26 | 3,421 | 70 | 1,499 | 30 | 4,920 |
Management's own MD&A states the split of the growth: "Revenu e from external customers increased $237 million, or 7%... Revenue from related parties increased $676 million, or 82%... primarily driven by a $591 million, or 141%, increase in license and other revenue". Of the $913m of FY26 revenue growth, $676m — 74% — came from the controlling shareholder's own orbit, and most of that was licence revenue rather than royalties. Reported growth was +23%; growth from unrelated customers was +7%.
Two things sharpen it and one softens it.
- Sharpening: the auditor's sole critical audit matter is "Revenue Recognition for Material Long-Term Revenue Contracts", whose description opens "The Company routinely enters into material long-term revenue contracts with customers, including related parties" and whose audit response includes "an evaluation of whether the transaction price and corresponding performance obligations were fairly presented for related party transactions". The auditor flagged the same line.
- Sharpening: the Shareholder Governance Agreement exempts SoftBank transactions existing at the IPO from related-party review altogether, and exempts ordinary-course transactions the audit committee judges to be on third-party terms below a $20m threshold.
- Softening: Q1 FY27 does not extend the pattern. External revenue was $901m, +24%; related parties $388m, +18%; the related-party share held at 30.1% rather than climbing. One quarter is not a trend, but it is the opposite of the FY26 shape.
Balance-sheet exposure to the same counterparties is material and mostly unbilled: contract assets from related parties $577m at 2026-06-30 ($646m at 2026-03-31) and receivables from related parties $281m ($270m) — together about 57% of a full year of related-party revenue.
Red-flag sweep
- Auditor: Deloitte; opinion clean, one critical audit matter as described above. No auditor change and no
NT 20-Fin the form census since 2023. - Going concern: not applicable — no debt, $3.1bn cash, positive operating cash flow.
- Related party: this is the related-party story; see above.
- Restatement / comment letters: the census shows 5
UPLOADand 5CORRESPitems, consistent with the 2023 IPO review period. Not verified by full-text search — see "Evidence that was thin".
S5 — Ownership
Control. "As of May 21, 2026, SoftBank Group beneficially owns approximately 86.4% of our total issued and outstanding share capital, representing a majority of the total voting power" (20-F, Item 3.D). Arm is a Nasdaq controlled company and relies on the exemptions; as an FPI it files no proxy and no domestic Section 16 reports. The public float is about 145m shares.
Institutional register. threespread.py holders ARM returned zero 13F filers for CUSIP 042068 205, so this came from the Equibles fallback (13F report date 2026-06-30, 1,058 rows across 985 institutions, 156,099,393 shares including option notional). Largest common lines: Invesco 12,816,588 (8.21% of institutional 13F shares); DZ Bank 6,504,147; Capital World 4,330,875; Arrowstreet 4,094,696; FMR 4,062,977. Nobody holds a controlling minority stake — SoftBank has it.
The option book is where the opinion is. Five of the top twelve 13F lines are options, and the second-largest line in the whole name is a put: Susquehanna 9,099,400 shares' notional, against a call line of 8,445,200. Jane Street (put 7,079,600 / call 4,311,800), Citadel (put 5,345,100 / call 3,319,900) and Barclays (put 2,967,200) show the same two-sided shape. In a stock with a 13.6% float, the derivative book is a large share of the tradeable opinion.
Short interest (FINRA via Equibles): 16,740,615 shares at 2026-08-31, down 1,523,490 on the settlement. That is ~1.6% of shares outstanding but roughly 11-12% of the public float, at 4.4 days to cover — several times MRVL's 1.4.
Insider transaction cadence: 94 Form 4 filings in the twelve months to 2026-09-21 (Arm files them despite FPI status), latest 2026-09-16, plus 35 Form 144s in the census since 2023. Filings, not transaction detail — a cadence reading only.
What hangs over it. No shelf: the EDGAR census since 2023 shows 20-F, 6-K, one SC 13G (2024), S-8s and the IPO F-1 family, and no F-3 or F-3ASR. There is therefore no registered secondary on file — but SoftBank's 86.4% is a standing overhang of a different kind, and the Shareholder Governance Agreement grants it registration rights.
