Arm Holdings plc (NASDAQ:ARM) built its business around helping other companies design processors. Selling its own data-center CPU gives it a chance to capture more revenue from each system, but changes its relationship with the same customers that license its technology. Investors need to decide whether that change enlarges the moat or makes it more expensive to maintain.
NVIDIA sold its Arm shares while continuing to use Arm’s architecture. Our analysis asks what that split between ownership and dependence means for Arm’s moat. The chip strategy also puts Arm closer to the hardware economics examined in our analysis of how abundant AI capacity could shift bargaining power toward chip customers.
For illustration purposes only. Photo from Arm Holdings
A bigger revenue opportunity brings different costs
In its fiscal 2027 first-quarter results, released July 29, Arm reported $1.29 billion of revenue. Royalties contributed $715 million, up 22%, and licensing and other revenue $574 million, up 23%. Data-center royalty revenue more than doubled. That gives the chip program a growing base rather than a stagnant business it must rescue.
Arm said its AGI CPU had been delivered to initial customers and that it had secured capacity for a $1 billion revenue opportunity across fiscal 2027 and fiscal 2028. Management described demand exceeding $2 billion over those two years. The latter figure is demand commentary, not recognized revenue or a guaranteed order book. Even the smaller capacity-backed opportunity spans two years, making it inappropriate to compare directly with one quarter’s revenue.
Licensing lets customers bear much of the manufacturing, inventory and product-commercialization burden. Selling chips can capture more dollars, but Arm must participate more directly in those costs. Revenue growth therefore needs to be tested against incremental gross profit and cash investment. A chip dollar and a royalty dollar are not interchangeable valuation units.
Insider Monkey’s hedge fund database showed 52 Arm holders in Q2 2026, compared with 46 in Q1, an increase of six. That historical participation says little about the eventual economics of the chip initiative.
A large multiple needs more than a successful launch
At the September 30 market-data snapshot, Arm traded at roughly 59 times enterprise value to trailing revenue and about 205 times trailing free cash flow. Those measures price in substantial expansion. The reported first-quarter revenue, annualized mechanically, would be $5.16 billion; it is a run-rate illustration rather than an annual forecast.
Against the approximately $306 billion enterprise value in the snapshot, a future $10 billion revenue business at a 30% operating margin would generate $3 billion of operating profit and still be valued at more than 100 times that profit. At $15 billion of revenue and a 40% margin, the equivalent multiple would be around 51. Both scenarios require assumptions about scale and business mix; neither is management guidance. They show why a new product opportunity cannot, by itself, explain the current valuation.
Arm had 15,521,694 shares sold short at the September 15 settlement, with 3.9 days to cover. Its relatively small publicly traded float makes the denominator important; a percentage calculated against all outstanding shares would understate public-float exposure.
The best bull case is broader than direct chip sales. Arm can earn royalties from customers’ processors while using its own product to accelerate the architecture’s adoption and demonstrate competitive performance. Continued growth in licensing and royalties would support that interpretation.
The bear case is a gradual erosion of neutrality. A customer may become less willing to share its roadmap or depend on a supplier that also competes for the same server budget. There is no basis here to claim that customers have already defected. The risk becomes investment-relevant if the chip business expands while licensing growth, customer breadth or royalty economics weaken.
Arm Holdings plc (NASDAQ:ARM) can expand its opportunity by selling competitive data-center systems. Its stock case improves when that expansion produces incremental cash profit while licensing and royalties keep growing. A larger chip business accompanied by weaker customer licensing or heavier capital demands would change the moat calculation.
