Start with the accounts, not the press release. Stanmore Resources ($SMR) (ASX:SMR) on 24 August put out a first-half 2026 result that looks, if you only read the headline loss, like a business in reverse. It is not. Saleable production was 6.5 million tonnes, unchanged on the prior corresponding period. Coal sales revenue was US$978 million, up 13 per cent. Underlying EBITDA was US$174 million, up 19 per cent. Operating cash inflow was US$176 million. The statutory loss after tax was US$44.2 million. That loss is depreciation doing its job — US$193.6 million of it — plus a royalty cheque that would have been a decent year's profit at a lot of other shops. [1][2]
The per-tonne gap is the only gap that matters in this trade. Average realised price rose US$21 a tonne, to US$153. FOB cash cost excluding royalties rose US$12, to US$101, inside the revised guidance band, with diesel and a stronger Australian dollar doing most of the damage. Subtract those two and you still have US$52 a tonne before the state. Then look at the P&L line the presentation is less eager to put in 72-point type: royalties expense, US$109.2 million, up from US$94.0 million. On 6.4 million tonnes sold that is about US$17 a tonne, an 11 per cent clip of revenue, and more than 60 per cent of the half's underlying EBITDA. Add it back onto the FOB cash cost and the all-in cash number the equity actually lives with is closer to US$118 a tonne against a US$153 realisation — call it US$35 a tonne of cash margin, before interest and the lease bill. [1][2]
FOB cash cost "ex royalties" is a cost. The royalty is also a cost. Printing the first without the second is how a US$101 tonne looks cheaper than the US$118 tonne the register actually funded.
That is the Queensland policy showing up in a specific company's books, not as a speech. The state's 2022 progressive coal royalty — 20 per cent above A$175/t, 30 per cent above A$225/t, 40 per cent above A$300/t, levied on revenue, not profit — is the highest coal-royalty regime going, as we set out in June. Stanmore's realised US$153 a tonne does not live in the 40 per cent headline band. It does not have to. The blended take was still US$109 million in six months, against a US$44 million loss for the people who own the shares. The state got paid. The equity did not. [6]
The other Queensland fingerprint is on the growth side, not the P&L. The Isaac Downs Extension — the life-of-mine extension that is supposed to replace Isaac Plains as that complex tapers ("value over volume," in the company's phrase) — had its Environmental Impact Statement submitted in June. The company's own timeline still has approvals as the bottleneck: estimated Environmental Authority and mining lease, then shovel-ready, with construction (box-cut, haul road, flood levees) after that. Eagle Downs and Lancewood sit further back on the same page, both waiting on state process. The presentation's own market slide is not shy: Australian met-coal production "in structural decline amid rising costs, regulatory headwinds, depletion and limited new projects." That is not a weather comment. [3]
On the register, as of a delayed Yahoo quote around midday Australian Eastern time on 1 September, SMR last printed A$2.87 on 901.4 million shares, for a market capitalisation of about A$2.59 billion. The 52-week range is A$1.73 to A$3.12. Trailing yield is 4.6 per cent, which is last year's fully franked 8.9 US cent final (US$80.9 million, paid 13 March) and nothing else — the board did not declare a 2026 interim, citing capital-allocation priorities, near-term growth and balance-sheet flexibility. Net tangible assets are US$1.771 a share. First-half EPS is a loss of US 4.9 cents. Cash on hand at 30 June was US$138 million; net debt on the company's own definition (term loan less cash, ignoring IFRS-16 leases) was US$72 million. A subsequent refinance upsized the term loan to US$250 million, killed US$70 million a year of scheduled repayments, cut the margin by a point to 3.50 per cent over SOFR, and stretched the undrawn revolver to March 2029. That is a real balance-sheet event. It is also, in our view, the reason the interim got skipped: they just bought themselves the option to fund the next pit without asking a met-coal-shy bank for project finance. [2][4][5][7]
The operational print underneath is not the problem. Poitrel ran ahead of its half of guidance. South Walker Creek was held back by first-half stripping and is supposed to carry the second half. Isaac Plains is fading on purpose. Mix remains a PCI house — 71 per cent PCI, 22 per cent coking, 6 per cent thermal — which is why the realised US$153 sits well under the premium hard-coking-coal tape and why the royalty does not print the 40 per cent political number. Guidance is reaffirmed: 12.8–13.4 million saleable tonnes, FOB cash cost (still quoted ex-royalties) US$98–103/t, capex US$85–95 million. Safety is better than the Queensland surface-mine average. None of that is in dispute. [1][3]
The Take (conviction 4/5). We said in June that Stanmore was a cash machine the polite lenders would not underwrite, and that Queensland's royalty was a tax on the act of mining. Six more months of accounts have not changed the diagnosis. The cash is still real. The statutory loss is still depreciation plus the state's take. The new information is the skipped interim and the US$109 million royalty line sitting there in black type, larger than the loss, larger than half the EBITDA, and levied whether or not the equity makes a dollar. Isaac Downs Extension will live or die on a Queensland approvals clock, not a coal-price clock. Own the tonne if the US$35 cash margin after royalty still pays you; do not confuse a refinance and a reaffirmed guidance range with a change in the political cost of being a Bowen Basin producer. That cost is the story. The accounts just finally said it out loud.