S8 — Valuation and sizing ⏳ OPEN (user)
Guidance evidence is two rows only (Equibles GetGuidance, from the 2026-07-29 call): Q2 FY27 non-GAAP operating expense $780M and Q2 royalty growth 13%. Arm does not put a revenue guide into the structured record the way Marvell does, so the model below leans on the reported trajectory, and its assumptions are mine, not management's.
| Model line | FY27E | FY28E |
|---|---|---|
| Revenue $m | ~5,900 (+20%; Q1 actual 1,289) | ~7,375 (+25%) |
| Gross margin | 97.5% (FY26 actual) | 97.5% |
| Gross profit $m | 5,753 | 7,191 |
| Non-GAAP opex $m | ~3,120 (Q2 guide $780m × 4) | ~3,588 (+15%) |
| Non-GAAP operating income $m | ~2,633 | ~3,603 |
| Net interest $m | +100 | +120 |
| Tax | 17% | 17% |
| Non-GAAP net income $m | ~2,268 | ~3,090 |
| Diluted shares m | ~1,085 | ~1,100 |
| Model non-GAAP EPS | ~$2.09 | ~$2.81 |
| Multiple at $275.61 | ~132x | ~98x |
For reference, reported FY26 GAAP EPS was $0.85 ($904m / 1,068m), so the stock is on ~325x trailing GAAP earnings, and the gap between the two columns is stock compensation running at ~$1.4bn a year against ~$0.9bn of GAAP operating income.
Bear / base / bull on the FY28 line: bear — royalty growth settles at the guided 13% and licence revenue normalises without the related-party step-up, revenue ~$6.3bn, EPS ~$2.10, 40x gives ~$84; base — the table, EPS $2.81, 60x gives ~$169; bull — AGI CPU and compute-subsystem royalties inflect, revenue ~$8.5bn, EPS ~$3.60, 80x gives ~$288. Only the bull case reaches today's price, and it needs both a 25%+ revenue path and a multiple that itself assumes one.
The honest summary: on any multiple this pipeline can defend, the price already contains the bull case. Arm's business quality is not in question — 97.5% gross margin, no debt, an architecture with no substitute — but at 98x a model FY28 that assumes 25% growth, the margin of safety is the multiple, and there isn't one.
Proposed buy price and proposed size are owner-only and are not stated here. Note for sizing: the 13.6% float, not the $1.9bn of daily turnover, is the real liquidity constraint — an 86.4% holder means the tradeable stock is a small fraction of the market capitalisation, and short interest is already 11-12% of it.
S9 — Monitoring
- Next earnings: Q2 FY27 results, expected early November 2026 — derived from filing cadence (Q2 FY26 results 6-K filed 2025-11-05), not from a calendar feed.
- Filing feed: the Tier 0 EDGAR sweep covers cik 1973239 (
watchlist.yamlcarries it), so 20-F and every 6-K arrive without further setup. Arm's 6-K stream is the only quarterly disclosure. - Leading indicator between quarters: the related-party revenue line on the face of the income statement in each 6-K, read as a percentage of total and against external growth. It is one number, disclosed every quarter, and it is what separates a licensing business compounding at 24% from a parent booking licences into a subsidiary at 86.4% ownership.
- Trigger that converts this to a buy: two consecutive quarters of external-customer revenue growth at or above 20% with the related-party share flat or falling, and royalty revenue growing faster than licence revenue, and a price at or below ~60x the model FY28 EPS — about $169 on today's model, roughly 39% below the 2026-09-18 close. Nothing in the business needs to break for that to happen; the multiple simply has to stop assuming the bull case.
Evidence that was thin, and why
- No earnings-call transcript was read, and Arm's structured guidance record holds only two rows. The gate-minimum spec asks for two transcripts and the Q&A; this run traded them for budget across three names. For Arm specifically the cost is higher than for Marvell, because the call is where a controlled company explains a related-party licence.
- The institutional register came from the Equibles fallback, not 3spread —
threespread.py holders ARMreturned zero filers for the ADS CUSIP (042068 205). The Equibles rows mix common and option positions and are shares-of-13F-total, not shares-of-outstanding, so no ownership percentage of the company can be read off them. - No
efts.sec.govfull-text sweep for restatement or auditor-change language; the auditor and comment-letter reads come from the form census and the opinion instead. - No analyst coverage count.
- Rene Haas's role is background knowledge, not read from a filing in this run — the node is marked
PROPOSED_UNRESOLVEDand no edge depends on it. - Next earnings date is derived from filing cadence, not a calendar feed (market data's earnings calendar needs a paid plan on this account).