Negative market reactions to higher AI-related capex are broadening, suggesting a mood shift against funding more AI infrastructure without early returns.

Wood compares AI economics to an airline model, implying sustained infrastructure spend may yield limited returns for funders if big winners don’t emerge.

Despite ongoing capex growth by hyperscalers, analysts have not yet consistently cut earnings forecasts, indicating a lag between investment and visible earnings impact.

The key question is who captures the economics of AI growth; demand may persist even as incumbents shift, and markets demand proof before financing further capex.

As AI-related capex persists, analysts caution that the payoff depends on proven revenue and cash-flow growth rather than mere infrastructure buildup.

CXMT’s 500% surge on listing highlights how investors are chasing AI beneficiaries even amid broader market weakness and shifting valuations.

In Korea, the AI/reliability investment cycle is unwinding: the Kospi tumbled from its peak, foreign selling surged, and leveraged ETFs saw heavy net outflows amid an AI-capex alarm.

The market is becoming selective about AI spending, demanding concrete evidence that it will drive revenue and cash-flow growth rather than just infrastructure buildout.

There is a risk of a broader AI-related selloff if expectations for hyperscaler spending slow, potentially dragging down tech equities beyond a minor correction.

The market is growing wary of a broad AI capex binge by hyperscalers, with warnings that spending may burn billions if investors stop rewarding large infra outlays without clear returns.

Microsoft sticks to its 2026 capex guidance around $175 billion, illustrating selective investor appetite depending on perceived ROI and the impact of accounting adjustments.

Investor scrutiny intensified as Meta’s free cash flow plunged by 91% in Q2 while it raised its 2026 capex guidance, signaling the cost of aggressive AI spending.