CROSSPOST: WILL LOCKETT: Starship Is Going Nowhere
Pre-IPO filings open SpaceX’s black box of financials. What spills out is a wildly expensive rocket, heroic technical risk, and a justified-equity-valuation story that leans heavily on very very...
Pre-IPO filings open SpaceX’s black box of financials. What spills out is a wildly expensive rocket, heroic technical risk, and a justified-equity-valuation story that leans heavily on very very distant cash flows. Google already had superprofits when it did its IPO, FaceBook had profits, Microsoft had profits. It was only Amazon among today’s Leviathans that did not—and Amazon had a clear path that penciled out, while SpaceX does not…
I do not know whether or not this from Will Lockett is right: I do not have the expertise to understand the technologies of spaceflight, and nobody really knows what the demand for what kinds of satellites will wind up being, either near-earth orbit or beyond, neither private nor public. I do know that the NASA budget is $25B a year, that Starlink’s revenues are $10B a year, and that Starlink’s competitors’ revenues are perhaps half that. So figure $40B a year in payments for rockets and satellites right now.
Then one (a) applies a growth rate g, (b) guesses at an appropriate risky required return on capital r, and (a) guesses that at most half of space spending could be profit, (c) guessing that only half of profits is available for pay-outs, (d) calculates that in order to meet a valuation of $1.5T with all of that budget transferred to SpaceX we need a payouts-valuation ratio of 150: that means a long-run growth rate of no lower than 0.67%/year below the cost of capital, with 85% of the preent value coming from profits to be paid to SpaceX after 2051.
Alternatively, if we suppose that in ten years SpaceX will be a “normal” business with a price-earnings ratio of 25, but that there will be no payouts until then, then at a real cost of capital of 4% per year profits then would have to be $900B—requiring an outer-space market of $1.8T or more, with the lion’s share going to Tesla.
Are either of these likely? No. Are either of these possible? I find myself more than just a little skeptical.
Let me give you some (brief) excerpts (because paywalled) from Will Lockett, who has at least a semi-informed view on this:
CROSSPOST: WILL LOCKETT: Starship Is Going Nowhere
<https://www.planetearthandbeyond.co> <https://www.planetearthandbeyond.co/p/starship-is-going-nowhere>
One of the few upsides to SpaceX’s rapidly approaching IPO is that we finally get to glimpse the murky world of its obscure finances…. Starship is critical to unlocking SpaceX’s future. Yet, these IPO filings only highlight that it is a biblically expensive mess…. It proves that Starship is nothing more than a hopeless money pit….
Reuters reported that… SpaceX… has spent more than $15 billion on Starship so far…. Falcon 9 cost just $400 million to develop, and the initial budget for Starship was $5 billion…. Musk has already blown the budget more than two times over. Yet, Starship isn’t even close to finished…. The only real targets it has achieved so far are completing a suborbital flight, reaching orbital velocity (but not orbit), landing and relaunching the booster (which is much easier than the upper stage), and conducting a splashdown landing of the upper stage….
Starship has two main use cases… launching Starlink satellites into LEO (Low Earth Orbit)… out-of-orbit flights like NASA’s Artemis missions and Musk’s Mars missions… [with] quite a complex and highly risky mission profile… a ‘depot’… Starship… sent into orbit, then multiple ‘tanker’ variants…shuttle fuel…. After this, the Mars/Moon-bound Starship launches to LEO, rendezvous with the depot, fully refuels itself from the depot, and then fires off to its destination….
Reach Orbit…. Starship needs to actually reach orbit. So far, the most it has achieved is take 16 tons on a transatmospheric flight at orbital velocity….. Increase Payload… transition from a max payload of 16 tons on a non-orbital flight to 100 tons…. Deploy Payload To Orbit…. Land & Catch Upper Stage…. Reuse Upper Stage…. Reduce Failure Rate…. …. Rapid & Long-Term Reusability…. Orbital Refuelling…. Develop, Test & Deploy Depot Variant…. Reduce boil-off to near zero in order to make this entire project viable…. Develop, Test & Deploy Tanker Variant…. Develop & Test Human Landing System…. It has taken SpaceX three years, (as of writing) 11 test launches and over $15 billion to meet four of its main targets…. But it still has 11 main targets to go, each one being significantly harder than the last….
SpaceX’s IPO is reportedly going to raise $75 billion… to be spent on building and launching orbital AI data centres as well as expanding Starlink into proper profitability…. If there is no Starship, then SpaceX can’t deploy AI data centres, or make Starlink truly profitable, as they both depend on… uber-cheap launch costs…. Starship costing over $15 billion to date undermines the narrative around SpaceX and its IPO. It should make any potential investor question where their cash is going and any pundit question Starship’s and possibly even SpaceX’s viability.
