sPEG Index editions

A 88-day follow-up to the February 13, 2026 AI Infrastructure Scarcity Index. Tracks how the market repriced scarcity-adjusted valuation across the ten companies controlling critical bottleneck layers of AI infrastructure — compute, fabrication, memory, lithography, power, network, design, and analog control. Index average sPEG fell from 0.50 to 0.47 despite a 47% average price gain, indicating the rerating gap has widened rather than closed.

The AI-native product market is splitting between commodity wrappers and judgment products. This sPEG Index ranks AI products by the scarcity, durability, and irreplaceability of the expertise behind them.

Tracks scarcity-adjusted valuation efficiency across companies controlling critical memory bottlenecks required for AI compute, including HBM memory, NAND storage, enterprise flash systems, and memory controllers. Memory has emerged as the primary physical constraint in AI scaling, as GPU deployment is limited by the availability of high-bandwidth memory (HBM). This index measures which companies control the most structurally scarce layers of the AI memory stack using the proprietary Scarcity-adjusted PEG (sPEG) ratio.

The AI Energy & Power Constraint Index measures scarcity-adjusted valuation efficiency across companies controlling the critical energy, electrification, and grid infrastructure layers required to sustain large-scale AI deployment. Unlike compute, which can scale through fabrication investment, energy infrastructure is constrained by physics, permitting timelines, and capital intensity. These constraints create structural scarcity. Developed through institutional M&A scarcity analysis, the index applies the proprietary Scarcity-adjusted PEG (sPEG) framework to identify valuation asymmetries across power generation, grid buildout, electrification, and energy management companies. This index establishes a baseline benchmark for evaluating how markets price the physical energy constraints underlying AI expansion.