SpaceX (SPCX) shares have lost half their value since listing, but the business behind that fall looks a lot healthier than the share price suggests.
The company trades at $114, having listed at $135 in June and briefly touched $225 before the selling started.
That’s a brutal round trip for anyone who bought at the top, and the question now is whether the price reflects a genuine rethink of what this business is worth, or simply the market digesting the biggest new listing in history.
The numbers actually look decent
Revenue came in at $7.8bn for the quarter, up 92% year on year, while the net loss narrowed to $541m from just over $1bn a year earlier. I’d rather see a business scale faster than its costs than the other way round, and that’s exactly what’s happening here.
Adjusted EBITDA nearly tripled to $3.5bn. The AI segment, which has been the sceptics’ favourite target, swung from a $276m loss to a $1.1bn profit, helped by $14.1bn of newly signed compute agreements.
Starlink remains the real engine, with subscribers doubling to 12 million and connectivity EBITDA climbing 64% to $2.6bn. Space, by contrast, is still bleeding cash on Starship development, with a $205m EBITDA loss.
What $5,000 becomes if the bulls are right
RBC, Bernstein, Macquarie and Cantor have all reiterated buy ratings this week, with targets clustered between $225 and $250, while Susquehanna sits far more cautious at $170 and Phillip Securities has gone further with an outright sell.
Taking the broker average of roughly $237, a $5,000 stake today would grow to around $10,400 if that target is hit, which tells you how wide the disagreement on this stock actually is.
Valuation remains the sticking point
The company trades on a price-to-sales ratio in the high 50s, down from the high 70s shortly after listing but still far above any established tech peer. For that multiple to look reasonable, growth needs to keep compounding at close to the current pace for years.
Capital spending backs this up. The company spent $18.4bn this quarter, with $15.8bn of that going into AI infrastructure alone, more than twenty times what it spent a year ago. It can clearly afford it, sitting on $100bn in cash and securities, but that spending has to eventually show up as earnings, not just ambition.
The lockup is the thing I’d actually watch
Roughly 911.5 million shares, worth close to $116bn, become eligible for sale on 6 August, more than the entire current public float, with further unlocks through the autumn.
My take
The operating business is improving on almost every measure, narrowing losses, a newly profitable AI segment, Starlink adding subscribers fast. Set against that is a valuation that leaves little room for disappointment and a wall of newly tradeable stock about to hit the market.
I wouldn’t call this stock cheap today, but I wouldn’t call it expensive relative to where the growth could take it either. The risk-reward only starts to look attractive once this week’s lockup has worked its way through the market.
