AI's grip on the S&P is total and Morgan Stanley's top analyst lays out case (finance.yahoo.com)

Morgan Stanley Wealth Management CIO Lisa Shalett warned that the U.S. equity rally is dangerously concentrated in an AI infrastructure story: the “Magnificent 7” plus ~34 AI-related firms now account for roughly 75% of S&P 500 returns since late 2022, 80% of earnings growth and 90% of capex growth. Shalett flagged a dependency on massive generative-AI data‑center buildouts—hyperscalers are spending nearly $400B annually—and pointed to large, intertwined deals (Nvidia’s headline $100B commitment to OpenAI, Oracle’s ~$300B data‑center pact) as creating circular financing and systemic risk. She warned this dynamic looks late‑cycle and could precipitate a sharp drawdown within 12–24 months if the capex narrative falters. The implications are both market‑structure and macro: AI capex has contributed roughly 100 basis points to GDP growth, but hyperscaler free cash flow has turned negative and may shrink ~16% over the next year, pressuring valuations and credit markets (Shalett said she’s watching CDS spreads on big borrowers). Counterarguments from some sell‑side analysts note deal tranches, performance‑based structures and that Nvidia’s actual funded exposure this year is far smaller than headline figures, suggesting potential for uneven “digestion” rather than outright collapse. Still, Shalett’s core message is that the rally’s breadth and leverage around AI infrastructure raise concentration, liquidity and credit risks that the AI/ML community and investors should monitor closely.