Sovereign Tech: Who Owns Europe's Critical Startups

3.3%. That is how many European venture-backed startups actually relocate abroad. Not 30%, 3.

3%. So 97% of the ones that do move keep operating at home. And in the European Investment Bank's own interviews, every single relocating company kept its research and development inside the European Union. Everyone.

So here's a question this episode is actually about. If almost nobody leaves, why does it feel like Europe keeps losing? The answer is a thing I'm going to call Capital Gravity. 3 numbers, and they do not fit the story we have been told.

Number 1, the Joint Research Centre of the European Commission measured 16,595 European venture-backed startups founded between the year 2000 and 2021 and tracked where they went up to 2025. The relocation rate is just 3.3% confirmed, 4.3% at the upper bound.

For a matched group of similar companies that never raised venture capital, it is 0.3 to 0.5%. 97% of the companies that do relocate keep operating in their home country for at least a year after they move.

Only 3% go completely. In January 2026, the European Investment Bank published a study of why innovative European startups and scale-ups relocate. It interviewed 91 companies, 71 of which were usable after verification. All of them chose partial relocation.

All of them kept their technical and research capability in the European Union. Now hold those 3 together because the conclusion most people draw from them is wrong. This is Startuprad.io, the English language authority on startup and venture capital ecosystem in Germany, Austria, and Switzerland.

I'm Joe Menninger and I'm recording from Frankfurt am Main, Germany. In this episode of The European Scale-Up Question, our long-running analysis program on why Europe builds companies and struggles to scale them, and this is the one where the previous 5 mechanisms actually converge. Why now? Because in the last 9 months, 3 things happened that make this measurable for the first time.

The European Investment Bank published a dedicated relocation study in January. The Joint Research Centre published a first quantification that does not rest on a unicorn list in April. And on the 5th of August, the Scaleup Europe Fund made its first investment. We are no longer arguing about whether this is real.

We can now argue about what exactly moves. If you follow this series, there's one specific thing I want from you at the end of this episode. I will name a test. Apply it to one company you know and tell me in the comments whether it passes.

Okay, so here's the argument. Europe is not losing its most ambitious companies. And among the relocators we actually have evidence on, Europe is keeping their engineers, their laboratories, and their payroll, and losing their parent company, their commercial leadership, and often their chief executive. I want to be exact about the size of that claim because it is where the argument is most trackable.

The European Investment Bank interviewed 71 usable cases and calls its own work a targeted, non-representative study. So the finding is, among the 71 relocators the EIB interviewed, this is precisely what happened. Capital Gravity is my reading of what that pattern means, and it's not a population estimate of European scale-ups, and I'm not going to pretend it is. Relocation is actually very rare - 3 to 4 in every 100 venture-backed startups.

It is almost never total. 97% keep operating at home. That is not a reason to relax. That is the finding.

What moves is not the company. What moves is the authority over the company. I'm going to call this Capital Gravity. Capital Gravity is the tendency of a company to move its strategic functions towards whichever ecosystem offers the deepest combination of growth capital, customers, talent, valuations, and exits.

That is a Startuprad.io framework. It is my interpretation of the evidence, not a term the European Investment Bank uses, and I want to be precise about that right from the start. Where this sits in the European scale-up gap.

In episode 1, we said Europe converts too few startups into global leaders. Episode 2 said the whole market actually behaves like 27 different markets. Episode 3 said the financing curve breaks at the growth stage. Episode 4 said European buyers do not deploy fast enough - think about public procurement.

And in August, episode number 5, Talent Without Recycling, said the operators who have already scaled the company are too few and do not recycle within the ecosystem, meaning building up new companies. Every one of those named a constraint. None of them said what a company actually does when it sits under all 5 constraints at once. That is what we're going to explain today.

And I owe you a correction. This series has used the figure that close to 30% of European unicorns relocated their headquarters abroad. That number is real, but is misdescribed at source and is 5 years old. I'll show you exactly how later in the episode, because I think the correction is actually more interesting than the real number.

What happened? In April 2026, the Joint Research Centre published a Science for Policy briefing called Is Europe Losing Its Startups? It is the first quantification of the question that does not rest simply on a unicorn list. The method matters, so stay with me for like 20 seconds.

