The S&P 500 achieved a 16.3% total return over April and May, with technology stocks surging 36.3%, according to a Wall Street Journal report cited in prior MSI coverage. Just 12 companies accounted for 12.2 percentage points of the index’s performance. The rally has persisted through the second quarter even as a growing number of analysts, academics, and market observers warn that the artificial-intelligence sector is in bubble territory.
The current investment frenzy will end the same way the dot-com boom did — with a crash — Mark Surman, president of the Mozilla Foundation, wrote in the Guardian in January. “If history is any guide, the current frenzy will end the same way the dot-com boom did: with a crash,” Surman wrote.
The scale of capital deployment and corporate borrowing in the AI sector has drawn comparisons to previous speculative episodes. The largest AI hyperscalers issued $159 billion in bonds globally this year, according to Dealogic data cited in prior MSI reporting. Alphabet announced an $85 billion equity raise, and four major tech companies are expected to spend over $670 billion on data centers and AI infrastructure this year, raising questions about whether the revenue will materialize to justify the spending.
Fortune reported in May that the AI economy could crash on mounting chip costs. Since many loans use existing chips as collateral, any default may flood the market with older processors, bringing their value down further in a cascading collapse. The token bubble compounds this risk: if enterprise customers begin capping or cutting AI usage — as Microsoft’s own license cancellations suggest is already happening, according to Fortune — revenue projections underpinning chip-collateralized debt may prove optimistic precisely when lenders need them most.
The risk extends beyond overcapacity in chip manufacturing. Further innovation in semiconductor chip design or major advances in quantum computing could immediately leave much of the hundreds of billions in data center infrastructure investment useless, according to Yale Insights.
In early June, Morgan Stanley Investment Management analysts Michael Mauboussin and Dan Callahan published a base-rate analysis of the AI sector, finding that the revenue growth rates baked into AI company valuations have no historical precedent. The analysis, titled “Bayes and Base Rates” and cited in prior MSI reporting, suggested that most investors in the sector will lose money.
Investor Michael Burry opened large bearish positions in Nvidia and Palantir last year, according to reports cited by MarketWise, adding to the growing list of cautionary signals from market participants. One analyst quoted by MarketWise wrote in a piece about coming AI-related layoffs that “AI is almost certainly a speculative bubble. But it can grow far more manic. And it likely will.”