Monopoly After Software
Software is commoditising and geopolitics is tightening. The durable companies of the next cycle will be capital-heavy, regulated, and built to work with the state.
Three days ago Anthropic blocked access to its most powerful AI models for all but U.S. citizens, citing a U.S. government export-control directive. It’s a sign of geopolitical reality and the increasing importance of powerful technology in global power politics.
The deeper signal isn’t about Anthropic; it’s about us. Europe runs on technology it doesn’t own. And a continent dependent on others makes for a weak partner.
The way the U.S. is pushing Europe to build its own capabilities is far from ideal, but if we hope to have a strong, unified West, it requires a strong Europe that’s an equal partner to the U.S., and technology is increasingly the lever that matters.
Software Commoditisation
At the same time as technology is increasingly becoming the determining issue for geopolitical power, software is getting commoditised and losing its moat beyond momentum. It seems the only strategy for the fast-growing AI application-layer companies is to run faster than the other guy. There will still be great new software companies, but the strategic value will increasingly accrue to the hardware layer, where the moats are built.
The History of Monopoly
The first great industrial monopoly was Standard Oil. Built in the 1870s by Rockefeller, it achieved dominance through the market rather than direct state control. That dominance came through horizontal integration (absorbing or crushing rival refiners), vertical integration (controlling pipelines, barrels, distribution), and leverage like secret railroad rebates rather than a royal charter or a government grant. Capital intensity was necessary but not sufficient. Railroads, built at roughly the same time, were even more capital-intensive and had monopolistic features, but they destroyed a lot of capital. It’s important to see which is which: capital intensity without a chokepoint destroys capital rather than protecting it.
With the emergence of software, the chokepoint moved from controlling physical resources to network effects and data. Pure software is a capital-light, zero-marginal-cost business. It’s easy to capitalise and tends to create fast profits compared to earlier monopolies. Everyone can think of Google, Facebook, Amazon, Microsoft and the like.
Now we’re seeing the market shift again. For the moment, the pendulum is swinging back.
The New Monopoly Shape
Global geopolitical reality and software commoditisation are creating the conditions for a new shape of company that will win over the next few decades. These companies will have larger capital requirements from the start, and the capex becomes a moat for those who can master it. They operate in the most important areas of our lives, like space, defence, manufacturing, health and energy, that tend to be regulated, have longer timelines to liquidity, and need to interface with the government as a regulator, a buyer, or both.
Consider the three largest IPOs happening this year: SpaceX, Anthropic and OpenAI. SpaceX has the clearest monopoly shape. It owns the critical chokepoint, which is launch capacity, and could extend that moat to Starlink, data centres in space and beyond. SpaceX is a capital-intensive business, but less so than you’d think after the initial investment, because its customers pre-finance it. Anthropic and OpenAI are less clear-cut. They are clearly beneficial for the world, just as railroads were, but it’s not clear they’re great investments. Right now the surplus flows to users (through collapsing prices) and to Nvidia (88% margins). Yet, because of the nature of the innovation, there is a chance one of them reaches escape velocity towards AGI and becomes the best investment in the history of the world. They are uniquely world-historical, and hence for many worth the risk. What all three share is the shape: heavy capex that turns into a moat, regulation due to the nature of the technology, longer timelines than software, and government as a regulator, a buyer, or both.
What This Means for Europe
During the past 20 years, most European entrepreneurs and investors have learned how to build and invest in software. Now we need to learn how to build and invest in companies that are comparatively capital-heavy, hardware-defined companies and operate in regulated markets, if we want Europe to prosper and become an equal to the U.S. Being capex-heavy and facing regulatory friction is not the point in itself, but many of the largest companies will have those features.
ASML is the best example we have. The company makes photolithography systems, the tools that print circuit patterns onto silicon wafers. It’s a single chokepoint for the largest technology build-out in history, and it follows the shape laid out above.
Has ASML been a good investment? Yes, I’d say so. ASML is one of the great compounders. Since its 1995 IPO it’s up roughly 142,000%, about 1,400x. Over the past decade the total return is around 1,500%, an annualised rate near 32%. Over its full history it has compounded at a pace that doubled investors’ money roughly every two years on average. Today it trades around $1,870 a share at about a $720 billion market cap, one of the most valuable companies in Europe. But it’s been a volatile one, because it sits at a critical chokepoint in the global supply chain for chips and geopolitics moves the stock. So if you invested, you needed to be prepared to hold for a long time.
I believe we can build more companies like ASML, and even SpaceX, here in Europe. We have historically had a strong industrial base, which is not as strong as before, but there’s still plenty of talent if we’re ambitious enough, decide to invest the capital, and align regulation accordingly.
Germany, Europe’s economic engine, has a deep history of building industrial companies, and just recently, here in Finland where I live, Oura and ICEYE have both crossed the €10 billion mark to become the country’s latest success stories.
These are hard companies to build, especially in Europe. But they are not impossible. If we optimise our efforts and capital for safe and familiar, we’ll never build the companies we need. Fortunately, it is exactly these companies that become the largest economic opportunities when they succeed.
The Stakes Are High
There’s no less at stake than the way of life we have built over the past 200 years in the West and everything we hold dear.
Both Europe and the US are weaker going alone. The alliance between Europe and the US that has been built over the past 70 years has made both stronger and ensured our prosperity and freedom.
The goal should be a strong Western alliance between Europe and the U.S., as well as with our other partners, instead of full sovereignty on everything. Ironically, building our own technology monopolies so we’re not reliant on others may be our best way to get to that alliance. So better get back to building hard things.