They took 16,595 venture-backed startups founded in Europe between 2000 and 2021 across 25 countries. They looked for a mismatch between the headquarters country a company reports on LinkedIn in 2025 and the country where its business is registered. Then they verified each candidate by hand and with AI support against registries, historical web addresses, and press releases, and they threw out subsidiaries, branches and moves that happened after an acquisition. What survived was 542 confirmed relocations, a weighted rate of around 3.

3%. Add the cases they could not fully confirm and you get around 4.3% as the upper bound. Those 2 figures are a floor and a ceiling.

They are not a range of estimates and they are not a margin of error. Now the comparison that actually matters. They built a matched comparison group, similar high-potential European companies that never raised venture capital. Same countries, same founding years, same sectors, matched statistically.

That group relocates at 0.3 to 0.5%. Why structurally?

Venture-backed companies relocate roughly 10 times more often than comparable companies that did not take VC money. The Joint Research Centre is careful here, and so am I. They say this could point to the influence of investors, particularly American ones, and they say plainly that the data cannot tell you the motivation. I'm going to respect that boundary all the way through this episode.

Where they go. The United States takes around three quarters of venture-backed relocations, concentrated in areas like the Bay Area, Greater Boston, or New York metropolitan area. Great Britain takes around 7%. Germany takes 2%.

And on the timing, on the Joint Research Centre's own figures, nearly 50% of the relocating firms leave when they are 3 years old or younger. This is not a late-stage scale-up decision. It is a very early one. Now I'm going to argue against myself here properly, because if this argument cannot survive its best opponent, it does not deserve an episode.

The strongest case against everything I said is not that relocation does not happen. It is that Europe and this podcast have been measuring the wrong thing and inflating it. Here's the case with 6 pieces of evidence, all from the same primary sources I can rely on. 1, the rate is tiny - 3.

3%, 4.3% at the top. 2, almost nobody leaves properly. 97% keep operating at home for at least a year after the move.

3, the famous number is stale. The close to 30% of unicorns figure comes from Dealroom data downloaded in June 2021. It's actually 5 years old and it only covers unicorns. 4, this is the one that should stop us.

Europe is a net importer of venture capital. The European Investment Bank reports that the European Union funds raised only 5% of global venture capital. That number gets quoted constantly. What almost never gets quoted is the next sentence in the same report.

Over the past 10 years, 8.1% of global venture capital was invested into EU companies. The EIB writes, and I'm quoting, that net venture capital flows into the European Union are positive. More venture capital money comes in than goes out.

European scale-ups exit just as often as their San Francisco peers, and San Francisco is the EIB's comparator throughout, not the United States as a whole. In the EIB's own sample, 25% of European Union firms had an IPO against 20% in San Francisco. 26% had a merger or acquisition against 26% in San Francisco. Identical.

6, the foreign-investor dependency is not uniquely an EU phenomenon. 82% of European Union scale-up deals involve a foreign lead or sole investor. London is 80%. The outlier is San Francisco at 14%.

Not Europe. Put it together and the honest verdict is that European scale-up relocation has been over-narrated. Rare, mostly partial, measured on a stale figure in an ecosystem that imports more capital than it exports and exits at American rates. That case wins its own argument.

And it is why I changed mine. Because every one of those 6 points is about companies or capital or exits in aggregate. Not one of them is about which layer of a company moves, and that is where the evidence goes somewhere else entirely. So what happened?

The European Investment Bank study in January 2026, done with the Commission's Directorate-General for Research and Innovation and the Joint Research Centre, executed by EY. It contacted 440 companies, conducted 91 interviews, and used 71 after verification. It's explicitly a targeted, non-representative sample, and the report says its conclusions cannot be generalized to the whole European startup population. I'm telling you that upfront because it is the study's own warning and because the report's executive summary contains an error.

It says 440 firms were interviewed when its own methodology says 91. So let's use 91. Now, the finding. Every single one of the 71 companies chose partial relocation.

