A loitering munition that a European army buys today costs about as much as a small car. Stark, the Berlin company that makes one, is now worth more than €3.5 billion. Hold those two numbers next to each other and the question answers itself: whatever investors are paying for, it isn't the drone.
Six months of funding rounds make that plain.
Start with the number everyone measures against. In May, Anduril closed a $5 billion round at a $61 billion valuation, roughly double its mark from a year before. The company reached $2.2 billion in revenue in 2025, which puts the round near 28 times sales. No airframe earns that multiple. The margin profile does.
Anduril reportedly targets gross margins between 40 and 50 percent, the kind of number that belongs to software, while the traditional primes like Lockheed and RTX live nearer 8 to 10 percent. Same customer, same war, a completely different business underneath.
Europe spent 2026 building its own version of that trade.
The airframe is a delivery mechanism
Helsing, the Munich company, is raising $1.2 billion at a reported $18 billion valuation, up from about $14 billion less than a year earlier. Read the coverage and count the sentences about aircraft. Almost none. The story is autonomy software, targeting, and the data loop that sharpens every time one of its systems flies. Helsing began as pure software and only later added the HX-2 strike drone. The drone is where the software lands. It was never the point.
The pattern repeats down the list, and it accelerated this month. On 2 July, Quantum Systems raised $1.2 billion at roughly $8 billion, co-led by Blackstone and Airbus, more than doubling its valuation in under a year. The company is already profitable, its founder calls it a "neo prime", and the capital is bound for production capacity and MOSAIC, the software layer that ties its drones, sensors, and counter-drone systems into one network. A week earlier, Stark raised €500 million at more than €3.5 billion, co-led by Sequoia and Founders Fund, with more than 80 percent of the money going to manufacturing and research. Portugal's Tekever crossed £1 billion to become the country's first defense unicorn.
Read those rounds together and one logic runs through all of them. Investors are underwriting two assets: software economics on one side, the capacity to build at wartime volume on the other. Ukraine taught every ministry that a clever prototype is worthless if you can only ship forty a month, so the ability to make the thing cheaply and fast has become a moat in its own right, harder to copy than the thing itself.
ℹ️The aircraft is the cheapest, most replaceable part of the stack. The money is paying for software margins and the capacity to build at scale.
Then there's counter-drone
Here the pattern breaks, and the break is the interesting part.
If you trust three years of headlines, detection is where the smart money should go. Every airport shutdown, every drone over a base, every Shahed swarm makes the case for knowing what's in your sky before it reaches you. Sticky infrastructure contracts, recurring revenue, a threat that grows each quarter. On paper it's the cleanest subscription story in the sector.
The valuations say otherwise. Over the past two years the entire counter-UAS category drew disclosed equity rounds from just 18 companies, at a median near $15 million, with a single deal above $100 million. That deal was Epirus, the American high-power microwave firm, which raised $250 million near a $1.35 billion valuation, and even that round priced below its previous one. Stark raised twice as much in a single afternoon.
$61B
$18B
$8B
€3.5B
$1.35B
The biggest pure-play in counter-drone is worth barely two percent of Anduril. Europe looks thinner still: the continent accounts for a healthy share of counter-UAS company formation but only a small slice of the capital, heavy on seed rounds and light on the growth money that turns a prototype into a production line.
Why the gap? Part of it is that offense photographs better than defense. A loitering munition has a clear story and a clear buyer. A detection grid is plumbing, and plumbing is a hard sell to a fund chasing the next Anduril. Part of it is fragmentation: counter-drone is a patchwork of radar, RF, optics, and effectors, and no single company has convinced the market it owns the full stack the way the autonomy players claim to own theirs. DroneShield, the listed Australian detection specialist, comes closest, with revenue up 276 percent to A$216 million in 2025 and software sales growing near 200 percent a year. It's a strong business. It's also a mid-cap, not a decade-defining one.
ℹ️When Quantum Systems raised its $8 billion, counter-drone wasn't a company it bought or a line it paid up for. It was one feature folded into MOSAIC, its software platform. That's the tell.
So the market has made a quiet call. It's pricing the moat in autonomy and mass production, and it hasn't yet decided that detection is a company rather than a feature.
The state money points the other way
The venture funds set the valuations. State capital sets the direction, and it isn't reading the same script.
The NATO Innovation Fund sits in Stark's cap table, one of several public and quasi-public vehicles moving into exactly the capital-heavy, slower-return corners that private growth funds underweight. Behind them sits ReArm Europe, a plan to mobilise up to €800 billion over four years, and a European Defence Industry Programme that earmarks money specifically for counter-drone and autonomous strike. That earmark matters. The people who actually buy this equipment in a crisis are saying, in budget lines, that they want detection capacity to exist, even while the private market prices it as an afterthought.
For anyone building in the Nordics, that's the more useful signal. Private multiples tell you where the hype is. Sovereign flows tell you where the buyers of the next decade want capacity to sit, and detection infrastructure fits that mandate far better than it fits a growth fund's model.
What the money is really saying
Strip out the noise and the verdict holds. The billions moving through European drone rounds this year are a bet on software economics and industrial scale, wearing an airframe. The drone is the part that matters least.
Counter-drone is the open position. It has arguably the strongest recurring-revenue case in the sector, and the market prices it like a rounding error. Either the smart money is early, and detection gets repriced the moment one company proves it owns the full stack, or the market has already ruled that seeing your own sky is infrastructure someone else bundles for free.
For a continent that spent 2026 learning it can't see its own airspace, that's not an academic question. It's the one worth watching.