Arm Holdings stock falls despite earnings beat
Arm Holdings tumbled Thursday on investor worries about the company's ability to secure sufficient supplies for its new AI chip to meet demand, despite beating earnings expectations.
Objective Facts
Arm Holdings released fiscal fourth-quarter earnings after market close on Wednesday, May 6, showing quarterly revenue of $1.49 billion (up 20% year-over-year and beating expectations of $1.47 billion) and adjusted earnings per share of 60 cents (also beating forecasts). CEO Rene Haas revealed that customer demand for the AGI CPU doubled to over $20 billion within six weeks of launch, though the company maintained a conservative outlook due to supply chain bottlenecks. The decline followed comments during the company's earnings call suggesting Arm has not yet secured enough supply capacity to satisfy an additional $1 billion in demand tied to its new AGI CPU. The broader market offered little offset, with the S&P 500 essentially flat at -0.05%, the Dow Jones down -0.04%, and the NASDAQ marginally higher at +0.11%. The combination of a supply-constrained near-term outlook, a royalty revenue miss, and a stretched valuation converged to drive the sharp decline despite the headline earnings beat.
Left-Leaning Perspective
On April 7, 2026, Morgan Stanley downgraded Arm Holdings from 'Overweight' to 'Equal-Weight,' slashing its near-term expectations and citing growing institutional anxiety over the company's ambitious pivot from a pure-play IP licenser to a direct manufacturer of high-performance AI CPUs. Morgan Stanley analyst Lee Simpson raised the price target to $150 but warned that Arm's strategic shift into designing its own AI chips faces significant execution risks, including soft demand, DRAM supply constraints, rising R&D costs, and potential conflicts with existing licensees. Seaport Research Partners analyst Jay Goldberg stated "It was a very tough setup for them - the expectations were just so high. They were good numbers, but not good enough," noting that royalty revenue was $671 million, compared with expectations of $697.1 million. The transition from a 98% margin IP business to a 45% margin chip manufacturer is a 'valuation reset' event that requires investors to change how they model the company's future, and while Arm remains the fundamental architecture of the modern world, its path to becoming a $25 billion revenue powerhouse by 2031 is clearly going to be more expensive and more volatile than previously hoped. Bearish analysts highlight that after a series of target hikes, Arm's valuation already prices in a lot of success, which can limit upside if execution or adoption of AI CPUs is slower than modeled, with some signals of concern that prior expectations, especially for the outer years of AI driven growth, may have run ahead of what can be supported by currently visible demand. The combination of OpenAI's missed growth targets, TSMC's full exit from its Arm stake, and profit-taking after prior gains highlighted how dependent Arm's story has become on confidence in sustained AI capital expenditure. Where targets have been reduced, the messages focus on execution risk, the timing of AI related revenue ramps and the possibility that competition or customer adoption patterns make the current P/E framework harder to sustain over time. Left-leaning analysis emphasizes that supply chain constraints reveal deeper operational vulnerabilities and that Arm's premium valuation is not justified given execution uncertainty in a crowded AI CPU market.
Right-Leaning Perspective
Bullish analysts point to AI related CPU workloads in data centers as a core driver for higher long term revenue assumptions, which feeds directly into higher fair value estimates and price targets. Susquehanna raised its price target on ARM to $210 from $170, maintaining a Positive rating, citing ARM's AGI CPU product as evidence that the investment case extends beyond smartphone royalties, noting the potential for earnings per share above $10 over the coming years even as smartphone royalty pressure persists. Evercore ISI lifted its price target on ARM to $227, retaining an Outperform rating, while projecting a path to $15 billion in FY2031 revenue and earnings per share above $9, with longer-term earnings scenarios reaching the low-$20s range if ARM executes on server CPUs and agentic AI workloads. Taiwan Semiconductor had recently sold its Arm Holdings stake, yet ARM stock had held firm as Wall Street kept betting on AI, data centers, and fresh chip ambitions—a dynamic that left the stock exposed once execution risk emerged. Right-leaning bullish coverage emphasizes that wealth management firm Cerity Partners partner Michael Ashley Schulman stated 'The market sees that as a party spoiler. They will likely get the supply but the market doubt hinges on whether it will be quick enough and then what happens when more demand arrives', suggesting supply constraints are temporary and solvable. One retail investor on Stocktwits highlighted how most chipmakers are choked by limited manufacturing capacities and opined that they should raise prices, framing the supply situation as a pricing opportunity rather than an existential threat. The AI semiconductor market is projected to grow from $150 billion in 2024 to $500 billion by 2028, offering Arm a vast runway, with its vertical integration strategy combined with a flexible licensing model positioning it to capture a significant share of this growth. Bullish analysis sees the sell-off as an opportunity and supply constraints as normal friction in scaling a new business, not evidence of fundamental problems.
Deep Dive
Arm's fiscal Q4 earnings announced that customer demand for its AGI CPU doubled to over $20 billion within six weeks of launch, indicating genuine and substantial hyperscaler commitment to Arm's new data center strategy. However, the company revealed during its earnings call that it has not yet secured enough supply capacity to satisfy an additional $1 billion in demand tied to its new AGI CPU, exposing a critical gap between design success and manufacturing execution. This discrepancy is not unique to Arm; Samsung, SK Hynix, and Micron—which collectively control the HBM market—have announced capacity expansion plans, but new fabrication facilities require 18 to 24 months from groundbreaking to production, pushing meaningful supply relief into 2026 at the earliest, with full market balance unlikely before 2027. What separates justified concern from over-reaction depends on context. The rapid expansion of AI infrastructure and workloads is exerting significant pressure on the memory ecosystem, with these AI workloads requiring large amounts of memory, and the shortage, in part, driven by a reallocation of manufacturing capacity away from consumer electronics toward high-margin memory solutions to support AI. This is structural, not cyclical. For Arm specifically, the question is whether the company can: (1) negotiate long-term supply contracts at favorable terms; (2) maintain profit margins as manufacturing costs remain elevated; (3) avoid alienating existing licensees (like Apple, Qualcomm) while competing with them in the CPU market through its own chips. A central concern raised by Morgan Stanley is the potential for Arm's entry into the chipmaking business to upset its licensee partners—a risk that transcends supply chain issues. Looking ahead, for investors, the coming months will be defined by 'execution tracking,' with updates on the Meta and OpenAI partnerships serving as the first 'canaries in the coal mine' for the AGI CPU's commercial viability. Success is not assured, but neither is failure. The stock's 7% drop reflects appropriate repricing of risk; whether it rebounds or falls further depends on quarterly progress securing supplies and demonstrating margin discipline. The market was pricing in flawless execution and unlimited manufacturing access—neither realistic for a company transitioning from pure-play IP licensing into hardware manufacturing during an unprecedented memory shortage.