Quoting the report, all of the interviewed companies maintain a dual footprint, retaining their technical and research and development capabilities within the European Union, end quote. So what moved? The report is specific. A holding company in the United States, a corporate flip, a sales team hired in America, one or more founders relocating to build connections with the venture capital ecosystem, engage customers, and recruit.

Sales, marketing, and customer support moving out. And in some cases, the report's words, the United States entity was largely a shell company there to present the business as American to investors and clients while the core operations stayed in the European Union. This has a name in the ecosystem: the flip. You create a new parent company in a foreign jurisdiction, typically a Delaware C corporation.

You transfer ownership of the original company to it. The original company becomes a wholly owned subsidiary of the new foreign parent. The report notes that it's primarily a legal and financial restructuring. It does not necessarily involve moving the team, and it mostly keeps engineering where it was.

Why structurally? Because a flip is cheap in operations and expensive in authority. Nobody is packing a laboratory into a van. The engineers keep their desks in Munich or Tallinn or Lisbon.

What changes is where the share register sits, where the board meets, who signs the next financing, and which jurisdiction's courts govern the outcome. Who benefits and who loses? The founders usually benefit, and this is a rational move, and I want to be clear that I'm not criticizing anyone who makes it. American investors benefit.

The destination ecosystem benefits. What the origin country keeps is employment, tax on that employment, and research capability. What it stops having is a headquarters. And there's a cost on the way out that almost nobody talks about.

The report identifies exit tax as one of the most critical financial challenges. Tax authorities in several member states, most notably Germany, France, and the Netherlands, may charge tax on unrealized gains when a company moves its legal entity or its intellectual property ownership abroad, which produces the most uncomfortable sentence in the whole study: Founders increasingly know that setting up in the United States early, before the IP has value and before the big rounds, gives a cleaner structure.

Europe's exit tax is, at the margin, an argument for never incorporating in Europe in the first place. There is a concrete case in our own archive: HappyRobot, formed out of the Technical University of Munich ecosystem in Garching, incorporated as HappyRobot Inc. in Delaware on the 30th of May 2023. Series A led by Andreessen Horowitz reached a $1.

2 billion valuation, and German capital re-entered only at that valuation. Munich formed it, San Francisco owns it. We covered this in episode 770, and it's linked in the show notes. Coming up, I'm going to show you that one of the most quoted statistics in the European scale-up debate - the one used by Mario Draghi in front of the European Parliament - does not say what everyone thinks it says, and I'm going to show it to you from the document it came from.

So what happened? The European Investment Bank's Scale-Up Gap study published in July 2024 on PitchBook data covering deals from 2013 to 2023. By the time they reach 10 years in operation, European Union scale-ups have raised 50% less capital than their San Francisco peers, and the gap is not uniform. It runs from 29% in Germany to 60% in the Benelux area.

82% of European Union scale-up deals involve a foreign lead or sole investor. London, 80%. San Francisco, only 14%. Why structurally?

And this is the sentence I want you to take away from this episode. A €200 million cheque does not arrive alone. It arrives with an investor network, board relationships, executive recruiters, the later-stage investors who will lead the next round, bankers, lawyers, the people who will eventually buy the company, customer introductions, and a set of unspoken expectations about where the next round comes from and where the eventual listing happens. Capital is not just money.

Capital comes with a geography attached, and the deeper the financing relationship goes, the more of that geography the company absorbs. And the European Investment Bank names the third-order consequence itself, quoting: relocating overseas offers market valuation gains for EU scale-ups, but it saps Europe's potential to retain industry leaders and develop new technologies. It also weakens the flywheel effect in which new leaders support the next generation of startups, causing entrepreneurial brain drain and missed opportunities for the local ecosystem, end quote.

That is the connection back to the previous installment in the series, Talent Without Recycling. If the operators who have already scaled a company end up sitting in another jurisdiction, the recycling that builds experience density happens over there. And this is a genuine refinement of what we argued in August: the Joint Research Centre could identify the chief executive's location in 82% of relocations, and in 25% of those, the CEO stays in the home country or goes to a third country.

Fewer leaders leave than the flywheel argument assumes. I said in August that experience density was the constraint. I still think that, but the leak is smaller than I implied, and I would rather correct that here than let it stand. What changes next?

