Arm: The Company That Owns Almost Nothing

In 1990, in a converted barn outside Cambridge, England, a small group of engineers set out to design a computer chip. The legend, often repeated, is that there were about a dozen of them, working in a building that had recently housed turkeys. They had a modest assignment: build a low-power processor for a doomed Apple handheld computer called the Newton. The Newton flopped. The chip did not.

That chip’s design, prized for sipping almost no power, would go on to become the most widely used processor architecture in the history of the world. Today it sits inside roughly ninety-nine percent of the world’s smartphones, and inside hundreds of billions of other devices, from cars to data centers to the earbuds in your pocket. And here is the strange and beautiful part: the company that owns that design, Arm, makes almost nothing physical at all. It owns no chip factory. It manufactures no processors. It is, in a sense, a company that owns almost nothing, and yet sits inside almost everything.

This is the story of the purest moat in technology, an empire built not on a factory but on a blueprint, and the opposite, in every way, of the make-everything colossus that is Samsung.

Born to Power a Failure

Arm’s origins trace to Acorn Computers, a British firm whose engineers, including Sophie Wilson and Steve Furber, had designed an elegant, efficient processor architecture in the 1980s. When Apple went looking for a low-power chip for its Newton, the answer was to spin that work into a new joint venture. In 1990, Acorn, Apple, and the manufacturing firm VLSI Technology together created Advanced RISC Machines, soon known simply as Arm.

The Newton, Apple’s stylus-driven handheld, was a commercial failure. But the chip architecture built for it had a quality that turned out to matter enormously: it used very little power. As the mobile-phone industry exploded through the late 1990s and 2000s, battery life became everything, and Arm’s energy-efficient designs became the natural choice. Arm went public in 1998, listing in London and on Nasdaq, and its architecture quietly spread into phone after phone.

The Business Model That Owns Nothing

Here is the genius of Arm, and what makes it the mirror image of Samsung. Arm does not make chips. It designs processor architectures and licenses them, the blueprint and the instruction set, to the companies that actually build the silicon. Apple, Samsung, Qualcomm, Nvidia, Amazon, and hundreds of others license Arm’s designs and either use them directly or build their own custom chips on top of Arm’s architecture. Arm collects a licensing fee up front and then a small royalty on every single chip its partners ship.

Think about what that means. Arm bears none of the staggering cost of building chip factories, the tens of billions that Samsung and TSMC must pour into fabs. It carries no inventory, runs no production lines, suffers none of the brutal capital intensity of manufacturing. It simply owns the design that nearly everyone else must license, and takes a tiny cut of an almost unimaginable flow: more than two hundred and fifty billion Arm-based chips have shipped, across a partner ecosystem of over a thousand companies. It is asset-light dominance, the ownership of a standard rather than a factory.

Decision Point

What Everyone Gets Wrong

Mistake #1: Thinking Arm makes chips. Its designs are in billions of processors. Reality: Arm makes no chips at all. It designs and licenses the architecture, and others, like Apple and Qualcomm, build the actual silicon. Arm owns the blueprint, not the building.

Mistake #2: Believing the asset-light model is weak. Owning no factory sounds like owning nothing. Reality: It is one of the strongest positions in technology. Arm captures a royalty on nearly every smartphone on Earth without bearing the colossal cost and cyclicality of manufacturing. Owning the standard can beat owning the means of production.

Mistake #3: Assuming Arm’s neutrality is timidity. Refusing to compete with customers looks unambitious. Reality: It is the core of the moat. Because Arm never competes with its licensees, fierce rivals like Apple and Samsung can all safely standardize on it. The restraint is precisely what made the architecture universal.

Mistake #4: Reading the failed Nvidia deal as a disaster for Arm. A $40 billion sale collapsed. Reality: Regulators blocked it largely to protect Arm’s neutrality, the very thing that makes it valuable. Arm then relisted publicly in 2023 in the biggest tech IPO of the year, independent and intact.

The Ownership Saga

For most of its life Arm was a quietly successful, independent British company. Then came a turbulent decade of ownership. In 2016, the Japanese investment giant SoftBank, led by Masayoshi Son, bought Arm for about thirty-two billion dollars, the largest acquisition of a European technology company at the time, and took it private as a bet on a future of connected devices.

