AI Has Delivered 'Zero Payoff' Outside Big Tech, Economist Warns Bubble Could Trigger Market Crash

Apollo Global Management chief economist Torsten Slok said aggregate financial metrics for the 493 S&P 500 companies outside the Magnificent Seven show no clear evidence of AI-driven financial gains since 2022, with revenue per worker and profit margins remaining broadly flat.

Separately, the International Monetary Fund (IMF) has warned that an AI-market correction could threaten global financial stability as falling valuations, forced selling and tighter credit spread losses beyond the technology sector.

Slok previously outlined a scenario in which AI disappointment would burst the bubble and halve the Nasdaq. Barclays estimates AI-related credit assets at about $680 billion (£510 billion), while the International Monetary Fund (IMF) and a Washington think tank have warned about the risks of an AI market downturn.

Non-Tech Firms Report Flat Performance

The Magnificent Seven, comprising Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla, have seen revenue per worker rise and profit margins expand. By contrast, margins at the remaining S&P 500 companies have stayed at about 10%, Slok said.

He argued that many corporate leaders increasingly view spending on AI tokens as a cost that squeezes margins rather than generates revenue. He also said business performance depends on human judgement, accountability, culture and trust, which AI models cannot replicate.

Goldman Sachs chief executive David Solomon said there is very little businesses do that requires frontier AI models. Slok argued that economic gains would remain limited while advanced AI spending was concentrated among a small group of venture capital-backed firms.

AI Borrowing Pushes up Credit Costs

Barclays estimated that $680 billion (£510 billion) in AI-related credit accounts for roughly 13% of duration-times-spread risk exposure in investment-grade debt.

Meta reported negative free cash flow of $2.12 billion (£1.59 billion) in the third quarter of 2026. Separately, it announced plans for equity offerings totalling up to $95 billion (£71.25 billion) as it funds its AI expansion. The company forecasts capital expenditure of $135 billion to $145 billion (£101.25 billion to £108.75 billion) in 2026.

Analysts warn that when several large technology companies issue bonds simultaneously, borrowing costs rise across the market. Yields on CCC-rated and lower-rated high-yield bonds reached 18.14% on 23 October 2026, the highest level since 2020, according to the ICE BofA data series published by FRED. The yield was 15.85% on 23 September.

The Open Markets Institute argued in its report, 'Brace for Impact: Protecting People and the Economy When the AI Bubble Bursts', that speculative data-centre debt was being repackaged and sold to pension funds, retirement accounts and university endowments.

The report compared the practice with the packaging of subprime mortgages associated with the 2008 financial crash. It said leading AI laboratories were accumulating debt while struggling to generate enough revenue to cover it.

The Open Markets Institute also argued that regulatory changes had weakened safeguards against financial-market disruption.

IMF Warns of Risks From Circular Financing

The IMF identified approximately $300 billion (£225 billion) in equity investments involving hyperscalers and weaker AI firms across 14 publicly disclosed transactions since 2023, alongside $650 billion (£487.5 billion) in debt and $1.4 trillion (£1.05 trillion) in future revenue commitments.

The IMF warned that a sharp decline in valuations could trigger margin calls and forced selling, reduce technology investment and tighten credit worldwide. Global exposure to US assets has reached a record high, leaving Asian markets particularly exposed.

The UK and euro area could be relatively insulated, although investors holding US technology shares would still face losses.