BNP Paribas reports that consensus estimates for 2026 hyperscaler AI capex have reached $725B following Q1 2026 earnings guidance, nearly double the $365B estimate from mid-2025; the bank notes capex is now rising faster than operating cash flow, creating external funding needs. UBS separately projects $511B in power-side spending alone through 2030 to support AI data-center load, while Evercore ISI places the capex figure closer to $800B concentrated among Alphabet, Microsoft, Meta, Amazon, and Oracle. For financial institutions, the OCF-capex gap flagged by BNP Paribas is a direct stress test of whether hyperscaler AI buildout remains self-funded, with implications for investment-grade debt issuance and credit exposure across the AI infrastructure supply chain.
314 other tracked developments touching the same claims.