Last year, hedge fund whisperer and founder of the firm Praetorian Capital, Harris Kupperman, issued a blistering financial analysis [1]on the data center frenzy sweeping across the United States.
Kupperman’s diagnosis was blunt: because data center components break down so quickly, the industry would need to see some $1 trillion in revenue across 2025 and 2026 just to cover the costs of their hyperscale facilities, let alone start turning a profit. With AI revenue today accounting for only a small fraction of that total, Kupperman’s argument is that the financial logic behind the data center buildout simply doesn’t hold up.
As Bloomberg reports, an analysis on data centers by Bain and Co, one of the largest management consulting firms in the world, found that the global AI industry would have to reach annual revenue of $6 trillion by 2031 to justify the amount of capital being injected into data centers.
The analysis[2] assumes that commercial AI tools will account for $1.8 trillion of revenue by that point — a pretty generous presumption, which requires global AI revenue to start growing almost exponentially. Even if the AI industry reaches that point, Bain’s analysis notes, the industry will still need to come up with $4.2 trillion.
Hitting that $6 trillion mark will take nothing short of a technological miracle, the report authors explain, a leap without precedent in the information age.
It would be interesting to know what exponent he/she/it is talking about.
Zero?
One?
Three?
Ten?
Suppose annual revenue grows only to $3 trillion, some will go broke and some will make money. If the Pareto Distribution is still functioning, that is the more likely outcome. I would bet on the data centers that don’t have to find external fee paying customers.
Tech titans gathered at the DBS Asia Future Forward Conference in Singapore on Monday and the consensus was clear: capital not hardware is the biggest bottleneck facing the industry.
Broadcom CEO Hock Tan said financing could become a bigger obstacle to AI expansion than shortages of semiconductors, memory chips or electricity, as the industry races to build the infrastructure needed to power increasingly sophisticated models.
″[What] I think might be the biggest constraint is capital,” Hock Tan said at the DBS Asia Future Forward Conference in Singapore on Monday.
Intel CEO Lip-Bu Tan echoed concerns about financing, pointing to the rising cost of capital and questions over whether companies investing heavily in AI will eventually generate sufficient returns.
Higher bond yields could make it more expensive to fund data center construction and chip purchases, potentially putting pressure on the economics of projects that may take years to generate returns.
Those off-balance-sheet obligations are real debt, and it trades. The structure works like this: a special purpose vehicle (SPV), mostly owned by an asset manager, borrows the money, builds the data center and leases it to the tech company. The tech company books the rent as a lease. The bonds sit with the SPV. Prices on those bonds have been sliding.
Banks have ramped up borrowing Year to date, these banks have borrowed $192 billion globally, up some 40% from the same time last year
Behind the banks sits private credit: loans made directly by asset managers like Blue Owl, Apollo, Blackstone, Pimco and Brookfield, using money from pension funds, insurers and, increasingly, individual investors in non-traded funds.
That’s what happens when an industry overbuilds by 10%. When they overbuild by 50% or 100% it’s an entirely different outcome. See: dark fiber, railroads, liars’ loans, Japanese assets, Roaring 20’s stocks, dot-com spending, etc.
It’s one thing when the builder put up a few too many houses and has to drop the price a little to move them. It’s quite another when there are thousands of them on the market and none are selling.
This bubble is particularly interesting to me, I have to say. It’s being pushed by some of the biggest, most profitable companies around: Microsoft, Amazon, Facebook, Google, Oracle, etc. I’m not to sure it isn’t a corporate case of “mine is bigger than yours” or “we can’t be left out even though we have no idea what this is going to turn out to be”, and at some point one of two things will happen:
Either all the dreams will come true and it will be ice cream and puppy dogs for everyone, or
While we look at this holding huge mountains of leased properties, the bagholders are the mid tier players.
As with small NG players when the stuff hits the fan, the big one’s buy them out for pennies on the dollar. There will be a lot of bad debt, but costs will come down to the big players.
Just a couple of days ago, I said the there is fraud going on beteewn the Tech Giants and Trump Administration. Where is the SEC, FBI, and Congress with fraud investigations? Is this to big to handle for them?
We’ve borrowed about $6 Trillion to fund tax cuts for the rich over the past 25 years and another $4 Trillion to fund useless wars in the Middle East. The stock market survived it all.
I expect the same result with the data centers whether they’re “good” or “bad”.
The two banks are leading the loan offering for Volta, a new artificial intelligence cloud company. The banks have underwritten a $1.3 billion term loan for Volta, while a $3.7 billion chip loan is not underwritten by the banks — a move to protect themselves from potential risks, according to people familiar with the deal.
Perhaps banks are now considering AI loans similar to the bundled housing mortgages–“A” rated but made up of subprime loans–made prior to the real estate crisis in 2008?
Broadcom CEO Hock Tan on Monday raised doubts over the record run up AI stocks over recent quarters, comparing investor speculation over the market to gambling. “A lot of investors have no clue what they are investing in,” he told a conference in Singapore. They “set very extreme expectations only to find out that it’s almost like they are in a casino.”
“Tech investors are getting uncomfortable with the amount of debt being used to fund the AI build-out,” says Dan Coatsworth, head of markets at AJ Bell. “There is a growing fear that big names in the tech space are running too fast, and they’re primed to trip up.”