Jefferies strategist warns hyperscaler AI capex could backfire

Economic Times reports that Jefferies' Christopher Wood says the AI capex cycle could end through market pushback rather than voluntary spending cuts. Wood points to large bond issuance by major hyperscalers and argues that borrowed money is increasingly funding the buildout. His warning frames the AI infrastructure race as a potential malinvestment cycle if returns fail to match the scale of Microsoft, Meta, Amazon, and Alphabet's spending.

Key details: Wood cited large 2026 bond issuance by the four major US hyperscalers, He warned that markets may challenge AI spending before companies cut it themselves, The note frames AI capex as a possible capital-destruction risk.

Why it matters: Investor scrutiny of AI debt and capex is becoming a pressure point on the compute buildout behind frontier models.