On Thursday, Monitoring Analytics, the independent market monitor for PJM, released its 2026 Quarterly State of the Market Report for PJM: January through September. The real-time load-weighted average locational marginal price in the first nine months of 2026 rose $24.47 per MWh, or 48.4 percent, from the first nine months of 2025 — from $50.51 to $74.98 per MWh. The total cost of wholesale power rose 40.9 percent, to $111.61 per MWh. And the monitor’s verdict on the energy market, the market that produced those prices, was that it was competitive.
That verdict is the story. A 48 percent price increase that a market monitor calls competitive is not a contradiction; it is a diagnosis. Prices rose because the system that sets them is under strain — more load, fuller wires, dearer fuel — and the market, by the monitor’s account, reported that strain faithfully rather than manufacturing it.
The nine-month bill
Start with the totals, because the composition is the argument. Energy was the largest component of wholesale cost at 63.5 percent of the total, capacity 17.4 percent, transmission 17.2 percent — 98.1 percent between the three. The total cost of energy per MWh rose $22.42, or 46.3 percent, from $48.47 to $70.89. The total cost of capacity per MWh rose $8.70, or 81.1 percent, from $10.72 to $19.42. Transmission barely moved: up $0.84, or 4.6 percent, to $19.16 per MWh. The total cost per MWh of wholesale power rose $32.42, from $79.19 to $111.61.
Two measures of energy price appear in the report, and they should not be confused. The real-time load-weighted average LMP — the price at which energy actually cleared, weighted by where load sat — was $74.98 per MWh. The total cost of energy per MWh — what the energy component of the wholesale bill worked out to — was $70.89. The gap between them is part of the story: uplift charges, the payments made to keep units committed and dispatched out of merit, more than doubled this year.
The capacity side of the bill is where last week’s coverage of this same report concentrated, and it deserves only a sentence here. The monitor found the capacity auctions for the 2025/2026 through 2028/2029 delivery years not competitive, primarily because forecast data-center demand was included in them, and put the cost of that inclusion at $29.4 billion, irreversible, with more to come in December. The energy side is a separate finding: the capacity market priced a forecast, and the monitor says the forecast was the problem; the energy market priced the power actually flowing, and the monitor says the price was, on the whole, honest.
What actually drove the $24.47
The report breaks the LMP increase into components, and the ranking is the most important fact in it. Of the $24.47 per MWh increase, $8.81 — 36.0 percent — was the transmission constraint penalty factor component of LMP. Fuel and consumables accounted for $5.73, or 23.4 percent. Emissions costs accounted for $3.32, or 13.6 percent. The market power components accounted for $1.71, or 7.0 percent.
Read that ranking again. The largest single driver of the energy price increase was not fuel. It was the penalty factor that LMP applies when transmission constraints bind — the administratively set price of scarcity on a congested interface, which rises as the system runs out of cheap ways to move power from where it is generated to where it is consumed. Fuel mattered, and the report notes that gas and coal prices both rose in the first nine months of 2026 against the same period of 2025, with ultra-low-sulfur diesel at New York Harbor up 58.6 percent. But the wires contributed more than the fuel did, by half again as much. That is a statement about the physical system, not about market design: a penalty factor is not a markup, it is the price the market assigns to a constraint that is genuinely binding. The monitor’s finding that prices were set, on average, by units operating at or close to their short-run marginal costs — “evidence of generally competitive behavior and competitive market outcomes” — sits alongside this: the system was expensive because it was constrained, not because sellers were withholding. The qualification matters too: high markups for some marginal units did affect prices, and the 7.0 percent market-power component — $1.71 of the $24.47 — is the quantitative version of it.
More load, on a system that was already full
The demand side is straightforward. The real-time hourly average load in the first nine months of 2026 rose 3.4 percent from the same period of 2025, from 93,683 MWh to 96,845 MWh — a large increase for a nine-month period in a region whose load growth was, until recently, close to flat, and it arrived concentrated where the wires were already tightest.
The generation that served it shifted in ways that underline the point. Generation from coal units rose 0.4 percent, from natural gas units 2.5 percent, from oil units 13.7 percent, from wind units 5.5 percent, and from solar units 20.1 percent, all against the first nine months of 2025. The oil figure is the tell: oil is the fuel of last resort in PJM, the units that run when nothing cheaper is available or deliverable. Solar’s 20.1 percent growth is the counterweight, but solar does not set the price in the evening hours when load is highest and the constraints bind hardest.
The geography makes the same point more sharply than any regional average can. The Delaware Division of the Public Advocate, comparing PJM zones over July through September, found that in September 2026 the average wholesale price in the Dominion zone — which holds the largest concentration of large-scale data centers in PJM’s footprint — was $93.89 per MWh, against $32.19 in the Delmarva Power zone. The Delaware analysis attributes the gap to hot weather from September 15 through 18 coinciding with maintenance outages affecting generation and transmission resources. That is a single month in two zones, but it is the mechanism in miniature: heat raises load, maintenance takes capacity and transfer capability out of service, and the price in the constrained pocket separates from the price a few hundred miles away.
The congestion ledger: $4.9 billion, and $7.6 billion not returned
When transmission constraints bind, customers pay more for energy than generation is paid to produce it. The difference is congestion revenue, and it is the report’s second ledger — the one most likely to be overlooked while the capacity debate absorbs the attention. Total congestion rose $2,674.0 million, or 119.7 percent, from $2,233.8 million in the first nine months of 2025 to $4,907.8 million.
