Fortescue's September quarter: the tonne that hasn't been sold is a commercial problem, not a mining problem

Fortescue's September 2026 quarter is easy to misread. Headline shipments of 46.8 Mt, down 6% year on year and 11% quarter on quarter, look like an operational stumble. They are largely not. Ore mined rose 4% year on year to 62.7 Mt, and the quarter carried the company's largest-ever port shutdown. The genuinely important number is elsewhere: sales of 42.9 Mt, 3.9 Mt below shipments, with another 4–5 Mt of inventory expected to accumulate in October because negotiations with China Mineral Resources Group (CMRG) remain unresolved.

Two different gaps

The 11% quarter-on-quarter fall in shipments and the 8.3% shipment-to-sales gap have separate causes and should not be bundled together.

Reported figures from Fortescue's September 2026 Quarterly Production Report, as compiled by SMM and SteelOrbis.

The sequential decline reflects the port shutdown — inload and outload, including the Train Unloader 1 tippler cell, major reclaimer and shiploader components — with further life-cycle port maintenance flagged for FY28–29. Processed tonnage falling faster than mined tonnage is the signature of a maintenance-constrained quarter, not a resource problem. The year-on-year decline is a different matter: it partly reflects the deliberate phase-out of West Pilbara Fines, which contributed 4.0 Mt a year earlier and nothing this quarter.

The shipment-to-sales gap is commercial. Fortescue sells into China through its Shanghai trading arm, holding port-side inventory there; title passes on sale, not on loading. A stalled pricing and term-contract negotiation with CMRG therefore converts physical tonnes into working capital rather than revenue. Net debt rose to US$2.8bn at end-September from US$0.9bn at end-June, driven principally by that inventory build. If October adds the indicated 4–5 Mt, cumulative unsold stock reaches roughly 8–9 Mt — about two-thirds of a month's shipment volume sitting in the chain.

The guidance arithmetic is tighter than it looks

Fortescue maintained FY27 shipment guidance of 197–207 Mt, including 11–14 Mt from Iron Bridge, explicitly "subject to" the CMRG outcome. Q1 delivered 22.6–23.8% of that range.

The remaining three quarters must therefore average roughly 50.1–53.4 Mt — between 3.3 and 6.6 Mt per quarter above the September run-rate. That is not unreachable: the company shipped 52.7 Mt in April–June 2026. But it leaves no room for a prolonged sales freeze, because production will eventually have to be trimmed to match stockpile capacity rather than mine capacity. Fortescue has said as much, warning that operations, "potentially including production levels", may be affected if talks are not concluded by 31 December 2026.

Iron Bridge is the tighter constraint. At 2.5 Mt, the quarter covered 17.9–22.7% of its 11–14 Mt guidance; the remaining quarters need 2.83–3.83 Mt each against a five-quarter high of 2.7 Mt. The asset is ramping, not yet at nameplate, and the guidance assumes a step-up that has not yet been demonstrated.

Costs and prices are the real earnings story

Hematite C1 cost of US$21.98/wmt is up 13.5% quarter on quarter and 21.0% year on year, and sits above the top of the unchanged FY27 guidance range of US$20.50–21.75/wmt. Higher diesel alone added about US$1.40/wmt year on year. Lower volumes over a largely fixed cost base did the rest.

Realised hematite pricing of US$79.79/dmt was 82% of the 61% Fe CFR index average of US$96.92/dmt, down from 89% realisation in January–March 2026. Part of that erosion is mix: the phase-out of West Pilbara Fines and the arrival of Fortune Fines at 1.5 Mt (3.2% of shipments, below the 5–6% of hematite shipments previously indicated) shift the portfolio toward lower-grade material. Part is the discount structure in a weak price environment. Iron Bridge concentrate, by contrast, realised US$111.09/dmt, or 115% of index — a reminder of why the concentrator ramp matters commercially even at modest volumes.

Product mix: better for furnaces, worse for realisations

Lump reached 3.8 Mt, or 8.1% of shipments, up from 4.6% a year earlier and the highest share in five quarters — Fortescue's strongest available lever on realised value, since lump commands a premium and displaces sinter feed in Chinese burden design. Super Special Fines (18.6 Mt) and Fortescue Blend (17.5 Mt) remain the core. The direction of travel is higher-value lump but also more low-grade fines, which raises slag volume and coke consumption per tonne of hot metal at Chinese mills relative to 61–62% Fe alternatives. That is precisely the kind of mix argument that becomes a negotiating point when term contracts are being reset.

Blacksmith: a bigger resource, a weaker grade

The Blacksmith resource increase deserves more scrutiny than the headline. The Mineral Resource rose 153% to 615 Mt, but grade fell from 59.3% to 56.0% Fe. The maiden Ore Reserve of 196 Mt is entirely Probable and averages 58.0% Fe as dry product; estimates are stated as at 30 June 2026 and were released with the FY26 results on 20 August, not with this quarterly.

A 2.5x tonnage increase that comes with a 3.3-point grade decline is an expansion of mine life, not an improvement in product quality. For a company whose realised prices are already slipping against index, adding a 58% Fe reserve base is strategically useful for sustaining volume as older deposits mature — and commercially neutral at best on realisations. Readers should treat Blacksmith as a life-extension asset rather than a margin lever.

Green metal: a genuine technical milestone, distant commercially

First hot metal at the Christmas Creek Green Metal Project, produced via Metso-supplied electric smelting furnace technology during commissioning, is a real achievement — it is the first operation of its kind in Australia using Pilbara ore and renewable-powered electric smelting. But the plant was designed for more than 1,500 t/yr at roughly US$50m of capital. It is a technology demonstrator. Its significance for Fortescue is option value on a decarbonisation pathway, not tonnage. It will not move earnings or iron ore markets in the planning horizon relevant to this guidance.

What to watch

The quarter's operational story is sound: maintenance executed, mined volumes up, lump share improving, Iron Bridge ramping. The commercial story is not. Fortescue has a customer concentration problem dressed as a negotiation, and its own guidance now embeds a deadline. Three indicators matter from here:

• Whether the October inventory build is reversed before 31 December, and whether Fortescue discloses any production curtailment;
• Whether Iron Bridge delivers the 2.83–3.83 Mt quarterly average its guidance implies;
• Whether C1 cost returns inside the US$20.50–21.75/wmt range as volumes normalise, or whether diesel and sustaining capital keep it structurally above.

A settlement with CMRG would resolve the second and third simultaneously. Until then, Fortescue is mining and shipping well, and being paid for less of it.