ARM Holdings: The Smartphone That Wouldn't Die

ARM Holdings: The Smartphone That Wouldn't Die

Welcome to another edition of Numbers & Narrative by Longwalk Research. In this series, we examine the current business reality behind market narratives using detailed data analysis rather than forecasting future outcomes. Each analysis questions prevailing assumptions by examining present-day metrics, competitive dynamics, and operational realities that may contradict popular investment themes.

The "ARM everywhere" narrative has captivated investors since the company's 2023 IPO. ARM's processor designs are powering the AI revolution, conquering datacentres, and revolutionising automotive computing. The smartphone, that ancient relic, is supposedly just legacy baggage.

The numbers tell a rather different story.

Smartphone Dependency Remains

ARM's latest figures reveal a company still tethered to the very market investors think it's escaping. Mobile smartphones generate £950 million in royalties—50% of ARM's total royalty revenue. This isn't some slowly declining legacy business; it's the beating heart of ARM's economics.

The smartphone market itself has been shrinking, down 7% from 2020 to 2024. Yet ARM's smartphone royalties have remained remarkably stable, suggesting the company is extracting more value per device even as volumes decline. That's actually impressive execution, but it also highlights just how dependent ARM remains on a mature, declining market.

Meanwhile, the "everywhere" part of the ARM story generates rather modest numbers. Infrastructure pulls in £380 million (20% of royalties), automotive £285 million (15%), and IoT £268 million (14%). These are real businesses, growing nicely, but they're still supporting acts to the smartphone main event.

The Datacenter Mirage

Datacentres represent ARM's most promising diversification story. The company has achieved a 12% market share, shipping 1.7 million ARM-powered servers in 2024. Amazon's Graviton processors have proven ARM's datacenter credentials, and major cloud providers are increasingly designing their own ARM chips.

But here's the rub: ARM generates roughly £225 per server in royalties. Compare that to £0.80 per smartphone, and datacenter silicon appears 281 times more valuable. Impressive, until you realise ARM needs to ship 281 times fewer servers to match smartphone revenue. Those 1.7 million servers generate perhaps £380 million in annual royalties—meaningful, but hardly transformational.

The datacenter opportunity is real and growing rapidly. But from a tiny base. ARM would need to ship over 10 million servers annually to match just its smartphone royalty income. That's aggressive growth from an already strong position, not a given.

RISC-V: The Real Threat Nobody Talks About

While investors obsess over ARM's diversification story, they're missing the real risk: RISC-V. This open-source processor architecture shipped 8 billion cores in 2024, primarily in microcontrollers and IoT devices where ARM traditionally dominated.

RISC-V's impact appears modest today—perhaps £50 million in displaced ARM revenue. But that misses the point. RISC-V is following the classic disruption playbook: start in simple, low-margin applications where "good enough" beats "best in class." Those 8 billion RISC-V cores are teaching thousands of engineers that they don't need ARM for everything.

The smartphone disruption won't happen overnight. Apple and Qualcomm aren't abandoning ARM for their flagship processors anytime soon. But in IoT, embedded systems, and eventually automotive, RISC-V offers customers something ARM never can: freedom from royalty payments.

Valuation Versus Reality

ARM trades at 40.8 times sales while Qualcomm, another smartphone-dependent semiconductor company, trades at just 5.8 times sales. That's a 7x premium for a company with similar smartphone exposure but smaller scale.

The premium presumably reflects ARM's superior diversification prospects and dominant market position. But the numbers suggest ARM's diversification is happening more slowly than the valuation implies. Smartphones still drive 29% of total revenue and 50% of the higher-margin royalty business.

The Narrative Versus The Numbers

ARM is genuinely everywhere—in cars, datacentres, IoT devices, and countless embedded applications. The company's processor designs are excellent, its moat is wide, and its market position is enviable.

But "ARM everywhere" and "ARM beyond smartphones" are different stories. The first is demonstrably true; the second is aspirational. ARM remains a smartphone company that happens to sell into other markets, not a diversified semiconductor platform that happens to do smartphones.

This doesn't make ARM a bad business. The company is executing brilliantly within its smartphone heartland while building promising positions elsewhere. But it does suggest the market is paying for a diversification story that's taking longer to materialise in the numbers.

For a company still generating half its royalties from a shrinking smartphone market, a 40x sales multiple looks rather optimistic. The future may well justify today's valuation, but the present numbers tell a story of smartphone dependency that's proving remarkably persistent.

Stance: Bearish - The "ARM everywhere" narrative doesn't match the smartphone-dependent reality, and the 7x premium to Qualcomm appears unjustified by the current business mix.

Thank you for reading this Numbers & Narrative analysis. This research approach examines current facts and realities about a business, rather than making point forecasts about the future. At Longwalk Research, we know there's no such thing as a crystal ball! For more contrarian perspectives on market narratives, visit our website at https://www.longwalkresearch.com/ where you can explore our full research library and subscribe to receive new analyses as they're published.

Longwalk Research provides independent analysis that questions market assumptions through detailed examination of current operational metrics and competitive realities.