Filling The Falcon 9-Shaped Hole: Stoke Space's Bet On Full Reuse

Space Exploration Technologies

Across the shows: 53 speakers in other episodes, 71 bullish and 25 bearish calls, past 30 days

• Devon Papandrew · Stoke SpaceUncertain

  Starship reaching high-cadence flight is not guaranteed and faces major technical challenges by next year.

  Papandrew says he doesn't bet against SpaceX, but he sees massive technical challenges between where Starship is today and flying at a very high cadence. He says he won't assume that happens by next year, as some people do. He notes SpaceX has a strong economic incentive to use the up-mass for its own Starlink and other constellations, so the capacity may not be available to outside customers.

  “I don't bet against SpaceX, but I I... I'm not gonna assume that's a foregone conclusion and it just happens by next year.”at 0:06

• Brett Winton · ARK InvestBullish

  SpaceX's booster reuse reliability has made reusability a clear winner, beating earlier cost spreadsheets.

  Winton recalls an ARK spreadsheet showing reuse only penciled out if landings succeeded about two-thirds of the time, a sufficient reuse rate. He says SpaceX's landing reliability has turned out much better than that, and that the reliability is what makes reuse clearly undeniably a winner. The cost estimate's large delta has been overtaken by how good SpaceX has been in practice.

  “As it turns out, you know, they're really good. And being really good, it's clearly undeniably a winner”at 17:55

• Devon Papandrew · Stoke SpaceBearish

  SpaceX's economics will push it to withdraw Falcon capacity, leaving a gap in the launch market.

  Papandrew says natural economic forces would push SpaceX to retract Falcon capacity as it shifts to Starship. Satellite operators have designed constellations around Falcon 9 and now face a supply gap. He says customers are already worried about access to space in the 2029 timeframe, and that the launch supply is constrained right now. Stoke's Nova Block 2 is positioned to fill that gap.

  “there'd be natural economic forces that pushed SpaceX to kind of retract Falcon capacity from the market.”at 20:26

• Brett Winton · ARK InvestBearish

  SpaceX's vertical integration into AI leaves rivals unlikely to rely on it for launch, opening room for others.

  Winton says SpaceX is vertically integrated through Starlink and also through its Grok AI system. Some well-capitalized competitors are challenging it in that stack. He argues those competitors are unlikely to depend on SpaceX as their launch supplier. He concludes that a large vertical that needs launch from a provider other than SpaceX probably has to be served by somebody else. Winton says this is an interesting point and is offering it as an opening for other launch providers.

  “SpaceX is not just vertically integrated through Starlink. They're also vertically integrated all the way through Grokbot to the end kind of like AI system.”at 44:09

Space economy

• Devon Papandrew · Stoke Space

  The space economy needs more independent launch providers, since only a very short list serves other companies.

  Papandrew argues launch is supply-constrained, and customers' access to space will limit the space economy's growth. He says the companies trying to procure launches for constellations are not the ones forecasting SpaceX's up-mass. Competitors like Blue Origin, ULA and Rocket Lab are tied to their own programs and Space Force or Amazon demand, leaving very few providers serving other growing space companies. He says more providers are needed to keep availability high and prices affordable.

  “We need a diverse set of independent launch providers who want to serve the broader space ecosystem.”at 0:00

• Devon Papandrew · Stoke Space

  Rocket industry will stay vertically integrated for a while, since integration controls schedule, cost and coupled-stage risk.

  Papandrew says vertical integration will last for at least a little while. Doing most work in-house lets a company control schedule and cost risk, and removes supplier risk in a capital-intensive industry where time is money. Rocket stages are also highly coupled, so an upper stage cannot easily be designed to mix with any bottom stage without losing mass, performance and structural margins. Horizontal integration could come later, once designs mature like aircraft.

