Vale's "10% Underwater" Claim: A Marginal Cost Signal Buried Under a Bear Market in Iron Ore

Singapore iron ore futures slipping below $90/mt in mid-October — a two-year low and roughly 15% lower year-to-date — is the headline. The more interesting number came from Vale CEO Gustavo Pimenta in a Bloomberg interview on 16 October: around 10% of global seaborne supply, some 160 million mt a year, is currently running at a loss, and only about 15 million mt has actually left the market.

That pairing of numbers is the analytical core of the story. It tells you where the marginal cost of seaborne iron ore now sits, and it also tells you how slowly the market punishes uneconomic tonnes.

What the arithmetic implies

The reported figures are round CEO estimates, not audited data, but they are internally consistent. If 160 million mt represents 10% of seaborne supply, the implied market is roughly 1.6 billion mt a year — in line with commonly cited global seaborne trade volumes. (This is my inference from the reported percentages, not a figure Vale disclosed.)

The withdrawal figure is the one that matters for price direction. Roughly nine-tenths of the identified loss-making pool is still shipping. Prices therefore do not bounce the moment marginal tonnes turn cash-negative; they grind lower until the pain becomes operational — credit constraints, depleted ore bodies, maintenance deferrals, contract renegotiations — and shut-ins accelerate. Pimenta's own framing reflects this: current prices "should be somewhat higher" to reflect freight and fuel, and further cuts should help rebalance the market, but the timing is uncertain.

The escalation is also notable. Vale had already flagged an update to its iron ore financial estimates based on a Middle East conflict model back in May, so the cost shock is being treated internally as a structural input assumption rather than a transient disruption.

Why Vale can afford to be philosophical

Vale's own cost position explains the tone. Its Q2 all-in costs rose 18% year-on-year to $61.6/mt, with C1 cash costs up 9% to $24.1/mt, driven mainly by Brazilian real appreciation. Even at sub-$90/mt prices, Vale sits comfortably above its all-in cost line.

The $37.5/mt gap between C1 ($24.1/mt) and all-in ($61.6/mt) is the single most instructive number in the disclosure. It captures freight, royalties, sustaining capital and other cash items — precisely the components most exposed to a fuel and shipping shock. When that wedge widens, the supply curve steepens quickly, and producers at the far end of the freight curve (higher-cost Indian, Iranian, or long-haul low-grade suppliers) fall underwater first. Vale, as a low-cost, high-grade Brazilian producer, is the beneficiary of that attrition, not its victim.

Its revised 2026 guidance — C1 of $22.5–23.5/mt and all-in of $58–62/mt — implies a modest easing from Q2 levels, leaning on currency and oil assumptions rather than volume leverage. Those two variables, the real and bunker fuel, are the levers to watch.

The counterweight: Simandou and Chinese demand

Pimenta's cost-support thesis collides with two disinflationary forces. First, weak Chinese steel demand. Second, new supply from Guinea's Simandou, which is ramping into a market already in surplus. Cost inflation from the Middle East conflict is cyclical; Simandou volumes are structural. The "floor" the CEO describes is therefore contingent — it holds only as long as freight and fuel stay elevated and Simandou ramps no faster than expected. His stated confidence in Chinese demand sits awkwardly against a two-year price low.

What to watch

• Freight and bunker fuel as the swing variable in the marginal cost curve — the fastest lever that would validate or undermine the "higher floor" argument.
• The pace of the remaining ~145 million mt of loss-making supply exiting, ideally via producer announcements and shipping data rather than price alone.
• Simandou ramp guidance from Rio Tinto and the Winning Consortium as the structural counterweight to cost support.
• Vale's realized premiums and pellet mix — high-grade, low-alumina ore tends to outperform when steelmakers chase productivity and emissions gains.

Caveat: the 10%, 160 million mt and 15 million mt figures are CEO estimates reported by trade press; the underlying Bloomberg interview and Vale's 2Q26 cost disclosures could not be independently verified here. Treat them as directional rather than audited.