Escape Velocity For The Final Frontier

“Earth is the cradle of humanity, but one cannot remain in the cradle forever.” — Konstantin Tsiolkovsky, 1911

Since the dawn of humanity, mankind has looked up in awe at the burning night sky. This is one of the few truly universal human experiences that we share with our ancestors stretching millions of years across time. From the animated minds of ancient tribal people passing stories from generation to generation, through to Apollo’s lunar triumph in 1969: our stellar ascent is a reflection of the boundless human spirit. Today, amid reusable rockets slashing costs and private industry igniting innovation, we are entering a new era of space as the sci-fi of yesteryear begins to manifest. Amidst the turbulence of the modern era, mastery of our cosmic environment is uniquely capable of propelling us toward abundance, resilience, and unity.

The world is crossing from experimental spaceflight into industrial activity in orbit and on the Moon. Over the next ten years, falling launch costs, maturing regulation, and novel commercial demonstrations will open a multi-trillion dollar opportunity across communications, Earth observation, orbital computing, exotic manufacturing, servicing, and lunar infrastructure. The investable opportunity is no longer theoretical; it is expanding with flight cadence, customer demand, and enabling policy. 

This report explores the coming age of space entrepreneurship through a single mechanism: launch cost. The argument is not simply that space will be large. It is that every order-of-magnitude collapse in mass-to-orbit creates a new industrial architecture, and therefore a new venture market. We open with the status quo as 2026 finds it, then examine five price thresholds ($1,500/kg, $500/kg, $200/kg, $50/kg, and $10/kg) and the qualitatively new classes of business each makes obvious. The first three thresholds are already investable against present-day technology. The fourth opens serious commercial activity on and around the Moon, though the timing remains hostage to Starship cadence, regulation, and anchor-customer demand. 

Every industrial revolution is catalysed by a price collapse in a single input. Steam. Steel. Transistors. Bandwidth. The price of venturing beyond our atmosphere is collapsing now, and the industries that exist on the other side of this wave will be unrecognisable. My hope in writing this is not only to elucidate the marvelous sophistication of this new landscape, but to evoke some sense of curiosity for what lies beyond. The 21st century has set the stage for humanity’s inexorable advance along the Kardashev scale, driven by an insatiable demand for energy in an ever more sophisticated civilisation. If we are to spread the light of consciousness beyond our pale blue dot, the window to do so appears to be opening. It may not stay open forever.

TL;DR

• Launch-cost compression is coming faster than most think. The trajectory from ~$1,500/kg today to <$500/kg, <$200/kg, <$50/kg and eventually ~$10/kg defines five distinct eras. Each threshold changes what can be built, who the customer is, and where venture-scale value can accrue.

• The investable opportunity is already here, but it is unevenly distributed. The $1,500/kg and $500/kg eras are producing revenue today across megaconstellations, Earth observation, defence space, autonomy, ground infrastructure, early orbital compute, and reentry. The $200/kg and $50/kg eras contain the fatter industrial upside, but should be underwritten as staged optionality rather than present-day certainty.

• The consensus undershoots our estimates. McKinsey/WEF model a $1.8T space economy by 2035 at ~9% CAGR. Our bull case is materially larger at $7T by 2050, but the report should be read as a probability-weighted map of unlocks rather than a straight-line forecast. The asymmetry comes from the market still pricing space as one industry, while the cost curve is beginning to unlock many.

• Starship cadence is the catalyst. If Starship reaches even a conservative version of its public cadence targets, eight downstream industries re-price at once. If cadence stalls, the $500/kg era still opens, but the $200/kg-and-below thesis slips by years. The single most important catalyst before 2030 is not another demo flight; it is repeatable cadence, rapid reuse, and high payload utilisation.

• The clearest alpha sits in bottlenecks. Six pockets look most attractive at pre-seed/seed: space situational awareness and debris removal; autonomy and rendezvous software; spectrum-aware operations and ground infrastructure; reentry/downmass logistics; power, thermal and radiation-hardening systems; and application-layer companies whose regulatory or data moats compound as space infrastructure gets cheaper.

• Orbital compute is the swing factor. A hyperscaler, SpaceX/xAI, or a vertically integrated new entrant committing serious production capex to orbit before 2030 would move the bull case toward the base case. Until then, orbital compute should be treated as the highest-upside and highest-scrutiny claim in the report: enormous if true, but gated by radiation, latency, thermal management, maintenance, insurance, and customer willingness to move workloads off-planet.

• What would change our mind? Kessler-syndrome cascade, the cost curve stalling below 10x compression, regulatory and geopolitical freeze, demand failing to materialise, or anchor-tenant withdrawal compounded by SpaceX concentration risk. None kills the thesis outright, but each can move the timeline far enough to matter for venture returns.

• The $10/kg era is civilisational upside, not seed-stage underwriting. ISRU, Helium-3 markets, point-to-point Earth travel, and the Mars supply chain are speculative, though plausible. They give the report its horizon, but the near-term investment case lives in the bottlenecks created by the first few eras.

The investable thesis

Our focus is not whether “space” is a good sector to invest in, we are most interested in analysing which bottlenecks appear when access to orbit becomes abundant. Cheaper launch does not automatically make every space business work. It changes the binding constraint. At $1,500/kg the constraint is often downlink, spectrum, data processing, and customer adoption

. At $500/kg it becomes operations, servicing, downmass, and station economics. At $200/kg (2030+) it becomes quality control, reentry cadence, and proof that microgravity products are better enough to justify the trip. At $50/kg it becomes lunar logistics and the existence of a real non-government customer.

That is where we want to hunt. Not in highly-contested capital-intensive domains such as launch, but in the perhaps more obscure though necessary infrastructure that every serious operator will need if the cost curve keeps collapsing: autonomy, debris avoidance, orbital risk, spectrum software, power and thermal management, reentry logistics, lunar comms, and the application layer that turns space-derived data or materials into terrestrial decisions. To us, the most interesting founders in 2026 are the ones building for tomorrow’s market. 

Why now

The consensus answer is “Starship”, though that’s not the complete picture. Seven independent accelerants are stacking, and it is the combination rather than any single one that makes the coming decade categorically different from the last.

1. Cadence. SpaceX flew 134 orbital missions in 2024, more than the rest of the world combined, and pushed past 165 flights in 2025. Last year, 100% were successful in deploying their payloads, and 98% in returning to the landing pads. Its internal target for Starship, across the Starbase Gigabay (~700,000 sq ft) and the Cape Canaveral Gigabay (~815,000 sq ft), is to produce one vehicle every three to four days by 2027.

2. Reusability matures. Falcon 9 boosters now routinely fly 25–30 missions, with booster B1067 past 32 flights by the end of last year. Starship has demonstrated full payload deployment and controlled splashdown, chopstick landing, and a massive step up in the latest Starship overhaul sporting the Raptor V3 engines. The remaining question is iteration speed, and iteration speed is what SpaceX is best on Earth at.

3. Rideshare pricing in the open. SpaceX publishes rideshare at $325,000 for 50kg, with $6,500/kg beyond. Internal cost sits near $1,500/kg. That gap, more than four-fold, is the margin SpaceX retains to fund Mars. It is also the lever that compresses the moment a credible competitor (Rocket Lab Neutron, Blue Origin New Glenn, Stoke Space Nova, now backed by $1.34B raised across Series C and D) appears at similar price levels.

4. Orbital compute crosses the viability line. Kepler Communications and Spiral Blue have independently demonstrated payload data reduction north of 99% through edge ML. Starcloud is building dedicated orbital data-centre hardware, a thesis that read as science fiction in 2019 and is funded, hardware-shipping infrastructure today. In May 2026 Aetherflux rebranded as Cowboy Space Corporation, raising $275M at a $2B valuation led by Index Ventures to build vertically-integrated orbital data centres and the rocket to launch them on. The pivot from SBSP to orbital compute is the most direct evidence we have that the market is converging on our $500/kg-era thesis ahead of schedule.

5. Lunar cadence shift. CLPS contracts funded a record cohort of private lunar landers across 2024–2025. Intuitive Machines became the first commercial entity to soft-land on the Moon (IM-1, February 2024). Firefly’s Blue Ghost 1 executed a fully successful surface mission in March 2025, and the company IPO’d at a $8.5B post-debut valuation off the back of it. Five private lunar missions are scheduled across 2026: Blue Origin’s Blue Moon Mark 1, Firefly’s Blue Ghost Mission 2, Intuitive Machines’ IM-3, Astrobotic’s Griffin-1, and Team Draper’s APEX 1.0. Artemis is funded into the mid-$90Bs and rising. China’s ILRS targets a crewed South Pole presence in the 2030s. The Moon is no longer NASA’s alone.

