VERDICT
PRICED FOR PERFECTION
The rockets are unambiguously real. The launches work. Starlink is generating genuine revenue. The xAI merger adds legitimate AI compute assets. But at 65x price-to-sales, $25B in negative free cash flow in the first six months of being public, and a stock that went from $135 to $225 and back to $143 in under 90 days — you're not buying a business at a reasonable price. You're buying a narrative at a premium. Whether that narrative is worth what you're paying depends entirely on assumptions about a future that nobody can actually model with confidence.
SpaceX went public on June 12, 2026 at $135 per share. By June 16 it had hit $225.64 — a 67% gain in four days. By late August it was trading near $143. The people who bought at $225 are down 37%. The people who got in at the IPO price are up about 6%. The people watching from the sidelines are trying to figure out what any of this means. Let me try to help.
What's actually real
SpaceX is a genuinely exceptional operating company. That part is not hype. The Falcon 9 is the most reliable orbital rocket ever built — over 400 consecutive successful launches as of the IPO. Starlink has more than 7 million paying subscribers globally and generates real, recurring revenue. The company completed the acquisition of xAI — Elon Musk's artificial intelligence company — in February 2026, making AI compute a third major business segment alongside launch services and Starlink. Revenue is reportedly doubling year-over-year. The operational achievements are historically significant. All of that is real.
The Goldman Sachs-led IPO was the largest in history. $75 billion raised. $1.77 trillion valuation at pricing. SpaceX dedicated 30% of the offering to retail investors — an unusual move that helped generate the FOMO-driven first-week price action. For retail investors who submitted indications of interest through Robinhood, Fidelity, SoFi, or Schwab and received IPO allocations at $135, the outcome so far is a modest gain. For everyone who chased the first-day pop above $160 and definitely above $200, the math is ugly.
What's priced in — and what that actually requires
At $143 per share, SpaceX carries a market cap of approximately $1.9 trillion. The price-to-sales ratio is 65. For context: Microsoft — which generates $245 billion in annual revenue with 45%+ operating margins — trades at roughly 12x sales. Nvidia — the AI chip company growing at 100%+ annually — trades at around 25x sales. SpaceX at 65x sales is pricing in a growth and margin trajectory that would need to be almost unprecedented in corporate history to justify at current prices.
The free cash flow situation is the number that deserves the most attention. In the first half of 2026 — the first reporting period as a public company — SpaceX burned $25 billion in negative free cash flow. Revenue is growing, yes. But the capital intensity of building out Starship infrastructure, expanding satellite constellations, scaling xAI compute, and funding ongoing launch development consumes cash at an extraordinary rate. A company burning $25 billion in six months at a $1.9 trillion valuation is pricing in a cash flow inflection that hasn't arrived yet. Maybe it arrives in 2027 or 2028. Maybe it takes longer. The gap between "will probably eventually be enormously profitable" and "is worth $1.9 trillion today" is where all the risk lives.
The Starlink argument
The bull case rests substantially on Starlink. If you model Starlink as a standalone satellite internet business with 7M+ subscribers, growing globally in underserved markets, with limited terrestrial competition in rural and remote applications, the valuation math gets more interesting. A mature Starlink at 50M subscribers generating $100/month average revenue per user would be $60 billion in annual revenue — and at utility-like margins, worth several hundred billion dollars on its own. That's a legitimate thesis. The question is the timeline and the competitive assumption embedded in it. Amazon's Kuiper constellation, OneWeb, and potential future competitors are not standing still. Starlink has a meaningful head start. Whether that head start translates into a durable monopoly or a first-mover advantage that erodes over time is the central uncertainty in the bull case.
The xAI merger adds a second major optionality layer. The combined company's AI compute assets — the Grok models, the training infrastructure, the inference capacity — position SpaceX as something more than a rockets-and-internet business. Whether xAI can compete credibly with OpenAI, Google DeepMind, and Anthropic for enterprise AI spend is a legitimate open question. The asset is real. The valuation it deserves is unknowable with any precision.
The IPO dynamic that retail buyers need to understand
SpaceX's decision to allocate 30% of the IPO to retail investors was unusual — and generated the demand surge that pushed the stock from $135 to $225 in days. But here's what that dynamic actually created: a massive supply of retail buyers who paid $150–$225 for shares in the first two weeks, and who are now underwater. Those investors face a choice — hold through a recovery that may or may not happen on a timeline they can stomach, or sell and crystallize the loss. That selling pressure is part of why the stock has drifted from $225 back toward $143.
The IPO mechanics matter here too. Lock-up periods for pre-IPO shareholders expire over the coming months, meaning additional supply will hit the market as employees and early investors can sell for the first time. Each lock-up expiration is a potential pressure event. The analyst who covered the IPO for a major research firm gave SpaceX a sell rating with a $115 price target at the $135 IPO price — calling it overvalued even at the starting point. At $225 it was being called generational. Now at $143 it's somewhere in between, and the only people who know where fair value actually sits are making wide-error-bar estimates at best.
The one thing to take away
SpaceX is not like other overhyped AI stocks or meme stocks — the underlying business is genuinely extraordinary and the operational track record is real. But "extraordinary company" and "stock worth buying at any price" are different statements. At 65x sales with $25B in first-half negative free cash flow, you're paying for a version of SpaceX that exists in 2030 or 2035. If that future arrives on schedule and the competitive position holds, the stock could be cheap at $143. If the timeline slips, the competition intensifies, or the capital requirements stay elevated longer than expected, $143 could still be overpriced. The honest answer is that nobody — including the analysts who modeled it — knows with confidence which version you'll get.