PJM Reliability Backstop: Hybrid Procurement Reorders the Merchant Power Opportunity

1. Executive Overview

PJM's reliability backstop proposal is a substantive acknowledgment that the current resource adequacy construct is no longer synchronizing large-load growth with financeable supply entry. The 2027/2028 BRA cleared 134,479 MW of UCAP, or 145,777 MW including FRR supply, and still finished 6,623 MW short of the reliability requirement at the FERC-approved cap of $333.44/MW-day. PJM's Board then directed staff to accelerate a backstop procurement that current tariff language would not otherwise trigger until 3 consecutive short auctions, and explicitly framed the mechanism as a transitional response while broader market reforms are evaluated.

The core investment implication is that PJM is moving, at least temporarily, from a pure scarcity-price model toward a hybrid framework in which new large-load demand is paired more directly with incremental supply through bilateral contracting and, if necessary, centrally administered procurement. The broad direction is clear even though the design is not yet final. As of April 11, 2026, PJM had scheduled an April 16-17 workshop, May 27 stakeholder voting, and a June FERC filing target, which means the market should treat the proposal as directionally important but still commercially unfinished. White House and governor support strengthens the political backdrop, but it does not make the redesign self-executing. Capstone described the January statement as policy signaling rather than binding reform, and any backstop auction still requires tariff revisions and FERC approval.

For equities, the most important conclusion is that the proposal favors companies that can monetize both existing scarcity and new contracted growth. Pure exposure to open merchant pricing remains valuable, but the bigger structural opportunity sits with companies that combine fleet optionality, brownfield development, commercial origination, transmission or fuel access, and credibility with hyperscale customers or LSEs.

2. Core Evidence: Why PJM Is Moving and What It Is Proposing

PJM's supply problem is not a narrow summer-peak issue. The system is absorbing a structural change in load shape. PJM's 2026 Load Forecast projects summer peak growth of 3.6% annually over the next 10 years, reaching 222,106 MW in 2036, winter peak growth of 4.0% annually over the next 10 years, reaching 204,650 MW in 2035/2036, and net energy growth of 5.3% annually over the next 10 years. PJM's shortfall paper states that data centers already represent more than 7% of energy use in the region, that more than 90% of their load is continuous, and that PJM now expects summer peak demand to climb by about 82 GW to 239 GW over 15 years.

The headline 15 GW figure is important but should not be treated as equivalent to the current reserve shortfall. PJM's proposal sets the initial procurement target using the 2026 Load Forecast for estimated 2029 summer large-load forecasts minus 2026 summer large-load forecasts, excludes FRR entities, and then allows EDCs to raise or lower their final targets based on bilateral contracting, BYONG arrangements, willingness to curtail, load PJM did not model, or load that should not be represented in the backstop. In other words, 15 GW is an initial growth-procurement target, not a fixed final number and not a direct synonym for the 6,623 MW 2027/2028 BRA shortfall.

The architecture is explicitly 2-phased. Phase 1 is a facilitated bilateral matchmaking process in which PJM and Charles River Associates act as confidential intermediaries between buyers and sellers. PJM is not a party to those contracts, will not provide pro formas, and is proposing a bilateral window from September 2026 through March 2027. Residual MW not solved bilaterally would move into a PJM-administered central procurement, currently planned to commence in March 2027, with PJM Settlements as counterparty on behalf of EDCs. This bilateral-first design strongly favors developers and incumbents that already have sites, customer channels, commercial credibility, and the ability to structure bespoke risk-sharing arrangements with hyperscalers or LSEs.

• The central procurement product is capacity-only UCAP rather than an all-in energy, ancillary services, and capacity contract.
• Suppliers can offer 2-15 year terms, and the central procurement is proposed as pay-as-bid rather than uniform-price.
• Each awarded resource would settle through a contract-for-differences against its weighted-average resource clearing price in future RPM auctions.
• Awarded resources would have a must-offer obligation at $0 into each future auction for the committed UCAP.
• PJM is proposing a daily shortfall charge equal to 20% of the commitment price for deficits against committed RBP UCAP, with no replacement MW allowed for committed MW while Connect and Manage remains in effect.

