🐂 ARM — Company Profile (Educational)¶
Educational site, not investment advice. Market cap / stock price / quarterly numbers change quickly, so we don't write them. Use SEC + Yahoo Finance links below to check current data.
📌 Basic Info¶
| Field | Value | Source |
|---|---|---|
| Primary business | CPU IP architecture licensing + physical IP (underlying ISA for most handset SoCs + embedded devices) — main competing path vs RISC-V | 20-F FY25 |
| Founded / HQ | 1990 / Cambridge, UK | wiki |
| CEO | Rene Haas (joined 2022/02) | company announcement |
| Primary listing | NASDAQ: ARM (IPO 2023/09; was SoftBank-owned private for 7 years prior) | — |
| Foreign filer | Yes (UK company, files 20-F annual report with SEC, not 10-K) | SEC |
| Major shareholder | SoftBank (SFTBY) holds ~87.1% (FY25 20-F, as of 2025/05/20) (free float only ~10% post-IPO) |
20-F equity disclosure |
| Fiscal year | Ends late March (FY26 = 2025/04 - 2026/03) | SEC EDGAR |
💡 Real-time market cap / share price / royalty + license mix / Armv9 penetration: see SEC primary + IR links below
🏭 Industry Chain Coordinates (Part 1-C6 framework)¶
Upstream (Who ARM depends on)¶
| Supplier | What flows | Key dependency |
|---|---|---|
| (Software company, no physical supply chain) | — | ARM is a pure IP company; upstream dependency is minimal |
SNPS + CDNS |
EDA + IP integration tools (bundled with ARM IP) | ecosystem partnership |
Downstream (Who depends on ARM)¶
| Customer | What flows | Concentration |
|---|---|---|
QCOM |
Snapdragon CPU IP licensing (post-litigation agreement) | major customer; QCOM's in-house Oryon core reduces some dependence |
AAPL |
A-series + M-series architectural ISA license (deeply customized) | major customer; Apple licenses ISA only, not ARM core IP |
MediaTek (2454.TW MTK) |
Dimensity SoC | major customer |
AMZN (AWS Graviton) |
In-house ARM data center CPU | key to ARM's data center penetration |
NVDA |
Grace CPU (ARM-based data center CPU) | data center |
MSFT Cobalt + META in-house ARM CPU |
In-house data center silicon | growth engine |
Samsung Exynos (SSNLF) / HUAWEI HiSilicon |
Various handset SoCs | long-tail diversification |
Competitors¶
- CPU ISA (direct, long-term): RISC-V (open ISA, royalty-free, primary path for Chinese domestic substitution) — main threat
- PC + data center x86:
INTC·AMD(CISC, full-stack competition with ARM) - GPU IP: Imagination Technologies (UK private) — direct rival to ARM Mali GPU
- Physical IP:
SNPS(Artisan business that ARM previously divested overlaps with SNPS)
💼 Business Model + Value Capture (Part 1-C7 lens)¶
- Revenue mix: License (upfront) ~50% / Royalty (per-chip shipment fee) ~50%; royalty is the long-tail recurring revenue engine
- Gross margin: GAAP 95%+ (IP company with intangibles, near-zero marginal cost); but R&D + SBC are high
- Pricing power: Armv9 royalty rate is ~2× Armv8; ARM negotiates pricing with hyperscaler in-house ARM CPU programs — very strong
- Replacement difficulty: High short-term (ecosystem + software + tools accumulated over 30 years); long-term depends on RISC-V adoption pace
(For license vs royalty / Armv9 penetration / data center royalty growth, see latest 6-K)
🏰 Moat Assessment (Part 3-C2 Buffett 5-step lens)¶
| Type | Strength (0-1) | Reasoning |
|---|---|---|
| Brand | 0.7 | "ARM" is essentially the ISA standard in embedded + handset ecosystems |
| Switching cost | 0.9 | 30 years of customer software + toolchain + engineering accumulation; massive effort to migrate to RISC-V |
| Network effect | 0.9 | Developers + compilers + OS (Linux/Android/iOS) + tools = full-stack platform effect |
| Scale + cost | 0.6 | Near-zero marginal cost (intangibles), but high R&D outlay |
Total: 3.1 / 4 → very strong moat (software ecosystem + switching cost double-block), but RISC-V is a structural long-term threat
⚠️ Key Invalidation Triggers (Part 3-C5 anti-thesis lens)¶
Monitor; if triggered, re-evaluate thesis:
- Trigger 1 (annual): RISC-V > 10% market share in embedded + data center (China + India + long-tail customers de-ARM)
- Trigger 2 (event):
QCOMOryon core /AAPLmove fully to ISA-only license → major royalty decline - Trigger 3 (medium probability): SoftBank concentrated ARM divestiture (SFTBY holds ~90%+; Stargate funding pressure could trigger)
- Trigger 4 (quarterly): Armv9 penetration growth < expectations (royalty rate uplift fails)
- Trigger 5 (low probability): U.S. / UK + China ARM IP export controls → China market royalty zeroed out
🔗 Cross-Chapter Teaching Links¶
- Part 1-C5 · Hardware Stack — CPU ISA layer complements GPU layer
- Part 1-C6 · Supply Chain 5 Roles — Pure IP company, value capture beyond physical supply chain
- Part 1-C7 · Value Capture — License + royalty dual engine
- Part 1-C9 · US-China Geopolitics — UK company + SoftBank ownership + China RISC-V strategy
- Patterns — #2 incumbent moat · #4 dual engine · #8 IP/intangibles model
📚 SEC Primary Public Filings (DYOR)¶
- Latest 20-F (foreign filer annual FY25) — full annual financials + risk factors + SoftBank related-party relationships
- Latest 6-K (quarterly — foreign filers don't file 10-Q) — quarterly license + royalty
- SEC EDGAR ARM all filings — complete filing history
- Investor Relations — earnings calendar + historical filings + Tech Day
Live quote: Yahoo Finance ARM
Educational · Numbers change, the method does not · Make investment decisions based on your own research + risk tolerance