The $626 Billion Space Economy: Where the Money Actually Sits in 2026

The space economy is worth $626 billion a year. The figure travels well. What rarely travels with it is an answer to a simpler question: where does the money actually sit?

Most of it sits on the ground. A large slice sits in government budgets that function as anchor tenancy rather than subsidy. And a thin, fast-growing layer is starting to sit in orbit itself.

Here is the map.

This radar sweeps the five places capital can actually land: connectivity, ground equipment, government demand, launch, and the in-orbit services layer being built on top of them. Figures come from Novaspace's 2025 Space Economy Report, ESA's 2026 space economy report and SpaceNews deal tracking unless noted.

Three numbers that frame the map

Growing: four currents pulling the market upward

Start with scale. A market compounding near 12% doubles in roughly six years, which means the difference between a good segment and a great one is measured in hundreds of billions by the mid-2030s. Novaspace calls 2025 a foundation year: the industry reorganised around security customers first, commercial ones second, and the capital followed that order.

Four currents carry most of the growth.

A digression worth thirty seconds. Remember how recent these economics are: putting a kilogram in orbit cost tens of thousands of dollars for decades, and the figure now sits under $3,000. Every business case above quietly multiplies by that ratio, which is why segments that looked absurd in 2015 raise half-billion-dollar rounds in 2026.

Falling: the slow leak out of geostationary orbit

The mirror image of that growth sits 35,786 kilometres higher. Traditional geostationary orbit (GEO) operators built their business on scarce orbital slots and multi-decade satellite lifetimes, and both advantages are eroding at once as low Earth orbit (LEO) constellations absorb traffic.

The consolidation wave tells the story plainly. Viasat absorbed Inmarsat. Eutelsat merged with OneWeb. SES is taking over Intelsat. When an industry's response to shrinking pricing power is serial mergers, the equity story becomes about cost synergies rather than growth, and cost-synergy stories rarely command premium multiples for long.

Iridium is the instructive exception: unglamorous, profitable, low churn, serving aviation, maritime and IoT niches the mega-constellations ignore. The lesson concerns pricing power. It accrues to whoever owns the customer relationship, and in satellite communications that relationship is migrating downstream.

Scale is now the entry ticket.

New: orbital compute wants its own line

SpaceX's June IPO filing turned a research-memo topic into a balance-sheet line item: data centres in orbit, powered by sunlight, cooled by vacuum. The pitch writes itself. The arithmetic is less cooperative.

Space Insider's analysis of the filing runs the numbers at current prices. A dedicated Falcon 9 flight carrying roughly 1,000 accelerators costs about $74M at list price, which works out to some $74,000 of launch per processor. The silicon itself costs $30,000 to $40,000. Launch would exceed the cost of the compute before a single watt of power or byte of storage is paid for.

The entire thesis rests on Starship reaching its advertised $100 per kilogram. Space Insider's read of industry figures places operational costs at $900 to $1,000. That gap is the whole story, and no operational track record yet validates closing it.

Treat the filing as a market signal rather than a blueprint. The landmark contracts it references cover ground-based capacity, not orbital assets, which tells you customers are buying optionality while the technology matures. Watch contract renewals rather than announcements: multi-year ground-capacity deals convert into orbital reservations only when launch costs cooperate. As we wrote in August when scoring the launch market, access to orbit is consolidating around a handful of providers, and whoever controls cheap mass eventually controls who gets to compute up there.

The public-market door finally opened

For two decades the sector's most valuable asset was famous for being uninvestable. That objection expired on June 12, when SpaceX listed as SPCX in the largest IPO on record, priced at $135 a share and valuing the company near $1.8 trillion at the open.

The trading history since is its own lesson in infrastructure repricing. The stock spiked, retraced more than 30%, dipped below its offer price, then recovered to roughly $1.84 trillion by mid-August, according to SpaceNexus's market reconstruction. A round trip that violent inside ten weeks tells you the market is arguing about cash-flow timing, not about whether the revenue exists.

Concentration risk travels with the ticker. Roughly 61% of SpaceX's 2025 revenue traces back to Starlink, which binds the sector's largest valuation to a single broadband business and its launch-cadence dependencies. Size the position accordingly; conviction does the rest.

The broader point is that access broadened in both directions at once. BlackRock opened a space technologies fund for European investors in June, packaging the theme without single-name risk. Rocket Lab became the pure-play alternative for direct exposure: profitable per Electron mission, with the medium-lift Neutron in development and shares up more than 400% from their 2022 lows. Private markets kept their own door open. Sovereign wealth funds anchored Axiom Space's $350M Series D at a reported $4B-plus valuation. Sierra Space closed $550M at an $8B valuation.

A sector that used to force a choice between diversified aerospace giants and illiquid venture checks now offers a full ladder: index exposure, pure-play equities, late-stage private rounds. The scarcity was never the technology. It was the entry path.

Reading the table: scale versus slope

The table rewards a specific habit: separating scale from slope. Satellite communications and ground equipment hold the wealth today. In-orbit services hold the growth rate. Government budgets hold the floor. A portfolio that mirrors the table rather than the headlines looks deliberately boring at the top and deliberately speculative at the bottom, with little in the middle. The middle is where 2026 did its real work: index funds, late-stage rounds and cash-generating operators now bridge a ladder that barely existed three years ago.