Brad here: Perhaps the sizzle being sold here really is that of Amazon? In 1997 Amazon went public very early in its corporate life, with tiny revenues and losses—the classic “sell the future, not the present cash flow” dot‑com era IPO. That story ultimately worked out spectacularly, but the equity value was overwhelmingly about hypothetical future dominance that somehow came to pass. In Amazon’s case it had very strong competitors in different segments of its businesses who could have but failed to grasp their opportunities (due, I think, to deep channel‑conflict and organizational problems). Walmart or Macy’s could not say: “we’ll destroy our margins for a decade to experiment with a new model that also cannibalizes our stores”. The money‑losing dot‑com’s advantage was that it found the tranche of investors most willing to delay cash flows in exchange for the option value of dominance.
But that is an exception: the historical template for mega‑cap tech IPOs is “already obviously profitable, but priced as if current profits are just a down payment on a huge, reasonably near‑term market” — not “loss‑making core project, with a valuation that only pencils out if most of the PV arrives after mid‑century.”
What government official would trust Musk, the erratic Nazi, with the country’s internet access. None. (Except corrupt Republicans).
What large corporations would trust musk with their web access? None.
So Starlink has retail customers only.
Is that enough of a service pool to make a profit?
My problem with Lockett is that if you’re writing a takedown on SpaceX and your first technical line item is a non-problem like Flight 10 not going orbital then you automatically question motive and/or knowledge.
Starship has two very serious, really obvious, totally un-demonstrated questions in the architecture that nobody knows the likely success of. Getting to orbit is not one of these problems. Lockett doesn't even mention them in the shared post.
I’m no fan of Musk but dislike for the man clouds rational discussion about these things (valuations included). The article doesn’t pass the smell test. Lockett makes the same strawman Flight 10 move multiple times at various levels of technical precision.
Here’s the move with Flight 10. Flight 10 was a test flight that could have committed to orbit but chose not to because they wanted re-entry data. There’s a difference between “failed to reach orbit” and “we decided not to bother because data is more valuable than showing off”. They don’t need to jump through hoops to meet milestones - SpaceX to this point has been a private company burning their own money on their own projects.
And it’s hard to really drive home just how weird a technology Starship actually is.
Musk’s talk about Mars is real. Starship looks like a rocket, like other rockets. Don’t let that fool you. Starship is a whole mass production system for reusable rockets unlike anything anyone has ever tried building.
The design decisions really are based around delivering the industrial tech stack of humanity to Mars. If you’re interested, it’s worth reading up on. They’re tooling up to build one Starship per day and fly thousands to mars. Currently they’re getting close to a 1 per week cadence. Nobody has ever built rockets this big. They’ve never built them this fast. It’s a colonization machine.
Of course, if you build something that is that far advanced, there are a lot of other uses. Maybe the DoD wants to deliver 200 tonnes of drones anywhere on the planet in 45 minutes. Or 1000 troops. Or a stack of tanks to Red Square in 30. It doesn’t need a strip to land on. If you’re dropping drones it doesn’t even need to land. And at a production cost of 100m who cares if you lose a few. These are sci-fi dreams but the Pentagon literally has a contract with SpaceX for this!!! The Rocket Cargo Vanguard program.
SpaceX stated goal is to soon be at 365 ships a year. Boeing at peak only made something like 90 747s per year. The current Starbase factory will soon have infrastructure to produce 1,000 Starships per year. This is no joke.
So what are the actually serious open questions about Starship viability?
The first is their heat shield system. Starship is only rapidly reusable as their heat shields are rapidly reusable and that’s not anything like demonstrated yet. If you remember that’s what killed the Space Shuttle reusable (and ultimately killed people on the space shuttle). If the system needs more than a few hours to check/refurbish then the core rapidly reusable premise goes out the door. Nobody has demonstrate this kind of tech and SpaceX keep the performance of their heat shields under wraps. Even SpaceX agrees this is a big open question for them.
Second, in space refueling is required by the architecture for anything beyond LEO ie. If you want to get to the Moon or Mars you need to be able to transfer fuel in zero g between tanks. This isn’t trivial and nobody has ever demonstrated in-space fuel transfer at any kind of scale beyond a few mockups.
While the technical are far more interesting than the IPO, go there we must.
The fundamental valuation question is a matter of what you believe.
