In September 2021, Lockheed Martin and Northrop Grumman invested in the same startup, in the same round. If you don’t follow defense contracting that sounds like nothing, so let me tell you why it isn’t.
These are the two largest defense-space primes in the United States, and they spend their whole commercial lives bidding against each other for the same programs. They almost never show up on the same cap table.
As far as I can tell this was the first time the two of them had ever co-invested in a venture deal.
The company was Orbit Fab. Its own marketing described it, plainly, as “gas stations in space.” At that point it had basically no revenue, no fuel in orbit, and a plan most serious people would have filed under science fiction.
A little earlier, Munich Re Ventures had come into its seed round, which is worth noticing too: Munich Re is one of the largest reinsurers on earth, and its entire business is pricing risk on things like rocket launches.
So you have a reinsurer that prices launches for a living, plus two arch-rivals who agree on almost nothing, all deciding at the same moment that this tiny company was worth a position. None of them needed the money. What they were buying looks to me much more like insurance. The interesting question is what they thought they were insuring against — and I think the answer tells you where the real money in this sector is going to end up.
I’ve watched the space sector for years, and most of what passes for analysis in it is a list of rockets and the people who build them. This edition is about the thing the rockets are quietly moving money toward, and away from.
The vehicle and the node
Start with one distinction, because everything here depends on getting it right.
In the history of shipping, the durable fortunes weren’t made by the people who built the ships. Shipbuilding was a brutal business — capital-hungry, competitive, and forever commoditizing, because a ship is ultimately a steel box that anyone with a yard and enough credit can build a slightly better version of.
The lasting money went to whoever controlled the places ships had to pass through: the Suez and Panama canals, the deepwater terminals, the handful of chokepoints traffic had no choice but to funnel through. A canal owner could charge a toll on the same chokepoint for a century while the shipbuilders downstream competed each other into thin margins.
I keep coming back to this because the space sector today is almost entirely a story about ships. SpaceX, Blue Origin, the launch providers, the satellite manufacturers — in this analogy they’re all shipbuilders, and some of them are extraordinary ones. They’re also in a commoditizing business, because the whole narrative of the sector is that the cost of getting to orbit keeps falling. That’s wonderful for the world. It’s slightly ominous if you’re hoping to earn a durable rent from building the vehicle.
So the question I care about is the one nobody seems to ask on stage at the conferences. Not who builds the best rocket — there are a dozen well-funded answers to that. Who ends up owning the equivalent of the canal.
Here it helps to be precise about the money, because the whole argument lives inside a split in the numbers. The cislunar infrastructure market — the thrusters, landers, aeroshells, propellant tankers, orbital transfer vehicles, all the hardware that gets built and flown — was estimated at roughly $13.84 billion in 2025, and is projected to reach about $24.83 billion by 2032, a compound growth rate near 8.7 percent. Respectable. It’s also, in my view, the wrong number to anchor on, because it measures the ships.
I think the value won’t stay with the ships. It will migrate, the way it did in shipping and the way it did in software, to a thin control layer sitting on top of all that hardware that decides who refuels, who docks, and on what terms. That layer isn’t in the $24 billion, because right now it barely exists. That’s exactly why the opportunity is large, and also why most of the companies chasing it are going to fail.
Why I’m writing this now
I wouldn’t put this in front of you in June rather than next June if nothing had moved. Three things have, and they’ve moved close together.
The first happened two days ago. On June 8, Quantum Space — a cislunar logistics company run by Jim Bridenstine, who used to run NASA — announced it’s going public through a roughly $1.2 billion merger with a blank-check vehicle, expected to close in the final quarter of this year under the ticker QSPC.
Forget the headline valuation for a second. What matters is that a former head of NASA looked at the calendar and decided the moment to put a more-or-less pure cislunar-logistics business in front of public investors is this year.
The second is that SpaceX itself is reportedly preparing a public listing around the middle of this year, at a valuation discussed in the trillions. When the biggest ship in the fleet lists, everything else gets repriced against it, whether or not the comparison makes sense.
The third is the quietest, and to me the most important. The U.S. Space Force has on-orbit refueling and orbital-transfer demonstrations scheduled across 2026 through 2028. Demonstrations are how a hypothesis becomes a funded program — or how it quietly dies. Either way they’re the events that decide whether any of this is real.
For most of the last decade this was an untradeable idea. The pure plays were private and the honest answer to any timing question was “someday.” That’s changing while we watch, which is the only reason this edition is worth your time this week and not later.
Why orbits become scarce even though nobody owns them
The idea underneath all this is easy to misread, so let me be careful about what I’m claiming.
I’m not claiming anyone will own an orbit the way they own a field. Nobody is getting a deed to a point in space. What I’m claiming is smaller and, I think, more durable: certain orbital positions become economically scarce even though no one holds legal title to them. The Lagrange points, where an object can hold station while burning almost no fuel, are the obvious case.
So are the corridors connecting low Earth orbit, geostationary orbit, and eventually the Moon.
Here’s the mechanism, because this is the part people skip. Scarcity in markets almost never requires ownership. It only requires that demand concentrate on a position with no easy substitute. The merchant who owned the warehouse beside the only deep harbor never owned the sea, and it didn’t matter — everything had to come through him anyway. An orbital chokepoint works the same way.
