The MIT Technology Review takes a look at the potential consequences of an AI bubble collapse, cautioning that major cloud providers might need to elevate their productivity nearly threefold by 2030 simply to break even on their massive infrastructure investments.
Jessica Wachter, a finance professor at Wharton, teamed up with a coauthor to calculate these projections using verified hyperscaler spending figures, estimating that Alphabet, Microsoft, Amazon, Meta, and Oracle will pour nearly $1.1 trillion into data centers through 2027. The stakes of this gamble extend far beyond the tech sector.
Inside the AI Bubble’s Productivity Math
Wachter, who formerly held the position of chief economist at the Securities and Exchange Commission (SEC), determined that hyperscalers need a 2.7-fold increase in productivity to achieve a break-even point by 2030.
Her mathematical model factors in asset depreciation, a 15% return, and the cost of capital. Without achieving that level of growth, Wachter and her research partner arrive at a severe prediction.
“The current buildout will be the largest misallocation of capital in history.”
— Jessica Wachter, Wharton finance professor,
Highlighting the financial strain, Alphabet reported a free cash flow deficit of $5.9 billion last quarter—marking its first such deficit since its 2004 IPO. As market concentration intensifies, investors are becoming increasingly anxious about the risks tied to artificial intelligence spending.
Debt Spreads the Risk Beyond Big Tech
According to Morgan Stanley’s projections, cloud giants intend to fund over half of their projected $2.9 trillion in data center expenditures through 2028 using external capital rather than relying entirely on existing cash reserves.
Illustrating the growing complexity of these funding structures, Meta recently offloaded an 80% interest in its Hyperion data facility in Louisiana to Blue Owl Capital, a private-credit firm.
Stijn Van Nieuwerburgh of Columbia Business School cautions that this type of borrowing is increasingly winding up inside private credit instruments and pension funds. He notes that many individuals fail to recognize how thoroughly this financial exposure has permeated their insurance and retirement portfolios.
Offering a speculative outcome, crypto strategist Arthur Hayes suggests that a bust in AI-related credit could compel the Federal Reserve to engage in money printing, potentially driving the price of 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,
While the exact timing remains unclear, Gary Gensler anticipates that a market correction will eventually occur. Whether the pullback unfolds slowly or happens suddenly could dictate the scale of permanent losses stemming from this trillion-dollar wager.
Frequently Asked Questions
How much are major tech companies expected to spend on data centers?
Alphabet, Microsoft, Amazon, Meta, and Oracle are projected to spend nearly $1.1 trillion on data center construction through 2027, with total planned spending reaching $2.9 trillion through 2028 according to Morgan Stanley.
What productivity gains do hyperscalers need to break even?
According to former SEC chief economist Jessica Wachter, hyperscalers will need to nearly triple their productivity—specifically growing it 2.7 times over—by 2030 to cover asset depreciation, the cost of capital, and a 15% return.
How are these massive data center projects being financed?
Morgan Stanley calculates that hyperscalers will fund over half of their $2.9 trillion in data center spending through external capital, utilizing debt and complex arrangements like Meta selling an 80% stake in its Hyperion data center to private-credit firm Blue Owl Capital.
How does this financial risk affect everyday people?
Columbia Business School’s Stijn Van Nieuwerburgh warns that the debt backing these projects flows into private credit vehicles and pension funds, meaning many people unknowingly carry exposure in their retirement and insurance savings.