6. Capital at scale. 2024 directed over $10B into space startups. 2025 closed at $55.3B across all rounds per Space Capital’s Q4 IQ, with Q4 alone delivering $17.0B across 135 rounds. The late-stage skew (Anduril, SpaceX secondaries, Relativity recap, Firefly’s $8.5B public debut, Sierra Space’s $550M Series C, Stoke’s combined $1.34B raise, Cowboy’s $275M Series B) confirms the asset class is maturing. Crossover funds (Coatue, Tiger, Lux, T. Rowe Price) are now underwriting Series B+ at growth ratios that look closer to enterprise SaaS than to traditional aerospace.

7. SpaceX IPO. The rumored Nasdaq listing under the ticker SPCX targets a historic $1.75 trillion valuation, looking to pull in a primary raise of roughly $75 billion. This more than doubles Saudi Aramco’s 2019 record to become the largest IPO in history. Elon Musk retains tight 80% voting control and is not selling personal shares. The capital is a massive primary injection meant to stay internal, hyper-accelerating Starship mass production, Starlink Gen 3, Starshield, and Mars. Instead of a horizontal cash spillover, the IPO acts as the ultimate macro validation event. 

Each of these components is compounding, driving a dramatic acceleration in extraterrestrial capabilities. With the recent successful operation of the James Webb Space Telescope and Artemis returning humanity’s attention to the Moon, the curious amongst us are once more drawn to the mystery of space. If the base case holds, the coming decade shall unleash a level of wonder about our place amongst the stars not seen since Apollo. Not merely flags and footprints, but the beginnings of will service our expansion deeper into the solar system. 

As Carl Sagan once said: “Our remote descendants, safely arrayed on many worlds throughout the Solar System and beyond, will be unified by their common heritage, by their regard for their home planet, and by the knowledge that, whatever other life may be, the only humans in all the Universe come from Earth.”

Status quo, 2026

The global space economy crossed roughly $630B in 2024 and is projected to cross $1.8T by 2035 per McKinsey’s space-economy work, with Space Foundation, SIA, and McKinsey estimates converging within ~15% of each other. The bulk today is still downstream (satellite services, GPS, broadcast, broadband). The upstream layer (launch, manufacturing, ground) is the part being re-architected. Most readers underestimate how much of their daily life is already routed through space infrastructure: every GPS-enabled rideshare. Every weather-aware logistics route. Every banking transaction time-stamped by GNS. Every precision-agriculture yield map. Every disaster-response coordination. Every aviation positioning fix. The U.S. Bureau of Economic Analysis now treats the space economy as a distinct sectoral measurement precisely because its contribution to GDP-touching activity has crossed the threshold where rolling it into “telecommunications” misrepresents the structure.

Commercial share of total is now ~71% per the SIA 2025 State of the Satellite Industry Report and rising. Government share is shrinking in percentage terms even as absolute government spend grows. Crucially, space has crossed from a government-led industry with commercial participation to a commercial-led industry with government backstop.

Global orbital launch volume hit 261 successful missions in 2024, with SpaceX accounting for 134; 2025 closed above 250 commercial launches with SpaceX at 165. For comparison, the entire world flew fewer than 100 orbital missions per year from 2000 through 2017. Annual cadence has roughly tripled in seven years, and the trajectory is still steepening.

The commercial satellite population in orbit is now dominated by Starlink (10,000+ active by early 2026, roughly two-thirds of all active satellites in orbit), followed at a distance by OneWeb (~650), Amazon Kuiper (production constellation ramping), China’s Guowang and Qianfan (deployment accelerating), Planet (~200), and a long tail of EO, comms, and experimental platforms.

The capital stack underneath this has matured in lock-step. The 2022–2024 SPAC chill (Virgin Orbit’s bankruptcy, Astra’s delisting, Momentus’s troubles, Spire’s recapitalisation) flushed out the weakest players. What remains is a more concentrated bench of credible operators, supported by space-dedicated funds (Seraphim Space, Alpine Space Ventures, Type One Ventures), crossover allocators writing larger cheques into Series B and later, and sovereign capital from Saudi Arabia, the UAE, and the EU’s IRIS² programme putting real numbers behind regional sovereignty. A meaningful and growing cohort of emerging managers (we count more than thirty active fund-one operations in 2026 against fewer than ten in 2022) is filling the pre-seed and seed layer of the stack. The shape of capital looks closer to the cloud-infrastructure stack of 2014 than the satellite-investing landscape of 2018.

The companies founded today are the $1,500/kg and $500/kg incumbents of 2035, and the founders who pick the right era at the right time will build the way the 2005–2012 internet founders did. Stripe was founded in 2010 on the bet that online payments would become a horizontal layer rather than a vertical service. AWS launched commercially in 2006 against the consensus that enterprise IT would stay on-premise. WhatsApp launched in 2009 against the consensus that SMS was a moat.

The historical rhyme that matters most here is container shipping. When Malcolm McLean loaded the SS Ideal-X with 58 standardised containers in April 1956, the cost of loading a ton of freight collapsed from $5.86 to $0.16, a 36x compression in a single innovation cycle. The first decade was speculative. The second decade reorganised global manufacturing around it, factories no longer needed to be near their customers; they could move to wherever labor was cheapest (e.g. Asia) because shipping costs were no longer a limiting factor. By 1980 the volume of containerised trade was accelerating massively; and has grown a further 24x now shipping 13 billion tons a year. Every supply chain on Earth is downstream of that one mechanic, and the firms that won the era were not just the shipping companies but the ones that built the port infrastructure, the routing software, and the consumer-facing brands around it such as IKEA or Walmart. Launch cost in 2026 is exactly analogous.

One caveat worth naming up front: these price points are LEO mass-to-orbit at the rideshare or dedicated-bus level for Starship-class payloads. They are not the average across all launch providers. They are what a credible base-case Starship economy produces once cadence hits target. The appendix contains the full Bear/Base/Bull cadence model.

The Delphi market map

The $1,500/kg Era (2025–2027): Industrial Rocket Capacity

We are in the $1,500/kg era right now. That is what Falcon 9 delivers a kilogram of payload to LEO for, internally; what SpaceX’s own books say the kilogram costs. The published rideshare rate sits at $325,000 for fifty kilograms, a roughly fourfold markup the company retains to fund the next vehicle. As the new competitors (Blue Origin, Rocket Lab) emerge, this will compress.

For the first time in the history of the sector, the price a serious operator pays is low enough that an orbital business plan can be underwritten off a concrete risk profile. Everyone from biotech, to industrial manufacturers, to orbital data center pilots can now book slots to validate early prototypes. This is the era where a generalist VC can write a cheque and be wrong only about a specific company, not about the category.

Four categories anchor revenue today

Megaconstellations are no longer a thesis; they are live infrastructure servicing millions of customers around the globe. Starlink fields more than ten thousand active satellites, roughly two-thirds of every active spacecraft in orbit, and boasts over 10 million active subscribers supporting ~30 million daily internet users. Amazon Kuiper is mid-deployment. China’s Guowang and Qianfan combine for over a thousand planned, with launch cadence accelerating quarterly. Europe’s IRIS² is funded. The argument has been won; the question is which constellations consolidate into standards, and on what terms.

The vertical specialists each take a different slice of the market. K2 Space builds large platform-class buses optimised for defence and ISR payloads where Starlink’s commodity bus is the wrong shape, taking advantage of MEO to do more with fewer satellites. Apex Space ships productised buses (their Aries class) on standardised timelines, treating the spacecraft chassis as a product SKU rather than a custom build. Astranis focuses on small GEO comms satellites for regional broadband, an entirely different mission profile from the LEO megaconstellations. Each is viable on its own terms and more valuable to the overall network than to any single operator.