The companion large-load framework matters as much as the procurement itself. PJM's Board and February stakeholder work contemplate a connect-and-manage pathway under which large loads that do not bring enough new generation can interconnect but would be curtailed prior to pre-emergency Demand Response. That is economically important because it means fast-track data-center load is not being offered firm service for free; it is being offered earlier access in exchange for explicit outage and backup-generation obligations.

The Expedited Interconnection Track is also more selective than broad market narratives often imply. PJM's February materials limited the track to shovel-ready new resources and uprates of at least 250 MW UCAP, opened it to only up to 10 projects per year for 2 years, and required support from a state Primary Siting Authority. The Board letter additionally made paired-load projects responsible for 100% of identified network upgrades. That should continue to favor brownfield incumbents and advanced developers, but it also limits how much supply can realistically convert on a short timeline.

Eligibility rules are narrower than the market's initial 'bring back old plants' framing implied. Only new resources can enter the central procurement, defined as resources that bring new ICAP and CIRs, have not already received future RPM commitments, and can reach COD by June 1, 2031. Eligible supply can include new builds, uprates, repowering or reactivation of currently deactivated generators, and certain DR/DER offers with identified sites and contracts. Delayed retirements, re-licensing, fuel switching, CIR-only uprates, and surplus resources are excluded. There is also no special interconnection track for RBP. Projects must proceed through the standard queue cycle, must include network-upgrade costs in their bid, and bear the risk if actual upgrade costs exceed what they bid.

PJM's selection logic further sharpens the advantage of incumbents and advanced developers. Stage 1 gating requires a critical path schedule, site control, financing plan, permitting plan, evidence of major equipment acquisition, prior similar project experience, fuel delivery arrangements where relevant, and electrical location in or firm transmission into PJM. Stage 2 then selects in least-cost order based on the lowest average cost of capacity per UCAP over the term. PJM is not proposing a broader subjective multi-factor screen at this stage, which should bias awards toward uprates, reactivations, brownfield combined cycles, certain nuclear uprates or restarts, and qualified DR/DER rather than higher-cost greenfield merchant builds.

3. Market Structure and Sector Implications

For PJM itself, the backstop is a meaningful institutional shift. The Board letter states that PJM does not view it as desirable in the long term for PJM to be the procuring authority for long-term commitments, yet PJM is nevertheless proposing a one-time centrally administered procurement because the current 3-year-forward, 1-year commitment structure is not reliably producing new financeable supply on the timeline that load growth now requires. PJM is therefore becoming, temporarily, more than a market administrator. It is becoming a reliability planner, contract facilitator, and, if necessary, a long-term capacity counterparty.

Cost allocation and jurisdiction sit near the center of the debate. The Board defined large-load additions at 50 MW or more at a single point of interconnection, while also stating that PJM is not creating a dedicated load queue or otherwise restricting new load interconnection at this stage. Instead, PJM is trying to preserve affordability by pushing incremental reliability costs toward the LSEs and EDCs associated with incremental load growth, with tradability of obligations through Capacity Exchange. At the same time, FERC has already ruled that PJM's tariff is unjust and unreasonable for co-located load because the rules lack clarity and consistency, and FERC required PJM to create clearer transmission-service options and revise its BTMG framework.

For generation providers broadly, the backstop is positive, but not in a uniform way. The biggest beneficiaries are not simply owners of existing capacity. They are owners that combine existing merchant leverage with some combination of brownfield sites, incremental MW opportunities, customer origination, fuel or transmission access, interconnection readiness, equipment procurement, and balance sheet credibility. The framework also expands the addressable opportunity set beyond thermal new build to include qualifying nuclear uprates or restarts, certain reactivations, and DR/DER with contractually identified sites.

A separate but critical overlay is the co-location and direct-supply regime. FERC's December order validates the need for tailored transmission-service options, but the March 2026 comment record shows the pathway is not yet commercially settled. PJM's proposal would create interim non-firm transmission, firm contract demand transmission, and non-firm contract demand transmission services, yet DCC argued colocated loads still would not have primary rights to colocated generation under interim NITS and would remain exposed to curtailment, while Vistra and Constellation criticized PJM's requested June 1, 2029 effective date as too slow. That means direct-supply normalization is directionally real but near-term economics and timelines remain contested.

