Brazil's iron ore cost squeeze reaches the domestic market: JMN halt shows freight is now a system-wide problem, not an export one

The fifth domino, and a more interesting one

When SteelOrbis reported on 15 October that JMN Mineração had halted its mine between Desterro de Entre Rios and Piracema, the significance lay less in the tonnage than in the identity of the miner. JMN is the fifth Brazilian iron ore producer to suspend or cut output in about six weeks, after Itaminas, CSN Mineração, Cabo Verde and Usiminas Mineração. But unlike the others, JMN does not ship to China at all.

Its customers are domestic. Yet it cited the Brazil–China freight rate as a principal reason for stopping. That is the analytically important fact in this story, and it deserves more attention than the bare announcement received.

Local trade press fills in what the international wire did not. The suspension began around 6–7 October (sources differ by a day), is of indefinite duration, involves about 49 redundancies, with some staff to be redeployed to sister company Ferro+ Mineração in the Congonhas–Ouro Preto area and a skeleton crew left on care and maintenance. There was no announcement of a restart trigger. The company said resumption would depend on "a combination of factors: prices, transport costs and operating margins" — a formulation that, read plainly, means it does not know when.

Why a non-exporter is exposed to a freight rate

The mechanism is straightforward once you lay it out. Brazilian domestic iron ore contracts, like most bulk raw material markets, are priced by reference to the international benchmark, usually on a CFR China basis with a netback adjustment for logistics. The domestic price is therefore an export-parity price: it is whatever a Brazilian mill would have to pay to replace that tonne with imported or exportable material, less the cost of getting it to market.

When freight rises, the netback collapses. At a fixed CFR China price, every dollar added to the Tubarão–Qingdao rate removes roughly a dollar from the FOB value of Brazilian ore. Domestic buyers, who know they are competing against import parity, pass that arithmetic straight into the price they will pay a local miner. The exporter and the domestic-only producer end up quoting against the same curve.

The squeeze is amplified by the shape of the price curve itself. High-grade ore has performed better than the benchmark as Chinese mills push blast-furnace efficiency and emissions compliance; the discount applied to lower-grade and lump material has widened. Smaller Minas Gerais producers concentrated in lower-grade or lump products, with short mine lives, older plants and thin beneficiation, sit precisely where the netback damage is greatest.

A rough illustration of the arithmetic (all figures analytical, based on reported benchmarks):

• Brazilian 65% Fe ore at $111/mt CFR China, freight $33.20/mt → implied FOB around $78/mt before port and inland costs.
• At the January 2026 freight level of roughly $20/mt, the same CFR price implies FOB around $91/mt.
• That is a $13/mt swing in realised revenue on a tonne that never leaves Brazil, which for a 2 Mt/yr operation is on the order of $26 million of annual revenue, against a cost base that has not fallen.

For a marginal producer with a large inland haul from Minas Gerais to port, plus rail and port charges, that is more than enough to turn a thin margin into a loss.

A freight spike that was both sharp and volatile

The trajectory of the benchmark within 2026 matters, because it explains why the halt came in October rather than earlier.

Two things stand out. First, the run-up was driven partly by bunker fuel costs linked to Middle East conflict, partly by seasonal Chinese restocking and tighter Pacific supply — a mix of cost-push and demand-pull that leaves the rate highly sensitive to geopolitical headlines. Second, the benchmark fell about $10/mt, or roughly 23%, in the two weeks to mid-October, yet producers were still halting.

That is the telling detail. If a 23% freight correction is not enough to restore viability, the problem is not only freight. It is that ore prices have fallen to their lowest since September 2024 amid a Chinese demand–supply imbalance, as SteelOrbis reported on the same day. The two curves have compressed the margin from both directions. Producers who halted in late September did so at the peak of the freight spike; JMN halted with freight already falling, which suggests the company judged the price side of the equation to be deteriorating faster than logistics was improving.

The volatility itself is a deterrent. A miner cannot restart a beneficiation plant and rehire a workforce against a benchmark that has moved 25% in a fortnight. Indefinite suspension is the rational response to a price signal that has stopped being a signal.

What the cluster of halts tells us about the market

Five suspensions in six weeks is no longer a set of individual business decisions. It is a supply response, and a discriminating one.