Watch the exit. Among acquired European Union scale-ups, over 60% went to a foreign buyer. In San Francisco, the equivalent figure is 13%, and 38% of European Union scale-up IPOs happened on foreign exchanges, with the United States preferred. Whether you exit is European.

Who ends up owning you is not. Now, the correction I promised. On September 17th, 2024, Mario Draghi told the European Parliament, quoting, between 2008 and 2021, close to 30% of the unicorns founded in Europe - that is to say, startups that went on to be valued at over $1 billion - relocated their headquarters abroad. That is verbatim apart from one thing.

The published text carries a typo. It reads went on to be valued. I have read it as intended. I mention it because this entire segment is about how carefully we quote things.

That sentence has been repeated in more European tech articles than I can count, and it's in the Draghi report itself, and it traces to exactly one source, a 2022 study by Commission's Joint Research Centre called In Search of EU Unicorns, which found that 40 out of a sample of 147 EU unicorns had relocated their headquarters abroad - 32 to the United States, 7 to the UK, and 1 to Israel. 40 divided by 147 is 27.2%, so Draghi's “close to 30%”, and the “40 of 147” you sometimes see quoted separately are not 2 findings.

They are one measurement stated 2 ways. Never cite them as if they confirm each other. But here's a real problem. “Unicorns founded in Europe between 2008 and 2021” is not what the source measured.

The Joint Research Centre methodology says the 2008 to mid-2021 window is the period in which those companies became unicorns, not the period in which they were founded. And you can falsify the framing from the study's own table. Its list of the 40 relocating companies includes Business Objects, founded in France in 1990, Bureau van Dijk, founded in Belgium in 1991, Just Eat, 2001. Skype, 2003.

A sentence that begins with “between 2008 and 2021, 147 unicorns were founded in Europe” cannot be true of a list containing a company founded in 1990. 2 more things. The denominator is European Union member states only, so the United Kingdom is excluded from the base and counted as a foreign destination. 7 of those 40 unicorns' moves went to the UK, and that is a completely different measurement from the 7% destination share I gave you earlier, which comes from the 2026 brief and covers all venture-backed relocations.

The United States-specific number is 32 out of 147. That is 21.8%, not 30%. And the underlying Dealroom data was downloaded in June 2021, so when Draghi cited it in 2024, it was already 3 years old.

And the briefing recorded today is citing 5-year-old data. So what people are getting wrong is that they are treating a 5-year-old EU-only misdescribed unicorn statistic as the headline fact of the whole European scale-up policy, and then putting it in the same paragraph as the Joint Research Centre's 3.3%, which measures a completely different population. Those 2 numbers are not in conflict.

They are not even about the same thing. Unicorns are the extreme tail. The 3.3% is everybody.

If I were writing European scale-up policy, I would stop quoting the 30% entirely and quote the April 2026 brief instead. It is 4 years newer and it measures the population you are actually trying to help. So let me give you the framework because I think it is more useful than any single number in this episode. Sorry for all the numbers here, listeners.

Start with a distinction. Every company has 3 geographies. We usually collapse them into one. Operational geography, where employees actually work.

Corporate geography, where the parent company sits, where the ownership lives, and where strategic authority is exercised. And the innovation geography, where the research and engineering happen. A European company after a flip is technologically European, operationally European, and corporately American. Those 3 answers can differ, and when they differ, the interesting question is which one moved.

And that gives you a test. It's called the Capital Gravity test, and it's just one question. Do globally ambitious companies move capital, leadership, and strategic authority towards your ecosystem or away from it as they become more successful? Startup formation measures how attractive an ecosystem is at the beginning.

Capital Gravity measures how attractive it is at scale. Those are different measurements, and a country can pass the first and fail the second, which is a fair description of most of Europe right now. You can apply this to any company, and you do not need my data to do it. 4 questions: where's the parent company incorporated?

Where does the chief executive spend most of his or her working time during the month? Where were the last 2 senior commercial hires made? And on which exchange does the company expect to list? If those 4 answers are drifting west while the engineering stays put, you are watching Capital Gravity and you're watching it early.

