The Ownership Test: Germany Builds the Winners. Who Ends Up Owning Them?
What Is This About?
This is the Q3 2026 review of the startup ecosystems of Germany, Austria, and Switzerland from Startuprad.io, covering July 1 through September 30, 2026. Every quarter, we score our previous predictions in public, distill the quarter's verified signals, and name the structural thesis. This quarter's thesis: Germany passed the Strategic Necessity Test — and Q3 posed the harder question we call the Ownership Test.
The Video Goes Live on Thursday, October 1st, 2026
The Audio Podcast
The audio podcast goes live earlier the same day.
Subscribe to our podcasts here, or on Spotify and Apple Podcasts.
Executive Summary
Q3 2026 was the quarter the structural rotation got priced. Helsing raised $1.8 billion at an $18 billion valuation on July 13; Quantum Systems raised $1.2 billion on July 2 — 3 billion dollars for two Munich defence companies in 11 days. Proxima Fusion closed €411 million at a €2.4 billion valuation with Google and RWE as strategic investors. And on September 5, Isar Aerospace's Spectrum reached orbit on its second flight — the first commercial space company from Europe to deliver satellites into orbit. At the same time, participation kept narrowing: EY counted 354 German financing rounds in the first half, down 11 percent, with 67 percent of capital in rounds above €50 million, and the Business Angels Club Berlin-Brandenburg reported Berlin pre-seed funding down 65 percent since 2022. Then the quarter's biggest outcomes raised the ownership question: Uber's $14.8 billion offer for Delivery Hero, Eli Lilly's agreement with AtaiBeckley, NVIDIA's roughly $12.9 billion acquisition of Hugging Face, and Cohere's definitive agreement with Aleph Alpha — while, according to KfW, more than three quarters of Q1 venture money into German startups came from abroad. Necessity attracts capital. Ownership is decided by infrastructure.
The Prediction Card
Accountability first. The H1 2026 review put five predictions on record with deadlines and confidence levels. Interim status as of this review:
Prediction 1 — KNDS formally restarts its IPO process by March 31, 2027 (confidence 7/10): live and volatile. KNDS postponed the listing on July 1. Bloomberg reported restarted investor meetings on August 26 — and on September 16, reported the owners were considering delaying again, possibly beyond this year. Process activity, yes; a formal restart, not yet.
Prediction 2 — at least two German strategic-sector companies announce priced equity rounds above €1 billion by December 31, 2026 (confidence 8/10): confirmed in 11 days. Quantum Systems ($1.2 billion, July 2 — just over €1.05 billion at that day's exchange rate) and Helsing ($1.8 billion, July 13).
Prediction 3 — SAP announces at least one further material AI acquisition beyond Prior Labs, n8n and Parloa by December 31, 2026 (confidence 6/10): open. SAP completed Prior Labs (closed July 16, announced July 17) and Dremio (July 6) inside Q3 — but both were announced in May, before the prediction existed. Completions are not new announcements.
Prediction 4 — DeepL announces a material strategic transaction, financing, or restructuring by June 30, 2027 (confidence 6/10): open. DeepL's roughly 25 percent workforce reduction was announced May 7 — the right direction, but a pre-existing signal cannot confirm a prediction.
Prediction 5 — a Profitability Cohort member (N26, Personio, Flink, FINN) announces an IPO mandate, filing, or formal sale process by December 31, 2026 (confidence 5/10): open. No mandate, no filing. Scalable Capital's €30 billion in client assets and Altersvorsorgedepot pre-marketing is adjacent — and exactly the speculation the prediction excludes.
Interim scorecard: one confirmed at the highest confidence rating, none missed, four on the clock. And one more from the monthly episodes: in June we predicted Isar Aerospace reaches orbit before December 31, 2026. Confirmed September 5 — with almost four months to spare.
The Rotation, Priced
In the Q1 review we said Germany's startup market had become a selection event. In the H1 review we formalised the selection rule as the Strategic Necessity Test: capital concentrates on companies that governments, industries, and supply chains need to exist. Q3 is the quarter that test got priced.
