US stock indexes, where seven tech groups accounted for half of all market earnings in 2025, posted double-digit gains and reached record highs.

European and Asian equities followed suit, lifted by AI excitement and monetary easing.

Expectations of further rate cuts have also benefited bond investors, giving US Treasuries their best annual performance in five years, even as inflation remains above the Federal Reserve’s 2% target.

For 2026, government stimulus in the US, Europe, and Japan, alongside continued AI investment, is expected to sustain global growth.

However, analysts warn this could reignite inflation, prompting central banks to halt rate cuts or even raise rates, slowing the flow of easy money into tech-focused markets.

“You need a pin that pricks the bubble, and it will probably come through tighter money,” said Trevor Greetham, head of multi-asset at Royal London Asset Management.

Tighter monetary policy would curb speculative tech investments, raise funding costs for AI projects, and reduce profits and stock valuations for major tech firms.

The rapid expansion of data centres by hyperscalers like Microsoft, Meta, and Alphabet is also driving inflationary pressures through rising energy and chip costs.

“The costs are going up, not down,” said Morgan Stanley strategist Andrew Sheets, noting that US consumer price inflation could remain above 2% until the end of 2027 due in part to heavy AI investment.

Other analysts echo the concern. J.P. Morgan’s Fabio Bassi said stimulus, rate cuts, and a strong labour market would keep inflation above target.

Aviva Investors highlighted risks from central banks ending rate cuts or hiking rates amid AI-driven and government-led spending.

Mercer’s Julius Bendikas added that inflation risk is underappreciated, prompting a cautious shift out of vulnerable debt markets.

Markets have already shown sensitivity to rising costs.

Oracle shares fell last month after disclosing higher spending, while Broadcom warned that high profit margins could be squeezed.

Rising memory chip costs are expected to pressure prices and profits for personal computer makers and AI firms later in 2026.

Deutsche Bank projects AI data-centre capital expenditure could reach $4 trillion by 2030, creating supply bottlenecks in chips and electricity that drive investment costs higher.

George Chen, a former Meta executive, warned that rising costs and inflation could reduce investor returns and curb the flow of money into AI-focused companies.