  “We have become very vertically integrated as an organization. ... one of the biggest benefits is being able to control schedule and cost risk.”at 13:06

• Brett Winton · ARK Invest

  Today's space economy is essentially Starlink plus defense, with broader commercial demand not yet here.

  Winton says the space economy can be rounded down to Starlink and Department of Defense, the two big vertically integrated users. He says commercial demand for access to space outside of those two is not really here yet, although Papandrew's customers want to be part of it. He asks what else exists beyond Starlink and defense national security, implying the broader market is still largely hypothetical.

  “I could round the space economy into Starlink and Department of Defense.”at 22:38

• Devon Papandrew · Stoke Space

  Demand for launch extends beyond Starlink, with communications and other constellations needing access.

  Papandrew argues Starlink is the big market today because it has solved the access problem. Communications broadly is a massive market, but others lack reliable access. Amazon has hundreds of satellites waiting on the ground for launch. Earth observation, navigation and timing, and signals intelligence companies used SpaceX rideshares to prove their models, and now want to scale from a few satellites to constellations of around a thousand.

  “Amazon has hundreds of satellites sitting on the ground waiting for launch.”at 23:46

• Devon Papandrew · Stoke Space

  Returning payloads to Earth is far more constrained than launching, and reusable return unlocks microgravity manufacturing.

  Papandrew says if launch access is constrained, access back down is 10 times more constrained. Materials and pharmaceuticals developed on the International Space Station get stuck because they cannot find a ride down, and life sciences payloads can die in orbit. A fully reusable rocket can carry payloads both ways. If return flights became routine, twice a week for example, orbital manufacturing could feed terrestrial supply chains in biotech, pharma and exotic materials, opening a large market.

  “access back down from spaces 10 x more constrained.”at 25:41

• Devon Papandrew · Stoke Space

  Satellites should not need redesign for reusable launch; Stoke keeps traditional fairings rather than standardising on Pez-style.

  Papandrew argues customers should not have to redesign satellites to benefit from full and rapid reuse. Stoke uses a traditional fairing that separates with a 180-by-360-degree deployment hemisphere, so customers can use existing designs and still gain access and lower costs. He says Stoke could accommodate a Pez-style dispenser if the industry standardises on one in five to ten years, but the reverse is hard, since existing satellites are optimised for their applications and a thin form factor doesn't suit high-power optics.

  “customers should not have to redesign their satellites to benefit from the economics of full and rapid reuse.”at 32:44

• Devon Papandrew · Stoke Space

  Reusable upper-stage reentry must be extremely reliable, given FAA scrutiny and overflight of the populated US.

  Papandrew argues that when an upper stage reenters, it will probably come back over the populated US. Even a failure in one in 10 or one in 50 flights would be unacceptable because debris could hit a large part of the country. He says the FAA would not allow a vehicle that failed that often to fly. He agrees with Brett Winton that reliable reentry matters most when returning valuable or human payloads. Practical reliability therefore has to be extremely high in any architecture that aims for full reuse.

  “If you're reentering an upper stage, you're probably coming back over the populated US... the reliability just from a practical standpoint has to be extremely, extremely high.”at 45:43

AI infrastructure & data centers

• Brett Winton · ARK Invest

  Orbital compute should beat terrestrial data centers once launch costs fall to around $300 per kg.

  Winton's spreadsheet estimates that the cost of serving compute in space will pierce the cost of building terrestrial data centers at a launch cost of about $300 per kilogram. He admits the figure will almost certainly prove wrong but says it points in the right direction. His logic is that solar power in sun-synchronous orbit is nearly free and continuous, and AI now provides a way to turn that power into economic output in orbit rather than bringing it back down. He sees in-space manufacturing as a largely undiscovered area that this could unlock.

  “The current spreadsheet we have that is almost certainly gonna turn out to be wrong, but is definitely in the right direction, is that the cost of serving FLOPS in space is just going to pierce past the cost of building terrestrial data centers on the ground at around the $300 per kilogram launch cost.”at 43:08