Short answer: not at this price. Here is the arithmetic, in about two minutes,
and a link to the full underwriting if you want to argue with it.
The call
Avoid at $115
SpaceX is a remarkable company. The stock is a different question. At $115.07 you are paying
79 times sales for a business with a −45% net margin that burned $19.8 billion of
cash in the last twelve months.
Traded 24 Jul
$115.07
IPO'd 12 Jun at $135
My fair value
$54
probability-weighted
Expected return
−10.9%
a year, from here
Would buy below
$60
and only after December
Three things the coverage isn't telling you
You're buying an AI company wearing a rocket costume. xAI merged into SpaceX on 2 February
and the financials were restated, so "SpaceX revenue" now includes Grok and X. That segment lost
$6.4B on $3.2B of revenue and consumed 61% of all capital spending in 2025. Starlink's cash
is funding a capex war against Microsoft, Google and Amazon — companies with far deeper pockets and no
rockets to pay for.
$1.52 trillion is a quote on 4.85% of the company. Only 639 million shares trade out of 13.17
billion. About 12% of that entire float changes hands every day. On 6 August, two days after
the first earnings report, 911.5 million shares unlock unconditionally — the float more than doubles in a
single session, then a larger tranche follows after the Q3 print. Real price discovery hasn't happened yet.
Starship hasn't done the two things the story requires. Thirteen test flights, zero
revenue-generating payloads, and orbital refuelling has never been attempted — which gates both the
NASA moon contract and the next generation of Starlink satellites. A scrubbed launch on 16 July erased
roughly $100 billion of market value in an afternoon. Meanwhile the actual rocket business grew revenue
at 6.5% a year from 2023 to 2025, because most of those launches carry SpaceX's own satellites.
What the price is asking you to believe
Run the valuation backwards. To justify $115 a share, SpaceX has to grow revenue 59% a year for six
consecutive years — from $18.7 billion in 2025 to roughly $307 billion in 2031. That is
approximately the size of Microsoft's entire business today, reached by a company that has never posted a
full-year profit at scale.
It is not impossible. It is not the expected case either, and today's price treats it as the expected case.
The other side, stated fairly
The bull case is real and I built it as strongly as I could: 95 million Starlink subscribers, Starship
achieving routine reuse, orbital data centres becoming an actual market. That version pays
+17.3% a year and would be an excellent investment. I put it at roughly
one chance in five. The other four outcomes lose money — which is the whole argument in one sentence.
Three dates that matter more than anything else
4 Aug
First earnings report as a public company
Watch one number: revenue per Starlink user. It has fallen from $99 a month to $66 while
the subscriber count quadrupled. If it stabilises, I'm too pessimistic and I'll say so.
6 Aug
911.5 million shares unlock
Unconditional — it happens regardless of price. Roughly $105 billion of stock meets a
market currently absorbing about $9 billion a day. If it clears without a break, my thinking on the float
is simply wrong.
8 Dec
The lock-up fully expires
By then about 40% of the company trades, up from under 5% today. That is the first point
at which the price means what a price normally means. This is why I'd wait.
If you're going to buy it anyway
Wait for all three: a price under $60, the December float, and Starship actually flying a
commercial payload plus refuelling in orbit. Any one alone isn't enough.
Size it as a venture bet you can lose entirely — 2% of liquid savings, not a core holding.
The payoff shape is one good outcome in five, and position sizing should respect the shape, not the average.
Don't short it. 17% of a tiny float is already borrowed. The squeeze risk runs against you and
Musk-driven momentum doesn't care about your thesis.
If you already own it, there's no obvious reason to hold a pre-earnings position in something you
wouldn't buy today.
The full version
Everything above is derived from a 25-slide underwriting: segment-by-segment forecasts to 2031, a
sum-of-the-parts cross-check, a sensitivity grid, all fourteen published analyst targets (they span $62 to
$800 — nobody knows), and the bear and bull cases each argued properly.
Best on a laptop — it's built as 1920×1080 slides. Arrow keys to navigate. Every hard
number is sourced on the slide it appears on, and anything I estimated is labelled as an estimate.