ASML Will Die

ASML Holding N.V. is the most valuable company the European Union has ever produced. The prediction is eclipse, not bankruptcy, and the Union’s own structure has already settled it.

The most valuable company the European Union has ever built will slide down the world rankings until it stops mattering. That much I believe firmly. The timing matters less to me than the cause, and the cause is the whole argument. ASML will not be brought down by bad engineering, a soft quarter, or some stroke of Chinese genius. It will go down because of where it lives.

One fact frames everything that follows.

The European Union has 450 million people. It has the deepest scientific and industrial lineage on the planet: the continent that produced the steam engine, the periodic table, quantum mechanics, the first programmable computer. World-class universities, an educated workforce, a single market the size of America’s. On paper this is the recipe for an economic superpower, and for decades the people building the Union told themselves it would be one.

Europe’s most valuable company is a Dutch maker of chip-printing machines, and it is a fraction of the size of the firms at the very top, perhaps a tenth of the largest American name. Take the top 50 companies in the world by total company value. Only ONE is in the EU. The summit of the table is American and Taiwanese. Europe is barely present on it.

This is not a bad year. The arrow has pointed the wrong way for decades, across boom and bust, through every colour of government and every fashion in economic policy. The EU hired Enrico Letta to diagnose the problem, then Mario Draghi, whose report ran to the better part of 400 pages and called the situation existential. Draghi’s own verdict is the useful part: the inputs are all there, he said, and Europe is simply failing to convert them into companies that can win.

That verdict eliminates the easy answers. If the man the establishment hired to find the problem rules out talent, and the decades rule out the business cycle, and the parade of governments rules out any single policy, then you are no longer looking at a mistake. You are looking at the design. Something in the structure of the thing is doing this.

What agency is, and why Europe hasn’t got it

When we call a state powerful we usually mean rich, or large, or advanced. Europe is all three. But power and agency are not the same, and the gap between them is the gap this whole essay lives in. Agency is the capacity to form an intention and force it on the world, including on people who would prefer you didn’t. It comes apart into three pieces.

The first is autonomy: the freedom to act without another power’s leave. France has it. It owns a nuclear deterrent it controls entirely alone, and in 2003 it looked Washington in the eye and refused to join the invasion of Iraq. That is what saying no to the hegemon, and meaning it, looks like.

The second is the willingness to defect: the stomach to step outside the rules when the rules have become a cage. Every serious power keeps this in reserve, the option to break the agreement, run the operation, do the thing that is not allowed. A state that has sworn off rule-breaking for good has handed a permanent lever to every state that hasn’t.

The third is unitary will: one actor that can decide and then act, quietly if it must, without routing the decision through forty committees and twenty-seven national vetoes first. A coalition that has to negotiate with itself before it can move does not have a will. It has a standing meeting.

The European Union is weak on all three at once. That was the founding choice, not a phase or an oversight.

The bet

After 1945, Europe looked at what unchained state agency had just done to it, twice in thirty years, and made a wager. The wager was that you could be powerful without being agentic: that rules, trade, markets and shared standards could stand in for armies, spies and the willingness to break a promise. Given the bodies in the ground, this was not a foolish bet. It may be the most morally serious decision any group of states has ever taken.

It pooled, on purpose, only the instruments that work through rules and markets: trade policy, competition law, the single market, the common currency, the great machinery of regulation. These are the tools of a referee. It refused, just as deliberately, the instruments that work through force and secrecy. No European army. No European intelligence service with the authority to run an operation. No shared appetite for defection. That refusal was a principle, not an unfinished to-do list. The Union did not run out of time to build a spy service with teeth; it declined to build one, because declining is what the project is.

The result is something new in the world: a regulatory superpower bolted onto a strategic dependency. Europe can dictate how the entire planet handles data privacy or chemical safety, the famous Brussels Effect, and in the same breath cannot defend its own companies without American permission.

