PUCO Adds Electricity Requirements for Data Center Developers

Ohio regulators (PUCO) are increasing electricity cost burdens on new data center developers by ordering First Energy to create a separate rate plan for data centers and barring them from AEP Ohio’s Transmission Cost Recovery pilot program. The move signals tougher utility cost allocation for large power users and could modestly raise operating costs for future data center projects in Ohio. The article is policy-focused and does not provide dollar or percentage impacts.

Analysis

This is a margin-compression event for the AI/data-center buildout, not a demand destruction event. The key second-order effect is that local utilities are trying to force hyperscalers and colocation developers to internalize grid upgrade costs earlier, which should raise project hurdle rates, slow speculative land grabs, and favor operators with balance-sheet flexibility, captive generation, or superior site-selection discipline. In practice, that shifts power from “any megawatt is good” developers toward incumbents that can pre-contract power or self-supply.

The most immediate beneficiaries are equipment and service providers tied to behind-the-meter solutions: gas turbines, switchgear, transformers, backup power, and modular generation. The losers are pure-play data-center developers and land banks whose economics depend on cheap, socialized interconnection; over a 6-18 month window, this can compress implied cap rates for assets in constrained grids and widen spreads between top-tier markets with available capacity and secondary markets where upgrade costs get pushed onto the tenant.

A bigger second-order risk is project delay, not outright cancellation. If permitting and rate-design complexity lengthens by even 2-4 quarters, the market could see a local bottleneck rather than a national one, which is bullish for utilities and grid suppliers but bearish for the most levered infrastructure names financing build-to-suit pipelines. Conversely, if state policy softens under AI competitiveness pressure, the market will quickly re-rate the entire “power scarcity” trade because many of these projects still have multi-year demand visibility.

The contrarian view is that this is not necessarily anti-AI; it may be pro-AI capex discipline. By making power more expensive and less subsidized, regulators may actually favor the largest hyperscalers, which can absorb or hedge costs and outbid smaller entrants, potentially increasing concentration in the sector. That means the headline may read as a setback for data centers, while the real winner is the handful of platforms with scale and procurement leverage.

Key Decisions for Investors

• Long grid/electrification beneficiaries on pullbacks: ETN or HUBB over 3-6 months, on the thesis that behind-the-meter and distribution spending accelerates as developers redesign projects around self-supply; target 10-15% upside with lower regulatory beta than pure data-center names.
• Short highly levered data-center developers/land plays where valuation assumes cheap grid access; use a basket/ETF proxy if single-name liquidity is poor, with a 3-9 month horizon and stop if state policy broadens cost-sharing relief.
• Pair trade: long utility-adjacent power infrastructure suppliers vs short data-center REITs over 6-12 months, capturing the spread between capex required to connect power and the higher financing cost borne by developers.
• Consider a call spread on natural-gas generation beneficiaries if you can source a liquid proxy, since on-site generation becomes more attractive when grid access is politicized; the trade works best if policy headlines persist for 1-2 quarters.
• Avoid chasing broad AI infrastructure strength until there is clarity on interconnection timelines; the near-term risk is estimate revisions for project starts, not end-demand, so the cleaner expression is selective long power bottlenecks rather than indiscriminate long data-center exposure.