I'll research NRG thoroughly before writing the preview. Let me start by gathering recent documents and data.Let me read the most recent quarter's transcript and release, plus check recent news.Let me gather the guidance details, the year-ago Q2 comparison, and recent stock performance.Let me check the PJM capacity auction 8-K and get the latest news/analyst sentiment.Let me get some external context on analyst expectations and summer 2026 power market conditions.I have what I need. Here's the earnings preview.

NRG Energy (NRG) — Q2 2026 Earnings Preview

Reports before market open, Tuesday August 4, 2026 | Conference call 9:00 a.m. ET


The one-line setup

This is the first "clean" quarter to judge the new NRG — the first full three-month period that consolidates the ~13 GW LS Power gas fleet and the CPower C&I demand-response platform (deal closed January 30, 2026), and the second quarter reported under new CEO Robert Gaudette. The debate is no longer whether NRG is a bigger, more gas-heavy, PJM-weighted company — it's whether management can (1) show the acquired earnings power flowing through, (2) reaffirm or nudge up the aggressive 2026 guide, and, above all, (3) put a hyperscaler/large-load contract on the board to validate the growth story.


Stock context going in


Numbers & the year-over-year distortion

Year-ago bar (Q2 2025): Adjusted EBITDA $909M, Adjusted EPS $1.73, Adjusted Net Income $339M, FCFbG $914M. (Note GAAP was a net loss of $(104)M / $(0.62) on non-cash mark-to-market hedge losses — expect similar GAAP noise this quarter; focus on adjusted figures.)

Segment Adj. EBITDA a year ago: Texas $512M / East $99M / West $43M / Vivint $255M.

Why Q2 2026 should look much bigger: Q2 2025 predates LS Power entirely. Q2 2026 carries a full quarter of the acquired East-heavy gas fleet, so a large reported YoY EBITDA increase is mechanical, not a "beat." The more useful lens: - Full-year 2026 guidance (reaffirmed at Q1): Adj. EBITDA $5,325–$5,825M (mid ~$5,575M), Adj. EPS $7.90–$9.90, Adj. Net Income $1,685–$2,115M, FCFbG $2,800–$3,300M — up ~45% at the EBITDA midpoint vs. the 2025 guide (~$3.7–4.0B), reflecting LS Power + organic growth. - Run-rate math: Q1 2026 delivered $1,080M Adj. EBITDA (down $46M YoY on mild Texas weather + Winter Storm Fern supply costs booked before the deal closed). That leaves ~$4.5B needed across Q2–Q4 to hit the midpoint. Watch the Q2 figure closely — it sets the market's confidence in the back-half seasonality and the LS run-rate.

What I'd watch on the P&L: the Texas segment (retail margin + June heat), the East segment run-rate now that the LS assets are fully owned and hedged, and continued Vivint Smart Home growth (customer count was +9% YoY in Q1, ahead of the 5–6% plan).


The five things that will actually move the stock

1. A large-load / data-center contract. This is the catalyst. Management has repeatedly framed a front-of-the-meter Texas gen-for-data-center deal (previously ~$90–95/MWh economics on a ~5 GW opportunity) and said flatly that to hit a 2029 COD they "have to get something done in 2026." Any signed MW — in ERCOT or via a PJM bilateral — is the single biggest upside trigger. Silence/further delay is the biggest disappointment risk. Note NRG already has 295 MW of long-term data-center retail agreements signed (announced Q2 2025) with a path to ~1 GW.

2. The PJM capacity print is now a known tailwind. On July 14, NRG disclosed it cleared 6,839 MW at an average $325/MW-day in the 2028/2029 PJM base auction — near the price cap and a rich forward capacity revenue stream (~$800M annualized gross) for the LS-heavy East fleet. Look for management to quantify the earnings contribution and connect it to the 2028+ trajectory.

3. Guidance — reaffirm vs. raise. They reaffirmed at Q1 despite a soft start and emphasized they're "even more confident on a free-cash-flow basis" (working-capital unwinds). With summer still ahead and PJM capacity now locked, watch for any language tightening toward the upper half — or a reiteration of the ≥14% five-year Adj. EPS / FCF-per-share growth CAGR (explicitly before any large-load or new-development upside).

4. Capital allocation / buyback pace. 2026 plan: $1.0B buybacks + ~$407M dividends (dividend $0.475/qtr, $1.90 annualized). Through April 30, NRG had already repurchased $817M (including 1.83M shares bought back directly from LS Power) at prices management called "well below" plan — implying per-share upside and a real possibility they exceed $1.0B if extra FCF materializes. Also track deleveraging toward the 3.0x net-leverage target (April refinancing retired the $1.5B Lightning notes, moved ~$1B secured→unsecured, and paves the way to remove acquisition ring-fencing).

5. New CEO's strategic fingerprint. Gaudette has signaled a tilt toward contracted, long-duration cash flows — new-build gas, data-center deals, and notably a willingness to partner with regulated utilities "from the Atlantic to the Pacific" who lack the capital/equipment/development capability NRG has. Expect updates on the GEV/Kiewit turbine-and-construction partnership and the up-to-2 GW of PJM uprate/conversion opportunities.


Growth/execution checkpoints to tick off


Risks & watch-items


Bottom line

Expect a large reported YoY EBITDA/EPS increase driven by the first full quarter of LS Power, a reaffirmed (possibly firmer) 2026 guide, continued aggressive buybacks, and a now-quantifiable PJM capacity tailwind. But with the stock down ~17% YTD and off ~25% from its highs, the print itself is likely secondary to narrative catalysts — specifically a first signed large-load/data-center generation deal and any raise/tightening of guidance. Bulls want proof the "platform for where the market is going" is converting into contracted MW; bears will point to soft ERCOT curves, delayed deals, and post-acquisition leverage. This is a quarter where the commentary outweighs the numbers.

Note: I was unable to pull sell-side consensus estimates for Q2 2026 from the available tools, so the comparisons above are framed against the year-ago quarter and management's reaffirmed full-year guidance rather than Street numbers.