Alliance Reopens the Export Valve — and Quietly Moves the Goalposts on Its Payout

For two years the bear case on Alliance Resource Partners was simple: a thermal coal miner with a shrinking customer base, structurally declining burn, and a distribution that had already been cut once. The third quarter of 2026 doesn't refute that case so much as route around it. Three things actually moved: a shuttered mine got a contract and a second life, the export market went from "we're not really looking at that" to a headline part of the sales book, and the oil & gas royalty segment posted records in its first full quarter owning AllDale III and IV. Underneath all of it sits a capital-allocation question management answered less clearly than it answered anything else on the call.

From "MC Mining only" to three million tons offshore

Start with the thing that changed most versus the company's own recent history: exports. On the February call, the message was unambiguous — “the only exposure we would have to the export market will be the MC mining tons” — Joseph Craft, CEO · 2026-02-02. Two quarters later, Alliance had sales volumes booked and priced of 3.7 million tons for export in 2026 and 3.0 million tons for 2027, and Joe Craft was describing the seaborne thermal curve as "very constructive," with API2 running from roughly $105 a ton at the start of the quarter to about $140 by the end of September. Recall the April call, when he framed the domestic-versus-export hurdle at “around $120 is where our preferred option would be on how the export market would compare to the domestic market” — Joseph W. Craft, CEO · 2026-04-27. At $140, that hurdle is cleared with room to spare.

The constraint now isn't price, it's logistics — and the company named a culprit you almost never hear on a coal call: “Barge capacity has become very, very tight due to aggregate movements that are necessary or you know, a mass project down in Louisiana” — Joseph W. Craft, Chief Executive Officer · 2026-10-26. That competition for Illinois Basin barges, plus higher ocean freight, is eating the incremental margin. The prize if it clears is explicit: “we are pretty confident that we could exceed our volume this year maybe by half million or a million tons” — Joseph W. Craft, Chief Executive Officer · 2026-10-26. Note that the barge businesses bottleneck now shows up as a first-class theme in the company's own keyword set — a supply-chain story masquerading as a demand story.

Mettiki: a write-down that turned into a contract

In the first quarter Alliance issued a WARN notice at Mettiki, suspended mining, and wrote the asset down hard. It now has a new multiyear supply agreement with its largest historical customer — 700,000 tons to Dominion in 2027, with the mine converting from longwall to two room-and-pillar sections across two shifts, keeping roughly 150 mineworkers employed and adding about 550,000 tons of 2027 production versus 2026. The multiyear coal supply agreement is the single most concrete datapoint on the call, and it comes with an optionality kicker: “by converting to a continuous miner operation, it will give us a little bit more opportunities to participate in the export met market because we can be a little bit more selective in mining, the seams” — Joseph W. Craft, Chief Executive Officer · 2026-10-26. Craft sized that at another 150,000–200,000 tons into met markets if the pricing is there. CFO Cary Marshall confirmed the carrying value stays put — no write-up, and no further balance-sheet adjustment expected.

The thesis the tape has stopped paying for

Craft's bull case rests on existing coal fleet longevity: electricity load growing in Alliance's served markets, permitting and interconnection slowing new build, and GE Vernova sold out of new turbine deliveries until 2031. He argued utilities are becoming more disciplined about counting only contracted data-center load, and that even then power demand growth is substantial — so coal burn should run harder for longer, with the visible impact landing "late 2027 and into 2028." There's supporting policy: the EPA finalized a rule repealing most 2024 greenhouse gas standards for power plants, and Craft cited an Indiana study finding 75% of the state's coal plants could generate more cheaply than new combined-cycle gas.

Here's the tension worth flagging. The market's own verdict on the AI-power complex has gone cold. The global AI data centers cluster is down roughly 5.5% over both the last 30 and 90 days (with 47 decliners against 6 advancers in the 30-day window), even though it's up over the trailing year. Alliance's stock, meanwhile, is up only 7.6% over 90 days and has spent the better part of 89 weeks going essentially nowhere on the long tape — a slow base, not a breakout. So the company is leaning into a secular story the tape voted on a year ago and has since been unwinding. That is not a reason to disbelieve the fundamental case; it is a reason the re-rating hasn't happened.

The goalposts moved

This is where the report gets interesting. Distribution coverage hit 1.76x — a three-year high — on distributable cash flow of $131 million, up about 29% year over year, with free cash flow of $199 million in the quarter. In April, management had been explicit about the bar: “We need to get our distribution coverage ratio more in line with an expectation of 1.2x to 1.4x on a going-forward basis before we would consider either of those” — Joseph W. Craft, CEO · 2026-04-27 — "either of those" being a buyback or a distribution increase. Asked directly about raising the distribution this quarter, the answer was different: “we want to stay at the 1.5x coverage ratio or greater. And I think that is a target that you know, we would want to sustain beyond 1 quarter” — Joseph W. Craft, Chief Executive Officer · 2026-10-26. The company has now cleared its own stated hurdle by a wide margin and replaced it with a higher one, while pointing the incremental cash at minerals M&A and debt. The AllDale IV acquisition was funded with roughly three quarters drawn on the revolver and a new “$150 million short term debt facility at Minerals. So that is a. that is an 18-month facility” — Cary Marshall, Senior Vice President and Chief Financial Officer · 2026-10-26 that amortizes quarterly out of mineral cash flow. That's a defensible use of capital. It's also a moving target, and unitholders should treat "1.5x sustained" as the new, softer goal line.

The balance sheet has room: liabilities at 39% of assets, interest coverage of 7.6x (still down about a fifth year over year), and total/net leverage of 0.89x and 0.78x on trailing adjusted EBITDA. Note the disclosure change — adjusted EBITDA is now defined to include non-controlling interests and then reported on an attributable-to-ARLP basis, a presentational shift that muddies comparability for anyone screening the series.

The filed numbers, which lag this print by a quarter, show the trough the company is climbing out of: revenue of $552 million, essentially unchanged year over year and 18% below the 2023 peak, with net margin back at 14.8% and a free cash flow margin near 19%. Capital discipline is doing real work here: capex came in at roughly 9% of revenue, down 25% year over year. All of that shows up in a price-to-free-cash-flow multiple under 9x, which is the entire point: the equity is cheap because nobody believes the demand tail lasts.

Two softer signals deserve a mention. The keyword momentum behind Gas Royalty segment fell sharply this quarter — precisely as that segment set record oil and gas royalty revenue of $59.8 million on record BOE volumes of 1.45 million. That's an artifact of the acquisition being absorbed into consolidated framing rather than a business deterioration, but it means the fastest-growing, highest-multiple part of Alliance is now less narratively visible, not more. And lower average pricing remains a persistent company theme: realized coal price per ton was $53.93, down 8.3%, and the Appalachia number fell 25% on an unfavorable Tunnel Ridge mix — cost per ton fell faster, which is why the spread widened, but the price line is still pointed the wrong way.

Bottom line

Alliance executed on the two things it could control: it restarted a mine that had been written off and it turned a dormant export channel back on with the price tailwind behind it. What it did not do is convert a 1.76x coverage ratio and 0.78x net leverage into a higher distribution to unitholders, preferring to redeploy into royalties and keep the stated leverage target in the back pocket. The stock's muted 90-day gain against a fading data-center-power tape suggests the market sees the same thing: a genuinely better business, still waiting for a reason to be re-rated.