PJM’s Independent Market Monitor says the inclusion of data centers in PJM’s capacity market has resulted in “an irreversible increase of $29.4 billion in capacity market costs for all customers to date, with more costs to come”. The finding is in the monitor’s 2026 Quarterly State of the Market Report for January through September, posted on 12 November; Bloomberg reported the rebuke of the largest US grid operator late that day.
The monitor, Monitoring Analytics, puts the effect at $9.95 per MWh: the total cost of wholesale power in PJM was $111.61 per MWh in the first nine months of 2026, and would have been $101.66 without data center load in the capacity market. That $9.95 is about 31% of the $32.42 year-on-year rise in the total cost, by our arithmetic, or roughly 1 cent per kWh.
Business Pill · Grid Cost Allocation
A short Business Engineer explainer on how grid costs are allocated. It teaches the general idea only and does not discuss this report.
The key insight: As we read the monitor’s figures, data center load accounts for about 31% of this year’s rise in the total cost of wholesale power in PJM, $9.95 of $32.42 per MWh, by our arithmetic, and that share counts the capacity market only.
The report says the results of the 2025/2026, 2026/2027, 2027/2028 and 2028/2029 capacity auctions “were not competitive, primarily as a result of the inclusion of forecast demand for data centers”. The energy market results were competitive, it says.
The total cost of wholesale power rose 40.9%, from $79.19 per MWh in the first nine months of 2025 to $111.61, according to Section 1 of the report. Capacity cost per MWh rose 81.1%, from $10.72 to $19.42. The monitor notes the $9.95 data center figure does not include the impact on energy or transmission costs.
PJM is also short of its reliability target. The shortfall grew from 208.7 MW in the 2026/2027 auction to 6,516.6 MW in 2027/2028 and 6,831.3 MW in 2028/2029, the report says.
Total cost of wholesale power in PJM, January to September, $ per MWh. The monitor estimates $101.66 without data center load in the capacity market, against $111.61 actual; the gap excludes energy and transmission effects.
The December Auction
The monitor says PJM’s Reliability Backstop Procurement filing, now delayed until after the December auction for 2029/2030, leaves most forecast data center load in that auction and “effectively ensures that the auction will clear at the maximum price and add another $5 to $10 billion of data center related costs to all customers”. Added to the $29.4 billion, that would be $34.4 billion to $39.4 billion, by our arithmetic.
It also says the 2029/2030 auction will use PJM’s 2026 load forecast, which includes data centers “that have already been canceled or are unlikely to reach projected demand levels at the target date”.
What the Monitor Recommends
The report says “PJM should postpone the December auction until it can remove all forecast data center load from the auction”, and run a reliability backstop auction for data center load without a bilateral contract for generation. Its preferred market solution is to require data centers to bring their own new generation, with a queue for large new loads that would not be connected until there is capacity to serve them.
The monitor writes that “Data center loads should pay their own costs and cover their own risks”. It says PJM’s current RBP and IRAS proposals fail to meet the objective, set out in the White House Ratepayer Protection Pledge and the governors’ principles, that data centers pay their own costs.
As we read the report, the monitor treats the cost as a policy choice rather than a market failure: it says PJM’s issues are a result of public policy decisions made by PJM, and that the markets have worked.
As we read the capacity figures, capacity cost per MWh rose 81.1% in a year, from $10.72 to $19.42, according to the report, while transmission cost per MWh rose 4.6%.
Monitoring Analytics, PJM Independent Market Monitor, 2026 Quarterly State of the Market Report, 12 November 2026
“Data center loads should pay their own costs and cover their own risks”
Three Implications
FOR PJM CUSTOMERS The monitor says the December auction will add another $5 billion to $10 billion of data center related costs unless forecast data center load is removed.
FOR DATA CENTER DEVELOPERS The monitor proposes that large new loads bring their own new generation, with a queue for loads until capacity exists to serve them.
FOR OUR COVERAGE We will report PJM’s response and the December auction result when they are published.
The framework says: “The structural bottleneck isn’t computing, it’s electrons.”
As we read it, the monitor’s report shows who pays for those electrons in PJM: capacity costs for all customers rose with forecast data center load. The framework does not discuss PJM.
What Is Not Established
The $9.95 and $29.4 billion figures are the monitor’s counterfactual estimates, not audited costs. The report does not convert them into a household bill figure in the section we read, and retail bills depend on each state’s utility rates. We have not seen Bloomberg’s article beyond its headline.
This Week In The AI Supercycle
The Great Convergence of Capital, Power, and Geopolitics
This week’s issue: for two years the AI build-out was priced as a supply story, and last week as a credit story. This week it was priced as a permission story, and the answer arrived as a cost, not a ban.
PJM’s market monitor puts the capacity cost of data center load at $29.4 billion to date and $9.95 per MWh of wholesale power this year, about 31% of the year-on-year rise by our arithmetic, and asks PJM to postpone December’s auction.
A note on sourcing. Figures and quotations come from Monitoring Analytics, the PJM Independent Market Monitor, 2026 Quarterly State of the Market Report for PJM: January through September, Section 1, posted 12 November 2026. Bloomberg is credited for its report of the rebuke. Shares, per-kWh and summed figures are our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
