III
Propulsion & propellants
Engines and storable propellants concentrate in a few qualified lines; hydrazine alternatives and Hall-effect thrusters have single-digit benches.
Propulsion is the part of the space supply chain that cannot be wished into existence with a render. An engine has to be poured, machined, hot-fired, and qualified — and the benches that can do this at flight quality are few. That scarcity is the chokepoint: storable propellants, qualified upper-stage engines, and electric thrusters all concentrate in single-digit production lines, and almost none of them scale on a software timeline.
Who owns it
The listed exposure splits cleanly between launch propulsion and in-space propulsion. On launch, L3Harris's Aerojet Rocketdyne unit holds a contracted backlog of "north of 150" RL10 upper-stage engines for SLS, Vulcan, and Atlas V, ramping toward 40 engines/year, and produces over 115,000 solid rocket motors annually at Camden, Arkansas. Karman Space & Defense (KRMN) has delivered 20,000+ solid-propellant systems and supplies boost motors for the Navy's Conventional Prompt Strike and the Army's Long Range Hypersonic Weapon — it grew revenue 37% to $471.5M in 2025 with backlog over $1B. On in-space, Rocket Lab (RKLB) builds its Archimedes methalox engine in-house and in April 2026 debuted the Gauss Hall-effect thruster on a line built for 200 units/year. Moog (MOG.A) builds chemical and electric propulsion modules in Niagara Falls and won an AFRL multimode (chemical+electric, single propellant) development contract in 2025.
What breaks it
The binding constraint is qualified throughput, not design. Electric-propulsion merchant supply is the tightest node: the most capable Hall and electrospray houses — Busek, Orbion, Phase Four — are private, leaving listed exposure to vertically-integrated builders (Rocket Lab, Moog) rather than pure merchants. Propellant chemistry is its own bottleneck: hydrazine is a REACH-restricted carcinogen, and the industry's pivot to "green" mono/bipropellants and methalox restarts qualification clocks from near zero. On the launch side, solid rocket motor capacity is the visible scarce asset — L3Harris is spending $215.6M on a Camden expansion targeting a six-fold increase in large-SRM capacity, which tells you how short the supply is.
What forces the reprice
Three levers. First, ownership: L3Harris agreed in January 2026 to sell a 60% majority of its Space Propulsion & Power business — RL10, in-space propulsion, power, avionics — to AE Industrial Partners for $845M, forming a new "Rocketdyne" (RS-25 and the SRM business stay). When that closes in H2 2026, LHX becomes far less of a space-propulsion play and far more of a missile-SRM one. Second, electric-propulsion qualification: Gauss flight heritage and Moog's multimode milestones convert backlog into recognized revenue. Third, constellation demand — proliferated LEO and missile-defense interceptors both pull on the same propellant and motor lines simultaneously.
Be honest about size and earliness
This is a small-revenue, high-narrative chokepoint, and you should size it that way. Of the listed names, only L3Harris ($23B+ revenue) and Karman ($471.5M) carry propulsion as material, audited revenue — and L3Harris's space-propulsion piece is being sold off. Rocket Lab's Gauss line is real hardware but had not announced a customer backlog at unveil; the Redwire "Valkyrie" thruster, developed under a strategic cooperation agreement with private Phase Four, carries a vendor revenue-potential figure that is a TAM slide, not a booked number — exactly the kind of figure to discount. And the elephant: SpaceX IPO'd June 12, 2026 (NASDAQ: SPCX) at a $1.77T valuation, but it is vertically integrated — 600+ Raptors built, ~1/day — and sells launch, not engines. SpaceX is the launch market; it is not a propulsion supplier you can underwrite as a merchant. Denser daily coverage exists (Payload, SpaceNews); this pack earns its edge only on the supply-chain specifics above — qualified bench counts, the AE Industrial carve-out, SRM capacity math — not on TAM. None of this is investment advice; it is a map of where the physical constraint sits and what events move it.
Who owns the choke
Karman Holdings Inc. (Karman Space & Defense)
Karman has designed and delivered 20,000+ solid-propellant systems and supplies boost motors for the Navy's Conventional Prompt Strike and Army Long Range Hypersonic Weapon programs. FY2025 revenue was $471.5M (+37%) with total backlog exceeding $1B as of March 2026; it acquired solid-propulsion specialist ISP in May 2025.
- Bull
- If propulsion concentrates in qualified lines, Karman has delivered 20,000+ solid-propellant systems and supplies boost motors for Conventional Prompt Strike and LRHW.
- Bear
- Karman's growth is tied to a few hypersonic-weapon programs whose funding and schedules are policy- and budget-sensitive, exposing it to defense-procurement cyclicality.