Helsing raised $1.8 billion at $18 billion in a round led by Dragoneer, with Lightspeed, ICONIQ, Growth Equity at Goldman Sachs Alternatives, JPMorganChase, CPP Investments, General Catalyst, Plural, and StepStone. Quantum Systems raised $1.2 billion at approximately $8 billion post-money, co-led by Blackstone, Noteus, Airbus, and Advent — and co-CEO Florian Seibel told the Financial Times that investors uncomfortable with a move into lethal systems had the opportunity to exit in the transaction. Proxima Fusion closed €411 million at €2.4 billion — what the company calls the largest European fusion financing to date — and RWE, which had already signed a February agreement to work toward a stellarator plant at the former Gundremmingen nuclear site, invested. The customer invested; that compresses the path from lab to grid.
Then September 5: Isar Aerospace reached orbit on Spectrum's second flight and deployed its payloads — the first commercial company from Europe to deliver satellites into orbit, and the first provider to hit the European Launcher Challenge orbital milestone ahead of the 2027 deadline, while Germany's government backs a reported €35 billion space build-out by 2030.
Note what the quarter's lead investors have in common: Dragoneer, Blackstone, Goldman Sachs, Google. Global institutional allocators, not European venture funds. Hold that thought — it is not as simple as it looks.
The Participation Collapse
The other side of the ledger. The August episode named it the concentration economy; at quarter scale it is a participation collapse. EY's half-year data: €5.3 billion invested in German startups, up 14 percent — across 354 rounds, down 11 percent, with 67 percent of capital in rounds above €50 million. Against the first half of 2021 (588 rounds), that is roughly 40 percent fewer financings with more money. According to the BACB's position paper — an advocacy document, but with concrete numbers — Berlin pre-seed funding is down 65 percent since 2022, back to roughly 2017 levels; the median angel ticket fell from €50,000 to €38,000; startup insolvencies nearly doubled since 2022 even as formations hit a record 3,053 in the first half, per the Startup-Verband. And in Bitkom's survey of 102 German tech startups — small and not representative, but directionally striking — 27 percent had forgone hires because of AI, 7 percent had cut jobs, 16 percent had added staff, and the average startup employs 12 people, down from 13.
Now look at the software companies that still won, and what they have in common. Camunda: annual recurring revenue nearing $200 million — the first $100 million took roughly ten years, the second is expected in about two — without a new venture round since 2021, valued above $1 billion since 2025. Pliant: past $100 million in annual recurring revenue. Moss: a unicorn at a €1 billion valuation on a Series C of just €35 million. In the H1 review we noted software was absent from the top funding tier; in Q3, the software story is not about round size at all. It is about proof. The casualty list from the same system: Sono Motors (insolvency proceedings opened July 31), Glow25 (roughly 120 job cuts, with AI-enabled automation named as a driver), Enpal (closing its Hamburg site while guiding to over €1.3 billion of revenue). Selection is not a phase of this market. It is the market.
The Ownership Test
The anchor statistic, from the September episode: according to KfW, more than three quarters of the venture money that went into German startups in the first quarter of 2026 came from abroad.
The outward-pointing evidence. Uber announced a voluntary takeover offer for Delivery Hero on July 16 — €41.50 per share, an implied equity value of $14.8 billion for 100 percent of the company ($13.7 billion adjusted for Uber's existing stake); announced, not closed, with the acceptance period running to November 5 and closing expected in the second half of 2027. AtaiBeckley signed a merger agreement with Eli Lilly dated July 15 — roughly $2.8 billion upfront plus up to $1 billion in milestones; announced, not closed. On September 16, Cohere and Aleph Alpha signed a definitive business-combination agreement — signed, not closed, but the flag on Germany's flagship AI lab is changing. NVIDIA announced the acquisition of Hugging Face on September 3 at approximately $12.9 billion. And HappyRobot, the $1.2 billion unicorn: formed in Munich, owned in San Francisco.
