The AI Infrastructure Bubble, Explained

That may turn out to be true. But “eventually” is doing a lot of work in that sentence, and the financial structure of the current buildout doesn’t have the luxury of waiting for eventually. The debts come due on a schedule. The interest payments are real. And the gap between what’s being spent and what’s being earned is not a rounding error. It’s an 18x chasm.

JPMorgan published the number in November 2025: for the current pace of AI infrastructure investment to earn a 10% return on invested capital, the AI industry needs to generate $650 billion per year in revenue. Not eventually. Sustainably. Every year. That’s the breakeven for the buildout to make financial sense.

Current generative AI revenue, according to Menlo Ventures’ 2025 State of Generative AI report, is approximately $37 billion per year. That’s not nothing. It’s real products generating real revenue. But $37 billion against a $650 billion requirement is a 17.6x gap. For context, the entire global video game industry generates about $180 billion. The AI buildout needs to generate 3.6x the entire gaming industry’s revenue just to break even.

The spending itself has become a macroeconomic event. Harvard economist Jason Furman calculated that AI infrastructure investment accounted for 92% of US GDP growth in the first half of 2025. Not 92% of tech sector growth. 92% of the entire economy’s growth. When a single capital expenditure cycle becomes nearly the entire growth engine of the world’s largest economy, the stakes of that cycle failing are no longer contained to the tech sector.

To fund the buildout, companies have turned to debt markets at a pace that has no precedent in tech history. Over $200 billion in AI-related debt was issued in 2025. Bank of America reported that hyperscalers issued $75 billion in bonds in September and October 2025 alone — twice the annual decade average — and $121 billion for the full year, more than four times the five-year average of $28 billion per year. This debt is secured against data center assets and future revenue contracts that may or may not materialize.

The credit markets are noticing. AI infrastructure credit spreads — the premium lenders charge above risk-free rates to fund data centers and GPU clusters — have been widening since mid-2025. This is our highest-weighted signal at 20% of the composite score, because credit markets have historically repriced risk before equity markets in every major cycle. In 2006, credit default swap spreads on subprime debt widened 18 months before equities peaked.

Meanwhile, the companies at the center of the buildout are making accounting choices that obscure the true cost. All five major hyperscalers — Alphabet, Amazon, Apple, Meta, and Microsoft — have extended the stated useful life of their server equipment, in some cases from 3 years to 6. This is a perfectly legal accounting decision, but it has the effect of suppressing depreciation expenses and inflating reported earnings. Michael Burry estimates the cumulative overstatement at $176 billion over the 2026–2028 period, based on his analysis of public 10-K filings.

The hyperscalers’ own spending ratios tell the story in simpler terms. Our Capex-to-Cash-Flow ratio signal tracks how much of their operating cash flow the Big Five are reinvesting into AI infrastructure. When this ratio approaches 100%, they’re spending everything they earn. Past 100%, they’re borrowing to build. That signal currently sits at 82 and scored into the red zone.

None of this means AI technology is failing. The technology is real and improving rapidly. But the financial structure built around the technology — the debt, the spending ratios, the revenue gap, the accounting choices — follows a pattern that has appeared before. The telecom fiber buildout of 1999–2001 saw the same dynamics: transformative technology, massive capital deployment, debt-funded expansion, and a revenue curve that couldn’t keep pace with the investment curve. The fiber got used eventually. The companies that built it mostly didn’t survive to see that happen.

The question the AI Infrastructure Stress Index tracks is not whether AI is valuable. It’s whether the money being spent today will generate enough return before the financial structure buckles under its own weight. Right now, the composite score sits at 68 out of 100. The data says: probably not.