A referee writes the rules and enforces them; a referee does not win the tournament; a referee who picks up the ball and starts scoring has stopped being a referee. The Brussels Effect broadcasts the rule to the world. It cannot broadcast a European winner, because producing a winner is the one thing a referee cannot do.

For fifty years none of this hurt, because the agency Europe declined to grow was being supplied for free by the American security umbrella. Europe got to be the referee because the United States agreed to be the muscle. The bet was only ever coherent under a benevolent protection that, for a long time, Europe did not have to think about. It is thinking about it now.

Why importance is a target

Any firm that becomes truly important, a chokepoint, something foundational to major industries in other countries, stops being a private commercial concern. It becomes an instrument of statecraft whether it likes it or not, something to be wielded and something to be aimed at. When this happens to an American company, the American state does two things at once. It conscripts the firm: enforce these sanctions, obey these export controls, deny your product to our rivals. And it covers the firm: diplomatic protection, retaliation on its behalf, and, when its own policy is what did the damage, a reflex of writing cheques. When Washington blocked grain sales to the Soviets in 1980 it cut its farmers a couple of billion dollars in compensation. When the trade war cost soybean growers their Chinese market it sent them tens of billions more. Obligation and protection, in the same hand.

Now run the same logic inside a polity with no agency to wield.

ASML makes the extreme-ultraviolet lithography machines that makers of all advanced computer and AI chips depend on. There is no second supplier. That makes it, by a reasonable argument, the single most geopolitically important company alive, the literal chokepoint of the technology war, and every great power would like to control what it does.

Washington conscripted ASML. It leaned on The Hague to stop ASML selling its most advanced machines to China, ASML’s largest growth market, in the service of American strategy and American ends. The Dutch government complied and enforced the block against its own crown jewel.

The second half did not happen, and structurally could not. Nobody covered ASML. The Netherlands has no power to refuse Washington, no power to trade compliance for protection, no power to retaliate if its firm is harmed. The Union has none of the extraterritorial weapons the Americans used, and has formally sworn them off. There is no European department standing by with a cheque for the market ASML was ordered to walk away from. It absorbed an act of American statecraft, paid for American strategy, and was shielded by no one: not the Dutch, not the EU, not the America whose policy it was carrying out. The Americans use it, the Dutch tax and regulate it, and nobody stands behind it, so it carries the most geopolitical weight a company can carry under the polity that cannot or will not defend it.

The two ways Europe kills its own

The Union endangers its own firms in two distinct ways, and they are different machines.

The first danger is the one the Union manufactures itself. Its own rules and policies raise the cost of operating, or simply legislate an industry out from under its feet. No foreign hand is needed. Sometimes this is diffuse: Europe’s energy policy has produced some of the highest industrial power prices in the world, and for anyone making steel or cars or running a data centre, where energy is a fat slice of the bill, that is a quiet tax on existing at all. Sometimes it is sharp and specific: a single regulation arrives and removes an industry’s right to operate overnight.

The sharp version has a body count. MiCA, the Union’s flagship crypto regulation, set a hard deadline of 1 July 2026, by which any firm serving EU clients needs a full and expensive licence. A pre-deadline population of well over a thousand registered firms is being compressed into a couple of hundred authorised survivors, of which only around fourteen can run a full trading platform. Roughly six in ten European crypto users are sitting on services that will simply be switched off.

One of the firms going dark is VNX, a Liechtenstein issuer of a euro token and a gold-backed token, suspending its exchange on 30 June, one day before the deadline. VNX is worth a second of your time. Not for its size, but because it is the opposite of the cowboy story you would expect: scrupulously, proactively compliant, audited, regulated under its national regime, the model of a good citizen. It is being sliced anyway. And the value it vacates does not stay in Europe. The clearest single winner of Europe’s own crypto law is Circle, a New York-listed American company, whose euro stablecoin has gone from a sliver of the market to something near half of it as the non-compliant are delisted. Europe wrote the rule, swept its own pitch clear, and handed the prize to an American firm. Nobody in Washington lifted a finger.