The counter-evidence is real. Helsing states it remained predominantly European-owned after its round — foreign lead capital is not the same thing as foreign ownership. Langdock reportedly re-domiciled its parent to Germany as an SE, the opposite of the outbound trend. SAP acted as a domestic acquirer twice and committed more than €1 billion over four years to build Prior Labs into a frontier AI lab in Europe. And September delivered the cleanest counter-pattern of the quarter in 48 hours: The Exploration Company raised $450 million on September 8, and on September 10 the European Space Agency signed it for the ALADDIN programme, worth up to €760 million including optional service missions — capital and customers back to back, with ownership staying home. HyImpulse raised more than €50 million on September 2, and on September 17 the German government and KfW started the next phase of the WIN initiative to mobilise domestic scale-up capital.
So here is the framework. The Strategic Necessity Test asks: does this company need to exist for governments, industries, and supply chains? Germany now passes it — the capital proves it. The Ownership Test asks the next question: when a strategically necessary company succeeds, who ends up owning it? Q3 did not settle that test; it exposed it. The quarter's biggest pending exits point outward, while its most strategic growth companies stay European-controlled on global capital. Necessity attracts capital. Ownership is decided by infrastructure — a domestic LP base at pension scale, a public market where a KNDS can actually price, and procurement as anchor customer, the way ESA just demonstrated. This connects directly to what Germany's Startup and Scaleup Strategy calls scaleup sovereignty: a country secures technological sovereignty by retaining ownership as companies scale — not by counting incorporations.
Operator and Investor Takeaways
For founders: the lead investors in this quarter's mega-rounds were American and global institutions — if you build in a strategic-necessity category, plan your Series B and beyond with that map, and decide before the term sheet what ownership outcome you are building toward. If you build software, the quarter's message is proof over narrative: Camunda, Pliant, and Moss mattered for revenue, not round size. And audit your hiring plan against AI capability — in Bitkom's survey, half of startups already changed their staffing because of it, in one direction or the other.
For investors — observations, not recommendations: the interesting ends of the market are the two the mega-rounds ignore. At the bottom, if the BACB's numbers hold, pre-seed competition in Berlin is the thinnest in a decade for whoever still writes angel tickets; High-Tech Gründerfonds launched fundraising for its fifth seed fund on August 26, but public capital alone will not refill that base. At the top, assets coming out of restructurings — Sono's solar business and IP is up for sale — may be priced very differently from their long-term technical value. And watch the Ownership Test as a return driver: if the Delivery Hero and AtaiBeckley transactions close in 2027, a significant wave of liquidity reaches German founders, employees, and angels. If.
If your company sells to founders, investors, or corporate innovators in Germany, Austria, and Switzerland, this analysis is the environment your buyers operate in — that is what a Startuprad.io growth partnership is built for.
Predictions for Q4 2026
Three new predictions, on the record, each with a deadline and a confidence level. One: KNDS does not complete its listing in 2026 — the IPO slips into 2027 (confidence 6/10). Two: Germany ends 2026 with at least 12 new unicorns for the year, up from 10 by late August (confidence 7/10). Three: at least one more DACH company valued above $1 billion agrees to an acquisition by a non-European buyer before December 31 (confidence 7/10). The four predictions still on the clock from the H1 card stay live; the annual review settles them all.
Conclusion
Q1 asked which companies survive: the selection event. H1 asked which companies attract capital: the Strategic Necessity Test. Q3 asks who ends up owning the winners: the Ownership Test — posed, not yet decided. The fourth question, for Q4 and beyond, is participation: whether an ecosystem of 354 rounds, collapsing pre-seed, and 12-person startups keeps producing companies worth owning at all. This review covers July 1 through September 30, 2026; content was finalised on September 28, and developments between then and September 30 are covered in the October news episode. The annual review goes live in early January and will score every prediction on this card.
