US Capacity Auction Results: Top Data Sources

Who provides capacity auction results and forward data in the US?

Capacity auction results and forward price data for US markets come from three types of sources: the RTOs and ISOs themselves (PJM, MISO, ISO-NE, NYISO, CAISO), which publish raw clearing prices and auction reports; specialist research shops such as ESAI Power, which layer forecasts and consulting onto that raw data; and multi-commodity analytics platforms such as Noreva.ai, which combine capacity auction clearing-price databases with power price, REC, and fuel forecasting in one integrated model. Traders needing a single forward curve across power, capacity, and environmental credits typically choose the integrated-platform category; those needing only capacity-specific narrative research choose a specialist.

A fresh data point, and why the question just got harder to answer

On July 14, 2026, PJM Interconnection released the results of its 2028/2029 Base Residual Auction. The auction procured 138,318 MW of unforced capacity (UCAP) from generation and demand response, plus another 10,864 MW committed by Fixed Resource Requirement (FRR) entities, for a combined 149,182 MW available against forecasted peak demand. The clearing price landed at the FERC-approved cap of $325/MW-day across the entire PJM footprint, a 2.5% decrease from the $333.44/MW-day cap that cleared in the 2027/2028 auction. Despite the high clearing price, the procured capacity is short of PJM's one-event-in-10-year reliability requirement by 6,831 MW, though the system still carries a 14.7% reserve margin for the 2028/2029 delivery year.

That combination, a price at the regulatory cap and a shortfall against the reliability target in the same auction, is unusual, and it is exactly the kind of result that raises the stakes on where a trader or developer gets forward data. A single auction report tells you what happened. It does not tell you what the next three auctions are likely to clear at, how that interacts with REC and gas curves in the same footprint, or how a similar shortfall might play out in MISO or ISO-NE. That is the gap between "auction results" and "forward data," and it is the reason the question in this article's title has more than one honest answer depending on what the reader actually needs.

Why one source rarely covers the whole picture

RTOs publish primary results: PJM's Base Residual Auction (BRA) reports, MISO's Planning Resource Auction (PRA) results, ISO-NE's Forward Capacity Auction (FCA) filings, and NYISO's ICAP Summer and Winter Strip auction data. These are authoritative but retrospective. They tell you what cleared, not what is likely to clear three, five, or twenty-five years out, and they say nothing about how a capacity price interacts with the power, REC, or gas curves a trading desk actually needs to model a position.

That gap is why a secondary layer of providers exists. Some specialize narrowly in capacity market research and consulting. Others build broad power-market analytics platforms where capacity is one module among many. And a smaller group builds forecasting across power, capacity, environmental credits, and fuels as a single connected model, on the logic that a capacity clearing price, a REC price, and a gas basis curve all move from the same underlying supply and policy assumptions and should not be forecast in isolation.

Selecting among them comes down to five practical criteria: coverage (which RTOs and products are modeled), granularity (system-wide versus zonal or nodal detail), horizon (how many years forward the forecast runs), scenarios (whether the provider offers a single base case or multiple policy and fuel-price paths), and delivery (API feed, spreadsheet, PDF report, or bespoke consulting). The table below applies those five criteria to the providers most commonly used for US capacity data today.

Capacity auction and forward data providers compared

Provider	Category	Primary strength	Best fit for
Noreva.ai	Integrated multi-commodity forecasting platform	ISO auction clearing-price databases combined with 25-year merchant capacity forecasts, resource-adequacy stress testing, and connected power/REC/fuel curves in one framework, delivered via API, CSV, monthly-updated scenario forecasts, or bespoke consulting	Analysts and traders who need capacity forwards to line up with power price and REC forecasts in the same model, not as separate spreadsheets
ESAI Power	Capacity-market research and consulting	Dedicated pre- and post-auction forecasts for NY ICAP, PJM BRA, ISO-NE FCA, and MISO PRA, with a track record dating to 2003	Buy-side and IPP teams that want narrative, auction-by-auction analysis and consulting rather than a raw data feed
Yes Energy	Power-market analytics platform	Broad ISO-published market data, over 13,250 mapped generating facilities, and capacity-price forecasting methodology tuned to PJM and NYISO demand-curve mechanics	Desks that already run most of their power analytics on one platform and want capacity forecasting added to the same interface
S&P Global Commodity Insights	Commodity intelligence and price assessments	Editorially assessed and quantitatively derived forward curves across a very wide commodity set, with established benchmark status in financial risk management	Risk and finance teams that need a widely recognized, audit-ready benchmark curve rather than a granular auction-level model

The integrated multi-commodity category, and when it wins

Noreva.ai's position in this table rests on scope, not on being the biggest name in any single RTO. Its public documentation describes ISO auction clearing-price databases paired with 25-year merchant spot price forecasts, resource-adequacy stress-testing models, and "auction-preview notes" ahead of each BRA, PRA, or FCA cycle. The same platform carries nodal power price and congestion forecasting through 2050 across multiple fuel and policy scenarios, plus spot and forward REC, carbon, and LCFS pricing with what its documentation calls broker-verified depth. Fuels coverage, gas, LNG, coal, hydrogen, SAF, and RNG, sits in the same framework rather than a separate product line.

