Get your free personalized podcast brief
We scan new podcasts and send you the top 5 insights daily.
Historically treated as slow, defensive income shelters, utilities became correlated with the Magnificent Seven tech trade as AI data centers demanded enormous electricity capacity. However, as bond yields spiked, utilities divorced sharply from mega-cap tech, proving that despite secular AI power narratives, utility valuations remain fundamentally bound by bond market discount rates.
Related Insights
While current financing focuses on data centers and chips, the next critical bottleneck requiring immense capital is energy and power. This demand will drive a huge wave of financing for utilities, which will use different instruments like junior subordinated debt and project finance due to their distinct balance sheets compared to tech hyperscalers.
The massive, direct, and geographically concentrated energy demand from AI data centers makes local U.S. power markets the most effective AI-related commodity trade. With 72% of data centers in just 1% of counties and a constrained grid, local power prices are poised to rise significantly, offering a targeted investment thesis.
AI companies are building their own power plants due to slow utility responses. They overbuild for reliability, and this excess capacity will eventually be sold back to the grid, transforming them into desirable sources of cheap, local energy for communities within five years.
Credit investors should look beyond direct AI companies. According to Victoria Fernandez, the massive infrastructure build-out for AI creates a significant tailwind for power and energy companies, offering a less crowded investment thesis with potentially wider spreads and strong fundamentals.
Unlike typical diversified economic growth, the current electricity demand surge is overwhelmingly driven by data centers. This concentration creates a significant risk for utilities: if the AI boom falters after massive grid investments are made, that infrastructure could become stranded, posing a huge financial problem.
Electricity prices have been on a consistent upward climb, contributing to inflation that directly impacts household budgets. A key driver behind this trend is the massive and growing energy demand from AI data centers. This suggests a new, structural source of upward pressure on utility costs that is just beginning.
Before AI delivers long-term deflationary productivity, it requires a massive, inflationary build-out of physical infrastructure. This makes sectors like utilities, pipelines, and energy infrastructure a timely hedge against inflation and a diversifier away from concentrated tech bets.
With 30-year U.S. treasury yields crossing 5.2% (an effective 9-10% pre-tax return), the incentive to hold volatile, high-multiple AI stocks diminishes significantly. Investors can now opt for a guaranteed high return from government bonds instead of gambling on short-term market sentiment, creating a major headwind for tech valuations.
Soaring power consumption from AI is widening the "power spread"—the difference between the cost to generate electricity and its selling price. This projected 15% expansion in profit margins will significantly boost earnings for power generation companies, creating massive value across the supply chain.
AI labs are flooding utility providers with massive, speculative power requests to secure future capacity. This creates a vicious cycle where everyone asks for more than they need out of fear of missing out, causing gridlock and making it appear there's less available power than actually exists.