Sources
Primary sources and key coverage for this review, as plain links:
Helsing Series E ($1.8B at $18B; “predominantly European-owned” — company statement): https://helsing.ai/newsroom/helsing-raises-1-8bn-in-series-e
Isar Aerospace reaches orbit, Sept 5 (Spectrum flight 2, Andøya): https://www.startbase.com/news/isar-aerospace-erreicht-beim-zweiten-flug-den-orbit-und-setzt-satelliten-aus/ · https://www.aerospacetestinginternational.com/news/isar-aerospace-delivers-satellites-to-orbit-on-second-flight.html · https://www.munich-startup.de/en/news/first-into-the-ocean-now-into-space-isar-aerospace-achieves-a-breakthrough
Uber–Delivery Hero voluntary takeover offer (announced July 16; offer document Aug 27; acceptance period to Nov 5, 2026): https://content-astro-web.finanzen.net/nachricht/aktien/eqs-wp-g-takeover-offer-target-company-delivery-hero-se-bidder-uber-technologies-inc-15801148 · https://live.deutsche-boerse.com/news/EQS-News-Delivery-Hero-and-Uber-to-Join-Forces-to-Deliver-More-for-Customers-Vendors-and-Riders-d350da37-544e-4bed-94f1-04a2a2a80213 · https://finviz.com/news/386013/uber-publishes-offer-document-for-its-takeover-offer-for-delivery-hero · https://www.wlrk.com/transaction/uber-in-its-14-8-billion-voluntary-takeover-offer-for-all-outstanding-shares-of-delivery-hero/
Also referenced on air and in the monthly episodes, by source and date: Quantum Systems $1.2B round (company announcement, July 2; co-CEO Florian Seibel in the Financial Times) · Proxima Fusion €411M at €2.4B (company announcement; RWE Gundremmingen agreement, February 2026) · The Exploration Company $450M round (Sept 8) and ESA ALADDIN contract worth up to €760M (Sept 10) · Cohere–Aleph Alpha definitive business-combination agreement (company announcements, Sept 16) · NVIDIA–Hugging Face acquisition, approx. $12.9B (announced Sept 3) · SAP–Prior Labs and SAP–Dremio completions (SAP announcements, July) · KNDS IPO reporting (Financial Times, CNBC, Bloomberg: July 1, Aug 26, Sept 16) · AtaiBeckley–Eli Lilly merger agreement (July 15) · EY Startup-Barometer, H1 2026 · KfW venture-capital analysis, Q1 2026 · BACB position paper (advocacy source, as labelled) · Bitkom survey of 102 tech startups · Startup-Verband formation data · HyImpulse round (Sept 2) · WIN initiative next phase (German government and KfW, Sept 17). Detailed claim-by-claim sourcing sits in the July, August, and September episode posts linked in this article.
About the Author
Joern 'Joe' Menninger joins from Frankfurt am Main, Germany. Joe is the founder and CEO of Startuprad.io, Germany's leading English-language startup media platform covering the DACH ecosystem since 2014. Connect with Joe on LinkedIn.
For AI retrieval and structured data, visit https://www.startuprad.io/llm
Transcript (AI-polished)
$3 billion in 11 days. That is what two Munich defence companies, Helsing and Quantum Systems, raised in the first two weeks of July. On September 5, Isar Aerospace reached orbit and deployed its payloads, the first commercial space company from Europe to deliver satellites into orbit. That was actually a historic event. And the same ecosystem from below: German financing rounds were already down 11% for the first half of 2026. Berlin pre-seed funding was down 65% compared with 2022, according to the city's own Business Angels association.
And a $14.8 billion takeover offer for Delivery Hero from Uber. Germany passed the Strategic Necessity Test. This quarter raised a harder question: who ends up owning the companies that pass it? I call this the Ownership Test. Hello and welcome, everybody. This is episode 780 of Startuprad.io, the review of the third quarter of 2026. I'm Joe Menninger, joining you from Frankfurt am Main, Germany. This is the quarterly review for the startup ecosystems of Germany, Austria, and Switzerland, covering everything that happened between July 1 and September 28, the day of recording. The factual claims are sourced; the interpretations and predictions are mine.