If that’s right, three kinds of operator could end up with pricing power that none of the current models contain. What stops me filing this under daydreaming is that all three already exist as real companies building real hardware, mostly in obscurity, while the cameras stay on the launch pad. I think that obscurity is part of the opportunity.
The first kind is the orbital tug — the last-mile carrier of space. A rocket drops a satellite wherever is convenient, and a tug collects it and takes it where it actually needs to go. Atomos Space, out of Loveland, Colorado, is building a vehicle called Quark to capture and relocate satellites with high-power electric propulsion. Impulse Space, founded by Tom Mueller — the propulsion engineer behind SpaceX’s earliest engines — is working on fast transfer from low orbit up to geostationary.
And it’s not only American: Japan’s NEC has its own transfer-vehicle program, sold as civil logistics, whose underlying skills — rendezvous, docking, on-orbit servicing, refueling — are inherently dual-use. That last point matters more than it looks, and here’s why: a node that sits on the civil-military line draws demand from two independent sources, which gives it a much harder floor than a purely commercial play.
The second kind is the propellant depot, and this is the one I’d point to if you asked me for the cleanest version of the whole idea. Whoever stores and sells fuel in orbit is selling access to the corridor itself. Which brings us back to Orbit Fab, the company those two rivals were insuring against.
Here’s the detail that made me take the whole thing seriously. Orbit Fab’s refueling port, which it calls RAFTI, was formally accepted in 2024 by U.S. Space Systems Command as an approved interface for refueling military satellites. The port costs about thirty thousand dollars. Sit with that pairing for a second, because it’s the entire argument in miniature. The moat in this layer was never going to be the fuel — fuel is a commodity anyone can eventually sell.
The moat is the standard: the physical interface every other vehicle has to adopt before it can refuel at all. A company that owns the fitting the rest of the industry is forced to build around owns something much closer to a toll booth than to a fuel business. And it owns it for a thirty-thousand-dollar part.
The third kind doesn’t run a single vehicle; it builds the whole transport graph. That’s Bridenstine’s Quantum Space, trying to build the foundational logistics layer for cislunar and national-security space rather than chase Earth-orbit satellites. Around it sit firms the public almost never tracks: TransAstra, with its transfer vehicles and asteroid-capture “optical mining,” and component makers like CisLunar Industries, building in-space power and metal processing. None of these are household names, and I think that’s the most telling thing about them.
Why this is structural, not wishful
It would be easy to read all this as science fiction with a spreadsheet attached, and there were stretches while I worked through it where I worried that’s what I was doing. What pulled me back each time is that every reader of this newsletter has already lived through the exact same transition once, in a sector none of us thinks of as exotic.
Think about the digital economy. In its first decade the value went to the people who made the devices and wrote the apps. In its second decade it moved underneath them, to the operating systems, the app stores, and the cloud that quietly taxed every application running on top of it. The device makers and app developers kept mattering. They just stopped capturing the economics, because someone had built a toll layer beneath them while they were busy shipping product. I don’t think that was an accident of software, and I don’t see why space would be exempt.
A refueling-and-transfer layer in orbit sits beneath every satellite, station, and lunar payload, in roughly the position the cloud occupies beneath the software it hosts, and it can charge all of them on the same logic. Standardized interfaces, the real estate at the depots, and the data layer of knowing where everything is and where it’s headed — together those are the pieces of a platform. And platforms are where the durable money in the last transition went.
Here’s where it stops being abstract, including for you. Most of the satellites you depend on without thinking about — the ones behind your phone’s location, the timing signal your bank uses to stamp transactions, the weather models, the imaging — don’t die because they break. They die because they run out of propellant, and they get written off as perfectly healthy spacecraft drifting with empty tanks. The whole premise of the depot layer is that a dying satellite isn’t garbage. It’s a customer who can’t yet reach a gas station.
Let me state my own view plainly, because this is where I think the consensus is wrong. The sector is being priced as a launch-and-hardware story — as a bet on the falling cost of reaching orbit. That’s a category error, and an expensive one. Launch is the part most likely to commoditize, and the market is paying up for it mostly because it’s the part that’s easy to see and easy to model. The defensible money, if there’s any here at all, is in the layer the market isn’t looking at yet. I might be wrong about the timing, and I’ll get to the ways I could be wrong about the whole thing. I don’t think I’m wrong about which layer the value wants to settle in.
🔒 The rest of this edition is for premium subscribers.
If you’ve read this far you have the thesis: the money ends up at the chokepoints, not the rockets. What you don’t have yet is the part that decides whether the idea makes you money or quietly costs it — how to actually hold it.
Below the line I work through why ordinary valuation falls apart here and what I use instead; the exact ladder of vehicles, from the liquid public names that give you the sector but not the chokepoint, through the one near-pure operator about to list, down to the private layer where the real toll booth lives; how I’d size any of it, which is the part most people get backwards; the four observable events that open and close the trade; and the four ways I think this thesis goes to zero — one of which is the very thing that makes the sector exciting to everyone else.