Space Exploration Technologies Corp. closed at $118.24 on July 23, roughly 12% below its June IPO price of $135 and 47% below the $225.64 intraday peak it hit four days after going public in the largest offering in history. The business underneath that chart is genuinely exceptional in places. So why has the stock been sliding almost since the moment it started trading?

Is Starlink actually as good as the hype suggests? By the numbers, yes. The satellite broadband unit generated $11.4 billion of 2025 revenue, growing roughly 50%, at a 63% EBITDA margin, with more than 10 million subscribers as of March. It is one of the best infrastructure assets built this decade, and the launch business behind it, flying Falcon 9 at an estimated internal cost of $15-20 million per mission, gives SpaceX a cost edge no competitor has matched.

Then why is the company losing nearly $5 billion a year? Because February’s xAI acquisition bolted a $6.4 billion operating loss and $12.7 billion of capex onto a group that was previously cash-generative. Strip out AI and the Starlink-plus-launch business is solidly profitable and self-funding. Include it, and the consolidated company burns cash at a pace that turns a crown-jewel asset into a harder valuation problem.

Is the stock at least cheap enough to compensate for that risk? Not yet. At roughly 83 times trailing sales, and closer to 136 times Starlink’s standalone revenue, the market is still pricing years of flawless execution across all three segments. Layered on top is a looming supply problem: an estimated $116-123 billion of IPO-locked stock becomes eligible to sell around August 6, with a larger tranche following in December, and Elon Musk’s roughly 6.4 billion shares remain locked only until June 2027.

So what would make this a buy? A probability-weighted fair value near $135 sits only modestly above today’s price, with a bull case near $200-230 if xAI losses narrow while Starlink keeps compounding, and a bear case near $70-85 if the AI burn balloons and lock-up supply forces a harder reset. Evidence that xAI’s losses are shrinking, or a price settling into the $80-100 range after the December unlock, would each meaningfully improve the setup.

The verdict: Watchlist, not Buy or Avoid. The assets are too good to bet against and the growth runway too long to dismiss, but the price already assumes the hardest parts go right. The disciplined move is to build the thesis now and wait for the lock-up supply to clear or the AI segment to prove it can pay for itself.

The complete segment breakdown, competitive analysis, and scenario model are available in the full SpaceX (SPCX) report on our Reports page.