So let's talk startups. Three numbers define this quarter. $18 billion: Helsing's valuation, by our count Germany's second decacorn after Celonis, and worth more than Lufthansa. 354: the number of German financing rounds in the first half of 2026, down 11%, with 67% of all capital in rounds above €50 million, per EY. And two: the number of flights it took Isar Aerospace's Spectrum to reach orbit when it deployed its payloads on September 5. A quick word on method. This review covers July 1 through September 28. We record a few days before quarter-end by design.
Anything that lands between this recording and September 30, of course, goes into the October news episode and into our Q4 and annual reviews. Yes, we do all of that for you. Announced deals are labelled announced. Completed deals are labelled completed. Company equity, secondaries, fund closes, and public grants are never added into one number. And where a figure comes from an interest group's own paper rather than neutral statistics, I will say so. Three things coming up. First, the prediction card.
The H1 review put five predictions on record with deadlines and confidence levels, and one confirmed faster than I expected. Second, what the software companies that survived this market actually have in common. And third, the question nobody in the ecosystem wants to answer: when strategically necessary German companies win, who ends up owning them? The prediction card. Accountability first. The H1 2026 review made five predictions. Here is where each one stands as of this recording. Prediction one: KNDS formally restarts its IPO process before March 31, 2027. Confidence was 7 out of 10.
Status: live and volatile. Bottom line, we're waiting. KNDS postponed the listing on July 1. The Financial Times and CNBC reported that the company struggled to convince investors of a valuation above €12 billion. On August 26, Bloomberg reported KNDS was restarting investor meetings, targeting a possible relaunch in the second half of September. On September 16, Bloomberg reported the company and its owners were considering delaying again, possibly beyond this year. So: process activity, yes. A formal restart, announcement, filing, or timetable? Not yet.
The clock runs to March 2027. The deeper signal is unchanged: private capital finances defence at record valuations while public markets keep saying not yet. Prediction two was the big one: at least two German companies in defence, robotics, space, quantum, or strategic infrastructure announce priced equity rounds above €1 billion before the end of 2026. Confidence: 8 out of 10, the highest on the card. Status: confirmed within the first two weeks of the half. Quantum Systems announced $1.2 billion on July 2, just over €1.05 billion at that day's exchange rate. Helsing announced $1.8 billion on July 13. Both German. Both priced equity. Both above €1 billion.
The prediction had until December. It needed only 11 days. Prediction three: SAP announces at least one further material AI acquisition beyond Prior Labs, n8n, and Parloa before December 31, 2026. Confidence: 6 out of 10. Status: open, with a nuance. SAP completed two AI acquisitions inside Q3: Prior Labs closed July 16, with completion announced July 17, and the data-lakehouse platform Dremio on July 6. But both were announced in May, before the prediction was made. Completions are not new announcements.
No genuinely new SAP AI acquisition was announced in Q3. The clock runs to December 31. Prediction four: DeepL announces a material strategic transaction, financing, or restructuring before June 30, 2027. Confidence: 6 out of 10. Status: still open. DeepL did announce a significant restructuring, a workforce reduction of roughly 25% in its shift to an AI-centric operating model, but that announcement came on May 7, in the second quarter, before this prediction existed. Pre-existing signals do not confirm predictions. What it tells us is that the direction of the call was right.
The qualifying event is still outstanding. Prediction five: at least one member of the Profitability Cohort, N26, Personio, Flink, or FINN, announces an IPO mandate, a regulatory filing, or a formal sale process before the end of 2026. Media speculation does not count. Confidence: 5 out of 10, deliberately the lowest on the card. Status: still open. No mandate, no filing, no process. The adjacent signal is Scalable Capital, €30 billion in assets under management and already pre-marketing its Altersvorsorgedepot, a German retirement-investment product. But Scalable is not in the cohort, and IPO chatter is exactly the speculation the prediction excludes.