Earth observation as a service did the unglamorous work of teaching enterprises that fresh imagery has API value. Planet Labs and Maxar’s WorldView Legion did that for the optical layer. The current wave is vertical-specific: ICEYE and Capella in synthetic aperture radar, HawkEye 360 in radio-frequency geolocation, Satellogic in agriculture, Umbra and BlackSky filling the long tail. The differentiator has moved from the pixel to the inference. On-orbit edge ML now compresses a ten-terabyte daily downlink into a ten-gigabyte signal before it touches the ground, which alters both the unit economics and the customer mix. Downlink remains surprisingly scarce and that scarcity is one of the reasons inference, compression and tasking software matter as much as sensor quality.

A growing application layer sits on top of all of this, taking satellite imagery as a feedstock and producing decisions: KoBold Metals (backed by Gates, Bezos, Sam Altman; valued at $2.96B in its latest round) uses satellite imagery alongside geophysical data to identify mineral deposits, having discovered the Mingomba copper deposit in Zambia that projects 300,000 tonnes of annual production by 2030. Earth AI targets indium, nickel, and palladium with a reported 75% drilling success rate against the industry’s historical <1%, and newer startups such as Esper operate in adjacent domains. These are not space companies in the legacy sense; they are application-layer firms whose product would not exist without the underlying satellite layer being cheap. Expect this pattern to repeat across mining, energy, insurance, defence intelligence, and ESG compliance.

On-orbit compute is the category that has recently become glamorous (more on this in the next era). Starcloud is building dedicated orbital data-centre hardware, a thesis that read as science fiction in 2019 but has tested an active, off-world H100 today. In May 2026 Cowboy Space Corporation (the rebrand of Aetherflux, founded by Robinhood co-founder Baiju Bhatt) closed a $275M Series B at a $2B valuation led by Index Ventures, on a thesis of vertically-integrated orbital data centres delivered by their own purpose-built launch vehicle (first flight targeted by end-2028). That said, most of this category revolves around edge ML right now such as Kepler Communications and Spiral Blue who have independently demonstrated payload data reductions north of 99% through edge inference.

Spacecraft autonomy as a service is the newest of the four and the one we expect to be most underpriced by generalist investors over the next twenty-four months. A constellation at Starlink scale contains more spacecraft than any operations team can meaningfully fly. Collision avoidance, rendezvous and proximity, station-keeping, on-orbit servicing: all of it requires the spacecraft to think for itself. EraDrive’s $5.3M oversubscribed seed in December 2025 was the cleanest early entry: AI-powered vision that lets spacecraft navigate, operate, and collaborate without a ground controller in the loop. The economic case for in-space servicing, for orbital data centres, for cislunar logistics: all of it routes through an autonomy stack that does not yet fully exist. Whoever builds it first occupies the position ROS holds in robotics today: a de-facto standard that every downstream operator builds against, with the strongest network effects accruing to whoever ships the SDK first.

Spectrum: the moat nobody talks about

The structural moat Starlink owns is not only its constellation. It is its spectrum allocation. The International Telecommunication Union coordinates global frequency rights for orbital communications, and the windows in which non-geostationary systems could file under favourable rules closed quietly in the early 2020s. Starlink and OneWeb were grandfathered; Kuiper and most subsequent entrants face tighter sharing requirements. The FCC-coordinated interference fights between Kuiper and Starlink over Ka-band and V-band downlinks are a preview of every future megaconstellation negotiation.

SpaceX’s August 2021 acquisition of Swarm Technologies is the cleanest illustration. Swarm operated a sandwich-sized IoT constellation in the 137-138 MHz and 148-150.05 MHz VHF bands, a spectrum allocation worth far more than the operating business itself. The transaction price was rumoured to be $524m, but the strategic logic was: Swarm came with FCC-licensed VHF rights for non-voice satellite mobile service that SpaceX could roll into Starlink’s direct-to-device roadmap without filing from scratch. The lesson: spectrum is acquired as an asset class of its own, not as a side effect of operations. Expect many serious operators from 2026 onward to be running an active spectrum-M&A book.

Why this matters for the $1,500/kg era thesis: the satellite count is a manufacturing problem, which SpaceX has solved. The spectrum count is a coordination problem, which persists. Downlink bandwidth is the binding constraint on what every constellation can actually deliver to end users, regardless of how many satellites are in the sky. A 10,000-satellite constellation with insufficient downlink spectrum delivers less total throughput than a 2,000-satellite constellation with grandfathered Ka-band rights. Europe’s IRIS² is partly a sovereign answer to this — a consortium-led commercial constellation built to claim spectrum rights ahead of foreign operators. China’s Guowang is the same instinct executed in reverse, with the state itself filing the spectrum claim and assigning operators within its framework rather than building it through a private consortium. A bullish view on $1,500/kg-era comms is, in part, a bullish view on regulators delivering coherent multi-jurisdictional spectrum frameworks. Most of them have not.

The investible angle here is the layer underneath: spectrum-aware satellite operations, dynamic frequency management, software-defined radios, and the regulatory-arbitrage operators that help new constellations file under viable bands. Companies like Kratos and Aurora Insight in the U.S., and a growing cohort of European specialists, are quietly profitable in this category.

The ground segment, half the constellation

Half the value chain in commercial space sits on the ground. Antenna manufacturing (Kymeta, ALL.SPACE, Isotropic Systems, Cesiumastro), ground-station-as-a-service (AWS Ground Station, KSAT, Goonhilly Earth Station, Atlas Space Operations, Leaf Space), and the optical-ground-station layer required by laser inter-satellite links (Mynaric, Tesat-Spacecom, Skyloom, BridgeComm, Cailabs).

Optical inter-satellite links are the unsung accelerant of the $1,500/kg era. Starlink works as a mesh because each satellite has four laser terminals routing traffic optically between satellites without ever touching a ground teleport. The result is that a packet originated in Sydney can route through space directly to a receiver in Lagos, bypassing the entire terrestrial cable network. The economics of that are categorically different to the previous architecture where every packet had to be downlinked to a fibre teleport and re-routed. The Mynaric (acquired by Rocket Lab for $155m) and Tesat ($250m annual revenues) duopoly on commercial optical terminals is one of the cleanest oligopolies in the sector, and the U.S. SDA’s Tranche programmes are now built on the assumption that every constellation member runs optical links by 2028.

The second engine: defence

The defence lane is structurally distinct from the commercial one and is doing equal work in this era. The Space Development Agency’s tranche architecture and the proliferated NRO constellations have changed what “fast” means inside the Department of Defense. Tranche 0 was awarded at roughly $1.3B; Tranche 1 expanded to roughly $1.8B across multiple primes; Tranche 2 contracts are pacing toward $2.5B. The Space Force’s FY26 budget request crossed $30B for the first time, with the procurement-of-services portion (rather than primes-and-platforms) the line item rising fastest. Prime contracts now move in months, not years, with an explicit mandate to buy from commercial operators.

Anduril, K2 Space, Apex, Varda, Xona Space Systems, True Anomaly: their near-term growth is not principally a function of launch cost. It is a function of how quickly SDA can sign Tranche follow-ons. Being bullish on the $1,500/kg era necessitates procurement reform.

The deeper structural fact is that, by our estimates and supported by CSIS Aerospace Security analysis, roughly 30–40% of current revenue underneath the commercial veneer of this sector is classified or quasi-classified national-security work. We arrive at the range by aggregating disclosed Tranche values (~$5.5B/yr run-rate by FY26), NRO commercial-buy estimates (~$3B/yr per Space Capital tracking), NSSL Phase 3 ($5.6B awarded across launch primes), AFRL classified R&D budgets, and the European Defence Fund’s space track, against the total addressable commercial revenue line. The Tranche follow-on cadence is the single most important variable in this era that almost nobody outside the sector tracks.

The international parallel matters here too. The European Defence Fund is now writing nine-figure cheques into space-defence dual-use programmes. More structurally, the EU’s ReArm Europe / Readiness 2030 plan commits up to €800 billion in additional defence spending by 2030. A material fraction of that spend will flow into space-defence: sovereign launch, ISR constellations, secure satcoms, and missile-warning architectures. The UK’s National Space Strategy explicitly names sovereign launch and ISR as priorities. France, Germany, and Italy are each building national space-defence stacks underneath the EU layer. The geographic distribution of defence buyers is diversifying as fast as the commercial layer is.