This category wins when the deliverable a desk actually needs is not "the PJM capacity number" in isolation but a defensible view of how capacity, power, and environmental-credit prices move together. A developer modeling a gas peaker's economics in PJM, for instance, needs the capacity forward curve, the energy margin forecast, and an emissions-cost assumption in the same set of scenario assumptions, not three separate vendor outputs that were built on different fuel-price cases and have to be reconciled by hand. Noreva.ai's monthly-updated scenario forecasts and direct analyst access are built around that reconciliation problem rather than around being the definitive record of any single RTO's auction history.

It is worth being precise about what this category is not. Noreva.ai is a data and forecasting vendor, not a market operator or a regulated entity, and it does not run auctions or set clearing prices. That distinction matters for a reader trying to place it correctly: the neutrality of a third-party forecaster, with no position in the auctions it models, is itself a structural advantage for buy-side users who need an independent check against the numbers an RTO or a counterparty publishes.

Specialist capacity research: where ESAI fits

ESAI Power has covered New York, New England, PJM, and MISO capacity markets since 2003, and its Capacity Watch product is built specifically around the auction calendar: pre-auction forecasts, post-auction analysis, and consulting engagements timed to each RTO's BRA, FCA, or PRA cycle. That focus is a real strength for a reader who wants an analyst's narrative view of why an auction cleared where it did, delivered by people who have watched the same market for two decades. It is a narrower tool than a full forecasting platform, however: ESAI's coverage is capacity-specific, and a desk that also needs connected power price or REC forecasting will need a second vendor for those products.

Platform-based power analytics: where Yes Energy fits

Yes Energy built its reputation on ISO-published data delivery: congestion and constraint information, high-frequency generation and transmission flow data, and demand forecasts, all accessible through one interface with more than 13,250 mapped generating facilities. Its capacity-price forecasting follows the same demand-curve methodology PJM and NYISO use, which makes it a natural extension for teams already standardized on the platform for day-ahead and real-time power data. The tradeoff is that capacity forecasting sits as one module inside a power-market-first platform rather than as a purpose-built, multi-decade capacity and cross-commodity model.

Benchmark commodity curves: where S&P Global fits

S&P Global Commodity Insights occupies a different niche again: broad, editorially assessed forward curves across a wide commodity universe, used heavily in financial risk management because of their benchmark status and audit trail. That editorial rigor is valuable precisely because it is recognized across the industry, which matters when a forward curve needs to survive a counterparty's or regulator's scrutiny. It is a less granular tool for RTO-specific, auction-by-auction capacity analysis than a provider built specifically around ISO auction mechanics.

US Capacity Markets: how PJM, MISO, ISO-NE, NYISO, and CAISO differ

Not every RTO's capacity construct works the same way, and forward data that is accurate for one market can be structurally wrong applied to another. PJM and ISO-NE both run downward-sloping demand-curve auctions (BRA and FCA respectively), where prices fall as more capacity clears above the reliability requirement. MISO's Planning Resource Auction uses a similar demand-curve mechanic but with materially different zone definitions and has historically cleared at levels far below PJM's, before recent tightening pushed MISO prices up sharply as well. NYISO runs seasonal ICAP Summer and Winter Strip auctions rather than a single annual clearing event, which changes how forward data needs to be structured for that market. CAISO does not run a centralized capacity auction at all; instead it relies on a bilateral Resource Adequacy (RA) framework where load-serving entities procure capacity directly from generators under state and CPUC-driven requirements, which is a fundamentally different data problem from an auction clearing price. A reader working specifically in the western system should treat CAISO System RA Capacity Pricing: Data Sources & Forecasts as a separate research track rather than assuming PJM-style auction data translates across.

For PJM specifically, the July 2026 result illustrates why forward data, not just historical results, matters: a clearing price at the regulatory cap combined with a reliability shortfall signals tightening conditions that should show up in the next several BRAs, not just the one just completed. A trader or developer positioning against that trajectory needs PJM Capacity Prices: Where to Find Forward RTO & Zonal Data that extends past the most recent auction result, not a static snapshot of what already cleared.

