PJM’s base capacity auction cleared at the maximum price its rules allow – the administratively set cap on its demand curve – while the amount of capacity still missing relative to the reserve requirement also grew in the same cycle. A price hitting its cap stops acting as a market signal and becomes a statement of scarcity the market cannot clear: PJM was willing to pay more, but its own system was designed not to. Because PJM serves roughly 65 million people and a large share of U.S. demand growth, the result will show up across rate filings, resource portfolios, and reliability planning for years, and it now makes further regulatory intervention in the market more likely, not less.
What a price cap appears: how PJM reaches the edge
PJM’s capacity market works on a three-year forward schedule. Load-serving entities must buy capacity certificates to cover their expected peak demand plus a required reserve margin, while generators, storage resources, and demand-response providers compete on price. The auction result is a single clearing price per region, often with multiple priced areas across the RTO. That price goes to all cleared resources, regardless of their individual cost, which is known as a clearing-out or uniform-price design.
The clearing price is not, however, infinitely flexible. It is structurally bounded by a “demand curve” – the variable reserve requirement curve – that caps the price at a level tied to the cost of new entry estimated for the market. The cap is intended as an explicit guardrail against monopoly pricing; it has been reached in previous auctions for specific localities and for certain sub-areas. But what the source describes – the capacity price hitting the price cap while the reserve shortfall at the same time grows – indicates the entire mechanism has no room left to expand rather than a case of limited turbulence in a narrow zone.
The result is a double break signal: the price has reached the top of the “allowed” range and the resource margin is still not sufficient. In other words, the shortage is real – the market cannot fill it regardless of price. The quote in the report is the best appreciation – Hoos of Aurora Energy: “the outcome demonstrates that the current system doesn’t work to bring online new capacity or stimulate demand response, the two things we need the most,” and the analyst warns of an “intervention doom loop.”
In structural terms, the concurrent outcomes are not contradictory. The price cap is a fixed ceiling, while the shortfall is a physical quantity: if the whole generation-by-committed material is insufficient, adding MW is available, the clearing due to capped can’t solve it. The market is telling the RTO and its member states that one year’s 2-3 year period does not have enough viable resource.
The conditions behind the auction: what a constantly ever-growing shortfall lowers
The causes, often described, are stronger overall: a larger group of well-known thermal plants retired, a very large interconnection queue with backlogged transmission upgrades for new gen including solar and storage, and retail rates whose load functions are on the verticality upward from data centers and the electrification of transport/buildings. On the demand side, the load of the PJM has grown far more rapidly than the base of capacity that was actually operating in the same years.
The branch also expands its supply side: an existing gas fleet is now dispatched more carefully under stressed weather and qualifications, which reduces amount of reliable capacity that PJM can count. The system allocates promptly, supply side: with an entire electric sector + several hours of storage not lasting cycle-as-counted basis, a large part of the queued power is not yet ready to be bid into the forward market at all. In that environment the auction is effectively looked at in a two-sided price-finding auction, but is in a “all’s first-come, first-served” that is undermined by policy-driven rollouts.
The role of demand response is a particular sign: it is supposed to be a resource that can respond to capacity pricing, but growth in actual DR participation has lagged the invention of capacity prices-persistent in the source’s analytical phrase-the need to “the two things we need.” The threshold, in the cap, is effectively “enough”, as the “cap” is insufficient to make demand shift materially.
Notably, if the auction is not actually “elastic” from the demand side, the result is heavyhandfully: all reductions in output do not reduce the price up to the supply that doesn’t exist; there is no new supply going in; price is at its ceiling. The mechanism – including the cap level – can be changed by the RTO and FERC. The question is how long the ceiling stays static while the load forecast grows, and the basis for this is that political intervention – probably along the line of “we jump the curve”-is already in others’ mind.
The “intervention doom loop”: parallel, other markets, and beyond the auction
The “doom loop” description belongs to Hoos, but in effect it is about a regularity: if the auction cannot clear physics though, it invites a series of emergency “fixes” – some already deployed in PJM, such as the 2023-2024 accelerator process that allowed certain plants to bypass the regular queue, or new transient wind-out mechanisms (rapid-entry, alignment or procurement) that were already-not yet published. Each discrete such fix typically assigns cash flows to a specific resource class and waits – and the natural result is that investors that “wait for the next fix”. The waiting itself prevents new entry, which continues the perception – that supported the original intervention, and that is why the “intervention doom loop” so neatly captures the present condition. In fact, an “intervention doom” is what happens if “market cap” is no longer the answer but the government is also not the answer. The figure is from the analyst’s prediction, not necessarily a statement from PJM, but it is consistent with the auction’s current outcome.
