PJM Capacity Auction Procures 138,318 MW of Generation Resources as Work Continues To Address Growing Electricity Demand

PJM’s 2028/2029 Base Residual Auction cleared capacity at the FERC-approved cap of $325/MW-day (UCAP), 2.5% lower than the prior $333.44 cap, for a total cleared value of $16.4B. However, total cleared capacity plus FRR resources fell short of the reliability requirement by 6,831 MW, the first time in PJM history the entire RTO missed the reliability target, prompting plans for a September FERC “Backstop Procurement.” PJM is also pursuing expedited and streamlined interconnection (including AI-enabled study timeline reductions) to restore supply as demand grows faster than generation.

Analysis

This is a bullish signal for existing dispatchable generation and a mixed signal for everyone trying to add load in PJM. The important mechanism is not the auction print itself; it is that the market is now formally signaling a durable reserve deficit, which should keep forward power spreads and ancillary-value optionality elevated for owners of existing nuclear, gas, and other high-accreditation assets over the next 3-12 months.

The second-order loser is incremental demand, especially data-center buildouts that rely on fast interconnection and cheap capacity. Even if hyperscalers can sign bilateral PPAs or self-supply, smaller colo operators and balance-sheet constrained developers will face higher all-in site costs, more deposit friction, and more incentive to migrate to less constrained grids. GOOGL is a relative beneficiary only at the margin: its collaboration on queue-study tooling is strategic optionality, not an earnings driver, but it does improve its ability to secure power for AI infrastructure versus weaker peers.

The contrarian risk is that policy tools cap the upside before scarcity can fully reprice. Collars, mitigation, and a potential backstop procurement mean this may translate into a slower grind higher in merchant margins rather than a blow-off move. What would falsify the scarcity thesis is a credible supply response by September—approved backstop MW, faster interconnection approvals, or a 2029/2030 auction that clears materially below the cap with a shrinking shortfall.

Key Decisions for Investors

• Go long CEG vs. short XLU for the next 3-6 months; CEG captures the scarcity premium through existing nuclear/merchant exposure while XLU is more exposed to regulatory lag and customer pushback. Risk/reward: ~2:1 if the forward curve stays firm; stop if PJM backstop procurement unlocks meaningful new MW or power forwards mean-revert.
• Add ETN or PWR on pullbacks for a 6-12 month grid-capex trade; the real follow-through from this auction is transmission, substation, and interconnection spend, not just capacity prices. Falsify if FERC slows filings or if utility capex guidance is cut.
• Use any post-announcement weakness in GOOGL only as a tactical add, not a thesis trade; the AI/interconnection tooling angle is small but strategically supportive for future data-center siting. Better entry is on broader AI-power concern selloffs, with a 12-18 month horizon.
• Avoid or underweight PJM-exposed load-growth stories that cannot self-supply power cheaply until the September FERC process clarifies cost allocation; the risk is higher site economics and delayed ramp, especially for smaller colo operators.
• Monitor the 2029/2030 auction and September backstop filing as the key catalyst window; if either shows materially improved supply, rotate out of merchant power beta and back into rate-based utilities.