This has happened before, more than once. The data-privacy regime concentrated the advertising market around the largest American platforms and is associated with a drop in European startup funding. The big financial-markets overhaul of 2018 hollowed out small brokers and thinned the research coverage of smaller listed companies. The chemicals regime priced smaller firms out of whole product lines. The post-crisis bank rules pushed the sector towards fewer, bigger institutions. The pattern is always the same: high fixed costs of compliance that the giants can absorb and the minnows cannot, and a field that ends up belonging to whoever already had scale, very often a bank. Whether you read this as negligence or as capture barely matters, and you do not have to prove either; the track record is the argument.

The second danger is the one we have already met with ASML: a foreign power imposes its will, and the home state cannot or will not defend the firm. That inability is not a separate failure tacked on; it is the whole of what makes this danger distinct. A foreign demand a strong state could swat away threatens no one. A foreign demand the home state has to obey is the entire problem. An American company cannot suffer this version at all, because the same Washington that conscripts it is the Washington that covers it. A European company suffers it precisely because there is no European hand to do the covering.

So you get two clean cases, one for each machine. VNX shows the EU destroying its own firm’s value with no foreigner in the room. ASML shows the EU unable to protect its firm’s value when a foreigner walks in. The VNX case is, if anything, the more damning of the two, because there is no hegemon to blame. Take the foreign power out of the picture entirely and the value still ends up abroad. That is as close to a controlled experiment as politics offers, and it points at the structure rather than at bad luck with a powerful neighbour.

Why money would not save it

The obvious objection is that this is overcooked. ASML is rich, superbly run, technologically untouchable; great engineering and deep pockets will carry it through. The objection has the mechanism backwards.

Run the experiment in your head. Take a European strategic company and give it everything the conventional wisdom says Europe lacks: a bottomless balance sheet, deep and liquid home capital markets, all the growth equity it could ever swallow. It still dies, because neither thing that kills it travels through the balance sheet.

The home-grown danger does not bill your bank account; it razes the ground you stand on. A war-chest does not exempt a crypto issuer from a ruling that cancels its licence, and it does not lower the energy price that sets a heavy manufacturer’s floor. Money can fund a fight. It cannot repeal the rule you are fighting. ASML, for what it is worth, survives this danger not because it is well funded but because its margins are freakish, and margin and capital are not the same thing; a fully funded company on thin margins bleeds out just the same.

The foreign danger is worse for the money thesis, because no balance sheet buys cover. Cash cannot make the Netherlands able to refuse Washington. It cannot conjure a European body to retaliate on your behalf. It cannot write you the cheque for the abandoned market, because there is no European hand with the authority to sign it. Cover is a property of the polity, not the firm, and you cannot buy it from inside a company however rich you are.

Which turns the whole conventional diagnosis on its head. The standard story is that the EU underperforms because its capital markets are too shallow, and the standard cure is to deepen them. But a polity this unwelcoming to business, offering this little protection, was never going to grow deep pools of patient, risk-hungry capital in the first place. Money pools where it is rewarded, and the EU is not a place that rewards it. The shallow market is downstream of the hostility, and building a grand Capital Markets Union would be irrigating a field that the climate still kills. VNX makes the same point from the far side: it was diligent, it was compliant, and diligence saved it from nothing. The missing ingredient was never money or care. It was a home government willing to back it.

The call, and the law underneath it

So, the concrete prediction: ASML gets relegated as the China premium bleeds out of the stock. The mechanism is one of the oldest results in economics, and almost funny in its reliability: wall a large enough market off and you don’t kill it; you hand it an overwhelming reason to build its own supply. Export controls are the finest subsidy for import substitution ever devised. China does not need to match ASML to hurt it; it needs only a domestic alternative its own market can live with, and the day it has one, two things leave ASML’s valuation together: the scarcity premium of being the only game in town, and the Chinese growth that was priced in. You do not have to beat a monopoly to re-rate it. You only have to give it a rival for the largest market on earth.

