NRG ENERGY, INC. (NRG) stock intrinsic value fair value estimate & owner earnings analysis

What the business is. NRG Energy owns power plants and sells electricity and natural gas directly to homes and businesses. Its core is Texas, where it is one of the largest retail electricity providers and owns roughly 12 gigawatts of gas, coal and other plants. It also sells power and gas to businesses in the Northeast and Midwest through Direct Energy (bought 2021) and runs Vivint Smart Home (bought 2023), a home-security subscription business with about two million subscribers. On January 30, 2026 it closed its largest deal: 18 gas and dual-fuel plants of about 13 gigawatts in nine states plus the CPower demand-response business, bought from LS Power for $6.4 billion in cash, 24.25 million new shares and about $3.2 billion of assumed debt, roughly $10.6 billion in all. That about doubles its fleet. NRG is not a regulated utility: nobody sets its return, and results swing with power and gas prices, weather and how its hedges perform. How we measure earnings, and a correction. We start from operating cash flow and remove stock-based pay, the spending needed to keep the existing plants running (NRG reports maintenance and environmental capital separately: $367 million in 2025) and the swings in receivables, payables, inventory and hedge collateral. Checking that figure against net income plus non-cash charges showed it about $780 million too high in 2025. The reason: two costs that recur every year were being removed along with the working-capital swings. Vivint and the retail business pay commissions and installation costs to sign customers; NRG spreads that cost over about five years ($510 million of amortization in 2025, up from $168 million in 2023) while the cash leaves through prepayments. Customer bad debts ($272 million in 2025) work the same way through receivables. Both are real operating costs, so we now charge them. Our previous estimates did not and are not comparable. On this basis owner earnings per share were $1.32 (2021, Winter Storm Uri), -$3.80 (2022, gas-price spike and hedge settlements), $5.80 (2023), $10.18 (2024) and $11.05 (2025). The large acquired-intangible and plant depreciation charges are added back, as sunk acquisition cost; the wearing-out of those assets is carried instead in how long we let earnings last. The earnings level. The business changed in 2021 (Direct Energy), 2023 (Vivint) and again in 2026, so older years describe a different company. 2021 and 2022 were distorted by Uri and the 2022 gas spike; we use those episodes in the bear case rather than in the normal level. The normal level is set from 2023-2025, about $2.2 billion a year of owner earnings for the pre-deal business, about $2.76 billion before interest (NRG paid $572 million of cash interest in 2025). Owner earnings grew about 2% in 2025 after per-share gains from buybacks; we assume no growth from here for the existing business, because that level already includes strong Texas and PJM capacity prices (NRG cleared PJM's 2026/27 and 2027/28 auctions at the price cap, about $121 million and $131 million) that are not guaranteed to last. The LS Power plants. These are bought, paid for and operating, so their earnings are in the bag, but NRG has not yet published their purchase-price accounting and we have less than one year of results. In the first half of 2026 operating income rose to $1.30 billion from $1.13 billion a year earlier, and interest expense rose to $595 million from $311 million. We credit the acquired plants with $700 million a year of cash earnings before interest in the base case, a modest return on $10.6 billion that reflects only what the half-year shows, $1.0 billion in the bull case and $400 million in the bear case. What NRG owes. At June 30, 2026 borrowings were about $23.4 billion, plus $650 million of preferred stock ranking ahead of the common, against just $162 million of cash: net claims ahead of shareholders of about $23.9 billion, roughly $113 per share on about 212 million diluted shares. Borrowings rose from $16.6 billion at year-end 2025 as the deal closed, and $1.5 billion is now due within a year. This debt is the main reason the cases are so far apart. Base case. Pre-interest cash earnings of about $3.46 billion a year ($2.76 billion existing business plus $0.7 billion from the acquired plants), held flat for 15 years with one severe Texas weather or gas-price event costing about $2 billion in year five, about what Uri and 2022 cost, then fading to half by year 30 as the gas fleet ages and is retired, ending at year 40. The business is worth about $45.1 billion; after $23.9 billion owed, about $21.2 billion, or $100 per share. Bear case. The 2021-2022 record repeated: a Uri-type freeze in year three costing $3.2 billion and a second gas-price shock in year eight costing $2 billion, the retail book earning near its 2023 level ($2.2 billion before interest) and the acquired plants contributing only $400 million. Earnings fade from year 10 to a third by year 25 and stop at year 30. The business is worth about $26.0 billion against $23.9 billion owed, leaving about $2.1 billion, or about $10 per share. In this case debt service still holds in most years, but the shock years would need revolver borrowing and buybacks would stop. Bull case. Retail at about $2.9 billion before interest, the acquired plants delivering $1.0 billion as capacity revenue already contracted continues, one smaller $1 billion shock in year seven, held flat for 20 years and fading to 60% by year 35. The business is worth about $55.9 billion, about $32.1 billion after debt, or $151 per share. We give no credit to new data-center supply contracts or new plants not yet built, because none is signed in the filings. What we left out. Market growth from data-center demand, the new-generation programs discussed by PJM and Texas, and further buybacks beyond today's share count: none is secured. Buybacks only redistribute value already counted. The margin-of-safety price. How sure are we of the earnings level? Moderately: three recent years agree within about 20% for the old business, but they were good years for power prices, and the acquired plants have under a year of history. The worst plausible outcome is a repeat of Uri or 2022 in the first year or two after the deal, while $1.5 billion of debt comes due and borrowings sit near $23 billion; the business would survive, as it did in 2021 and 2022, but shareholders could be diluted or see payouts suspended. If we are wrong on the acquired plants (worth $400 million a year rather than $700 million) and on capacity prices, the shares are worth roughly $45-60. NRG pays about $2 a share in dividends while waiting, and the base-case owner earnings after interest are about $10.50 a share. The margin-of-safety price is $50: from that price the dividend plus normal earnings retained (or used for buybacks) imply close to 20% a year in the base case, and a partial bear outcome implies shares near that price rather than wiped out. The margin-of-safety price applies only once the first full-year report on the LS Power plants is available, because that is the number we are least sure of.