Then, in 2020, came an even bigger twist. Nvidia, the rising power in chips, agreed to buy Arm from SoftBank for about forty billion dollars, in what would have been the largest semiconductor deal in history. But the proposed deal alarmed the entire industry, precisely because Arm’s value rests on its neutrality. If Nvidia, a chip company, owned the architecture that all the other chip companies depended on, could it really stay neutral? Regulators in the United States, the United Kingdom, and Europe moved to block it, and in February 2022 the deal collapsed. Instead, SoftBank took Arm public again, relisting it on Nasdaq in September 2023 in the largest technology IPO of that year. Arm was independent once more.

The Honest Present

Arm enters its next era from a position of extraordinary strength and one real question. Its architecture is utterly dominant in mobile, and it is now expanding aggressively into the markets that matter most for the future: data centers, automotive, and the chips that power artificial intelligence, where energy efficiency, Arm’s founding advantage, is suddenly the most precious quality in computing. Partnerships with Nvidia, Amazon, Apple, and others have made Arm a foundation of the AI era, and its market value has soared well past a hundred billion dollars since relisting.

The questions are about concentration and ownership. SoftBank still controls the large majority of Arm’s shares, tying the company’s fate partly to one investor’s grand ambitions. A meaningful share of Arm’s revenue has historically come from China, exposing it to geopolitical risk. And the smartphone market that built Arm is mature. But the core position, ownership of the architectural standard inside nearly everything, remains one of the most enviable in all of business.

Why This Matters to Investors

Arm is the purest illustration in this collection of a truth that runs against intuition: that owning the standard can be a deeper moat than owning the factory. Samsung pours tens of billions into fabs to make the physical chips. Arm spends none of that, and collects a royalty on a vast share of the chips made by everyone, Samsung included. The asset-light owner of a ubiquitous standard occupies one of the most durable and profitable positions there is. The Greatest Companies thesis is that durable competitive advantage builds lasting wealth, and Arm’s advantage may be the deepest of all the moats in this book, because it is built on two reinforcing forces: switching costs so high that an entire industry’s software is written for Arm’s architecture, and a neutrality so credible that even bitter rivals trust it. The investor’s lesson is to look past the obvious, capital-heavy players and ask who quietly owns the standard, the toll-road, the blueprint that everyone else must license, because that owner often earns the most durable returns while bearing the least cost. None of this is investment advice; it is a way of reading history.

Lessons in Order of Depth

Method: own the standard, not the factory

Arm’s method is to own the design that everyone must license, and to leave the brutal, capital-intensive work of manufacturing to others. It captures value at the most defensible point in the chain, the architecture, while avoiding the cost and cyclicality of the fab. The method is to identify the layer of an industry where the standard lives, the part that, once adopted, everyone is locked into, and to ask whether owning that layer, rather than the visible, physical product, is the stronger and lighter position.

Money: a small royalty on an enormous flow

Arm collects only a modest fee on each chip, but those chips number in the hundreds of billions, shipped by a thousand partners across nearly every device on Earth. A tiny cut of an almost universal flow is one of the most powerful and least appreciated money machines in business. The trader’s parallel, the same one Visa and BlackRock teach, is that the steady toll-taker on an enormous flow often out-earns and outlasts the player making big, visible, risky bets.

Mind: the power of strategic restraint

Arm’s defining choice was what it refused to do. It never competed with its own customers, never tried to grab the fat margins of building chips or devices, and that restraint is exactly what made it universal. The mindset lesson is that discipline and self-limitation can be a competitive weapon, that the willingness to stay in your lane, to be the trusted neutral party rather than the greedy competitor, can build a position far stronger than aggression ever could.

The deepest question: is it better to own the blueprint or the building?

Arm and Samsung are the two poles of how to win in silicon. Samsung owns the factories, bears the cost, captures many layers, and survives any storm. Arm owns the blueprint, bears almost no cost, captures one thin but universal layer, and depends entirely on others to build. Arm’s returns on capital are extraordinary precisely because it owns so little; Samsung’s resilience is extraordinary precisely because it owns so much. The deepest question Arm poses is whether, in the long run, it is better to own the design that everyone needs or the means to produce it, and the honest answer is that both can be among the greatest businesses in the world, by opposite routes.