Interim scorecard: one confirmed at the highest confidence rating, zero missed, four on the clock. And from the monthly episodes, one more confirmation to log. In June, we predicted Isar Aerospace would reach orbit before the end of 2026. They did, on September 5. On its second flight, Spectrum delivered its payloads to orbit from Andøya. Sorry, it's in Norway; I butchered the name, I know. Confirmed with almost four months to spare. In the Q1 review, we said Germany's startup market had become a selection event. In the H1 review, we formalised the selection rule as the Strategic Necessity Test.
Capital concentrates on companies that governments, industries, and supply chains really need to exist. And Q3 is the quarter that test got priced. The evidence: Helsing, $1.8 billion at $18 billion, in a round led by Dragoneer with Lightspeed, ICONIQ, Growth Equity at Goldman Sachs Alternatives, JPMorganChase, CPP Investments, General Catalyst, Plural, and StepStone. Quantum Systems, $1.2 billion at approximately $8 billion post-money, co-led by Blackstone, Noteus, Airbus, and Advent. Co-CEO Florian Seibel told the Financial Times that investors uncomfortable with a move into lethal systems had the opportunity to exit in the transaction. Proxima Fusion: €411 million at €2.4 billion, what the company calls the largest European fusion financing to date, with Google and RWE as strategic investors. RWE had already signed an agreement in February to work toward a stellarator plant at the former Gundremmingen nuclear site before it invested. The customer invested. That compresses the path from lab to grid.
And on September 5, Isar Aerospace reached orbit on Spectrum's second flight and deployed its payloads, the first commercial company from Europe to deliver satellites into orbit and the first provider to hit the European Launcher Challenge orbital milestone ahead of the 2027 deadline. Germany's government is backing a reported €35 billion space build-out by 2030. The Strategic Necessity thesis does not get cleaner evidence than a rocket. Note what the lead investors of the quarter's biggest rounds have in common: Dragoneer, Blackstone, Goldman Sachs, Google. Global institutional allocators, not European venture funds. Hold that thought, but hold it carefully.
It is the door into segment three, and it is not as simple as it looks. Participation collapse. Now the other side of the ledger. The August episode named it the concentration economy. At quarter scale, I want to be more precise: it is a participation collapse. The baseline numbers from EY's half-year data: €5.3 billion invested in German startups in the first half of 2026, up 14%, but rounds fell 11% to 354, and 67% of all capital went into rounds above €50 million.
Compare the peak: the first half of 2021 had 588 rounds. Roughly 40% fewer financings than five years ago, with more money. Those are half-year figures. The Q3-specific count is not yet published, and we will score it in the annual review. At the base of the funnel, the picture darkens, and here the source matters. This is the Business Angels Club Berlin-Brandenburg's own position paper, an advocacy document, but with concrete numbers. According to the BACB, Berlin pre-seed funding is down 65% since 2022, back to roughly 2017 levels.
The median angel ticket is down from €50,000 to €38,000. Startup insolvencies nearly doubled since 2022, even while formations hit a record 3,053 in the first half, per the Startup-Verband. And labour is concentrating too. Bitkom surveyed 102 German tech startups, a small sample and not representative, but directionally striking. 27% had forgone new hires in the past 12 months because of AI. 7% cut jobs because of it. 16% added staff because of it. The average startup in the survey now employs 12 people, down from 13. The ecosystem grows in value without growing in participation.
Now look at the software companies that still won this quarter and what they have in common. Camunda, the Berlin process-orchestration company: annual recurring revenue nearing $200 million. The first $100 million took roughly 10 years; the second is expected in about two, without a single new venture round since 2021. Valued above $1 billion since 2025: a hidden unicorn built on revenue, not a raise. Pliant: past $100 million in annual recurring revenue.
