In just a couple of months, SpaceX could finally go public.
Investors have been waiting years for a chance to own a piece of Elon Musk's famed space company. At one point, we thought we'd have to settle for owning SpaceX's Starlink subsidiary alone, but by all accounts, Musk has decided to IPO the whole shebang -- and throw in social media site X and artificial intelligence company xAI for good measure.
But this is not good news for investors.
Image source: Getty Images.
Starlink is the best part of SpaceX
SpaceX is a profitable company today. But multiple media reports on the upcoming SpaceX initial public offering confirm that both X and xAI are losing money -- maybe losing so much money that they're going to turn the entire IPO unprofitable.
If, and it's a big if, SpaceX manages to return to profitability in the future, though, it will be because of just one thing: Starlink.
With more than 10,000 Starlink communications satellites in orbit, SpaceX operates by far the largest and most profitable satellite constellation around Earth. In 2025, Starlink accounted for $10.4 billion of the $15 billion in sales SpaceX generated. In 2026, space research firm Payload estimates Starlink sales will surge 80% to $18.7 billion, and account for nearly 80% of all revenue at SpaceX.
Best of all, Starlink is profitable. Internal SpaceX documents forecast an operating profit margin as high as 60% from the business -- a stark contrast to the negative margins at X and xAI.
Image source: Amazon.
Starlink has competition -- from Amazon
But here's the thing: As anyone who's read Adam Smith's Wealth of Nations can tell you, high profit margins attract competition from rivals who covet those margins. Price wars ensue, profit margins decline, and everyone winds up a little bit poorer in the end. (Except for consumers. Competition is great for consumers!)
This is the risk that Amazon (AMZN +0.55%) poses to SpaceX.
As you may have heard by now, Amazon has its own plan to build a satellite communications network to rival Starlink: Amazon Leo. From a standing start last year, Amazon has used rockets from launch providers United Launch Alliance, Arianespace, and SpaceX itself to launch some 300 Leo satellites into orbit over the past year. In April, the e-commerce giant announced it will grow its constellation by another couple of dozen satellites by buying Globalstar (GSAT +0.75%) for $11.6 billion.
That's not yet enough to initiate commercial satellite internet service; Amazon is believed to need nearly 600 satellites in orbit for that to happen. (For context, SpaceX launched 800 satellites before opening up beta service for Starlink.) Presumably, Amazon can hit its mark after another 10 to 12 launches, which could conceivably allow it to announce beta service by the end of 2026.
And at that point, Amazon Leo will begin competing with SpaceX Starlink... and prices for satellite internet service might begin to fall, taking SpaceX profits down with them.
NASDAQ: AMZN
Key Data Points
Caveats and provisos
One caveat to that, however: Counting the units it will soon acquire from Globalstar, Amazon has about 330 satellites in orbit as of early May. Under the terms of its FCC license permitting it to operate the Leo constellation, however, it's supposed to have at least 1,618 satellites in orbit by July 30, en route to a total constellation size of 3,236.
Launching anywhere from 20 to 30 satellites at a time, and launching at the rate of about once per month, Amazon's certainly going to miss its July target. What happens then?
At its present launch cadence, Amazon will need more than four more years to hit 1,618 satellites in orbit. Amazon realizes this. So does the FCC.
Amazon has asked the FCC to extend its deadline to reach the halfway mark, and I believe the FCC will agree to the request. It would promote price competition for there to be two big satellite communications companies in existence. If Amazon can demonstrate it's making every effort to complete its constellation ASAP by hiring every available launch provider (Ariane, Blue Origin, Rocket Lab, ULA, and yes, SpaceX, too) to put its Leos in orbit, then FCC should find a way to promote that effort.
It might not happen soon; it might not even happen this year -- but competition is coming for SpaceX and Starlink.
