Arm is executing well as it expands from mobile into cloud AI, PCs, automotive, and physical AI, with revenue and royalties accelerating. But the stock already reflects a lot of that optimism, leaving limited upside at current levels.
Arm Holdings (ARM) owns one of the semiconductor industry's most valuable compute platforms, and its fiscal 2026 revenue reached $4.92 billion after three consecutive years of growth above 20%. The investment case rests on a clear shift: Arm is moving from a mobile-focused CPU intellectual property supplier toward a broader platform spanning cloud AI, smartphones, PCs, automotive, robotics, and physical AI.
The operating evidence is strong. Q1 fiscal 2027 revenue rose 22% year over year to $1.29 billion, royalty revenue increased 22% to $715 million, licensing revenue grew 23% to $574 million, and non-GAAP EPS rose 29% to $0.45. Data center royalty revenue more than doubled, while Neoverse shipments surpassed 1.5 billion cores.
The problem is price. At $241.56, ARM carries a trailing P/E of 311.9x, a forward P/E of 151.5x, an EV-to-revenue multiple of 68.2x, and a free-cash-flow yield of 0.6%. Those figures leave little room for execution errors, smartphone weakness, supply constraints, or a slower adoption curve for the Arm AGI CPU. For a moderate-risk investor with a medium-term horizon, ARM earns a Hold rating: the business is strategically strong, but the stock already prices in a substantial portion of that strength.
Arm Holdings plc, headquartered in Cambridge, United Kingdom, researches, develops, licenses, and markets CPU intellectual property, graphics processing unit IP, neural processing unit accelerators, systems IP, compute subsystems, software, tools, and related services. The company was founded in 1990, has 9,584 employees, trades on Nasdaq under ARM, and operates as a subsidiary of SoftBank Group.
Arm's core model has two main revenue streams. Customers pay license fees for access to Arm technology and design rights, then Arm earns royalties when those customers ship chips that use the licensed technology. This model lets Arm participate in device and infrastructure growth without carrying the capital burden of owning large-scale wafer fabrication facilities.
The company is also expanding into more integrated products. Its March 2026 introduction of the Arm AGI CPU marked a move into production silicon, while compute subsystems and chiplets extend Arm's role beyond individual CPU cores. That expansion raises revenue per design opportunity, but it also introduces supply-chain, margin, execution, and customer-channel risks that were less prominent in the traditional licensing model.
Arm reports its business primarily through licensing and royalty revenue rather than conventional operating segments. In fiscal 2026, royalty revenue was $2.61 billion, or 53.1% of total revenue, while license and other revenue was $2.31 billion, or 46.9%. Total revenue increased from $4.01 billion in fiscal 2025 to $4.92 billion in fiscal 2026.
Royalty revenue is the more durable growth engine because it scales with customer shipments after a design wins adoption. Q1 fiscal 2027 royalty revenue reached a record $715 million. Cloud AI was the largest driver, data center royalties more than doubled year over year, and higher royalty rates from Armv9 and compute subsystems helped offset weaker smartphone unit demand.
Licensing revenue remains strategically important because it leads future royalty revenue. Q1 licensing and other revenue reached a record $574 million, up 23% year over year. A SoftBank technology licensing and design-services agreement contributed $193 million, and management described the remaining fiscal-year quarterly run rate for that agreement as approximately $200 million. Annualized contract value rose 13% year over year, giving the licensing stream a stronger underlying indicator than any single quarter.
The end-market mix is broadening. Data center royalty growth is accelerating, edge AI royalty revenue continued to grow despite a soft smartphone market, and physical AI contributed to royalty growth through advanced driver-assistance systems and autonomous applications. This mix shift matters because it places more of Arm's revenue behind higher-value compute markets rather than relying solely on smartphone volumes.
Get AI research on any stock
Instant reports, daily intelligence, and an AI analyst in your pocket.
The Arm AGI CPU is the most important new product in the investment case. Arm introduced it in March 2026, delivered initial products to multiple customers, and secured manufacturing capacity for an opportunity initially estimated at $1 billion across fiscal 2027 and fiscal 2028. Management later said demand had risen above $2 billion, including customers in the United States and China.
The current AGI CPU uses a 128-core design. Arm's management also pointed to Amazon's 192-core Graviton5 as evidence that higher core counts are the direction of travel for agentic workloads. The product is designed for traditional servers, head nodes, and agentic applications, giving it exposure to several parts of the AI infrastructure stack rather than one narrow workload.