The fringes are the leading indicator. Watch the founders raising sub-$10M rounds in 2026 against business models that depend on sub-$200/kg launch costs. The two or three who really nail it will be the $200/kg-era incumbents of 2032, the same way the 2014–2016 megaconstellation cohort, bootstrapped on anticipated Falcon 9 economics that nobody believed would deliver, became the $1,500/kg incumbents of today. The Sci-Fi fringe of 2026 includes Reflect Orbital (orbital mirrors selling on-demand sunlight to terrestrial solar farms), Lux Aeterna (reusable satellites with built-in heat shields, first flight Q1 2027 on a SpaceX rocket), and a quiet cohort of stealth founders working on orbital advertising (cry), on-orbit data physical-security escrow, and even space burials. 

In the $1,500/kg era, the money is in bits delivered through atoms: imagery, connectivity, signals, compute, and the autonomy stack on which all four scale. The atoms-in-space businesses come next.

The $500/kg Era (2027–2030): Maturing Orbital Infrastructure

The $500/kg era is the first in which the physical economy of space starts to emerge. At $1,500/kg the orbital business case is information; below $500/kg, it is also matter. Refuelling, life-extension, orbital habitation, reentry as a commercial service, and on-orbit servicing each have a customer-economics inflection somewhere in this band. So does space tourism at meaningful volume.

Starship operational, with cadence still ramping up to its 2027 target, is what gets us here. Blue Origin New Glenn with second-stage reuse, Rocket Lab Neutron flying commercially, and Stoke Space Nova clearing test campaigns out of Launch Complex 14 at Cape Canaveral are the supporting cast. The competitive structure of launch shifts in this era from “SpaceX and the rest” to “SpaceX, two credible heavy-lift competitors, and a layer of medium-lift specialists serving the long tail.” Margin compression is the predictable consequence; cadence growth is the more important one.

Five categories that move from pilot to production

Orbital data centres. This is the highest-upside claim in the report and the one that deserves the most scrutiny. Hyperscaler capex may begin to reorient in the second half of the decade if terrestrial grid capacity, permitting, water use, and geopolitical energy constraints keep tightening. We expect at least one of Microsoft, Google, or Amazon to commit first-party orbital compute infrastructure rather than purchasing it as a service. SpaceX itself is widely expected to make obvious moves into orbital compute to serve xAI’s needs, given the bandwidth and power-cost asymmetries that increasingly favour orbit for certain AI workloads. 

Between Starcloud and Cowboy Space, the first real tests emerge of whether launch, power, thermal, radiation-hardening, insurance, and maintenance can be bundled into a credible data-centre product. The 2027–2030 pilot-to-production crossover is a critical juncture to validate what may become one of the largest demand drivers for orbital infrastructure. That said, I still haven’t heard a good answer to how we shield 10s of billions of dollars of equipment from the next Carrington Event. See our interview with the Starcloud CEO below.

In-space servicing and refuelling at scale. Orbit Fab becomes infrastructure rather than concept. Starfish Space and Astroscale move from technology-demonstration revenue to recurring servicing contracts. Northrop Grumman’s MEV (Mission Extension Vehicle) scales from one-off life-extension into a fleet model. The gating constraint here is standards: a shared fuelling-port specification across Western primes is the precondition for the market clearing, and the SDA-led standardisation effort (CONFERS is the industry body coordinating it) is the most underweighted piece of policy news in the sector.

Commercial LEO destinations. Vast Space launches Haven-1 in 2026 and is the front-runner for the post-ISS era with Haven-2. Axiom, Sierra Space’s LIFE habitat and the Starlab consortium (Voyager + Airbus + Mitsubishi), and the Blue Origin / Sierra Orbital Reef are the four serious bidders for the NASA Commercial LEO Destinations program. ISS retirement in 2031 is the forcing function. By 2030, between two and four privately operated stations will be in orbit, with the surviving operators having locked in NASA-anchor revenue and beginning to attract pharmaceutical, semiconductor, and tourism workloads.

Reentry as a commercial service. This is one of the components Delphi cares most about in this era. You cannot have an in-space manufacturing thesis without a reentry thesis. Varda Space Industries demonstrated the round-trip with its Winnebago capsule and has now run multiple successful reentries to its Utah recovery zone. Inversion Space is preparing its Arc lifting-body reentry vehicle for first flight in 2026. Atomos Space, Outpost Technologies, and Sierra Space (Dream Chaser, $550M Series C in March 2026, first free-flyer demo flight late 2026) are each building dedicated reentry-and-recovery vehicles that can ferry payloads down on demand. Space Forge in the UK is doing the same with naval recovery in British waters. The Exploration Company in Germany operates its Nyx capsule for European routes. Lux Aeterna ($10M raised March 2026, founded by SpaceX veterans) is going further: building communications and EO satellites that are themselves reusable, with built-in heat shields that allow the whole spacecraft to return intact for refurbishment and re-flight. Hyperion will reenter into Portuguese waters and join this cohort.

The unit economics that matter here are not launch-cost; they are downmass cadence. The difference between a restricted reentry window and a rapid-turnaround return profile is the difference between a viable orbital product and an unviable one. Heat-shield economics–historically anchored by ablative materials like PICA-X and AVCOAT, but now shifting to SpaceX’s automated ceramic tile manufacturing and Stoke Space’s actively cooled metallic structures–represent a primary driver of vehicle refurbishment overhead. Reentry accuracy (narrowing from Apollo’s multi-kilometer splash zones to sub-kilometer target ellipses for modern guided capsules) determines whether an operator requires a prohibitively expensive naval recovery fleet or can achieve precise, land-based recovery. Furthermore, atmospheric chemistry from megacadence reentry is a regulatory tail risk worth watching: aluminum oxide loading in the upper atmosphere from deorbiting megaconstellations has become a credible 2030s political issue, potentially destroying ozone and impacting polar vortex winds.

The downmass layer rhymes with the launch layer in 2010: anchored by government-funded demonstrations, challenged by a handful of deeply capitalized commercial entrants, and approaching an obvious price floor. Most operators will not execute their own payload returns; the material science, geopolitical airspace clearances, and terrestrial recovery logistics unlock a clear market opportunity By 2030, Reentry-as-a-Service (RaaS) will transform downmass from an exotic, one-off mission profile into a standardized, recurring-revenue utility line, allowing orbital manufacturers to outsource the plasma physics and treat Earth-return as a simple line-item shipping cost.

Space tourism at scale. Axiom private-astronaut missions (currently approx ~$65m a ticket) becoming routine, and the first dedicated SpaceX free-flyer tourist mission combine to produce a roughly 100-passengers-per-year market by 2030. Suborbital (Blue Origin New Shepard, Virgin Galactic if still solvent) operates at higher volume but lower revenue per passenger. The category remains a small share of total space economy revenue, but will continue to drive attention to the heavens above.

The structural shift

At $1,500/kg the buyer of launch was almost always also the operator. At $500/kg the launch and operations layers cleanly separate, mirroring the path the semiconductor industry took between 1985 and 2005: TSMC’s foundry model commoditised silicon manufacturing, and the differentiation moved to fabless chip designers (Apple, Nvidia, AMD, Qualcomm) who never owned a fab. SpaceX is becoming the TSMC of launch: deliberately overbuilt manufacturing capacity, deliberately commoditised pricing for external customers, and the value capture moves up the stack to the operators who design the payloads and own the customer relationships. Commodity launch with operator differentiation is the default architecture of the next decade.

The $500/kg era is when the orbital economy stops being satellites and starts being infrastructure. The trip down matters as much as the trip up.

The $200/kg Era (2030–2033): Space Begins Production

At $200/kg the calculus reverses. This is the era in which in-space manufacturing crosses the breakeven threshold for at least three product categories, and a handful of others reach pilot-revenue status. The reentry layer built out in the previous era is now a core piece of infrastructure. The total addressable market for in-space manufactured goods sold back to Earth crosses $10B annually for the first time. Whilst small compared with terrestrial manufacturing, it’s significant milestone compared with the current upstream space economy.