The worry is not just about PJM. Across the US, other RTOs have run similar peaks. In CAISO, the resource adequacy markets have been replaced with bilateral contracts; in ERCOT, an energy-only scarcity price cap was far and was raised in stages after extreme-events and cost-hysteresis, to be followed by new-era reliability off systems beyond the price signal. There has also been a growing use of “reliability procurement” inside – almost the extra-market tool. The PJM full cap-state is not anomalous, but from this structural economic principle: capacity markets, insofar as they are actually cleared, are always one or two rounds away from an administrative threshold. The market has the least chatter because that threshold now is the outcome.
A second cross-connection: storage and demand response are both cardinal resources that cannot show up in numbers as simple peaker MW – and both have arguably been under-accepted. If the market is always below, the price at the cap goes – the phenomena in “capacity-price delivers” – that still doesn’t cost new storage. The problem is that the price cap is too low to pass through the cost of a new resource – but a generator new storage build can rely on a capacity rate at cap only in part because storage often collects capacity for only a few hours, and a price cap allows capacity-value signal to be monopolized by incumbents. That is a structure interpretation, not a reported fact.
Data-center loads make another brace. The interconnection request queue… “digital loads grow” both legally determine this as contract with “fixed” capacity. PJM cannot forward price the currently large data-center capacity it will have to serve; the price cap is just a static boundary. The elasticity of demand, including a large margin of load that is actually programmable for giving a higher price – the cap is not transparity; so more of it will be served at fixed price, and the reserve frequency will remain much shorter in time than the targeted margin.
Who will be most affected
- Utility planners and load-serving entities – Plan the next rate cycles with capacity costs at the top of the curve, not at the average of history: budget “cap” as the base case, and expect higher retail open-access pricing based on capacity, not energy.
- Generation and storage developers – The high cleared price is a tailwind for new peak capacity, but only if the cap stays up; a merchant storage or gas peaker contract should reflect the fact that a cap creates a “hurry” premium, not a long-run sustainable structure; if the cap is eventually raised to the true marginal cost, there is some room for up – after bullish-, but the standard returns also remain tagged to timing risk through the “fix-waiting” cycle.
- Demand response aggregators and DR providers – The result strengthens the long-term “intervention” narrative needed to make demand-side resources dispatchable; even if the cap limits prices, a default shortfall is an invitation for DR-based commitments (including price-responsive capacity) and it is characterised by the fact DR actually within the market in the cap signal has served.
- State/reg/ policymakers – This is the hard data point for interventional planning: it should not be taken as an argument to “cost$ RTO design “but” – rather, a reason to set or modify retail obligations (temporary LOLP) and to closely monitor the capacity commission as it “corrects the solution”.
Next watch context
- Next base auction. Watch if next clearing remains at the cap – that would indicate the demand curve is no longer functioning as a “demand incorporation” but is the price ceiling – or whether PJM raises the case (or the reserve margin), which is equivalent to an administrative price increase.
- Queue and nearby transmission. The grid is slowed by interconnection processing; track PJM’s queue metrics and few projects that actually wind up in “completed source” for the youngest vintage.
- New build commitments. An auction price at cap should translate into 3-year out-period new-build commitments – not only “commit to file” – so you want to track actual tariffed new gas peaking, storage, and hybrid projects started over the next 18-24 months.
- PJM load forecast updates – The shortfall line literally depends on load growth assumption; next-year forecast which accounts for data centers and existing co-location contracts will be among the earliest “soft” news about whether the gap remains structural.
Bottom line
The capacity settled that at the cap is no longer a sign of strong price – it means the market cannot close a reservoir; the capacity gap remains. If the next auction shows the same combination – price cap plus expanded shortfall – the grid will be an administrative product: the market pricing on its own looks broken, and the “intervention doom loop” is already a real analytical outcome. That is the world now: investors and utilities commit resources regardless of the auction, and the auction’s demands to the alarm but not to the schedule. The important action is therefore not just what the cap number was, but whether it stays at the top for another delivery year.
Read the full report at Utility Dive.
Note: facts and figures attributed above to Utility Dive reflect that outlet's original reporting. Broader context, cross-sector connections, and forward-looking scenarios reflect independent analysis by our editorial team.
About this article: Drafted by Energy Ai with AI-assisted research and writing based on public reporting, then reviewed under our editorial process before publication.
Leave a Reply