The product's economics require patience. Management expects first-generation gross margins in the high-30% to low-40% range, with a path toward 50% over the following years as Arm brings more work in-house. Production also depends on wafers, substrates, testing capacity, and memory. Demand above $2 billion is encouraging, but supply capacity determines how much of that demand becomes revenue.
Arm's strongest advantage is the combination of architecture, installed base, and software compatibility. The company says its technology has been deployed in more than 350 billion chips and more than 99% of smartphones. Those figures give chip designers and software developers a familiar foundation across mobile, embedded, automotive, cloud, and edge markets.
The developer network reinforces that position. Arm's software ecosystem supports more than 22 million developers, while its Performix optimization product has support from Microsoft, MongoDB, Redis, and SAP. The Arm MCP server passed 10,000 Docker downloads during the quarter. These are concrete signs that Arm is investing in tools that make its architecture easier to deploy in AI workloads.
Arm also offers customers several levels of engagement: individual IP, compute subsystems, or production silicon. That flexibility can reduce adoption friction because a customer can use Arm without designing every component internally. The tradeoff is strategic tension. The 2026 20-F warns that Arm's move into silicon could cause customers to seek better terms, reduce collaboration, or pursue alternative architectures.
Arm's traditional licensing model is asset-light, but the AGI CPU changes the operating profile. The product requires coordination across wafer suppliers, substrate providers, memory vendors, and test capacity. Management described the market as tight across all four areas and said supply expansion remains central to the fiscal 2027 and fiscal 2028 opportunity.
The company is increasing research and development investment to support next-generation architectures, compute subsystems, and the AGI CPU family. Q1 fiscal 2027 non-GAAP operating expense rose 18% year over year to $733 million, while management said the spending plan for the year remained intact. This investment supports product breadth, although it places pressure on near-term operating leverage.
Execution has been strong in the reported quarter. Q1 operating cash flow reached $902 million and free cash flow reached $694 million, compared with $365 million and $181 million, respectively, in the December 2025 quarter. The cash generation gives Arm room to fund product development and supply commitments without relying on heavy borrowing.
The semiconductor market is expanding around AI infrastructure, memory, networking, edge inference, automotive systems, and industrial automation. Gartner estimated worldwide semiconductor revenue at $793 billion in 2025 and forecast $1.56 trillion in 2026 and $1.94 trillion in 2027. Gartner also reported that AI semiconductors, including processors, high-bandwidth memory, and networking components, represented nearly one-third of semiconductor sales in 2025.
Arm's own 2026 investor materials frame its addressable market at $535 billion in fiscal 2026 and more than $1.5 trillion in fiscal 2031. The framework divides that opportunity into cloud AI, edge AI, and physical AI, with cloud AI identified as the largest expansion area. Arm's Q1 data center royalty growth provides operating evidence that this market is translating into current demand.
The market is not uniformly favorable. Higher memory prices have weakened smartphone demand across low-, mid-, and upper-tier products. Arm's Q2 guidance calls for royalty growth in the low teens, below the pace management had previously associated with the broader royalty business. Cloud AI has offset that pressure in recent results, but the mix makes Arm more exposed to the timing of data center deployments.
Like what you're reading?
Get full access to AI-powered research reports, market analysis, and portfolio tools.
Arm serves semiconductor companies, original equipment manufacturers, cloud service providers, and organizations building products for smartphones, consumer electronics, industrial IoT, embedded systems, cloud data centers, networking, automotive, robotics, and physical AI. That customer range gives the company access to many device categories while preserving a common architecture.
The customer base includes major platform companies. NVIDIA brought the Arm-based Vera CPU into production, Google uses Axion as a host CPU for its TPU systems, AWS plans to deploy tens of millions of Graviton5 cores, Microsoft expanded Azure Cobalt 200 virtual machines, and Qualcomm announced the Arm-based Dragonfly C1000 for AI data centers.
Concentration remains a material risk. Arm's fiscal 2026 filing says its five largest customers represented approximately 57% of revenue, while Arm China represented approximately 16%. Institutional ownership was 96.3% of shares, and tracked institutional activity showed 13 holders increasing positions versus 6 decreasing positions. High institutional ownership supports liquidity, but it can also amplify price movements when growth expectations change.
Arm competes at the architecture level with x86, primarily through Intel and AMD, and with the open RISC-V ecosystem. It also competes with proprietary CPU teams inside large cloud companies and semiconductor firms. NVIDIA, AMD, and Intel are adjacent competitors in heterogeneous compute because GPUs, accelerators, ASICs, and CPUs increasingly work together in AI systems.
