Iron ore traded in a narrow range through the first half of May 2026, but the news flow underneath was anything but quiet. Property completions in China continued to drag, infrastructure stimulus delivery slipped against the announced schedule, and steel mill margins in Hebei tightened to the point where a few smaller operators idled blast furnaces.

For Australian producers, the question is whether this is a soft patch or the start of something structural. My read, after a fortnight of conversations with people on the marketing desks, is that it’s mostly the former — but the case for “structural” is no longer dismissible.

What the spot price isn’t telling you

Spot 62% Fe was hovering around the mid-90s USD per dry metric tonne CFR through the second week of May. That’s a respectable level by any historical standard and well above the cost curve for the major Pilbara producers.

The number that matters more is the 65% premium. It’s been compressing. When the premium narrows, it tells you Chinese mills are blending down — accepting lower-grade ore because their margins don’t justify paying up for higher iron content. That’s a demand-side signal, not a supply-side one.

The other indicator I’m watching is the lump-fines spread. Lump has been trading at smaller premiums than typical for sinter-feed-driven mills. Again, mills choosing the cheaper option because they can’t recover the cost in finished steel prices.

The producer response so far

BHP, Rio Tinto, and Fortescue have all signalled in their quarterlies that volume guidance is unchanged. That makes sense — at current price levels, the majors are still printing money and there’s no operational reason to throttle back.

The conversation is different further down the cost curve. Mid-tier producers operating in the high-70s to low-80s break-even range are now running their cash flow forecasts on $85 ore, not $100. A couple have quietly delayed expansion CapEx that was due to start in H2 2026.

I’d expect to see this in the August quarterly results from the second-tier producers. None of them will say “we’re worried about demand” — they’ll talk about “phasing decisions” and “capital discipline” — but the meaning will be the same.

Three things that could shift the picture

A genuine Chinese property stimulus. Beijing has signalled patience and targeted measures over broad-based intervention. If that posture changes, the price response would be quick. I don’t see catalysts for a change before the autumn plenum.

Supply disruption in Brazil. Vale’s wet-season volumes have been okay but not exceptional. Any meaningful disruption would tighten the seaborne market quickly given Australian volumes are running close to nameplate.

Decarbonisation policy step-change. The longer-term thesis on iron ore is that direct-reduced iron processes will gradually push high-grade pellet feed premiums up and 62% fines down. The transition is happening at a pace measured in years, not months, but every six months there’s another small data point — a new green steel pilot in Sweden or the Middle East — that nudges the structural case along.

What I’d be doing if I were a non-major producer

Stress-test the budget at $80 ore for the back half of 2026. Not because that’s my forecast — it’s not — but because the asymmetry of risk matters. If you’re built for $100 ore and the world delivers $85, you’ve got problems. If you’re built for $80 and it delivers $95, you’ve got optionality.

Push hard on operating cost. The cost-out projects that mid-tier miners deferred during the 2022-2024 boom are exactly the projects you want running now. Predictive maintenance programs, fuel-burn optimisation on the haul fleet, ore-grade reconciliation tightening — all the things that look incremental in isolation but compound at scale.

Be honest with the board about timing on growth projects. The right move in mid-2026 is probably to keep the FEED studies progressing while pushing financial commitment to mid-2027. Optionality has value when the demand picture is this uncertain.

The headlines to ignore

Anything claiming a “rebound” or a “crash” in the next six weeks. The market is range-bound for a reason — supply is steady, demand is steady-but-soft, and neither side has a near-term catalyst for a big move. The story for the back half of the year is whether the structural demand picture is degrading slowly or stabilising.

The data point I’ll be watching most closely between now and August is Chinese steel exports. If they keep running at 100+ million tonnes annualised, that’s a sign mill demand for ore is being sustained by external markets rather than domestic construction. That’s the same volume, but a different story — and a more fragile one.