Sunrun Inc. (RUN) stock intrinsic value fair value estimate & owner earnings analysis

Business. Sunrun installs rooftop solar and home batteries and, for most customers, keeps ownership and sells the power under 20 or 25 year leases and power purchase agreements. In 2025 revenue was $2.96 billion: $1.82 billion from those customer agreements and incentives and $1.14 billion from selling systems and products. The company owns $16.8 billion of energy systems, funded mostly by about $14.0 billion of non-recourse debt (repayable only from the assets and contracts it is secured on), plus outside tax-equity investors who receive part of each project's cash. Recourse obligations at the parent are smaller: about $474 million of convertible notes and $238 million on a credit line. What the numbers say. Cash from operations has been negative every year for a decade, including $(421) million in 2025 and $(766) million in 2024, because the company pays for sales, marketing and administration up front (2025: $709 million and $278 million) while customer cash arrives over 20-plus years. After removing the swing in working capital (mostly a recurring build in prepaid and other current assets, $614 million in 2025) and the cost of stock compensation ($108 million), owner earnings were about $(2.49) per share in 2021, $(2.21) in 2022, $(3.40) in 2023, $(2.52) in 2024 and a small positive $0.42 in 2025. We measure from 2021, after the October 2020 Vivint Solar merger; the middle of those years is about $(2.5) per share. We treat the roughly $2.5 billion a year spent on new energy systems as growth spending rather than maintenance, because the existing fleet needs almost no capital to keep earning; the earnings figure is therefore not reduced for it. Reported profit of $450 million in 2025 to common shareholders is not a guide: it is driven by accounting allocation of losses to outside investors and a tax benefit, and cash flow shows no such profit. A $3.1 billion goodwill write-off in 2024 and $1.2 billion in 2023 were non-cash and are correctly ignored, but they show past acquisitions did not earn what was paid. How we value it. New installations are paid for with newly raised outside debt and investor money, not with cash the business generates, so we do not credit growth from them. The 2025 improvement in owner earnings is also not extrapolated; it follows three years of negative results and one year is not a trend. We value only what is contracted: the filings show $37.3 billion of contracted revenue not yet recognised, with about 5% (roughly $1.87 billion) recognised each year and little change expected for ten years because the fleet averages under six years old, then declining as 20 to 25 year contracts expire. Against that we subtract fleet operating costs (about $0.55 billion of the $1.28 billion cost of customer agreements is cash cost), a slimmed administrative cost, distributions to tax-equity investors (about $0.25 billion a year), and then interest and repayment of the roughly $14.9 billion of debt. On those figures the discounted value of the contracted fleet cash flows, roughly $9 to 10 billion at a 6% rate, falls well short of the debt raised against it. Debt on individual solar funds is non-recourse, so shareholders are not personally liable for the shortfall, but it leaves no evidenced value for the shares. Bear case (about -$3 per share). The federal solar credit for leased residential systems ends for systems placed in service after 2027 unless construction began by July 4, 2026, and new restrictions on foreign-sourced components apply. If new installations slow, the fixed sales, administrative and interest costs fall on a shrinking base, the funds have nothing left over, and the parent must still honour about $0.7 billion of convertible notes and credit-line debt. The company would owe more than its shares are worth by about that amount, roughly $3 per share. Base case (about -$1 per share). Contracted revenue is received as scheduled, the funds service their own debt but leave little or no residual, and the parent's own borrowings are only partly covered by whatever the funds release. Nothing in the filings names a source of value above that, so growth, margin gains and contract renewals beyond the initial term are left out. Contract renewals are left out because the fleet is too young for there to be any renewal record. Bull case (about +$2 per share). The company's own levers work: systems already under construction or in inventory ($181 million at year-end) are installed while credit rules still apply, tax-credit sales already contracted for 2026 are completed, and enough customers stay beyond the initial term that some funds begin to release surplus cash. Even then the contracted flows barely cover the debt, so the surplus is small. What would make this wrong. If fund debt is much more amortising than we assume, or if customers renew at high rates beyond year 20 to 25, the shares would be worth materially more. We tested the view against the specific figures: 2025 owner earnings turned positive, but only by $110 million, on $9 billion of new debt and investor money raised over 2023 to 2025 that funded the build-out, and the debt sits at $14.0 billion non-recourse against a fleet whose contracted revenue nets down to well under that amount. We could not reconcile the size of recurring prepaid and other current asset growth from the filings alone, and treated it as working capital. Because owner earnings have been negative in nearly every year, there is no price at which we currently see a margin of safety.