Congestion revenue, the monitor says, is paid by customers, belongs to customers, and should be returned to customers. The mechanism for returning it is the auction revenue rights and financial transmission rights market. But only 72.8 percent of total congestion paid by customers for the first four months of the 2026/2027 planning period was returned through the ARR and self-scheduled FTR revenues offset. Customers have received $7.6 billion less in congestion revenues than they should have since the 2011/2012 planning period, the monitor calculates, as a result of flaws in the ARR/FTR market design. Only 66.4 percent of congestion revenue was returned to load over that period. That is a design failure with a direct relationship to the price surge: the same constraints that drove the $8.81 penalty-factor contribution to LMP also generated $4.9 billion in congestion revenue, and more than a quarter of the congestion paid in the first four months of the current planning year was not returned to the customers who paid it. Total energy uplift charges tell the same story from the other side: up $744.2 million, or 112.5 percent, from $661.6 million to $1,405.7 million.
The investment signal is working — which is the awkward part
There is a finding in the report that complicates any simple story of market failure. Theoretical energy market net revenues — the measure of the incentive to invest in generation — rose sharply in the first nine months of 2026: up 105 percent for a new combustion turbine, 76 percent for a new combined cycle, 119 percent for a new coal plant, 38 percent for a new nuclear plant, 313 percent for a new diesel unit, 36 percent for a new onshore wind installation, 41 percent for offshore wind, and 21 percent for solar.
The energy market is paying new entry substantially more than it did a year ago. That is, in the monitor’s framework, the market doing its job: high prices are the signal that is supposed to summon new supply, and the signal is now loud. The awkward part is the same one this desk noted last week: the two markets are doing different jobs, and only one of them is broken. And the signal, however loud, cannot conjure a plant. Turbine backlogs are measured in years — GE Vernova reported a 116 GW gas-turbine backlog in the second quarter of 2026 — and transmission takes longer still. The net revenue figures are theoretical, for a new unit that does not yet exist.
What it means for buyers
Wholesale costs are not retail bills — hedging, rate cases, and allocation rules delay and redistribute wholesale costs before they reach a household — but a 40.9 percent increase in the total cost of wholesale power does not stay in the wholesale market. It moves through, on a lag, into the rate cases that set retail rates. The monitor’s own counterfactual makes the scale concrete: without data-center load and its impact on capacity costs for the 2025/2026 and 2026/2027 delivery years, the total cost of wholesale power would have been $101.66 per MWh instead of $111.61 — a $9.95 per MWh adder, or 9.8 percent, covering the capacity market only and excluding any data-center effect on energy or transmission costs. That adder is smaller than the first-half figure of $11.11 per MWh, the arithmetic of a fixed charge spread over more megawatt-hours. But the energy increase it sits on top of is larger in dollars than the capacity increase, and it is driven by load, constraints, and fuel — the physical system, not the auction design.
The practical consequence is that the two costs behave differently from here. The capacity cost is set by auctions that have already cleared, and the next one, for the 2029/2030 delivery year, closes December 15 — the fourth and final auction under the price collar, capped at $337.83 per MW-day, against a reliability requirement of 158,712 MW, 2,699 MW higher than the year before. The energy cost is set hour by hour, and it will move with the winter. A buyer can hedge the energy price and cannot hedge the capacity forecast. That asymmetry is the shape of the risk now.
December, and the redesign that cannot wait for it
The report’s warning about December is the sentence that should frame everything that follows, and it is the one place where the energy story and the capacity story converge. The monitor expects the December auction to add significantly to the $29.4 billion: the auction will price a demand curve that still includes more than 20,000 MW of forecast data-center load, against a reliability requirement that has risen again, in a system where the energy market is already pricing scarcity through constraints. The workshop PJM convened on November 13 to discuss redesigning the market is a multi-year project by PJM’s own timeline. The next auction closes in a month.
That mismatch is the report’s real subject, on both sides of the ledger. The energy market did what a market is supposed to do: it priced scarcity, stayed competitive on average, and is now paying new entry far more than a year ago. The capacity market did what a forecast-driven market does when the forecast is wrong: it locked in payments for load that may never arrive. The energy surge is the cost of a system running at its limits. The capacity cost is the cost of buying insurance against a future that was overstated. Both are now on the same bill, and the bill is $111.61 per MWh.
What to watch
December 15: the 2029/2030 Base Residual Auction closes — the last auction under the price collar, and the first test of whether the monitor’s warning about a significant increase in the $29.4 billion total is borne out. Read the energy numbers alongside the clearing price: the winter LMPs, the congestion total, and whether the ARR/FTR return rate improves from 72.8 percent.
The Large Load Registry, which took effect October 12, faces its first test in the same auction: whether the December demand curve reflects the roughly 4,017 MW of canceled and delayed data-center load the monitor has documented, or prices the old forecast one more time. And the constraint ledger: the monitor’s $7.6 billion ARR/FTR shortfall dates to the 2011/2012 planning period, and in a year when congestion rose 119.7 percent, the design flaw that returns only two-thirds of congestion revenue to the customers who paid it is no longer a footnote.