ESA prices three rocket startups on milestones, not flights

Divide each award by the payload its vehicle is advertised to carry and one programme resolves into three quite different prices. Isar's €197.8 million against 1,000 kg to low Earth orbit ^[4][6] is about €198,000 per advertised kilogram; Rocket Factory Augsburg's €186.9 million against 500 kg to a 500 km sun-synchronous orbit ^[8] is about €374,000; PLD Space's €158.9 million against 540 kg ^[9] is about €294,000 ^[21]. ESA is paying roughly 1.9 times as much per promised kilogram at RFA as at Isar ^[22], which reads less as a price for lift than as a price for keeping a second and a third entrant in the field.

Totalled, the three contracts come to €543.6 million ^[19], or about $640 million at the rate implied by Isar's own conversion of €197.8 million into $233 million, which is 0.16% of the roughly $400 billion at which SpaceX has been valued ^[14][24]. At that ratio, the money is doing less to fund a rival than to buy an option on having one, and Isar's 36.4% of the pot ^[20] is the largest single position ESA took.

The structure is the part worth studying. Nothing is paid up front and funds release as milestones are met ^[3], which is a tranched round in every respect except that the investor takes no equity and no board seat, and Andreas Schwarz of the German parliamentary budget committee has already said governments should weigh taking stakes in providers alongside long-term launch commitments ^[16]. Non-dilutive money is the cheapest capital on any cap table, but milestone cash arrives after the spend that earns it, so the working capital between a 30-second flight ^[5] and an orbital launch before 2028 ^[2] remains equity's problem, priced by investors working to a deadline they did not choose. Put the headline number over the flight record and you get roughly €6.6 million for every second Spectrum has spent off the pad ^[23], which is a price for a schedule rather than for a vehicle.

Isar's contract is funded mainly by Germany with contributions from Austria and Norway, and PLD's mainly by Spain with German support ^[7][9], so the competition Géraldine Naja described ^[13] runs between entrants drawing on separate treasuries rather than bidding against each other for the same euro. The fourth award, to MaiaSpace, has not closed; ESA says the process is nearing completion and expected to resume in the coming weeks ^[10]. One oddity in the account: proposals were due April 2026 and appraised on business case, engineering quality, resources, schedule and compliance ^[11], leaving under 21 months between appraisal and the deadline ^[26].

The version where this works has Isar's second Spectrum reaching orbit with payloads ^[6] and demand waiting behind it, chiefly the Airbus, Leonardo and Thales satellite combination carrying about €6.5 billion of annual revenue and due to begin operating in 2027 ^[15], against which all three launcher awards are 8.4% of one year of that revenue ^[25]. The version where it does not has all three slipping and ESA writing fresh milestones, and that is where the venture analogy fails, because a fund marks the position down while an agency reschedules it. This is probably wrong, but Lucia Linares saying that getting to space is "not easy and never guaranteed" ^[12] sounds like a deadline with give in it. What would prove it wrong: ESA letting an award lapse in 2028 rather than extending it, or a priced round at one of the three that treats milestone paper as booked revenue instead of a schedule obligation.