Global markets face renewed turmoil as AI debt and war fears rise

Financial markets have entered a period of renewed instability as escalating conflict in the Middle East and concerns over artificial intelligence investment fuel alarm among investors. The optimistic sentiment that characterised the summer, driven by a rally in US equities and the promise of an AI revolution, has given way to caution. With the war in Iran intensifying without a clear resolution, the global economy faces mounting pressure from soaring government bond yields and a potential slowdown in the AI sector. This shift has raised fears that the current market conditions could precipitate a broader crash, particularly as the cost of borrowing reaches levels not seen since 2007.

The immediate trigger for this anxiety is the sharp rise in US government borrowing costs, which have climbed to their highest level in nearly two decades. This increase carries significant implications for households, businesses, and governments worldwide, as it reflects growing fears that the ongoing conflict is driving inflation higher. Additionally, investors are concerned about the sustainability of US public finances, with Washington’s debt levels exceeding $40tn. The combination of these factors, alongside oil prices rising above $100 a barrel, has created heavy selling pressure in the bond market, raising the question of whether equities will be the next asset class to suffer a significant decline.

Central banks are responding to these inflationary pressures with aggressive monetary policy. The US Federal Reserve defied political pressure to raise interest rates for the first time since 2023, while the European Central Bank and the Bank of Japan have also moved to tighten policy. The Bank of Japan raised its policy rate to a 31-year high, and financial markets now anticipate four rate hikes from the Bank of England before the end of next year. The rationale behind these measures is to prevent short-term inflation from becoming entrenched, although this approach risks weighing on an economy already struggling with a cost of living crisis. Historically, US recessions have followed the first rate rise by approximately three to three and a half years, and markets often decline before such downturns begin.

A significant portion of investor concern centres on the valuation of the US stock market, which is heavily weighted towards technology stocks. The S&P 500 index remains close to its all-time high, with the combined value of the so-called magnificent seven tech companies exceeding $20tn. However, the cyclically adjusted price-to-earnings ratio, or CAPE ratio, has risen to its highest level since 2000. At nearly 41 points, this ratio is more than double its long-term average and approaches the record high seen just before the dotcom crash in 1999. This suggests that the US market is unusually highly valued relative to its profits, a condition that increases vulnerability to a correction.

The sustainability of the AI investment boom is under scrutiny, with analysts warning that the current spending may not be justified by future returns. Research indicates that for the AI sector to turn a profit, related sales would need to increase by between $600bn and $800bn within two years. One consultancy estimates a 30% chance that the AI bubble will burst next year, citing the lack of economic justification for the current capital expenditure. This sentiment is echoed by a large number of investors who registered for a recent analyst call discussing the risks of reckless development in the sector, drawing parallels to the dotcom era when infrastructure was built ahead of demand.

Stress is also evident in credit markets and retail investor behaviour. The spread between risky and safe high-yielding debt has widened, indicating greater caution among lenders. In South Korea, a wave of margin calls hit 1.2 million retail investors who had borrowed to buy AI-linked chip stocks, forcing many to sell their positions. Similarly, the shares of Oracle, a major player in AI infrastructure, have halved since a peak last year as investors worry about the company’s borrowing levels to fund datacentres. These developments highlight the fragility of the current market structure, where high leverage and aggressive spending could lead to a rapid unwinding of positions if confidence wanes.

Despite the turbulence, some analysts argue that a full-scale crash is not inevitable. They suggest that markets may be overestimating the inflationary impact of the geopolitical conflict and that the productivity gains from AI could eventually justify current valuations. The US economy has shown signs of improved productivity, and the UK economy has outperformed expectations in the first half of the year. However, former Bank of England chief economist Andy Haldane warns that while an outright collapse is unlikely, a slow release of pressure from the AI sector could slow the global economy. The situation remains fragile, with the outcome dependent on how well the market absorbs the combined shocks of war, inflation, and technological uncertainty.