According to an analysis by MIT Technology Review, major cloud providers could need to nearly triple their productivity by the end of the decade simply to break even on their massive, trillion-dollar artificial intelligence infrastructure investments.
Based on confirmed outlays from Alphabet, Microsoft, Amazon, Meta, and Oracle, Wharton finance professor Jessica Wachter and a coauthor calculated that data center spending will approach $1.1 trillion through 2027. This immense financial gamble carries implications extending far beyond Silicon Valley.
Inside the AI Bubble’s Productivity Math
Wachter, a former chief economist at the Securities and Exchange Commission (SEC), determined that these hyperscalers must boost their productivity by a factor of 2.7 to achieve a break-even point by 2030.
Their mathematical model factors in capital expenses, asset depreciation, and a targeted 15% return. Without that substantial increase in productivity, Wachter and her coauthor arrive at a sobering assessment.
“The current buildout will be the largest misallocation of capital in history.”
— Jessica Wachter, Wharton finance professor,
Alphabet recently reported a $5.9 billion free cash flow deficit for the quarter, marking its first such shortfall since its 2004 initial public offering. As market concentration intensifies, investors are becoming increasingly anxious about the financial hazards tied to AI spending.
Debt Spreads the Risk Beyond Big Tech
Data from Morgan Stanley indicates that hyperscalers intend to fund more than half of their projected $2.9 trillion in data center expenditures through 2028 using external capital rather than relying strictly on existing cash reserves.
For example, Meta offloaded an 80% interest in its Hyperion data center located in Louisiana to private-credit provider Blue Owl Capital, highlighting the growing intricacy of how these tech giants secure funding.
Stijn Van Nieuwerburgh of Columbia Business School cautions that this type of borrowing increasingly winds up inside private credit portfolios and pension funds. He notes that the average person remains largely unaware of how deeply these financial vulnerabilities have penetrated their insurance policies and retirement accounts.
Meanwhile, cryptocurrency strategist Arthur Hayes has suggested a potential fallout, theorizing that a collapse in AI credit might compel the Federal Reserve to print currency, which could drive bitcoin (BTC) as high as $1 million.
“A parlay bet by the capital markets and the economy.”
— Gary Gensler, former SEC chair and MIT Sloan School professor,
Though the exact timeline remains unclear, Gary Gensler anticipates that a market correction will eventually occur. Whether that adjustment happens swiftly or over an extended period could dictate the ultimate scale of losses from this trillion-dollar wager.