The five largest hyperscalers (Amazon, Microsoft, Alphabet, Meta, Oracle) are on track for $700-900B in combined capex in 2026, a 36% increase over 2025, with roughly $450B tied directly to AI infrastructure. Sequoia's David Cahn estimates a ~$600B annual gap between AI infrastructure spend and AI ecosystem revenue, with Allianz Research noting the capex-to-revenue divergence at 46%, exceeding the 32% seen in the 2001 telecom cycle. CreditSights documents that aggregate capex now exceeds projected cash flows for the group, which raised $108B in new debt in 2025, and Morgan Stanley and JPMorgan project the sector may require $1.5T in new debt issuance over several years. For financial institutions, this signals a structural shift in hyperscaler credit profiles, with capital intensity (capex as a percentage of revenue) reaching 45-57%, reducing the asset-light, high-FCF characteristics that historically justified premium equity multiples.
508 other tracked developments touching the same claims.