AI investment cycle shifts pricing power toward memory suppliers

Jefferies (Christopher Wood) argues the AI investment cycle is shifting pricing power toward memory suppliers.

• Main announcement/action: Jefferies’ Christopher Wood argues pricing power is moving from hyperscaler cloud providers to memory producers (notably SK Hynix and Micron); Jefferies estimates memory prices rose roughly 50% last quarter, and notes new fabrication facilities require capital outlays in the tens of billions of dollars, with some memory manufacturers asking customers to help finance expansion in return for assured supply.
• Background and other details: Jefferies highlights energy constraints for AI growth — OpenAI urged the U.S. to add 100 gigawatts of new power capacity per year, Jefferies warns the U.S. grid is aging (~70% over 25 years old), and cites Google’s $4.75 billion acquisition of Intersect (a wind and solar developer) this week as an example of a hyperscaler securing power directly. This article is an analysis/opinion piece adapted from Jefferies’ “Greed & Fear” newsletter and references recent announcements (Google acquisition, OpenAI letter) rather than being a primary press release.