On April 30, the Federal Energy Regulatory Commission approved a mechanism that will prevent the wholesale electricity price in PJM from jumping $225 per megawatt-day. That may sound like an obscure grid operator decision, but it affects the power bills of 67 million people across 13 states and the District of Columbia, and it reveals something uncomfortable about how America's electricity grid now works: we have to manually prevent market prices from blowing apart because one industry, data centers, is consuming capacity so aggressively that the normal auction process cannot handle it.
PJM, which operates the largest competitive electricity market in the United States, was facing a straightforward market problem. In its most recent capacity auction for the 2027-28 delivery year, prices hit the FERC-approved cap of $333.44 per megawatt-day. Without intervention, the auction for 2028-29 would have broken that cap entirely, likely clearing at $550 per megawatt-day or higher. That is a 69% price spike in a single year. The data center boom in PJM's footprint is the direct cause: the region's peak load forecast for 2027-28 is 5,250 MW higher than the forecast for 2026-27. Of that 5,250 MW increase, nearly 5,100 MW comes from data center demand alone. FERC's April 30 order extends a price collar, a cap at $325/MW-day and a floor at $175/MW-day, through the 2029-30 delivery year, locking in a price band that prevents explosive volatility.
The economics here matter because they illustrate why price controls, which economists usually dislike, are sometimes the cleaner answer to a temporary supply constraint. PJM's grid operator, using sound statistical methods, determined that new generation and storage are being built and connected at a steady clip, but there is a two-to-three year lag between when a developer decides to build and when that asset can physically inject power. The data center rush means demand is jumping faster than supply can follow. FERC approved the collar extension on April 28, with all 13 governors in PJM's territory, the White House National Energy Dominance Council, and the U.S. Department of Energy backing the move. This is not a close call: every political and regulatory stakeholder agreed that letting capacity prices spike to $550/MW-day right now would be unnecessarily destructive and solve nothing because new supply cannot materialize quickly enough to satisfy the incentive anyway.
Why now? Because two things collided. First, the data center industry moved from a speculative narrative to a real, structural load driver. PJM processed over 170,000 MW of new generation requests since 2023, and the interconnection queue still has 30,000 MW of generation waiting to be studied. Second, FERC and PJM learned from the 2026-27 and 2027-28 auctions that the old price caps were already binding, prices were hitting the ceiling. The 2027-28 auction cleared at exactly the FERC-approved cap, $333.44/MW-day. Without extending the collar, the next auction would have broken free and spiked to $550, creating price whiplash that would have made long-term investment planning nearly impossible for utilities and forcing ratepayers to absorb the shock all at once. Extending the collar gives the interconnection queue time to clear and gives FERC room to implement other reforms, like the expedited transmission and co-location rules it adopted in late 2025 to unlock grid capacity faster.
The collar extension benefits residential and commercial ratepayers directly, it prevents their power bills from experiencing a sudden 69% jump in the wholesale capacity component. Utilities that already have customer protections, like regulated vertically-integrated utilities, are largely insulated from the spike anyway; the hit would have fallen heaviest on restructured markets and the commercial customers exposed to spot prices in those markets. The approval also signals something important: FERC will intervene in energy markets when the underlying constraint is clearly temporary and structural intervention is cheaper than the chaos of an unconstrained price. But there is a hard tradeoff. Battery storage developers, who have been warning that price caps dampen their incentive to enter the PJM market, lose some margin in their investment calculus. The data shows developers are adding 24 GW of utility-scale battery storage to the U.S. grid in 2026, up from 15 GW in 2025. But the price collar may mean fewer of those storage projects land in PJM, where the peak load volatility is actually highest. That is not a trivial cost.
This is the right call for the next three years, but it is also a stopgap. The real story here is that PJM's capacity market is fundamentally broken as currently structured because it assumes supply can track demand in a one-year auction cycle. Data centers are killing that assumption. The interconnection queue is the release valve, but it is clogged. FERC is trying to unclog it with transmission interconnection reforms and co-location fast-tracks, but those take time. What the collar extension actually tells you is that FERC knows the grid operator cannot let prices spike unpredictably until those reforms clear the backlog. It is acknowledging that the market mechanism is temporarily dysfunctional and choosing stabilization over volatility while the plumbing gets fixed. That is honest. But if the queue does not clear fast enough, or if data center demand continues to grow faster than supply can follow, FERC will face an ugly choice between letting prices spike again or maintaining permanent price controls, and nobody wants to be regulating energy markets forever.
Watch three things over the next 18 months. First, the July 7 auction closing for the 2028-29 delivery year, does it clear at the new $325 cap, or well below it, suggesting supply tightness is easing? Second, FERC's progress on the interconnection queue and transmission co-location rules it ordered in late 2025, are utilities and developers actually using the expedited tracks, and are projects moving faster? Third, PJM's data center demand update in its next load forecast, does the 5,100 MW figure hold steady, accelerate, or begin to level off? If supply is catching up and the queue is clearing, the collar becomes moot and prices will naturally decline. If the queue remains clogged or data center demand keeps accelerating, FERC will face real political pressure to either make the collar permanent or let it expire and absorb the price shock. The next 18 months will determine whether this is a surgical emergency measure or the beginning of a much longer regulatory intervention.