Business condition attached to the margin of safety
The first trains are operating and at least four quarters of reported distributions to NextDecade clearly exceed the interest on its roughly $3 billion of holding-company borrowings plus its corporate overhead.
NextDecade is building the Rio Grande LNG export terminal near Brownsville, Texas. It has no revenue yet: 2025 brought a net loss of $430 million and operating cash flow of negative $169 million, while about $4.8 billion was spent on construction. Owner earnings, which is cash from operations less working-capital swings and stock compensation, were about negative $0.94 per share in 2025 and negative in every year we could compute (2022 to 2025). We treated construction spending as development rather than upkeep, since there is no operating plant yet to maintain. What is real. The company has signed long-term LNG sale agreements with 14 buyers for about 25.3 million tonnes a year from Trains 1 to 5, with a weighted average term of 19.5 years. About 23.75 million tonnes are priced off Henry Hub with fixed fees of roughly $3.0 billion a year before inflation adjustments, and buyers owe the fixed fee even if they skip a cargo. Bechtel is building Trains 1 to 3 (Phase 1), Train 4 and Train 5 under fixed-price contracts. Phase 1 (about $18.0 billion) was 39.8% complete at year-end, first LNG is expected in the first half of 2027, and the company says it is ahead of the guaranteed schedule. Trains 4 and 5 (about $6.7 billion each) are guaranteed for the third quarter of 2030 and the second quarter of 2031. All three are financed. Why that does not translate into value for these shares. The fixed fees belong to the plants, and NextDecade owns only part of each. It is entitled to up to about 20.8% of Phase 1 distributions, and most of that share is paid only after its financial partners reach an agreed distribution threshold. It receives 40% of Train 4 distributions (60% after the partners reach a set return) and 50% of Train 5 (70% after). Ahead of any distribution, each plant must pay its operating costs and service its own project debt; Phase 1 alone has about $11.9 billion of committed bank and note facilities, and Trains 4 and 5 add about $3.9 billion and $4.1 billion. Separately, NextDecade has borrowed at the holding-company level to fund its $2.4 billion of equity commitments to Trains 4 and 5: a $1.2 billion term loan at 13%, a corporate loan of about $0.3 billion at 8% to 13.5% (the 8% portion is exchangeable into shares at $9.50), and a bank facility of up to $1.5 billion at SOFR plus 3.5%. That is roughly $3 billion of borrowings, costing about $0.35 billion a year in interest once fully drawn, on top of overhead (2025 general and administrative costs were $202 million consolidated). Some interest can be paid in kind for a period, which defers but does not remove it. The annual report does not disclose plant operating costs or the partners' distribution thresholds, so we built our own estimate from the disclosed fees, debt and ownership shares. Once all five trains run, that estimate puts distributions to NextDecade at roughly $0.1 billion to $0.4 billion a year depending on plant costs and debt repayment, before the ownership step-ups on Trains 4 and 5. Against about $0.35 billion of holding-company interest and roughly $0.05 billion of overhead, that leaves nothing we can evidence for shareholders, and in the more cautious versions falls short. Contracts run about 20 years from each train's start, and we assume nothing after they end. Bear, base and bull. In the bear case, Phase 1 distributions stay behind the partners' threshold and plant costs run above our estimate, so NextDecade's cash barely touches its debt costs. The base case has the contracted fixed fees flowing from 2027 to 2031 as scheduled, but distributions to NextDecade only roughly cover its holding-company costs. The bull case adds lower plant costs, the Train 4 and Train 5 step-ups and the early-cargo margins already sold, and still lands near break-even after the holding-company debt. We value all three at $0 per share. Because much of the borrowing sits in subsidiaries, we are not asserting that shareholders would owe more than they invested; the point is that no positive value can be traced to evidence. What we left out and why. The company has sold over 175 trillion BTU of early cargoes for 2027 and 2028 at a margin above $3.00 per million BTU, about a third of open volumes through early 2029; that is real but one-time and sits mostly in Phase 1, where NextDecade's share is small and subordinated. Trains 6 to 8 (about 18 million tonnes a year) have no financing or construction decision; the regulatory filing for expansion is expected in mid-2026 and is not yet approved. Uncontracted plant volumes and any growth in LNG demand or prices are not evidenced for this business, so they are excluded. We tested our conclusion against the loan terms: the holding-company loans are secured by pledges of NextDecade's equity in the plants, so lenders are paid before shareholders. What would make this wrong. Plant operating costs far below our estimate, the partners' return thresholds being met early (triggering the higher ownership shares), or a financing of the expansion trains on terms that add distributions would each raise value. An earlier auditor's report carried inside this filing describes going-concern doubt for the 2023 statements; the 2025 statements do not, because the plants were financed in 2023 and 2025.