ARM's Supply Problem Just Made RISC-V's Best Pitch in Years

ARM traded at a trailing P/E — how many years of past earnings the stock costs of about 251 as of May 7, 2026. That's not just expensive — it's a valuation partly engineered by scarcity. With SoftBank keeping the vast majority of shares off the market, the limited float mechanically inflated the multiple. Every new buyer was bidding against a thin pool of available stock.

The stock had rallied 48% in the month leading into earnings. Even after the 10% drop, ARM sat at $213 — far closer to its 52-week high of $239.50 than its low of $100.02.

When a stock with a thin float runs up nearly 50% in a month, the exit door works in reverse on bad news. ARM didn't just disappoint — it told investors it might not be able to meet demand for its hottest product. On double the normal volume, the narrow float turned an orderly sell-off into a stampede.

The company's most recent quarterly revenue was $1.24 billion, up from $983 million a year earlier. The business is growing. But the stock was priced as if growth would be flawless — and flawless just left the building.

The companies most exposed are the hyperscalers pouring tens of billions into AI infrastructure. Microsoft, Google, and Amazon have all announced custom silicon programs built on ARM-licensed designs. Amazon's Graviton server chips and Google's Axion processors both rely on ARM architecture. When ARM flags supply constraints on its newest designs, those datacenter build timelines don't just pause — they get re-sequenced.

Earnings season is still rolling, with more semiconductor companies on deck. If ARM's supply issues are unique to ARM, the damage stays contained. If other chipmakers echo similar constraints, this domino gets much bigger.

The real question: can the big cloud companies stick to their AI spending plans — hundreds of billions of dollars over the next two years, combined — if the core architecture they're building on can't keep up? ARM's warning is the first crack in that assumption.

ARM's earnings flagged that memory is becoming expensive for smartphones. That's a problem layered on top of the supply constraints for ARM's newest chip designs.

Smartphone makers now face rising costs on two critical components simultaneously. They can either absorb the hit — which compresses their profit margins — or pass it along to consumers, which risks slowing unit sales.

Neither option is great. This squeeze hits during a period when handset makers were counting on AI features to drive an upgrade cycle. If the components that enable those features cost more and arrive slower, the upgrade story gets harder to sell.

As of May 7, 2026, ARM's market cap — the total value of a company's outstanding shares was approximately $226.5 billion. A 10% drop in a single day wiped roughly $25 billion off that figure. For SoftBank, which has positioned ARM as the centerpiece of its AI ambitions, the loss is more than paper-thin. It's a credibility dent.

Just weeks before the earnings stumble, SoftBank Group International gave ARM CEO Rene Haas an expanded role, effective April 20, 2026. The timing is awkward. SoftBank elevated ARM's leader to signal confidence in the company's strategic importance — and then the stock posted its worst day in months.

SoftBank has bet heavily that ARM will be the foundational architecture for the AI era. The sell-off doesn't invalidate that bet. But it does remind investors that even the best-positioned companies can trip when execution doesn't match the narrative.

ARM's business model depends on licensing its chip designs. Every time a company builds a chip using ARM's blueprints, ARM gets paid. That's a phenomenal business — until customers start asking whether they have other options.

This isn't just a thought exercise. Google has invested heavily in RISC-V through its backing of SiFive, a startup designing RISC-V-based processors. Alibaba has already deployed its XuanTie RISC-V cores in production hardware. These aren't science projects — they're active hedges against ARM dependence by two of the world's largest technology companies.

If ARM's supply constraints persist, companies that license ARM designs have a concrete reason to accelerate those hedges. Switching chip architectures takes years, not quarters. But the economic case for diversification just got stronger.

ARM had $2.15 billion in remaining performance obligations as of December 2025, with about 31% expected to convert to revenue over the next 12 months. That backlog is a sign of strong demand. But it's also a measure of how much is riding on ARM's ability to deliver. If customers shift even small workloads to RISC-V, ARM's long-term licensing edge gets a little weaker. That edge is the entire reason the stock trades at a premium.