Arm (ARM) Q4 2026 Earnings Call Transcript

Arm reported record Q4 revenue of $1.49 billion, up 20% year over year, with full-year revenue rising 23% to $4.92 billion and non-GAAP EPS hitting a quarterly record of $0.60. Management said demand for the new Arm AGI CPU has doubled to more than $2 billion across fiscal 2027-2028, while fiscal Q1 2027 guidance calls for $1.26 billion in revenue and roughly 20% growth. The company also highlighted data center royalty revenue more than doubling and operating margins near 49%, reinforcing a strong AI-driven growth outlook.

Analysis

The key read-through is that ARM is no longer just a royalty compounding story; it is morphing into a platform-control story where silicon, subsystems, and IP reinforce each other. That matters because the company is effectively trying to reprice the CPU layer in AI infrastructure, and the demand inflection suggests hyperscalers are willing to pay for power efficiency and software continuity even if they already have internal silicon roadmaps. The second-order effect is that ARM becomes a toll collector on a larger portion of the AI stack, which should support a higher multiple if execution holds.

The market may be underestimating how much this changes the competitive set. The immediate pressure is not just on x86 incumbents; it is on any merchant silicon strategy that depends on generalized CPU attach rates inside AI racks. If ARM’s rack-level solution gains traction, it can pull share from bespoke in-house programs by reducing deployment friction, which also benefits ODMs and certain hyperscaler ecosystem partners, while raising the bar for Intel’s data center relevance over time.

The main risk is not demand, but supply and narrative slippage. Management is explicitly constrained by foundry, memory, packaging, and test capacity, which means the next 1-2 quarters could show strong bookings but muted revenue conversion, creating a classic “great demand, modest beats” setup. There is also a valuation risk: once investors extrapolate a 2031 target into the present, the stock can become hypersensitive to any hint that AGI CPU adoption is back-half weighted or that licensing normalization offsets the silicon excitement.

Contrarian takeaway: the consensus may be too focused on whether ARM can win chip share, when the bigger question is whether it can expand the total addressable spend per deployed AI rack. If that thesis is right, the winners are not only ARM but also ecosystem names that sell the picks-and-shovels for Arm-based deployments; if it is wrong, the stock will be vulnerable to a sharp multiple reset because the market is already discounting a very long runway.

Key Decisions for Investors

• Long ARM on a 3-6 month horizon into supply-chain updates; best entry is on any post-print consolidation, because the near-term downside is more about timing than demand. Risk/reward is attractive if investors continue to price only the royalty base and not the silicon option.
• Pair trade: long ARM / short INTC for 6-12 months. The thesis is structural share loss in data center CPUs versus ARM’s expanding attach to AI infrastructure; the short is cleaner if x86 demand remains soft and ARM’s execution stays on track.
• Add exposure to ecosystem beneficiaries on pullbacks: AMZN, GOOGL, MSFT, NVDA, and META over the next 1-2 quarters. These names benefit from lower rack power and potentially faster AI deployment cycles if ARM-based orchestration becomes standard.
• Use FFIV as a tactical long if you want a secondary beneficiary of AI/networking rack rebuilds; this is a lower-beta way to express broader Arm-enabled infrastructure spend over 3-6 months.
• Avoid chasing ARM after sharp gap-ups; instead, sell cash-secured puts or use call spreads to express upside while limiting multiple-compression risk if the market starts discounting the 2031 target too aggressively.