Skip to content

AI investment bubble could burst in 2027-2028, triggering 36% S&P 500 crash

AI bubble could burst in 2027-2028, triggering 36% S&P 500 crash
  1. The AI investment bubble to burst in 2027 or 2028, potentially triggering the worst stock market crash since the global financial crisis.
  2. His year-end 2027 forecast puts the S&P 500 at 5,000 points, implying a 36% decline, while Europe’s Stoxx 600 could lose more than 30%.
  3. Rising borrowing costs, persistent inflation and depleted cash flows at major technology companies threaten AI infrastructure investment, although Klement expects equities to continue gaining through 2026.

Panmure Liberum expects the artificial intelligence investment boom to unravel in 2027 or 2028, potentially triggering the most severe global stock market crash since the financial crisis.

Global equities have climbed to record highs this year, supported partly by expectations that accelerating investment in AI infrastructure will drive technology sector earnings. The rally could reverse sharply as mounting financing costs put pressure on the companies funding that expansion.

Major cloud computing companies, known as hyperscalers, have largely exhausted their free cash flows, leaving them increasingly dependent on borrowing at a time when debt financing is becoming significantly more expensive.

The S&P 500 will finish 2027 at 5,000 points, representing a 36% decline from current levels. Analysts projection is substantially more bearish than those of seven other strategists tracked by Bloomberg, whose forecasts imply average potential gains of 14%.

Europe’s Stoxx 600 index to fall to 430 points, more than 30% below its current level.

The forecast represents a major reversal for a strategist whose bullish position on European equities last year proved successful. As recently as mid-September, Klement’s central assumption was that the S&P 500 would reach 8,300 points by the end of 2027.

Outlook reflects concerns that persistent inflation and the resulting increase in borrowing costs could undermine the capital expenditure supporting AI infrastructure development.

According to Bloomberg Intelligence estimates, hyperscalers could spend $713 bn on data-center capital expenditure in 2026, more than twice the amount invested last year.

Spending is expected to increase again in 2027, although at a slower rate. Those investment projections have supported earnings forecasts for major US technology companies.

Other market strategists remain more optimistic. Citigroup strategists said this week that strong corporate earnings in 2027 could sustain further gains in global equities, even as higher interest rates and geopolitical risks weigh on the investment outlook.

The Stoxx 600 finished last year close to his above-average forecast, and he remains the most bullish strategist tracked by Bloomberg for the European benchmark through the end of 2026, predicting an advance of approximately 10%. Outlook for US equities over the remainder of this year is similarly positive.

Despite anticipating a substantial market correction, analysts is not recommending that investors immediately reduce their equity exposure. Instead, they advise clients to prepare contingency plans and develop market-timing indicators that could help identify when a potential crash is beginning.