4. Company-Level Positioning

Constellation's upside from the PJM shift is substantial, but the mechanism is more indirect than direct central-RBP participation. The strategic edge is its ability to sell clean, firm, 24/7 output into premium long-duration contracts, not just to harvest spot merchant scarcity. Its disclosed growth vector includes about 2,300 MW of long-term structures at natural gas plants that value capacity and reliability, an 835 MW Crane restart backed by a 20-year Microsoft PPA and a DOE loan guarantee of up to $1.0 billion, a 1,121 MW Clinton contract including an uprate, and more than 1,100 MW of data-center development agreements. Constellation has also stated that it still has 147 million MWh of annual nuclear generation available to contract at premium pricing. The limitation is that most of the existing nuclear fleet is not central-RBP eligible because the proposal is limited to new resources, so the direct RBP upside is likely selective and project-based rather than broad-based. A fresh caveat from the February 2026 IMM answer to CEG's rehearing request is that existing capacity cannot simply be sold twice. If FERC or PJM ultimately require stricter delisting or other economic separation when existing generation serves co-located load, the upside from direct-supply structures using legacy nuclear output could be less accretive than the most bullish framing implies.

Vistra arguably has the cleanest combination of existing PJM scarcity leverage and already-demonstrated bilateral monetization of that scarcity. The company's East fleet totaled 22,254 MW within a 43,641 MW total capacity base at year-end 2025, and management has emphasized that its 22 GW modern combined-cycle gas fleet can run at higher utilization as load grows. In January 2026, Vistra also signed 20-year PPAs with Meta covering more than 2,600 MW across PJM nuclear facilities, including 2,176 MW from operating plants and 433 MW from uprates. The company then added further optionality through the announced acquisition of roughly 5,500 MW of modern gas generation from Cogentrix, following the earlier 2,600 MW Lotus portfolio purchase. Vistra therefore has exposure to 3 distinct earnings vectors: higher scarcity rents on open PJM fleet, bilateral contracting on existing nuclear, and incremental MW from uprates or brownfield gas that could fit either bilateral deals or central procurement.

Talen Energy has the highest direct exposure to the data-center power thesis in PJM, but the direct exposure is more to bilateral and direct-supply normalization than to the central procurement itself. Talen owns about 13.1 GW of power infrastructure, including 2.2 GW of nuclear, and has already established the market's most visible hyperscaler-linked structure in PJM. Its expanded Amazon agreement covers 1.9 GW and about $18 billion of revenue under a 17-year contract, with full quantity of 1,920 MW through 2042 and a ramp to full volume no later than 2032. Talen also cleared 8,745 MW in the 2027/2028 BRA at $333.44/MW-day, retains direct exposure to current PJM scarcity, and has expanded gas optionality through an announced January 2026 acquisition adding about 2.6 GW of natural gas capacity in western PJM. The principal limitation is that existing contracted or legacy-supported assets are not the cleanest fit for direct central-RBP eligibility, and the company remains more concentrated than peers around a smaller set of assets and customer relationships. The March 2026 co-location comment record reinforces that point. If colocated load lacks dedicated rights to paired generation, remains exposed to curtailment, or must give up more PJM market economics than bulls assume, TLN's direct-supply upside remains large but less frictionless than headline enthusiasm suggests.

NRG is the company whose strategic position may have changed the most because of PJM's shift. The January 2026 completion of the LS Power acquisition added 18 natural-gas-fired facilities totaling about 13 GW and CPower's commercial and industrial VPP platform, taking NRG's generation fleet to about 25 GW. NRG's competitive generation mix moved from 16% PJM pre-acquisition to 39% PJM post-acquisition, and management has explicitly tied the deal to a BYOP strategy for large loads. In February 2026 earnings materials, NRG highlighted at least 1 GW of contracted data-center opportunities supporting BYOP solutions, more than $2.5 billion of annual recurring adjusted EBITDA opportunity associated with 6 GW of data-center power agreements, and more than 6.4 GW of incremental natural-gas supply opportunity, including 5.4 GW of ready-to-build CCGTs with equipment and EPC secured plus at least 1 GW of PJM uprates from the LS Power portfolio. NRG also retains immediate PJM scarcity exposure through a reported capacity position of 5,770 MW for planning year 2025/2026, 6,005 MW for 2026/2027, and 6,015 MW for 2027/2028, with forecast PJM capacity revenue of $644 million in 2026 and $729 million in 2027.