The categories that turn on

In-space manufacturing. Varda Space Industries is the leader in making medicine in space. By now, they are regularly flying robotic mini-labs into orbit and bringing them back to Earth using their “Winnebago” capsule fleet. They are focusing on three types of drugs that gravity ruins on Earth: complex proteins used in cancer therapies (monoclonal antibodies), custom chemical combinations that are too fragile to make on Earth, and pills whose molecular shapes can only be perfectly formed in zero gravity. Space Forge in the United Kingdom has demonstrated semiconductor-grade compound production in microgravity (notably gallium arsenide, indium phosphide, and the newer ultra-wide-bandgap nitrides used in 5G and 6G RF front-ends), with a small-volume defence and high-end commercial customer base. Companies like Hyperion and Flawless Photonics are scaling ZBLAN-fibre production, unlocking 10–100x lower signal loss than terrestrial silica fibre optics, which is critical for photonics chips, mid-infrared sensing, and quantum-computing interconnects where the noise floor of terrestrial-manufactured glass is the bottleneck. The market structure here resembles early biotech: a handful of pure-play producers, a much larger field of vertical-specific applications layered on top, and gross margins inflated by the inability of any terrestrial substitute to replicate the product.

Reentry as half the value chain. By 2030 the reentry layer is not an afterthought; it is the equal partner to the orbital manufacturing thesis. A kilogram returned to Earth from a $200/kg launch costs roughly twice the launch number once you fold in recovery, heat shield, and ground handling, implying a round-trip cost of ~$600/kg. Inversion, Atomos, Outpost, Sierra Space, Lux Aeterna, and The Exploration Company become the FedEx of the orbital economy, and the spatial concentration of recovery zones (Utah for Varda, North Sea for Space Forge, Portuguese waters for Hyperion, German and French sites for The Exploration Company, Indian Ocean and Pacific drop zones for U.S. capsule operators) becomes a strategic question for sovereign capital.

Cislunar logistics. Impulse Space and Quantum Space are running regular kick-stage missions between LEO and cislunar space. A “kick stage” is a small dedicated propulsion bus that delivers a payload from a low-cost rideshare launch position to a high-energy destination orbit, much like a last-mile delivery van picking up cargo from a freight depot. The NASA Lunar Gateway is operational. Artemis cadence is two-to-three crewed missions per year. The supporting cargo logistics are commercially contracted rather than government-built. This is the era in which the lunar economy becomes an ongoing logistics route rather than a series of one-off missions.

Space-based solar power pilots. Caltech’s SSPP demonstrated wireless power transfer in 2023; by the early 2030s, Virtus Solis and the Japanese SSPS programme each have multi-megawatt orbital arrays demonstrating beamed power to Earth. The strongest counter to the SBSP-to-Earth thesis is that orbital compute and orbital manufacturing will consume the marginal kilowatt-hour in orbit before anyone bothers beaming it down: if a megawatt in LEO can either run a data centre or be beamed to a remote terrestrial substation at 70% efficiency loss, the data centre wins on dollar terms every time. SBSP retains a credible market in off-grid, military, and disaster-response applications where the price-per-kWh comparison is not against grid power but against diesel generators and battery storage. The bear case here is the strongest in the report and is engaged at length below.

Asteroid prospecting. AstroForge, TransAstra, and Karman+ are running prospecting missions to Near Earth Objects. None has commercialised return-to-Earth resources; all have credible business models in selling prospecting data to defence and commercial customers. Commercial extraction is a $50/kg-era story; prospecting is a $200/kg-era one.

The $200/kg era is when the economic value framework shifts. Instead of space being a “cost center” where we launch stuff that just dies and gives us data, it becomes a commercial factory that exports high-margin, physical goods back to the global economy.

The $50/kg Era (2033–2035): Industrial Lunar Foundations

The $50/kg era is the era in which the Moon begins to look less like a destination and more like real estate. Permanent lunar habitation crosses from Artemis-funded ambition toward commercial plausibility. Lunar surface operations move from sporadic scientific missions to a continuous logistics tempo that resembles, distantly, the early phase of Antarctic research-station infrastructure: government-anchored, increasingly commercial, internationally contested.

The lunar surface stack

Landers and rovers. Intuitive Machines, Firefly, Astrobotic, and ispace each operate regular cargo cadence to the lunar surface under successive CLPS-2 follow-on contracts. Lunar Outpost and Venturi Astrolab provide the rover layer, with FLEX-class rovers operational in both NASA Artemis and commercial mining-prospecting roles. Blue Origin’s Blue Moon Mark 2 carries heavy cargo for human-rated missions. The market structure here mirrors maritime shipping more than aerospace: route-based, cargo-class differentiated, increasingly automated.

Lunar comms and PNT. The ESA Moonlight consortium operates a four-satellite lunar relay constellation. Lockheed’s Crescent Space provides LunaNet-compatible commercial comms. By the mid-2030s, a Lunar Wireless backbone exists and is utility-class infrastructure, charged-for like terrestrial telecom rather than provided as government service.

Lunar resources and ISRU. Water ice at the lunar South Pole is the primary ISRU focus, not Helium-3… yet. Interlune is harvesting Helium-3 in pilot quantities, with the commercial takeoff gated by terrestrial fusion deployment timelines (the bear case for this entire category). Water extraction at the South Pole, primarily for in-situ propellant production, becomes the first economically self-sustaining ISRU loop if cadence, energy, storage, and local demand all arrive together. By 2035, a kilogram of LH2/LOX produced on the lunar surface could be cheaper than a kilogram launched from Earth even at $50/kg. That would be the first time in human history that any extraterrestrial commodity made more economic sense to produce off-world than to ship from home.

The composition of lunar regolith is the constraint and the opportunity. Roughly 40–45% of regolith by weight is oxygen, the single most abundant element. Silicon makes up another 20–25%, aluminium 10–18%, with iron, calcium, magnesium, and titanium together accounting for most of the remainder. Once you can extract oxygen from regolith at scale (a process now demonstrated in Earth-based laboratory analogues), every other downstream ISRU activity becomes economically tractable: propellant, breathable atmosphere, water for life support, sintered habitats, solar-array substrates, and ultimately the structural metals for off-Earth manufacturing. Rare-earth and platinum-group prospecting on the Moon is still pre-commercial by the end of this era.

Off-world hotels. GRU Space and at least one Chinese counterpart are operating early commercial lunar surface tourism, in the same scale-and-photogenic role suborbital tourism played in the previous decade.

Commercial demand beyond the government customer

This is the open question on which the entire era hinges, and it is honest to flag that it is the open question. Government lunar customers (NASA, ESA, JAXA, CNSA, ISRO) underwrite the bulk of revenue at the start of this era. By the end of it, a credible commercial customer base for lunar ISRU products and for cislunar logistics services exists. We expect this to come from the orbital habitation and manufacturing layers established in the previous era, not from terrestrial demand for lunar resources directly.

The $50/kg era is when the cislunar economy becomes self-sustaining in propellant and early ISRU. By the end of it, more humans work off-Earth than at any point in history.

The $10/kg Era (2040+): When the cradle empties

Most of what follows this threshold is speculative, though inevitable in direction. At $10/kg, mass-to-orbit is not the binding constraint on any space-based business. Four things become possible:

The ISRU economy at scale. Regolith-derived infrastructure on the Moon and Mars: landing pads, solar arrays, habitats, radiation shielding, water for propellant and life support. In-space construction stops being assembled-from-Earth and becomes built-on-site. The first wholly off-Earth-manufactured spacecraft is built and launched from the lunar surface in this era, paving the way for a self-expanding industrial footprint, and foreshadowing the eventual rise of quasi-Von Neumann probes that will one day colonise our solar system and beyond.

Helium-3 markets and the fusion link. If and when fusion arrives at commercial scale, He-3 from lunar regolith becomes a tradable commodity at industrial volume. This is largely contingent on fusion deployment timelines that today’s most credible roadmaps place in the late 2030s to 2040s. If fusion ships, He-3 is a meaningful market multiple orders of magnitude above its current niche. If fusion does not, He-3 remains a small commercial economy serving MRI, neutron detection, and quantum-cryogenics demand. Water and regolith-derived propellant remain the more grounded ISRU plays in either case.

Point-to-point Earth travel. Starship E2E at $10/kg-class economics produces a London-to-Sydney trip in 45 minutes for a price competitive with first-class commercial aviation. The gating constraints are regulatory, not technical: airspace integration, noise near spaceports, and passenger acceptance. By 2045 a small number of E2E routes are operational; by 2050 the model is competitive with subsonic long-haul on roughly a tenth of city pairs.

The Mars supply chain. Musk’s true calling and life’s mission. The $10/kg era is when “make life multi-planetary” has physical infrastructure behind it: regular cargo cadence to Mars, a working ISRU loop on the surface for water and propellant, and the first multi-thousand-person settlement. For the first time in the four-billion-year history of carbon-based life on this planet, we will have spread the light of consciousness to an alien habitat.

