To hear Elon Musk tell it, Starlink -- the satellite internet business that SpaceX has been building for nearly a decade now -- is generating cash and well on its way to profitability. SpaceX's President, Gwynne Shotwell, backs him up on this, and says that SpaceX as a whole began generating cash as far back as 2022, and reported positive profits at least once in 2023.

But don't believe it, says Bloomberg.

In a column that's been getting a lot of attention on the internet lately, Bloomberg argued last week that Elon Musk isn't disclosing all of the costs of running Starlink, and that this makes his profitability claims "suspect." Contrary to the billionaire's boast, argues the news agency, Starlink "is still burning through more cash than it brings in."

Constellation of three satellites in orbit over Earth.

Image source: Getty Images.

Bloomberg bursts Elon Musk's bubble

Citing undisclosed "people familiar with [SpaceX's] finances," Bloomberg says Starlink has "lost hundreds of dollars on each of the millions of ground terminals it ships." But does this claim hold water? Let's consider:

At last report, SpaceX Starlink had some 2.6 million customers around the globe. Bloomberg didn't say precisely how many "hundreds" of dollars it thinks SpaceX is losing on each of these customers' terminals, which sell for $499 apiece. But let's say each terminal sold loses $200 for SpaceX. Times 2.6 million, that would work out to a total loss of $520 million on hardware since Starlink began operating. This year, our friends at Payload Research forecast that Starlink will add a further 1.5 million customers, implying ongoing losses on hardware of another $300 million.

That's not an insubstantial sum. But according to Payload's Starlink forecast for 2024, the company will generate more than $6.8 billion in revenue this year -- up more than 50% from $4.2 billion in 2023. And Starlink is hoping to generate operating profit margins of 60% on its revenue -- $4.1 billion.

Subtract $300 million from $4.1 billion and... well, I don't know about you, but it seems to me that Starlink should probably earn quite a tidy profit from Starlink this year, even if it does lose $300 million on just its hardware.

You see, what SpaceX is essentially running here is a razor and blades business model. SpaceX may lose a bit of money on each "razor" it sells (i.e., the terminals), but make up that loss many times over by selling "razor blades" (i.e., internet service) to make the razors useful. And a business model that's served Gillette (for example) well for more than 100 years -- and that serves internet companies like Comcast and Verizon pretty well when they give away modems for free today in order to charge for internet service tomorrow -- I think it will work out quite well for SpaceX and Starlink going forward.

Any other objections?

Now you may be thinking: Fine. So SpaceX sells terminals at a loss and makes back its losses by selling Starlink internet service. But surely Bloomberg had other reasons to argue that SpaceX's profitability claims are "suspect," right? They couldn't have hung their hat entirely on just this one single objection, could they?

And you're right. Bloomberg does have another criticism. Namely, it argues that in communications with investors, SpaceX often "strips out" the cost of launching Starlink satellites to orbit, making its profits look better than they otherwise would. And it's true that this would be a pretty big line item to ignore, cost-wise, except for one simple fact: This is SpaceX we're talking about. And unlike most satellite communications companies, SpaceX owns its own rockets and can set whatever prices it wants to launch them.

Furthermore, SpaceX rockets are reusable. And because the cost of launching reusable rockets multiple times eventually falls to approach the cost of just refueling the rocket (about $200,000 per launch, or less than $10,000 per satellite launched), those launch costs arguably don't amount to much when spread across a user base of 2.6 million (soon to be 4.1 million) paying customers.

Long story short, it may be true (as Bloomberg claims) that calculating the profitability of Starlink is "more of an art than a science." But it's also true that if Starlink turns into a $6.8 billion-a-year business in 2024, the costs that Bloomberg is complaining about are going to amount to mere rounding errors on a terrifically profitable satellite internet business.

Granted, we're going to have to wait for Starlink to officially IPO and release its financials for public examination to be 100% sure how profitable it is. But for the time being, everything I've read tells me that, when Elon Musk says Starlink has reached breakeven, and that SpaceX is on the path to profitability -- he's probably telling the truth.