The Legendary Scorecard

Founder Vision	8
Innovation	9
Execution	8
Moat	10
Capital Allocation	7
Wealth Creation	8
Durability	9
Historical Importance	9
Overall	8.5

The overall figure is an editorial verdict, not a weighted average. Arm earns a perfect ten for moat, because ownership of the architectural standard inside nearly every phone, defended by enormous switching costs and credible neutrality, is about as deep an advantage as exists in technology. It scores lower on capital allocation, reflecting the turbulence of its ownership saga under SoftBank and the abandoned Nvidia deal, factors that have little to do with the brilliance of the underlying business.

At a Glance

Timeline

• 1990: Arm is founded in Cambridge as a joint venture of Acorn, Apple, and VLSI to build a low-power chip for Apple’s Newton
• 1990s: the energy-efficient Arm architecture becomes the standard for the rising mobile-phone industry
• 1998: Arm goes public on the London Stock Exchange and Nasdaq
• 2007: the iPhone and the smartphone era cement Arm as the architecture inside nearly every phone
• 2016: SoftBank acquires Arm for about $32 billion and takes it private
• 2020: Nvidia agrees to buy Arm for about $40 billion
• 2022: the Nvidia deal collapses under regulatory pressure
• 2023: Arm relists on Nasdaq, the largest tech IPO of the year, and refocuses on AI and data centers

Key Numbers

Related Reading

Arm is one pole of a great contrast in silicon, and its counterparts sit alongside it. Read how Samsung took the opposite path, owning the entire physical stack from memory chip to finished phone. See how Nvidia, which tried and failed to buy Arm, built its own dominance in AI chips on top of, and alongside, Arm’s architecture. And study how Apple, a co-founder of Arm in 1990 and now its most famous licensee, designs the custom chips that made the modern iPhone and Mac. For the full collection, see our Greatest Companies of All Time hub.

Go Deeper

Arm is a study in the asset-light moat, in the power of owning a standard rather than a factory, and in the competitive strength of strategic restraint. Those are the same forces that separate investors who find the durable, low-cost toll-taker from those who chase the flashy, capital-hungry bet, and the book teaches you the discipline to look for where the real, defensible value in a system lives, and to recognize that sometimes the strongest position is the one that owns the least.

This article is part of the Greatest Companies series, adapted from the book Greatest Companies, now available on Kindle.

Frequently Asked Questions

When and why was Arm founded?

Arm was founded in 1990 in Cambridge, England, as a joint venture between Acorn Computers, Apple, and VLSI Technology. Its first task was to build a low-power processor for Apple’s Newton handheld. The Newton failed, but the energy-efficient chip architecture it produced became the foundation of the mobile-computing era.

What does Arm actually do, if it does not make chips?

Arm designs processor architectures and the instruction sets that run on them, and licenses those designs to other companies. Firms like Apple, Qualcomm, Samsung, and Nvidia license Arm’s architecture and build their own chips on it. Arm earns a fee for the license and a small royalty on every chip its partners ship, while owning no factory.

Why is Arm’s architecture so dominant?

Arm’s designs are extremely energy-efficient, which made them ideal for battery-powered phones just as the mobile era exploded. Over decades, an enormous amount of the world’s software was written for Arm’s architecture, creating huge switching costs, and Arm’s refusal to compete with its own licensees made it safe for rivals to adopt. Today it is in roughly 99 percent of smartphones.

Why did Nvidia’s attempt to buy Arm fail?

In 2020 Nvidia agreed to buy Arm from SoftBank for about $40 billion. Regulators in the US, UK, and EU moved to block it, largely because Arm’s value depends on being neutral, and a chip company owning the architecture that all its rivals license raised serious competition concerns. The deal collapsed in February 2022, and Arm relisted publicly in 2023.

Who owns Arm now?

After the Nvidia deal failed, SoftBank took Arm public again on Nasdaq in September 2023, in the largest technology IPO of that year. SoftBank still owns the large majority of Arm’s shares, so while Arm is publicly traded, it remains effectively controlled by SoftBank.