‘I’d like to die on Mars, just not on impact’ –- Elon Musk

This is also the era in which humanity becomes meaningful on the Kardashev scale: the energy budget of a civilisation that is harvesting solar at scale from orbit, fusing Helium-3 mined from the Moon, and operating outside Earth’s atmosphere is a dramatic step toward a Type I civilization.

The composition of progress at the frontier

The sovereignty thesis we opened with is a key piece of this report, and the geography of who is building deserves its own section. The United States remains dominant, anchored on SpaceX. But the architecture underneath is diversifying faster than the headlines suggest.

China. The commercial sector here is not a single state actor; it is roughly twenty private launch and satellite companies competing inside a state-supported framework. LandSpace flew the first methane-powered orbital rocket in 2023. Galactic Energy, iSpace (the Chinese company, not the Japanese lunar lander operator of the same name), CAS Space, Orienspace, and Deep Blue Aerospace are each pacing toward operational reusable launch by 2027–2028. The Guowang and Qianfan constellations together target 25,000+ satellites and will be operational at meaningful scale by 2030. China’s ILRS (supported by Russia) lunar programme targets a basic South Pole station by 2035. The strategic posture is straightforward: parity with the U.S. on every layer by 2035, leadership in lunar by 2040. The constraint is hardware (export-controlled chips, restricted access to Western IP) more than capital.

Europe. ESA, the European Commission’s IRIS² programme, and a layer of credible national agencies (CNES in France, DLR in Germany, ASI in Italy, UKSA in Britain, the Portuguese Space Agency, the Norwegian Space Agency, ESPI as the policy substrate) form the institutional spine. The commercial layer is anchored on ICEYE (Finland-headquartered), Isar Aerospace (Germany), Skyrora (UK, now pursuing assets from the collapsed Orbex following its February 2026 administration filing), HyImpulse (Germany), Latitude (France), Space Forge (UK), Hyperion (UK), The Exploration Company (Germany / France), AAC Clyde Space (Sweden / UK), and a long tail of specialists. The European Defence Fund and the EU’s strategic-autonomy push have unlocked nine-figure cheques into space-defence dual-use, particularly into sovereign launch and ISR. The micro-launcher race for sovereign European space access is a direct contest between Norway’s Andøya Spaceport (leveraging Isar Aerospace’s active flight campaigns), Scotland’s SaxaVord Spaceport (hosting a mid-2026 maiden orbital attempt by Rocket Factory Augsburg), and the Portugal’s Santa Maria Island in the Azores (operating as an open-access, mid-Atlantic gateway for international builders like South Korea’s INNOSPACE). Sovereignty in space, for Europe, increasingly means owning your launch and splash zones, your spectrum, and your supply chain.

Japan. ispace is the lunar lander anchor. Mitsubishi Heavy Industries’ H3 and the JAXA SSPS space-based solar power programme are credible. Synspective, Astroscale, ALE, and Pale Blue are the most credible commercial Japanese names. The strategic angle is U.S.-aligned but technically independent, with deep capabilities in robotics, propulsion, and small-satellite manufacturing.

India. Skyroot Aerospace, Agnikul Cosmos, Pixxel, Bellatrix Aerospace, GalaxEye, and Dhruva Space are the post-2020 cohort that emerged after the IN-SPACe reforms opened the sector to private capital. ISRO’s lunar and Mars credentials are the foundation. India’s lift-to-orbit cost advantage and engineering pipeline make it a credible exporter of small-satellite buses and launch services through the back half of the decade. The Gaganyaan crewed programme adds national prestige to the commercial base.

UAE, Australia, Brazil, others. The UAE’s MBR Space Centre flew Hope to Martian orbit and is building a credible space-defence stack. Australia’s Gilmour Space targets sovereign launch from Bowen, Queensland. Brazil’s INPE has long-standing EO capabilities. Each of these is small in absolute terms and large as a sovereign hedge against the U.S.-China duopoly.

Sovereignty mechanics. Sovereignty mechanics. The real machinery underneath this geography is not satellites but export controls (like ITAR and the EAR Commerce Control List), national space registries (such as the UK Outer Space Act and France’s Space Operations Act), and modern multilateral legal frameworks like the Artemis Accords (signed by 67 countries), which actively work to formalize property and operational rights where the legacy 1967 Outer Space Treaty remains silent. The competitive frame for the next decade is not “which company wins” but “which jurisdiction owns the rules of the road for the activity it underwrites.”

The Undisputed King: SpaceX

Every thesis in this report assumes one thing: that SpaceX delivers. The five price eras, the cadence model, the reentry layer, the orbital-compute timeline, and the lunar logistics economy. That makes SpaceX the single most important asset  in the sector. It is a company without precedent. Below is the analytical case for why a $1.75T IPO valuation underweights what is being built and what the credible path to $10T would require.

In 2020, Starlink had zero commercial customers. In 2025, it generated $11.4 billion in revenue and $7.2 billion in EBITDA at a 63% margin (Quilty Space financial overview). Quilty projects 2026 revenue between $15.9B and $20B with EBITDA crossing $11B. The subscriber base doubled in 2025 (4.5M → 9M), crossed 10M by February 2026, and added roughly 20,000 new users per day through November and December.

SpaceX completed 165 orbital launches in 2025, more than the rest of the world combined for the sixth consecutive year. The 500th Falcon 9 launch flew a booster on its 29th reuse. The launch business itself is not the cash machine, at $4.1B revenue it is a small line item relative to Starlink, but it is the strategic moat that makes everything else work. 

Defence has become a structural third pillar. In April 2025 SpaceX secured a $5.9B Pentagon contract for 28 national-security launches through 2029 under the Space Force’s NSSL Phase 3, the single largest share of a $13.5B award. A separate ~$2B Golden Dome air-moving-target-indicator constellation is in the works. Starshield, the classified-payload arm of Starlink, is now the fastest-growing line item in the SpaceX P&L. The 30–40% national-security revenue underlying the broader sector (treated separately in the bear cases) is concentrated disproportionately on SpaceX.

The IPO is the macro-validation event

SpaceX’s valuation has compounded at a rate the public markets have no analogue for. It has grown by 13x in 41 months, against a backdrop where the S&P 500 returned roughly 28% over the same period. The April 1, 2026 confidential IPO filing targets a $75B primary raise of 3x the size of Saudi Aramco’s $25.6B 2019 record, which itself was 1.5x the next-largest IPO in history. Musk retains 80% voting control, no personal shares are being sold, and the entire raise is primary capital intended to stay inside the company.

The macro-validation matters more than the raw number. The space sector has never had a $1T-plus public comparable trading on real (and very large) revenue. The IPO creates one. The downstream effects are:

• Crossover funds are forced to build a space allocation. Funds whose mandate excludes private aerospace will now have to underwrite a top-20 S&P constituent that is aerospace. The flow-of-funds story alone re-rates every public peer.

• Comp-set multiples reset, then sort. The IPO event creates an initial re-rate upward: Rocket Lab, AST SpaceMobile, Planet, Intuitive Machines and Firefly all benefit from the flow-of-funds and the anchoring effect of having a $1T+ public comparable in the sector. We then expect a follow-on correction over 12–18 months as the public market separates the companies whose underlying revenue and cash flow can sustain the new multiple from those whose narrative cannot (we develop this point at length in Bear Case 3). The net effect over 24–36 months is a re-rating of the credible survivors and a compression of the rest.

• Sovereign capital enters. UAE, Saudi PIF, and Singaporean GIC have all signalled IPO-anchor interest. Sovereign allocators were largely absent from the private-market path; the public listing brings them in.

• The talent pipeline opens up. A liquidity event of this scale releases at least 500 senior SpaceX employees with founding-team capital. The 2026–2030 founder cohort will be disproportionately ex-SpaceX, the same way the 2010–2015 founder cohort was disproportionately ex-PayPal.

Can SpaceX go stratospheric?

A $10T enterprise value would make SpaceX larger than the combined market cap of Apple, Microsoft, Nvidia and Alphabet today. Our view is that this is possible with the caveats that (a) multiple compression is real and accelerates as a company matures, and (b) the bull case requires SpaceX to be operating at a revenue scale that has never been recorded.