Do you think SpaceX is:
A) just another satellite launch provider, with Musk flavoured meme stock fairy dust
B) a company that will radically re-orient human relationship with space and unlock the Werner von Braun fantasies of the late 1960s
I hate to be both-sideist… but there’s reasonable arguments on both sides.
I think not a. However, b is rediculously execution dependent, and space is a graveyard of ambitious companies. The really crazy money is in Jevons paradox multipliers. Does SpaceX suddenly own all the aluminum, containers and microchips at once? Can Musk launch the space economy dreams of fellow traveler Werner von Braun?
Lockett looks at SpaceX as a launch services company that’s priced like a software company and hates it. Maybe that’s the prospectus, but the prospectus is not the valuation. Value it like that if you want and come away annoyed, but the investment proposition is more a VC roll of the dice on SpaceX being THE vertically integrated space infrastructure company with a technology moat that is 5-10 years ahead of the competition.
Maybe it is, maybe it isn’t. But SpaceX has demonstrated enough technology not to be dismissed out of hand no matter how problematic Musk is. People laughed at Falcon 9 trying to land and blowing up. Now everyone is trying to copy them. Full disclosure I own Baillie Gifford funds that own SpaceX, I’ve had them for a long time and I think the valuation of SpaceX is silly.
So here’s how you get to a 1.75T valuation for someone who isn’t just in it for the meme lolz thinking “expensive but I’ll roll the dice”. Do I think the following story is likely? No. Is it possible? Yes.
Bear scenario.
Starship fails completely. SpaceX is just another launch provider. Musk's an idiot.
15% chance valuation goes to 400b
-77% return
Base scenario
Starship works like a cheaper Falcon-class replacement 500$/kg to LEO. Costs reduce but not transformative. The addressable market says roughly the same. Musk is... Musk.
20% chance valuation goes to $1.0T
-43% return
Conservative Bull
Starship works but weekly reusable or something not daily. They make up in volume launches per day, not true sci-if style aircraft reusable turnaround. Launch costs drop by orders of magnitude. Musk is... a lot richer.
30% chance valuation goes to 1.6T
-8.6% return
Aggressive Bull
Starship works with rapid reusable but they never quite get it as cheap as their goals of 10$/kg. We still get to mars. More capcacity and lower costs increases demand. SpaceX is the leading provider with large amounts of capacity, the whole space business is a rising tide. SoaceX is the AWS of space infrastructure. Musk is Andrew Carnegie.
+129% return
25% 4T valuation
The von Braun Dream Space Economy
Musk colonizes Mars. Multi planetary species-dom is a thing. The addressable market explodes, and SpaceX is front and center. We mine asteroids, we have factories on the moon. SpaceX is the spine of it all. They end up in a monopsony with Blue Origin and some Chinese companies building O’Neil cylinders, space elevators, and silicon fabs using in space mined minerals. The rest of Earth citizens worry a lot about rods from God and mass drivers. Musk is Eldon Tyrell.
10% 15T.
+757% return
Probabilities sum: 100%
Probability-weighted expected value: $3.24T
Probability-weighted expected return at $1.75T entry: +85.1%
Median scenario: Conservative Bull at $1.6T (close to IPO ask)
Probability of negative return: 65% (bear + base + conservative bull)
Probability of clear loss greater than 10%: 35% (bear + base)
The question then is how do you discount when the real upside is 15-25 years in the future.
Am I pulling these numbers out of my ass. Yes!
But the point is to show what this might look like if you want to consider it not silly.
The real SpaceX investment return was probably in private equity anyway!
The interesting thing is if Musk can get to (or close to) $10/kg goal of internal marginal cost (not market price). It’s not impossible. Unlikely but they have a chance if it all goes to plan. SpaceX can then profit like crazy because no one else has tech remotely similar and still do speculative things in space (orbital data centres, Mars cargo, asteroid missions, lunar facilities, mass drivers, space stations) for low single-digit millions in launch cost. Competitors pay 50-200x more. Build infrastructure for internal demand, create markets, then collect rent off old and new. That’s the aggressive bull.
The really interesting thing is the tail upside. If the Musk dream of multi-planetary von Braun sci-fi economy in 2050 comes to pass then the 1.75 is cheap and you get that tail upside for free. 10% is probably high on the von Braun dream. However, if you give it any meaningful possibility then SpaceX is undervalued at 1.75T.
If you’re more conservative… bear base at 40%, conservative bull 35%, aggressive bull 20% von Braun at 5% and knock off a discount, you get something moderately overpriced, not epically overpriced.
But who am I kidding. No-ones thinking this through. YOLO meme stock. Space! BUY!
(not financial advice)