Moss: a unicorn at a €1 billion valuation on a Series C of just €35 million. In the H1 review, we noted software was absent from the top funding tier. In Q3, the software story is not about round size at all. It is about proof. My read: the door for software is open, but the currency at the door changed from story to revenue. Augustus is the other face of selectivity: a $180 million Series B at a $1 billion valuation, raised while progressing toward a US national-bank launch, with preliminary conditional approval from the OCC and further US regulatory approvals reported since. Regulated infrastructure is the fintech lane that still prices.
And the casualty list from the same system inside the quarter: Sono Motors, insolvency proceedings opened July 31. Glow25, roughly 120 job cuts, with management naming AI-enabled automation as a driver. Enpal, closing its Hamburg site, affecting up to 85 roles, while guiding to over €1.3 billion of full-year revenue. Selection is not a phase of this market. It is the market. The Ownership Test. Now the thread that ties the quarter together, and the uncomfortable one. I want to be precise here because the evidence points in more than one direction, and that is exactly what makes it interesting. First, the anchor statistic from the September episode: according to KfW, more than three quarters of the venture money that went into German startups in the first quarter of 2026 came from abroad.
Keep that number in mind. Now the outward-pointing evidence. Delivery Hero: Uber announced a voluntary takeover offer on July 16, €41.50 per share, an implied equity value of $14.8 billion for 100% of the company, or $13.7 billion adjusted for Uber's existing stake. Announced, not closed. The acceptance period runs until November 5, and closing is expected only in the second half of 2027. But if it closes, Berlin's biggest platform company belongs to San Francisco. AtaiBeckley: merger agreement with Eli Lilly dated July 15, roughly $2.8 billion upfront plus up to $1 billion in milestone payments. Announced, not closed.
Aleph Alpha, once the face of German AI sovereignty: on September 16, Cohere and Aleph Alpha signed a definitive business-combination agreement. Signed, not closed, but the flag on Germany's flagship AI lab is changing. Hugging Face, French-founded and US-headquartered, the open-model infrastructure half of Europe's AI stack depends on: NVIDIA announced the acquisition on September 3 at approximately $12.9 billion. And HappyRobot, the $1.2 billion unicorn we profiled in episode 770: formed in Munich, owned in San Francisco. And now the counter-evidence, and it is real. Helsing raised from Dragoneer, Goldman, and JPMorganChase.
The company itself states it remained predominantly European-owned after the round. Foreign-led capital is not the same thing as foreign ownership. Langdock, which we flagged in July for tension over its US parent, reportedly did the opposite of the outbound trend, re-domiciling its parent to Germany as an SE, with the large majority now owned by founders and employees in the EU. SAP acted as a domestic acquirer twice, Prior Labs and Dremio, and committed more than €1 billion over four years to build Prior Labs into a frontier AI lab in Europe. September delivered the cleanest counter-pattern of the whole quarter in 48 hours: on September 8, The Exploration Company, the German-parent space startup, raised $450 million. On September 10, the European Space Agency signed it for the ALADDIN programme, worth up to €760 million including optional service missions.
Capital and customers, back to back, with ownership staying home. HyImpulse raised more than €50 million on September 2. On September 17, the German government and KfW started the next phase of the WIN initiative, the push to mobilise domestic scale-up capital. When Europe is the customer and the LP, ownership stays close. That is the pattern to build on. So here is the framework. The Strategic Necessity Test asks: does this company need to exist for governments, industries, and supply chains? Germany now passes it; the capital proves it. The Ownership Test asks the next question: when a strategically necessary company succeeds, who ends up owning it? Q3 did not settle that test. Q3 exposed it.
The quarter's biggest pending exits point outward, Delivery Hero and AtaiBeckley, while its most strategic growth companies, Helsing, Proxima, and Isar, stay European-controlled on global capital. Necessity attracts capital. Ownership is decided by infrastructure, and that is the test Europe has not yet built for. What people are getting wrong: this is not a sellout morality tale, and it is not an argument against American capital. Dragoneer and Blackstone financing Munich defence companies is the system working. My thesis is that the gap sits on the European side of the table: a thin domestic LP base at pension scale, a public market where a KNDS could not price and no German tech listing of comparable scale happened, and SAP as the quarter's most visible domestic acquirer, carrying that role largely alone. Ownership follows infrastructure. Europe built the companies before it built the balance sheets to keep them. That is an interpretation, but it is the one the quarter's evidence keeps pointing at.