We have watched this exact film with the roles swapped. When Washington barred its own champions from selling advanced chips to Huawei, it did not cripple China; it forced China to build, at speed, an industry it might otherwise never have bothered with, and Huawei has all but thanked it for the favour. American firms lost a fortune in sales and helped midwife their own competitor. Then comes the asymmetry. When those controls bit the American firms too hard, the American state could loosen them, and did: it granted licences, carved out exceptions, and in 2025 reopened the China market to its own champions outright. ASML has the identical wound and no such relief, because the rule that cut it off is not its home state’s to repeal. The Hague enforces a control authored in Washington; it cannot dial back what it did not write.

One distinction turns a single obituary into a law. Look at ASML as it stands today and it is a victim of the foreign danger only; it has already eaten the home-grown danger and survived, because its margins let it swallow European energy and tax and regulation that would have drowned a thinner firm. That survival is the giveaway. The home-grown danger is a thinning machine, and it is much of why a continent with every advantage fields so few firms near the top at all. The scarce European names in the global rankings are its survivors. ASML is the greatest survivor, and the foreign danger has come for the survivor.

Now imagine the next one. Imagine a founder building tomorrow’s chokepoint, the next thing the great powers will fight to own. They cannot assume ASML’s margins, the kind thinner businesses can only envy, so they cannot assume they will survive the home-grown danger the way ASML did. And if they somehow do survive it, surviving only qualifies them for the foreign one, with no one behind them when the powers come to collect. For that founder the exposure is genuinely double: clear the first danger to reach scale, then find that clearing it was merely the entry fee for the second. ASML is the alarm; the next founder, and the next funding round, are who it is for.

The part with no exit

You might think the answer is obvious: integrate more. Build the Capital Markets Union, pool the budgets, finish the project. This is roughly Draghi’s prescription, and it is a category error.

There are two different things people call integration. The first is economic: deeper capital markets, joint borrowing, a banking union. That addresses the shallow-capital symptom we have already set aside, it does nothing to the machines that kill the firm, and even on its own terms it would not hand Europe autonomy, or a willingness to defect, or a single will: it’s the wrong layer entirely.

The second kind of integration is the one that would supply agency: a real army, a spy service with operational teeth, a shared readiness to break the rules when it suits, an executive that can act without twenty-seven vetoes in the way. That would do it. But it is precisely the integration the founders refused on principle, and a Europe that built it would have had to abandon so much of what it was founded to be that it would no longer be the European Union. The renunciation of force and secrecy and defection is not a hole in the project. It is the project. An entity that thinks and acts with France’s unitary nerve is not a more finished EU. It is a different animal wearing the name.

That is the trap. The thing that would save the strategic firm is the one thing the Union cannot become without ceasing to exist. So “just integrate more” is either beside the point or self-abolishing, and there is no remedy on the shelf that does not dissolve the patient who needs it. I am not providing an answer. I am explaining the disease. I don’t know what would cure it.

Which is why this is a cautionary tale and not just a stock call. The lesson generalises past one company. Importance inside an agency-less polity is not an asset; it is a liability, a target your own success paints on your back. The better a strategic company does in Europe, the more dangerous its position becomes, and the more certain it is that no one will be standing behind it at the end. A founder who understands this does not build there. Not because the engineers are missing; they are the best in the world. Because the value created there gets coerced away or quietly captured elsewhere, while the continent keeps the laboratory and the tax receipts and loses the trillion-dollar company and the seat at the table.

So the next company gets built in Europe and listed in New York. The talent stays where it trained; the value is captured elsewhere; the cover was never on offer. ASML at number 20 is the EU’s most valuable name, on a list it was built to lead and is starting to fall off.