Earnings Recap — Q2 FY2026
CY Q3 2026 · Reported August 4, 2026 · Beat 4 of last 7 quarters
The Buildout is a website built on agentic AI tools, currently in beta, and could have factual errors.
NRG's BYOP model directly addresses the AI infrastructure buildout by pairing new generation with hyperscaler demand, reducing grid strain and transmission needs. The 1.2 GW project with a leading cloud/AI customer, backed by GE Vernova turbines and Kiewit EPC, demonstrates a scalable template for serving data center growth. The company's secured 5.4 GW of turbine capacity and 2 GW of PJM upgrades position it to capture further AI-driven power demand.
NRG delivered solid Q2 results with adjusted EBITDA of $1.2B, up 34% YoY, driven by the acquired LS Power portfolio and higher PJM capacity values. Texas EBITDA declined $131M due to lower load and power prices (Houston around-the-clock averaged $33/MWh, 8% lower YoY and well below the $52 planning assumption). East EBITDA increased $370M on the LS Power contribution, partially offset by higher retail supply costs and a $70M RGGI cost in Virginia. The company announced a 1.2 GW BYOP project with a leading hyperscaler, with a 15-year term, $3.2B investment, and expected $500M annual EBITDA at full operation. Management reaffirmed 2026 guidance and the $1B annual share repurchase commitment.
Management reaffirmed 2026 guidance, expecting to land within the range but below the midpoint due to softer Texas prices and Winter Storm Fern costs. They announced a landmark 1.2 GW Bring Your Own Power (BYOP) project with a leading global cloud/AI hyperscaler, with principal commercial terms aligned and a customer capital commitment. The project targets late 2029 COD, with $500M annual adjusted EBITDA and $375M annual FCF before growth at full operation, and a pretax unlevered IRR within the 12%-15% target range. The $3.2B investment will be funded via operating cash flow and balance sheet capacity, with 2026 spend of $721M largely equipment-related and redeployable. Management reaffirmed the $1B annual share repurchase commitment and 3x net leverage target, while noting the project extends the deleveraging timeline to 2029. They see a large pipeline of 5.4 GW of secured turbine/EPC capacity and 2 GW of PJM upgrade opportunities, with the BYOP model expected to be repeatable and the industry standard.
“Put simply, we're paid for the megawatts we build and make available, not for how much the data center runs.”
on BYOP commercial structure
“We will not trade discipline for scale. Each project must stand on its own, meet our risk-adjusted return thresholds and be supported by the commercial and credit protections appropriate to the capital we deploy.”
on Capital discipline
“We have the commercial structure, the equipment and the capabilities to deliver it at scale.”
on BYOP model
How are you thinking about the timeline for expansion to 2.4 GW and whether the build multiple is the new norm?
Returns are expected to be consistent across projects, with the 12-15% IRR target. Timing depends on turbine deliveries and site readiness; the GEV/Kiewit structure allows for serial 1.2 GW blocks each year after 2029. The 1.2 GW site could expand, but other sites could also support larger projects.
What drove the confidence to announce principal terms, and what approvals are outstanding?
We announced because we reached a material point with commercial alignment. Remaining items include final negotiations, land matters, and internal approvals. We will update when we have another material piece of information, not on a set timeline.
Is the PPA equivalent now north of $90 or $100 per MWh given higher CapEx?
The dollar per MWh metric isn't the focus due to the structure. If you assume expected utilization, it's in the $85-90+ range, but the structure provides certainty and flexibility for both parties.