• It is concentrated in the small and mid-tier Minas Gerais segment, not in Vale's system or in the major pellet feed producers. The companies named — Itaminas, Cabo Verde (a 600 kt/yr lump operation at Catumbi), JMN, plus production cuts at CSN's Pires complex and Usiminas — are the high-cost, low-volume end of the Brazilian curve.
• It is being transmitted through the price mechanism, not through export volumes. Brazil's iron ore exports to China have continued to grow; the Chinese customs data cited by SteelOrbis showed January–September imports up 4.1%. The tonnage lost from JMN and its peers is small against national output. What is being cut is the marginal, high-cost tonne whose economics no longer clear.
• It signals a cost curve that has steepened. When freight doubles, the effective landed cost of every Brazilian tonne rises, but the netback damage falls hardest on producers furthest from port and lowest in grade. That is a classic marginal-supply shakeout, and it tends to persist until either prices recover or the marginal capacity is permanently retired.

The arithmetic tension in the reported figures

One number in the reporting does not reconcile and is worth flagging. Local press, citing the company, said JMN represented about 10% of the J. Mendes group's iron ore output. That is hard to square with a 2 Mt/yr JMN against a ~7 Mt/yr Ferro+: on a nameplate basis JMN would be roughly 22% of combined capacity, and if Ferro+ mine capacity is closer to 9 Mt/yr (as Brasil Mineral reports), the share falls to about 18%. A 10% share would imply group output of roughly 20 Mt, which no source supports.

The most likely explanation is that JMN's actual output was well below its 2 Mt/yr nameplate — a common situation for a smaller, older operation — or that the "10%" refers to a different metric, such as revenue or a specific product line. Either way, readers should treat the 2 Mt/yr figure as an estimate and the 10% claim as company-sourced but internally inconsistent. The difference matters: a 2 Mt suspension is a rounding error in a 400 Mt/yr national market; a persistent 10%-of-group cut is a signal about the group's own strategy.

The strategic read-across

The J. Mendes group has been investing in exactly the direction that the market is now rewarding: high-grade pellet feed. Its Ferro+ beneficiation expansion, targeting around 5.5 Mt/yr of pellet feed from compact and friable itabirites, is a bet on the premium end of the curve, where Chinese demand is most durable and where freight is a smaller share of landed value.

JMN, by contrast, is the group's older, lower-grade unit. Suspending it while running Ferro+ at full tilt is not distress; it is portfolio triage. The group is concentrating its capital and workforce on the asset with the better margin and letting the marginal one wait. The redeployment of workers to Ferro+ makes that explicit.

For the wider market, the message is less comfortable. The Brazilian supply response to higher freight and weaker prices is being delivered by exactly the producers who cannot afford to wait — which means the tonnage will return only if both the freight rate and the ore price cooperate, and probably not before Chinese steel margins and restocking behaviour turn. Until then, expect more of the same: small suspensions, production cuts at marginal complexes, and a domestic market where the seaborne benchmark sets the price even for miners who never see a ship.

What to watch

1. Tubarão–Qingdao freight stability. A rate that holds below about $30/mt for a month is the precondition for restarts; another spike above $40 would extend the suspensions into 2027.
2. Chinese rebar and hot-rolled margins. The ore price weakness is demand-led; a Chinese construction recovery is the only durable fix.
3. The low-grade and lump discount. If the spread between 65% Fe and 62% Fe keeps widening, the suspended producers have no economic case to return.
4. Whether larger producers follow. CSN and Usiminas cut low-grade output first. If they extend cuts to higher-grade product, this becomes a national supply story rather than a marginal-producer story.

Reference points

• SteelOrbis, "Fifth Brazilian iron ore miner halts ops on high Brazil-China freight rates; low ore prices", 15 October 2026 — link
• Minera Minas, on the JMN suspension and 49 redundancies, 15 October 2026 — link
• Mais Minas, on the 7 October halt and dismissals, 15 October 2026 — link
• SteelOrbis, "Brazil-China iron ore freight rates reach multi-year high amid strong fundamentals", 3 September 2026 — link
• SteelOrbis, "Third iron ore miner in Brazil reduces production amid higher Brazil-China freight rates", 29 September 2026 — link
• SteelOrbis, "Fourth Brazilian iron ore miner hit by Brazil-China freight rates and low ore prices", 2 October 2026 — link
• J. Mendes, Mineração business pages (Ferro+ ~7 Mt/yr) — link
• Revista Brasil Mineral, "As Maiores Empresas do Setor Mineral 2025" (Ferro+ capacity data) — link

All price, freight and capacity figures are as reported by the cited sources; the margin and netback calculations, the interpretation of the 10% output share, and the characterisation of the halt as portfolio triage are the author's analysis and estimates.