AI Power Grid & Infrastructure

Targets the critical bottleneck in the AI revolution — power. Every data center, every GPU cluster, every reshored factory needs massive electricity, and the US grid can't keep up. More than half of planned data centers face delays because transformers and switchgear are in critical shortage. This strategy holds 11 stocks across the power supply chain: independent power producers with 10-20 year contracted revenue from hyperscalers like Microsoft, Amazon, and Meta (Vistra, Constellation Energy, Talen Energy), grid equipment makers building the transmission and distribution infrastructure to deliver that power (Eaton, GE Vernova, Quanta Services, Hubbell), and specialist suppliers of data center switchgear, cooling, and connectivity (Powell Industries, Vertiv, Amphenol). XLU regime filter de-risks the speculative tiers when utility sentiment weakens, while Tier 1 power producers stay invested through downturns backed by locked-in purchase agreements. $1.4 trillion in committed utility capex through 2030 and $650B+ in hyperscaler AI spending in 2026 alone ensure the demand pipeline extends for years.