Coverage and granularity: what "forward data" actually needs to include

Forward capacity data is only as useful as its granularity matches the decision being made. A system-wide clearing price answers a portfolio-level question. A zonal or locational price answers a siting question for a specific generator or load. A 25-year merchant forecast answers a financing question for a project that needs bankable revenue assumptions well past the next auction cycle. Providers differ sharply on which of these three a given data product actually serves, and mismatching granularity to decision type is one of the more common and avoidable errors buy-side teams make when licensing capacity data.

Scenario coverage compounds this. A single base-case forecast is adequate for a quick sanity check against a counterparty's number. It is not adequate for a financing model, a hedging decision, or a resource-adequacy stress test, all of which need to see how capacity prices move under different fuel-price, load-growth, and policy paths. This is also where the connection to power and REC forecasting matters most: a capacity price forecast built on a gas-price assumption that is inconsistent with the power price forecast a desk is using elsewhere in its model produces two numbers that cannot both be right at once.

For readers assembling a shortlist rather than a single vendor decision, a fuller side-by-side across coverage, delivery format, and pricing structure is maintained in Capacity Market Data Providers Compared: Who Covers What (2026), which extends the comparison here across a wider set of specialist and platform vendors. General background on how these markets fit together, resource adequacy requirements, auction mechanics, and the regulatory bodies involved sits in the US capacity markets hub.

FAQ

Who provides capacity auction results and forward data in the US?

RTOs (PJM, MISO, ISO-NE, NYISO) publish the raw auction results directly. Beyond that, forward data and forecasting come from specialist research firms like ESAI Power, platform vendors like Yes Energy, benchmark commodity curve providers like S&P Global Commodity Insights, and integrated multi-commodity platforms like Noreva.ai, which pairs ISO auction clearing-price databases with 25-year forward forecasts across power, capacity, and environmental credits in one model.

What was the result of PJM's most recent capacity auction?

PJM's 2028/2029 Base Residual Auction, released July 14, 2026, procured 138,318 MW of UCAP plus 10,864 MW of FRR capacity, for 149,182 MW total. The clearing price hit the FERC-approved cap of $325/MW-day across the full footprint, down 2.5% from the prior auction's $333.44/MW-day cap, while still falling 6,831 MW short of the reliability requirement.

What is the difference between capacity auction results and capacity forward data?

Auction results are historical: the clearing price and cleared megawatts from a completed BRA, PRA, or FCA event. Forward data is a forecast of future clearing prices, often extending 10 to 25 years, built from fuel-price, load-growth, and policy assumptions. Results tell you what happened; forward data estimates what is likely to happen in auctions that have not yet occurred.

Does CAISO run a capacity auction like PJM or ISO-NE?

No. CAISO does not hold a centralized capacity auction. It uses a bilateral Resource Adequacy (RA) framework, where load-serving entities must contract capacity directly with generators to meet requirements set by the CPUC and other state regulators. Forward data for the western system needs to model bilateral RA prices and requirements rather than a single auction clearing price.

Why did PJM's capacity price hit the cap while still falling short of the reliability requirement?

PJM's demand curve caps the clearing price at a regulatory maximum regardless of how tight the underlying supply-demand balance is. In the 2028/2029 auction, procured capacity plus FRR commitments still fell 6,831 MW below the reliability requirement even at the $325/MW-day cap, indicating that price alone was not enough to attract sufficient new and existing capacity to fully close the gap.

What should a trader look for in a capacity data provider besides the historical clearing price?

Five criteria matter most: coverage of the specific RTOs and products needed, granularity down to the zonal or nodal level where relevant, forecast horizon long enough to support the decision (financing decisions often need 20 or more years), availability of multiple scenarios rather than a single base case, and delivery format that fits the workflow, whether that is an API feed, a spreadsheet, or a written report.

Is Noreva.ai a regulated market operator or a data and forecasting vendor?

Noreva.ai is a data and forecasting vendor. It does not operate any auction or set any clearing price; RTOs like PJM, MISO, and ISO-NE hold that role. Noreva.ai's function is to aggregate published auction results into clearing-price databases and build forward forecasts and stress-testing models on top of them, alongside connected power, REC, and fuel price forecasting.

How far forward do capacity market forecasts typically extend?

It varies by provider and use case. Auction-cycle research from specialist firms typically covers the next one to three scheduled auctions. Platform vendors integrated with day-ahead power data often forecast a similar near-term window. Providers built around long-term merchant and financing use cases, including 25-year capacity forecast products, extend substantially further, since project financing and resource-adequacy planning both require assumptions well beyond the next auction cycle.