L3Harris Technologies, Inc. (Aerojet Rocketdyne)
L3Harris's Aerojet Rocketdyne holds a contracted backlog of 150+ RL10 upper-stage engines (SLS, Vulcan, Atlas V) and produces 115,000+ solid rocket motors/year at Camden, AR, where a $215.6M expansion targets a six-fold large-SRM capacity increase. In January 2026 it agreed to sell 60% of its Space Propulsion & Power business (RL10, in-space, power, avionics) to AE Industrial Partners for $845M, retaining RS-25 and the SRM business.
- Bull
- If propulsion is the upstream choke, L3Harris's Aerojet Rocketdyne holds 150+ RL10 backlog and is six-fold expanding large-SRM capacity at Camden.
- Bear
- L3Harris is actively shrinking the exposure, agreeing to sell 60% of its space propulsion business for $845M, and space is tiny inside the prime.
Moog Inc. (Class A)
Moog builds integrated chemical propulsion modules at its Niagara Falls facility (two completed in 2025 for ESPA-Grande-class national-security buses), unveiled a Hall-effect thruster family for small/medium spacecraft, and in 2025 won an Air Force Research Laboratory contract to develop a dual chemical-electric multimode propulsion system using a single propellant tank. A new propulsion clean room is due to complete by summer 2026.
- Bull
- If Hall-effect and multimode thrusters scale, Moog builds integrated propulsion modules, launched a Hall-effect thruster family, and won an AFRL multimode-propulsion contract.
- Bear
- Space propulsion is a small piece of diversified motion-control firm Moog, and its new clean room and thruster lines are still ramping toward production.
Redwire Corporation
Redwire is co-developing the Valkyrie Hall-effect thruster (based on NASA's H71M design) in a strategic JV with private propulsion firm Phase Four, targeting high-volume DoD smallsat supply; the company has cited up to $150M annual revenue potential, which is a market-opportunity figure rather than booked backlog. Redwire reoriented in 2025 toward defense tech via the Edge Autonomy acquisition.
- Bull
- If smallsat propulsion scales, Redwire's Valkyrie Hall-effect JV with Phase Four targets high-volume DoD supply, with cited up to $150M annual revenue potential.
- Bear
- The $150M is a market-opportunity figure, not backlog; Redwire's revenue increasingly reflects Edge Autonomy drones, a textbook space-washing dilution of the thesis.
Rocket Lab Corporation
Rocket Lab acquired SolAero in January 2022 for $80M; SolAero is one of only two US space-grade solar cell producers (the other is Boeing's Spectrolab). Q1 2026 revenue was $200.3M with a $2.2B backlog, and the company signed preliminary CHIPS Act terms for up to $23.9M (plus $25.5M from New Mexico) to expand compound-semiconductor output ~50%. In February 2026 it added silicon arrays targeting gigawatt-scale space data centers.
- Bull
- If a second reusable launcher and space-grade supply scale, Rocket Lab owns SolAero (one of two US cell makers), $2.2B backlog, Neutron, and Mynaric optics.
- Bear
- The reusable Neutron has not flown after a tank rupture, so much of the launch upside remains option value dependent on proving booster recovery.
SpaceX (Space Exploration Technologies Corp.)
SpaceX priced its IPO at $135 and debuted on Nasdaq June 12, 2026 under SPCX, closing its first session at $161.11 (up ~19%) in the largest IPO in history, making it one of the largest US public companies. Starlink, with thousands of satellites in orbit and ongoing launches, is the dominant single source of demand for space-grade solar power.
- Bull
- If launch stays one company's market, SpaceX flew ~50% of 2025 launches and 80%+ of mass, and Starlink is the dominant single source of solar-power demand.
- Bear
- The IPO floated only ~4.3% of shares, so SPCX trades on a sliver of float, and the 2025 Space segment ($4.1B) carried an operating loss.
Catalyst calendar
- 2026-07-23L3Harris Q2 2026 earnings (before open)mediumFirst print since the $845M AE Industrial majority-sale agreement — watch for closing timeline and how much RL10/in-space propulsion revenue is being deconsolidated versus retained SRM/missile propulsion.
- 2026-08-06Rocket Lab Q2 2026 earningsmediumLooks for Gauss thruster order intake and Archimedes engine qualification progress — the bridge from a 200-unit/year line to recognized propulsion backlog.
- 2026-12-31Expected close of L3Harris / AE Industrial 'Rocketdyne' majority sale (H2 2026)mediumOn close, RL10 and in-space propulsion move to a 60%-AE-Industrial entity, structurally shrinking LHX's space-propulsion exposure and re-rating it toward missile SRMs.
- 2026-08-2340th Annual Small Satellite Conference (SmallSat 2026), Salt Lake CitylowPrimary venue where electric-propulsion merchants and constellation buyers set thruster procurement terms — the demand-side signal for the tightest node in this chokepoint.