Arm Holdings (ARM) Stock Forecast: What Could Drive It in 2026
Short answer
What is actually driving Arm Holdings (ARM) right now is Armv9 royalty uplift: Arm's newest architecture generation, Armv9, carries royalty rates roughly double those of the older Armv8 it replaces. Revenue (FY2026, ended March 2026) is ~$4.92 billion. If that keeps playing out, the setup is favourable; the risk to it is the clearest risk is valuation: Arm trades at a forward P/E near 190 and a trailing multiple in the hundreds, so the price already assumes years of rapid, uninterrupted growth, and any wobble in AI or smartphone demand can trigger sharp drops (the stock fell sharply in late 2025 on such fears). No one can predict where ARM trades, and Walnut does not publish targets, so treat this as a scenario, not a price target or prediction.
What could drive Arm Holdings (ARM) higher?
Armv9 royalty uplift
Arm's newest architecture generation, Armv9, carries royalty rates roughly double those of the older Armv8 it replaces. As phone makers and chip designers migrate to Armv9 and to Arm's Compute Subsystems, Arm collects more per device even if unit volumes stay flat. This mix shift is a structural tailwind that has lifted royalty revenue without Arm needing to ship a single chip itself.
Data center and AI demand
AI infrastructure is becoming a major royalty source, with data-center royalties more than doubling year over year in fiscal 2026. Arm-based CPUs from hyperscalers and AI accelerator vendors are displacing some traditional x86 server chips, and management frames agentic AI as driving a large step-up in CPU cores per data center. This pushes Arm into a higher-value market than mobile.
Own-silicon ambitions
In 2026 Arm moved beyond licensing IP to selling complete chips, launching the AGI data-center CPU with up to 136 Neoverse cores on TSMC's 3nm process, with Meta as the first customer. Selling finished silicon can capture far more revenue per design than a royalty, though it also puts Arm in partial competition with its own customers. Management has outlined a multibillion-dollar chip-revenue target over the rest of the decade.
What could weigh on ARM?
The clearest risk is valuation: Arm trades at a forward P/E near 190 and a trailing multiple in the hundreds, so the price already assumes years of rapid, uninterrupted growth, and any wobble in AI or smartphone demand can trigger sharp drops (the stock fell sharply in late 2025 on such fears). SoftBank's roughly 86 to 90% ownership means a thin public float and a persistent overhang if SoftBank ever sells more shares. Royalty-free RISC-V is a long-term structural threat backed by large customers seeking to avoid Arm fees. And revenue is concentrated among a handful of large customers, while Arm's new chip business risks competing with the very licensees it depends on.
How to think about a ARM forecast
Rather than chasing a price target, it tends to help to weigh the drivers above against the risks, decide how long you are willing to hold, and size the position so a wrong call is survivable. A “forecast” is really a probability-weighted view of those drivers playing out, not a number.
For the full picture, see the ARM guide and whether ARM is a buy. In Walnut you can pressure-test the thesis against your real portfolio.
The bottom line on the ARM outlook
The bottom line: what is driving Arm Holdings (ARM) is Armv9 royalty uplift, with revenue (fy2026, ended march 2026) at ~$4.92 billion. If that keeps playing out the setup is favourable; the risk is the clearest risk is valuation: Arm trades at a forward P/E near 190 and a trailing multiple in the hundreds, so the price already assumes years of rapid, uninterrupted growth, and any wobble in AI or smartphone demand can trigger sharp drops (the stock fell sharply in late 2025 on such fears). No one can predict the price, so treat any ARM forecast as a scenario, not a target or prediction, and decide from your own thesis and time horizon. Walnut is not an investment adviser.
Build a basket around ARM with Walnut
Use Arm Holdings as one constituent in a thematic basket Walnut's AI helps you assemble. Describe a thesis you believe in, the AI proposes the holdings and weights, and you approve before any broker order.
FAQ
What is the forecast for Arm Holdings (ARM)?
+
No one can reliably predict where ARM will trade, and Walnut does not publish price targets. What is more useful is the setup: the drivers that could push Arm Holdings higher and the risks that could weigh on it. This page lays out both so you can form your own view. Not a recommendation.
What could drive ARM higher?
+
The main growth drivers are Armv9 royalty uplift; Data center and AI demand; Own-silicon ambitions. Whether they play out is the real question, not a guaranteed path.
What are the risks to ARM?
+
The clearest risk is valuation: Arm trades at a forward P/E near 190 and a trailing multiple in the hundreds, so the price already assumes years of rapid, uninterrupted growth, and any wobble in AI or smartphone demand can trigger sharp drops (the stock fell sharply in late 2025 on such fears). SoftBank's roughly 86 to 90% ownership means a thin public float and a persistent overhang if SoftBank ever sells more shares. Royalty-free RISC-V is a long-term structural threat backed by large customers seeking to avoid Arm fees. And revenue is concentrated among a handful of large customers, while Arm's new chip business risks competing with the very licensees it depends on.
Will ARM stock go up in 2026?
+
Nobody knows, and anyone who says they do is guessing. Arm Holdings's direction depends on whether the drivers above outweigh the risks, plus the broader market. Focus on the thesis and your time horizon rather than a single-year call.
Is ARM a buy?
+
That depends on your thesis, time horizon, and what you already own, not on a forecast. See the ARM "is it a buy?" page for a framework. Walnut is not an investment adviser.
Walnut is informational, not investment advice. This page describes drivers and risks; it is not a price forecast, target, or recommendation. Markets are uncertain and past performance does not predict future results.