Jefferies sees a 2008-style credit angle in the AI boom as hyperscaler RPO backlog nears $2.3tn
SumitGup · x · 2026-10-13
Jefferies argues credit risk is now a real issue in the AI boom because "it can no longer be assumed that large language models will ever be profitable," drawing a parallel to 2008 real estate credit. An accompanying chart (Bloomberg data) shows hyperscalers' remaining performance obligations—Microsoft, Amazon, Alphabet, Oracle—climbing from $0.3tn in early 2022 toward $2.3tn by Q2 2026, with Oracle and Microsoft expanding fastest.
More from Venture
- China's $165B AI buildout rests on $11B model revenue and 4% of Nvidia's chip output — demian_ai · 2026-10-13
- subquadratic's New Model Beats Jev in 6 of 7 Production Use Cases Before Launch — ryanallred · 2026-10-13
- Palantir sets Q3 earnings date for November 2 — BrettKrieger12 · 2026-10-13
- Meta Muse DAU fell 62% in three weeks as insiders made 172 trades with no buys — richardttang · 2026-10-13
- Polymarket prices AI bubble bursting this year at just 6% — Polymarket · 2026-10-13
- Flexport CEO Channels Munger: In AI, Your Competitors Are All Geniuses — a16z · 2026-10-13