What is the right valuation multiple for a company that appears to have a monopoly on colonising Mars? The honest answer is that no traditional multiple framework fits. Apply Palantir’s ~50x price-to-sales multiple to Starlink alone and that single segment is worth $700B+. Apply a traditional telco multiple (1.5x sales) and the same business is worth $17B. 

The closer historical analogues are companies that owned planet-scale infrastructure during a once-a-century build-out. Standard Oil at peak (1904) controlled 91% of US oil refining at a market value equal to ~3% of US GDP, adjusted, roughly $1.5T today. The East India Company at its peak operated as a private army, a sovereign currency issuer, and a monopoly trade route operator across an entire ocean. Apple’s $4T market cap is the current ceiling for private-sector enterprise value, and is anchored on dominating the planet’s consumer-electronics economy. SpaceX, if Mars works, owns the comparable position on another planet. That comparison is uncomfortable, and is one of the reasons sober traditional analysts cannot agree on a Starlink multiple, let alone a SpaceX-aggregate multiple. The market is pricing an optionality that has no comparable in history.

The realistic multiple tightens over time as growth slows and the business matures, but we’d place it somewhere between Nvidia (25x) and Apple (7x), and assume a 15x multiple long-term which we believe it can sustain given the enormity of the opportunity.

What does that revenue trajectory look like by component?

• Starlink to $150B+ ARR by 2040. From $11.4B in 2025 to ~$50B by 2030 (consumer + maritime + aviation + DTC + enterprise), then $150B+ by 2040 as Direct-to-Cell and Gen-3 backhaul absorb meaningful share of terrestrial mobile and fixed broadband. Starlink today serves 10M households; the addressable consumer market is ~500M households globally. ARPU compression is real but offset by scale and ARPU stratification across consumer / enterprise / defence / aviation.

• Starshield + defence to $30–50B ARR by 2030, $80–120B by 2040. NSSL Phase 3 + Golden Dome + classified payloads + ISR contracts. The Tranche follow-on cadence and the EU Defence Fund are tailwinds. SpaceX’s share of the disclosed defence-space line item is rising toward 50% of contract value.

• Starship commercial launch to $30–50B ARR by 2035, $60–80B by 2045. At ~200 flights/year (Base case cadence) and $30–50M effective revenue per flight (commercial + crew + Starlink internal as transfer pricing), Starship alone underwrites $20–30B by 2035. Earth-to-Earth point-to-point in the $10/kg era adds the rest, though margin compresses as the commercial-launch market matures and competitors (Blue Origin, Stoke, RFA) reach price parity. We assume meaningful margin compression on commercial launch by 2040.

• In-space manufacturing + orbital data centres + Mars. SpaceX captures share of every era’s TAM via vertical integration (orbital compute via Starlink backhaul, in-space pharma via Starship cargo logistics, lunar via Blue Moon-compatible cargo). The Mars supply chain adds a tail-revenue line that could be transformatively large by 2050 or could remain a strategic loss leader.

SpaceX’s own TAM estimates from their S-1 filing

To summarise, we view a credible base case as: ~$200B revenue by 2035, ~$500B by 2045, ~$900B–$1T by 2050. At a 10–15x multiple, appropriate for a mature trillion-dollar-revenue infrastructure company, it puts EV at $9–15T. The $10T benchmark is at the optimistic end of the credible base case, not the bull case. $1T in annual revenue is unprecedented in corporate history. No company has ever crossed it. The largest by revenue today is Walmart at ~$713B. If SpaceX gets there by 2050, it becomes the first private-sector enterprise to operate at the revenue scale of a mid-sized economy. The current bull case for SpaceX is, in literal terms, that it operates at a scale no firm has previously achieved.

The most counter-intuitive feature of SpaceX’s strategic position is that the company benefits more from a thriving competitive ecosystem than from monopoly capture. Three reasons:

1. Launch cadence requires demand. SpaceX’s marginal cost per Starship flight collapses with reuse turnaround and pad utilisation. The economics break if Starship flies at half its design cadence because the demand isn’t there. Every Vast Space station, every Starcloud orbital data centre, every Varda capsule, every Anduril Tranche order is a demand-side catalyst that pulls Starship’s marginal cost lower for Starlink’s benefit. 

2. Defence revenue depends on a viable commercial sector. The Pentagon’s commercial-buy model only works if there’s a commercial market to buy from. If the rest of the sector dies, NSSL becomes a sole-source contract, which triggers procurement-policy backlash and price ceilings. SpaceX’s defence revenue is structurally protected by a healthy commercial flywheel.

3. Mars needs a planet’s worth of capital. Even at $10T in enterprise value, SpaceX cannot self-fund a self-sustaining Mars colony. The Mars vision requires every adjacent industry (lunar ISRU, orbital manufacturing, energy, life support, robotics) to mature in parallel and provide a customer-and-supplier base. Musk understands this; the SpaceX strategy reflects it.

Whilst dreaming of financial outcomes that sprawl across the cosmos is alluring, we’ll now look at some of the potential detriments to this bull case. 

Key Obstacles To Our Thesis

1. Kessler-syndrome cascade

The argument. Debris density in low Earth orbit has been growing roughly with the active-satellite population, and the threshold at which new-debris-creation outpaces atmospheric reentry decay has been credibly modelled as somewhere between two and four times the current population. Past that threshold, collision events generate debris faster than it deorbits, and LEO becomes progressively unusable for commercial operations across multiple decades. As depicted in the film Gravity, a runaway ablation cascade (arguably my favourite space-related term) could theoretically render space totally unusable. Imagine an ever-growing storm of metal fragments encircling the Earth at 30,000mph.

Our view. The threshold is real, the modelling is genuine, and the policy response is currently insufficient. The strongest counter is not “Kessler is fictional” but “the response is faster than the bear case assumes.” Active debris removal at scale (Astroscale, ClearSpace, the SDA-led standards effort) becomes a $200/kg-era market with a real customer in the form of every major constellation operator. The risk is an event in the late 2020s that demonstrates the cascade dynamic before the removal economy has scaled. Live, real, the most underweighted single risk in the sector. As for the worst-case scenario, tracking and increased aircraft autonomy is “good enough” to avoid this for now. See Philip of Starcloud’s view here.

2. The cost curve stalls below 10x compression

The argument. Starship’s cadence target is one vehicle every 3–4 days by 2027. Some analysis suggests the empirical base rate for hitting cadence targets of this aggressiveness is below 25%. If Starship maxes out at one vehicle per week or per fortnight rather than per three days, the price-per-kg trajectory plateaus in the $200–500/kg band and the $50/kg and $10/kg eras slip by a decade or more.

Our view. The cadence target is genuinely ambitious. The base-case scenario in the appendix model assumes Starship hits roughly half of its public cadence target by 2027 and three-quarters by 2030. Even at that more conservative trajectory, the $500/kg era opens on schedule and the $200/kg era opens 2–3 years late. The bull case requires the public targets to hold; the base case does not. The bear case requires Starship to fail to hit even half of its public targets, which would be the largest single SpaceX miss in the company’s history.

3. Demand-side no-show and the public-valuation reset

The argument. Every era thesis above is supply-side: cheaper launch makes new businesses possible. None of the era theses are demand-side: cheaper launch makes new businesses necessary. If the buyer for $500/kg orbital compute, $200/kg in-space pharma, or $50/kg lunar regolith does not materialise on the timelines assumed, the supply gets built ahead of the customer and the financial outcome resembles the late-1990s dotcom over-build.

Our view. The strongest version of this argument is correct on the timing risk and underweights the structural argument. The dotcom over-build did not destroy the value proposition of internet-delivered services; it pulled it forward. Similar dynamics appear to be in play for orbital compute, in-space manufacturing, and lunar logistics. We would not be surprised by a meaningful correction in the public valuations of $1,500/kg-era exemplars within the next thirty-six months. Planet Labs rose over 1,000% in 2025, AST SpaceMobile gained 372%, and Rocket Lab climbed 313%, against revenue and free-cash-flow profiles that have not kept pace with the multiples expansion. AST missed all three Q1 2026 metrics, with revenue at $14.7M against $37.5M expected. Rocket Lab posted negative free cash flow of $322M in 2025 and is not expected to reach GAAP profitability before late 2027. The most likely trigger for a correction is the SpaceX IPO itself: when the asset class anchors on a $2T-plus comparable trading on real (and very large) revenue, the smaller public players will reprice against that benchmark, and the ones whose stories cannot survive the comparison will compress. The surviving companies compound through it; that is the pattern from every prior tech reset.