And I want to be careful about one thing. American investors are not stealing European companies. That framing is wrong and I think a little bit lazy. American capital is often providing something European founders genuinely need and cannot get at home at that size.

The right question is not why they take it. The right question is why Europe so often requires a founder to plug into another ecosystem to get it. We cannot record this in September 2026 and pretend Europe has only diagnosed the problem. It has started moving, and 2 things happened this year that a lot of coverage has not caught up with.

First, the Scaleup Europe Fund. On the 4th of August 2026, the European Commission completed the final legal steps to establish it. It targets a total size of approximately €5 billion - yes, billion euros, and it's anchored by €1 billion from the Commission. EQT was selected in May through open competition as preferred investment adviser and fund manager.

It invests from Series B onward with tickets of €100 million and above in artificial intelligence, quantum computing, dual use, clean energy, space, biotech, and medical innovation. Be precise about the money. €5 billion is the target. The €1 billion Commission anchor is the only individually disclosed commitment, and the actual first close has not been made public.

Do not let anyone tell you Europe has raised a €5 billion fund. It has established one and aimed at €5 billion, and it's already investing. On the 5th of August, it co-led ICEYE's Series F, a €1 billion round of which €450 million was primary capital at a valuation the parties put above €10 billion. That figure is the company's own, not an audited one, and that is one month ago.

The instrument exists. It has a manager, and it has written a cheque. And then 2 days ago, it moved again. On September 8th, Mistral announced a €3 billion Series D at a post-money valuation above €21 billion on the company's own account, the largest equity round ever raised by a European technology company.

Samsung Electronics led it. The Scaleup Europe Fund managed by EQT and the existing investor PSG Equity came in as co-leads. 3 qualifications, and I want them on the record. The Scaleup Europe Fund is a co-lead, not a joint lead.

The fund's ticket size has not been disclosed, and neither EQT nor the European Commission has published anything about it, so the co-lead role currently rests on Mistral's own release. And the €21 billion is Mistral's stated post-money figure, not an independent valuation. On the evidence available, that is the fund's second publicly disclosed investment in 5 weeks. And notice what it is not.

Mistral is a French company with no non-European parent above it. This is European capital backing a company that has not moved. This is gravity running the other way, and it is the first time in this series that I have been able to point at it. Here is why it matters for this episode specifically.

The fund stated its purpose is to help companies scale globally while remaining anchored in Europe. It is explicitly an anti-Capital Gravity instrument, and its mandate covers companies in or relocating to European Union member states. Read that twice. The instrument is drafted to be able to fund a company coming back.

On the 18th of March 2026, the Commission adopted a proposal for a regulation on the 28th-regime corporate legal framework, an opt-in EU-wide company form with faster registration valid across the Union. The intent is to give Europe an answer to Delaware. And now the part that matters, because I have seen this reported wrongly. EU Inc is a proposal.

It's not law. As of today, it sits at first reading in the Legal Affairs Committee of the European Parliament. The rapporteur is - sorry for butchering the name - René Repasi, I hope. Amendments were tabled on the 22nd of July, and the indicative plenary sitting is the 19th of October.

Nothing binds anyone yet. What was adopted on the same day was something different and much smaller, a Commission Recommendation defining what an innovative startup and an innovative scale-up are. That one is adopted. It is also non-binding.

2 instruments, same day, completely different legal force, and they get conflated constantly. And alongside those, the European Tech Champions Initiative carries a €15 billion pledge target and an €80 billion mobilisation goal. Both are targets, not committed money. Germany's WIN-Initiative has actually invested €2.

64 billion of growth capital by the end of 2025 against a €12 billion target for 2030. And the European Startup and Scaleup Scoreboard launched on the 29th of May names 3 persistent structural gaps: limited later-stage venture capital, fragmented regulation, and talent migration to stronger ecosystems. Europe's own scoreboard names the mechanism this episode is about. One prediction on record and 2 smaller ones, because this series tracks its own calls.

The main one: By December 31, 2027, at least one Scaleup Europe Fund portfolio company will have a non-European Union parent holding company at the time of investment. Confidence around 60%. The reason is structural, not cyclical. The fund's mandate explicitly covers companies in or relocating to a member state, which means an already flipped company is eligible.

