Why are the most profitable companies raising so much money?
Originally written in Chinese. Translated with AI, reviewed by Gary. Read the Chinese original →
SpaceX had only just IPO'd, raising roughly US$75 billion, and less than two weeks after listing it went to the bond market to issue bonds. My first reaction, naturally: didn't you just get all that money? Why borrow more?
When I looked into it, I found SpaceX isn't the only one doing this. Alphabet recently did an equity raise of close to US$85 billion, Amazon issued about US$54 billion of bonds, Oracle plans to raise US$45 to 50 billion, and Meta and Nvidia each issued about US$25 billion of bonds. Roughly added up, the money these few companies have recently raised through equity, bonds and debt is in the hundreds of billions of dollars.
What's most striking is that, apart from SpaceX, which is still in a very heavy investment phase, all of these companies are extremely profitable.
Alphabet, Amazon, Meta and Nvidia all have very strong cash flow. It's not that they lack money, or that banks won't lend to them. Quite the opposite: because they're so strong, the market is very happy to lend to them.
So why, with business this good, are they still issuing new shares, selling bonds and taking on debt?
I think this shows that the logic of funding today is very different from startup funding in the past.
Then it was users, now it's compute
Startups used to raise money because there was lots of hot money in the market and everyone wanted to grab market share fast, subsidise users fast and grow revenue fast. That's why that era loved talking about unicorns, valuations and revenue multiples. Many companies weren't making money yet, but as long as growth was fast enough and the story big enough, the market would give them a high valuation.
Funding in that era was essentially: I haven't proven I can make money, but I'll prove I can get very big.
This AI round of funding is different. This time it isn't a bunch of small startups burning cash to grab a market. It's the world's most profitable, most cash-rich companies raising money at scale. Because AI is no longer just software. It looks more and more like a heavy-asset industry. It needs data centres, GPUs, power, land, cooling, networks and long-term energy contracts.
Tech companies used to be asset-light: write software, sell ads, sell cloud, with very high margins. But AI has now turned into an infrastructure war worth tens and hundreds of billions of dollars. Whoever gets the GPUs first, locks in power first and builds data centres first has a chance to hold a position in the next round of AI competition.
These companies aren't raising money today just to "have money". They're racing for the infrastructure of the next 5 to 10 years.
I think it's a bit like startups grabbing market share in the past, but the foundation is completely different. Back then it was a race for users; now it's a race for compute. Back then you burned cash to buy growth; now you use capital to lock up infrastructure.
And these companies raising money now doesn't mean they're bearish on AI. Quite the opposite: they're very bullish, which is why they dare to bet this big. But they also know the funding environment won't necessarily stay this good. Today the market still believes in AI, the bond market is still buying and valuations are still high, so from a CFO's point of view, locking in long-term money now makes a lot of sense.
It isn't pure optimism, and it isn't pure pessimism. More precisely, they're very bullish on AI's future, but they know this war is extremely expensive, so while the market window is still open, they're securing the money they'll need to fight for the next few years.
Using capital to buy time
Funding today is very different from startup funding in the past.
Funding used to be about using capital to buy speed. Today it's about using capital to buy time and certainty.
Before, it was: I'll grab market share first and figure out how to make money later.
Today it's: I'm already very good at making money, but this AI war is so big I can't fight it slowly on my own cash flow.
The raise itself isn't the point. The point is what the money eventually becomes.
If the money only makes the books look nicer, or only lets the company keep losing money, then funding just postpones the problem. But if the money turns into faster turnover, lower costs, higher efficiency and bigger infrastructure, then funding is leverage.
That's what's really different about the AI era today.