Arm's data center position has improved through customer validation. Management said NVIDIA's Vera delivers up to 50% higher CPU performance and 2x greater energy efficiency than comparable x86 systems. Google, AWS, Microsoft, and Qualcomm have each attached Arm-based CPUs to their own cloud or AI infrastructure plans, creating a broad set of reference designs.
RISC-V remains the clearest structural threat because its open model can appeal to customers seeking architectural control and lower licensing dependence. Arm's 20-F also identifies proprietary architectures as a competitive risk. Arm's defense is the installed base, the software ecosystem, mature tools, and the ability to move the same architecture from cloud infrastructure to consumer and physical systems.
AI capital spending is the main macro tailwind. Gartner expects the AI data center ecosystem to rise from 36.5% of semiconductor revenue in 2026 to more than 53% by 2030. IDC also reported that spending on Arm-based accelerated server platforms nearly doubled over the prior two quarters and surpassed x86 platforms.
The macro risk is concentration in a capital-intensive investment cycle. Arm's Q1 results benefited from cloud AI overperformance, while smartphone royalties faced weaker unit demand from higher memory prices. That split shows how a data center slowdown and a consumer recovery would affect different parts of the company in opposite directions.
Geopolitical exposure is substantial because Arm serves customers in the United States, China, Japan, Taiwan, Korea, and other markets. The 20-F identifies trade restrictions, sanctions, export controls, and changes in government policy as business risks. Arm China's approximately 16% contribution to fiscal 2026 revenue makes the China relationship especially relevant to the medium-term risk profile.
Arm earns an A on balance sheet health, reflecting a strong financial position that gives it flexibility as it expands into more integrated products.
Unlock the full analysis
Premium members get the complete breakdown — pick rationale, financial metrics, and recent earnings detail.
Get Full Access →Q1 fiscal 2027 revenue rose 22% to $1.29 billion and non-GAAP EPS climbed 29% to $0.45, showing solid operating momentum.
Unlock the full analysis
Premium members get the complete breakdown — pick rationale, financial metrics, and recent earnings detail.
Get Full Access →Annualized contract value increased 13% year over year, while management said demand for the Arm AGI CPU has risen above $2 billion.
Unlock the full analysis
Premium members get the complete breakdown — pick rationale, financial metrics, and recent earnings detail.
Get Full Access →Arm trades at 311.9x trailing earnings, 151.5x forward earnings, and 68.2x EV/revenue, with a free-cash-flow yield of just 0.6%.
Unlock the full analysis
Premium members get the complete breakdown — pick rationale, financial metrics, and recent earnings detail.
Get Full Access →At $241.56, Arm sits above our $240 fair value estimate, so the stock earns a Hold despite strong AI and royalty growth.
Unlock the full analysis
Premium members get the complete breakdown — pick rationale, financial metrics, and recent earnings detail.
Get Full Access →Arm is becoming a broader compute platform at the same time that AI is expanding across data centers, PCs, smartphones, vehicles, and robotics. Fiscal 2026 revenue of $4.92 billion, Q1 fiscal 2027 revenue growth of 22%, data center royalties more than doubling, and Neoverse shipments above 1.5 billion cores establish a credible growth foundation.
The investment decision is therefore less about business quality than valuation discipline. ARM has the cash, ecosystem, customer validation, and product expansion to remain a market leader, but its 311.9x trailing P/E and 151.5x forward P/E make execution expectations unusually demanding. For a moderate-risk investor, Hold is the appropriate balance between Arm's platform strength and the price already attached to it.
Get AI-powered research reports, daily market intelligence, and a personal analyst in your pocket.
Get Full Access →Not ready to subscribe? ·
Cancel anytime
Dylan Jovine of Behind the Markets is promoting NVIDIA’s “Secret Supplier,” also branded “One Chip to Rule Them All.” We trace the clues and test the royalty claims.
Arm Holdings plc American Depositary Shares (NASDAQ: ARM) drops after the Qualcomm royalty trial begins, pressuring the stock alongside a weaker semiconductor tape. The move reflects legal risk to a key revenue stream, even as Arm’s long-term AI and licensing story remains intact.
Arm Holdings plc American Depositary Shares (ARM) rises 6.2% as investors stay focused on its role in agentic AI infrastructure. The move follows recent NVIDIA-related commentary and positive fund notes, but light relative volume suggests this is a rebound rather than a confirmed breakout.