The single thread of Q3 2026: the structural rotation completed and got priced. And the moment it was priced, the question moved. Q1 asked which companies survive: the selection event. H1 asked which companies attract capital: the Strategic Necessity Test. Q3 asks who ends up owning the winners: the Ownership Test, posed, not yet decided. Three quarters, one arc, from survival
to necessity to ownership. The fourth question, for Q4 and beyond, is participation: whether an ecosystem of 354 rounds, collapsing pre-seed, and 12-person startups keeps producing companies worth owning at all. For the framework being applied here, go back to two episodes: the H1 2026 review, 'Europe's Startup Recovery Was a Structural Rotation,' where the Strategic Necessity Test was defined, and Unicorn Atlas number one on Helsing, where the defence-capital thesis started. Links in the show notes. Takeaways: operators, investors, and ecosystem. If I were a founder or operator: three things. First, the lead investors in this quarter's mega-rounds were American and global institutions. If you build in a strategic-necessity category, plan your Series B and beyond with that map and decide before the term sheet what ownership outcome you are building toward, because that choice gets made whether you make it or not. Second, if you build software, the quarter's message is proof over narrative.
The software names that mattered, Camunda, Pliant, and Moss, mattered for revenue, not round size. Third, run the AI headcount audit. In Bitkom's survey, half of startups already changed their staffing because of AI in one direction or another. If I were an investor, and these are observations, not recommendations, the interesting ends of the market are the two the mega-rounds ignore. At the bottom, if the BACB's numbers hold, pre-seed competition in Berlin is the thinnest in a decade for whoever still writes angel tickets. HTGF launched fundraising for its fifth seed fund on August 26, but public capital alone will not refill that base. At the top, assets coming out of restructurings, Sono's solar business and IP is up for sale, may be priced very differently from their long-term technical value.
That is a market-inefficiency observation, the same one we made in August. And watch the Ownership Test as a return driver. If the Delivery Hero and AtaiBeckley transactions close in 2027, a significant wave of liquidity reaches German founders, employees, and angels, the recycling this ecosystem has been waiting for. If. If I were building the ecosystem, policy, institutions, infrastructure, my thesis from this quarter's evidence is that the binding constraint is shifting from company formation and capital attraction toward retention of ownership. The levers are known. The Altersvorsorgedepot launches January 1, 2027 and could become a domestic capital channel over time; a public market where a KNDS or a DeepL can actually price; and procurement as anchor customer, the way ESA just demonstrated with The Exploration Company.
Every quarter those levers stay unpulled, the Ownership Test keeps getting answered abroad. Our Q4 predictions, on the record. We have three new predictions, each with a deadline and a confidence level. Number one: KNDS does not complete its listing this year; the IPO slips into 2027. Confidence: 6 out of 10. Prediction two: Germany ends 2026 with at least 12 new unicorns for the year, up from 10 by late August.
Confidence: 7 out of 10. Prediction three: at least one more DACH company valued above $1 billion agrees to an acquisition by a non-European buyer before December 31. Confidence: 7 out of 10. These are forecasts, not facts. Check me in the annual review. And the four predictions still on the clock from the H1 card, KNDS restart, SAP's next AI acquisition, DeepL, and the Profitability Cohort, stay live. The annual review settles them. That is Q3 2026, the quarter the rotation got priced, participation kept narrowing, and the Ownership Test was posed.
The annual review goes live in early January and will score every prediction on this card. I'll see you next quarter. If this episode sharpened how you see the market, send it to one founder who is still operating on last cycle's assumptions. This is Startuprad.io. That's all, folks. Thank you very much. Find more news, streams, events, and interviews at www.startuprad.io. Remember, sharing is caring.
Created with the assistance of AI.
Comments