5. Relative Positioning

Relative exposure is not identical across the four names. Talen Energy has the highest direct leverage to the large-load direct-supply and co-location theme because its Amazon-Susquehanna structure sits at the center of the policy debate. Vistra has the strongest combination of large existing PJM merchant exposure and already-proven bilateral monetization through long-duration nuclear PPAs, plus incremental gas and uprate optionality. NRG appears to have the strongest brownfield gas, BYOP, and DR/DER execution platform for PJM's proposed bilateral-first framework. Constellation has the strongest clean-firm contracting franchise and substantial upside from broad PJM scarcity, but less direct central-RBP torque because much of its advantage sits in existing nuclear output and select incremental projects rather than a broad catalog of net-new backstop-eligible MW.

• Talen is the highest-beta expression of direct large-load contracting and co-location normalization.
• Vistra offers the cleanest blend of merchant scarcity leverage and demonstrated bilateral monetization, with incremental gas and uprate optionality on top.
• NRG appears most aligned with the current draft's bilateral-first logic because it now combines customer access, brownfield gas, ready-to-build projects, uprates, and DR/DER through CPower.
• Constellation is the strongest clean-firm contracting franchise, but it is less of a pure central-backstop award vehicle than the headline narrative might imply.
• This is not a ranking of quality. It is a map of which part of the structural shift each equity expresses most directly.

6. Risks and Disconfirming Evidence

The most important reason not to overstate the thesis is that the design is still proposed, not final. The workshop, stakeholder vote, and expected FERC filing could materially alter the bilateral-first architecture, the central-procurement price cap, the term structure, the performance penalties, or the treatment of different technologies. If the final design is materially less financeable than the current draft, the headline strategic shift would remain real, but the near-term earnings conversion for public equities would be smaller.

There is also a meaningful disconfirming scenario in which forecast tightening, EDC target adjustments, BYONG arrangements, bilateral contracts, and adjacent reforms collectively reduce the central-procurement need far below the market's current 15 GW mental model. In that outcome, bilateral contracting could still validate the theme, but the direct award opportunity for new brownfield and restart projects would be smaller than many investors now expect.

The direct-supply/co-location path also deserves more skepticism than the current market narrative often applies. PJM's February 23 proposal prompted March filings from DCC, Vistra, Constellation, and the IMM arguing that the design remains commercially restrictive, overly reliant on curtailment, and too slow to solve immediate load growth. Under the interim NITS construct criticized by DCC, colocated load would not have primary rights to colocated generation and could still face curtailment, while PJM reportedly sought a June 1, 2029 effective date because of system changes. That pushes against the view that co-location is already a near-term de-risked earnings bridge.

• If PJM or FERC ultimately narrows direct-supply and co-location economics, the most thematic bilateral structures could become more transmission-cost intensive or less attractive than investors currently assume.
• EPSA/P3 and PJM commentary also point to incremental auctions, tighter load vetting, and more than 12 GW of expected new generation as reasons the eventual residual central procurement may be smaller than the headline 15 GW starting point suggests.
• If the backstop becomes very large and successful, future RPM prices could face pressure because awarded RBP MW must offer at $0 into subsequent BRAs, which would reduce the scarcity premium on existing open fleets.
• If the backstop under-delivers, existing fleets may benefit from higher scarcity rents, but the probability of stronger political and regulatory intervention would also rise.
• Company-specific execution matters. NRG must convert a large brownfield and DR/DER platform into actual projects and contracts. Talen remains more concentrated around fewer assets and customers. Constellation's direct central-procurement eligibility is narrower than its broad strategic relevance. Vistra still faces the trade-off between monetizing scarcity and inviting a longer-term normalization of capacity pricing if the backstop works too well.

7. Catalysts and Watchlist

The broader investment conclusion is that PJM has crossed an important threshold. The region is no longer debating whether large-load growth is material enough to reshape the supply stack. PJM is now redesigning reliability procurement around that premise. For merchant generation providers, that is supportive in 2 distinct ways. Existing open fleets gain from continued scarcity in energy and capacity, while incremental projects gain from a new pathway to contracted or semi-contracted cash flows. The best-positioned public equities are the ones that can arbitrage between those 2 earnings models rather than rely exclusively on either one.