The $1.5 Trillion “Speculative” Side of the AI Datacenter Boom

The $3 trillion AI datacenter boom is really two booms in one. The first is the “healthy” $1.4 trillion boom, funded by the cash flow of “hyperscalers” like Google, Microsoft, and Amazon. The second is a $1.5 trillion “speculative” boom, funded by risky debt, and it’s the part that has economists worried.
This $1.5tn “funding gap,” identified by Morgan Stanley, is being plugged by “private credit,” a “shadow banking” sector that has alarmed the Bank of England. These lenders are funding projects “without their own customers,” according to Gil Luria of DA Davidson, who calls them “speculative assets.”
The chair of Alibaba, Joe Tsai, has also warned of a “bubble” in these types of projects. The concern is that these lenders, “eager to deploy capital into AI,” are ignoring the “very quickly depreciating assets” they are financing. A hedge fund founder has warned these datacenters will depreciate twice as fast as their revenue.
This is all happening as an MIT study shows 95% of businesses are getting zero return on generative AI pilots. This raises the question of who will pay for this $1.S 5tn in speculative capacity.
If these “unproven” projects fail, the “hundreds of billions of dollars” in debt they hold could go bad, posing a “structural risk to the overall global economy.” While the “healthy” boom builds the future, the “speculative” one could be building the next financial crisis.