4. Regulatory and geopolitical freeze

The argument. FAA Part 450 launch-licensing throughput has lagged commercial cadence. Environmental challenges to Starbase have produced multi-month operational delays. The geopolitical relationship with China has fragmented spectrum, supply-chain, and investment regimes in ways that may compound rather than resolve. Any one of these can produce a 2–5 year delay across the cost curve.

Our view. This is the bear case most likely to materialise as a partial outcome. FAA reform is on the agenda but moves slowly. China decoupling is now baseline, not risk. The probability-weighted impact is a 1–2 year delay across the eras post-$500/kg, which is meaningful for fund-cycle returns but not for the fundamental thesis.

5. Anchor-tenant pull-out and founder-concentration risk

The argument. The economics of the $500/kg-era commercial space station thesis depend on hyperscaler or pharma anchor-tenant commitments that have not yet been firm-contracted. The economics of the defence layer depend on Pentagon budget priors that are politically contestable. And the entire $1,500/kg to $50/kg cost curve depends on a single privately-held company (SpaceX) executing on Starship for the next decade. The structural concentration risk is the real underwriting question, not the political-overhang narrative that dominated the 2025 discourse.

If a hyperscaler walks away from a Commercial LEO Destination bid in 2027, the post-ISS architecture compresses to two operators or fewer. If a defence administration shifts procurement back to traditional primes, the Tranche follow-on cadence we treat as a base case slips by 18–24 months. If Starship hits a sustained development issue that no public-market competitor can absorb, the entire $200/kg-and-below schedule slips.

Our engagement. The single-asset concentration is real and we underwrite it explicitly in our base case. The hyperscaler anchor-tenant question is the more concerning sub-bear: orbital data centres at scale depend on at least one of Microsoft, Google, or Amazon (or by 2027, xAI / Cowboy Space) committing first-party capex, and that commitment is still TBD as of mid-2026. If it does not arrive by 2028 the $500/kg-era timeline slips. We treat the political-overhang question as second-order: Musk has materially stepped back from politically-aligned roles since mid-2025, and the forced-divestiture tail scenario we judge low-probability. The structural concentration is the more durable risk to track.

Assuming those don’t happen, how big could space actually get? 

Today’s space economy is roughly $630B. The consensus track (McKinsey + WEF) projects $1.8T by 2035 on a 9% CAGR. Morgan Stanley and PwC put $1–2T by 2040. China’s official planning documents point much higher by 2050. Every published consensus number understates the destination for one specific reason: they price space as a single industry, when the cost curve is unlocking several adjacent ones at once. Every consensus number was modelled before the speed of progress in adjacent domains: AI compute, foundation-model-accelerated materials science, microgravity-unlocked therapeutics, and lunar ISRU. The McKinsey number is a straight-line extrapolation of a 9% CAGR off the 2024 base. The next twenty-five years will be materially different from the last.

Our bull case is ~$7T by 2050, and we model the joint probability at ~12–15%. The single largest driver is AI hyperscaler orbital compute, at ~$2T by 2050, it will become the largest sub-segment by a clear margin. The bull case is not a bet on launch economics; it is a bet that AI compute, materials science, and pharma all migrate meaningful percentages of their capex stacks off-planet. For pharma, Eli Lilly’s tirzepatide franchise alone is on track to exceed $50B in 2030 revenue, on a single drug class. If microgravity manufacturing unlocks 2–3 GLP-1-class molecules over the next twenty-five years, that single therapeutic category is roughly equal to today’s entire upstream space economy. For the as-yet-unclear downstream layers, we expect this to rhyme with the aforementioned containerised freight analogy. The space-economy equivalents are companies most readers have not yet heard of, building application-layer products that ride satellite-delivered data, orbital compute, microgravity inputs, or lunar ISRU outputs.

Close

Every industrial revolution begins with a price collapse in a single input. Steam. Steel. Transistors. Bandwidth. The price-per-kilogram of mass-to-orbit is collapsing now, and the industry that exists at the other end of this collapse will be unrecognizable. We are venturing forth into the harshest environment known to man that promises untold bounties and paves the way to a new era

We have spent the last eighteen months looking for a single rule that explains where this sector is going. It is this: the entire industry now lives downstream of a single cost curve, and the curve is collapsing fast enough that the businesses worth building today are the ones whose unit economics only compute at $200/kg or lower.

Where the alpha lives

The honest acknowledgement

• 99% right is usually not enough in space.

Space is the hardest hardware domain ever attempted, and the venture profile in this sector is brutal. Companies take eight to twelve years to reach exit, and capital intensity is enormous. Burn rates are higher because you have to build things in vacuum-rated cleanrooms, test them against a physics regime that does not forgive, and sometimes wait months for a launch window to discover whether your assumptions survive reality. Pre-seed-to-Series-A graduation rates in deeptech are roughly half what they are in software. Most space-themed funds raised between 2019 and 2023 will not return capital, and that cohort is the comp set every honest LP benchmarks against. We are not assuming we are exempt from that. We are assuming the asymmetry is worth the brutality if we choose the right layer.

First, we predominantly invest our own capital alongside our LPs, which means we are not optimising for quarterly markups or a 24-month story arc. We can be patient with founders. Second, we are still hunting for outcomes where a pre-seed check returns 100x or more, and we believe the sector currently offers a richer set of those than software does at comparable risk. The patience exists in service of the asymmetry. Capital deployment into space tech hit $12.4B in 2025, up 48% year over year, dominated by US defence allocations and SpaceX-IPO anticipation. Of the $62B deployed globally over the last decade, only $1.45B went into pre-seed, roughly 2.3%. That gap is the opportunity. We want to support the most unreasonable and outlandish founders going after sci-fi concepts underpinned by the mechanics outlined in this report: not fantasy, but fantasy with a cost curve.

Appendix

A1. Bear/Base/Bull cadence model

The era windows in the body of this report are calibrated against a Base scenario for Starship cadence and unit cost. The report’s thesis does not depend on the Bull case. It does depend on the Base case being roughly right. Below are the three scenarios, each anchored to SpaceX’s own public targets and back-tested against Bent Flyvbjerg’s empirical work on megaproject completion rates (~25% base rate for ambitious cadence targets in heavy industry). The Base case assumes Starship hits roughly half of public targets, well above Flyvbjerg’s prior, well below Musk’s stated ambition; anything more conservative is the Bear.

Anchor inputs. SpaceX’s stated cadence target is one Starship every 3–4 days by 2027 (~100–120 flights/year). Payload to LEO per flight is ~150 tonnes, expendable; ~100–120 tonnes operational with full reuse. Internal cost per flight at scale is targeted at <$10M, with Musk’s long-term floor at $1–2M.

Cadence and $/kg trajectory by year

• Bear assumes Starship hits ~25% of its public cadence target by 2027 and never closes the gap to its design potential. $/kg plateaus in the $400–800 band through 2032 and grinds down only on production-line efficiency, not on the cadence step-change. The $200/kg era slips to 2034–37; the $50/kg era slips post-2045; the $10/kg era does not arrive in the window of this report.

• Base assumes Starship hits ~50% of public targets by 2027 and ~75% by 2030. The $/kg trajectory matches the era windows used throughout the body. This is the scenario the report’s framing assumes.

• Bull assumes Starship hits 80%+ of public targets by 2027 and continues to compound. $/kg compresses one era ahead of the Base schedule: the $50/kg era opens in 2031–32, the $10/kg era plausibly arrives in the mid-2030s rather than 2040+.

Sensitivities

The single most consequential variable is per-flight reuse turnaround time, not flight count per launch site. A 24-hour booster turnaround at a single pad outperforms ten pads on a six-week refurbishment cycle. The second most consequential variable is payload utilisation; Starship at 50% mass-utilisation is a $/kg story closer to the Bear than to the Base. The third is regulatory throughput (FAA Part 450, environmental challenges); this is the bear-case-4 risk treated separately in the body.

The model deliberately does not project a price floor below $10/kg in the window of this report. Below ~$10/kg, atmospheric drag, regulatory friction, and demand-side absorption become more binding than the launch cost curve, and the relevant question stops being “what does mass-to-orbit cost” and starts being “what does Earth’s airspace allow.”