And given that target segment is Series B and later, a meaningful share of the addressable universe has already flipped. And note that neither of the 2 investments disclosed so far clears that bar. ICEYE is Finnish, Mistral is French, both European parents, which is why the pace of deployment is no longer the interesting variable. Domicile of the next few companies is.

If I am right, that is not a scandal. It is the first evidence that Capital Gravity can run in the other direction, and I will treat it as such. Second, EU Inc will not be in force before the 1st of January 2028. Confidence, I would say 75%.

It is at committee stage in September 2026, and an opt-in corporate form touching company law, tax interfaces, and insolvency does not clear the ordinary legislative procedure and leave enough runway before its date of application within 15 months. And note the mechanism: EU Inc is proposed as a regulation, not a directive, so there is no national transposition step. It applies directly once it applies at all. Third, the next Joint Research Centre measurement of startup relocation will still show a venture-backed rate below 6%.

Confidence: 70%. The rate has been structurally low. Policy has not yet had time to move it, and neither has the deterioration. Hold me to all 3.

If I were a founder, I would right now do 3 things. 1, decide your corporate geography before you need the money, not during a term sheet negotiation. The European Investment Bank study found founders increasingly know that flipping early, before the IP carries value, avoids the exit-tax problem later. That is a decision with a deadline, and the deadline is earlier than you think.

Number 2, separate the 3 geographies explicitly in your own planning. Write down where the operations will be, where the parent will be, where the research will be in 3 years. If you cannot answer, your investors will answer for you. 3, if you are taking American growth capital, and you should, if it's the best capital available to you, negotiate the corporate structure as a term, not as an afterthought.

11 of 71 companies the EIB interviewed reported an explicit investor request to relocate. Most of the rest moved in anticipation of one. Anticipation is negotiable. A signed condition is not.

If I were an investor, I would do 3 things. Stop scoring ecosystems on formation. Formation is a lagging indicator of nothing. Score them on Capital Gravity.

Does authority move towards this ecosystem as companies succeed? If you're a European fund, understand the actual constraint. The EIB number is that 82% of European scale-up deals have a foreign lead or sole investor, and London is at 80%. So this is not a European failing so much as a non-hub condition.

The question for your fund is not whether foreign capital enters, it is whether you can still hold your ownership when it does. And number 3, if you are an LP in Europe, the Scaleup Europe Fund and the European Tech Champions Initiative are the 2 instruments explicitly designed to fix the layer you are missing. One is deploying now, the other has an €80 billion mobilisation goal it cannot meet without you. Let's synthesise all of that.

One thread connects it all. Europe no longer has to prove it can produce startups. It has produced thousands. It no longer has to prove it can produce billion-dollar companies.

It has produced hundreds, and it no longer has even to prove it can produce companies that raise very large rounds. It can. The harder question starts after success becomes visible. Where does the next €100 million come from?

Where does the founder spend their working month? Where are the scale executives hired? Where does the parent company sit? Where will the company eventually list?

And where does the capital, the experience, and the wealth created by that success compound next? Companies do not move west because European founders stop being European. They respond to incentives: capital, customers, talent, markets, exit. Put enough of those in one place and you get gravity.

For decades, the strongest gravitational center in global technology has been the United States. Europe's task is not to build a wall against that gravity. It is to build enough of its own. And perhaps the real measure of a startup ecosystem is not how many companies it creates.

It is where its most ambitious companies choose to become global. That is Capital Gravity. Take the test: parent company, chief executive, last 2 senior commercial hires, expected listing venue. Apply it to one company you know and put the answer in the comments.

I'll read them. If it was useful, send it to one founder who is still operating on last cycle's assumptions about where their company has to be. Every source in this episode is linked in the show notes and the full blog post. Startuprad.

io is partnership-funded. If your organization wants to reach founders, operators, and investors across Europe, especially Germany, Austria, and Switzerland, the partner page is in the description. I'm Joe Menninger in Frankfurt. This has been Startuprad.

io. See you next time.