Vale (VALE)

The Cost Curve Is the Equity Story

Iron ore producers sell an undifferentiated product into a Chinese-dominated seaborne market, so the analysis reduces to two questions: where does the producer sit on the cost curve, and what does its jurisdiction do to the cash that position generates. Vale is the anchor case for both.

In calendar 2025 Vale produced 336 Mt of iron ore, more than any other producer, and sold 314.4 Mt. Its fines C1 cash cost, the direct cost of mining, processing, rail and port before royalties, freight and capital, was $21.3/t. That puts it third on the disclosed big-4 curve, comfortably below the marginal seaborne tonne but above the Australian cost leaders.

Each major defines C1 differently (Vale strips out third-party ore purchases, BHP excludes royalties and freight, Rio and Fortescue report wet tonnes), so the curve is a ranking with caveats rather than four numbers on one basis. Our iron ore cost curve guide works through what each definition includes and how to line them up.

From $21.3 to $54.2: The Bridge That More Than Doubles the Cost

Vale is the teaching case for why C1 is not a breakeven. The company discloses an all-in cost of $54.2/t for 2025 against the $21.3/t C1 headline. The roughly $33/t in between is freight to the customer, government royalties, sustaining capital and corporate costs, none of which appear in C1 and all of which come out of the same revenue line. Vale publishes an all-in cost rather than a named breakeven, which is unusually honest disclosure; most producers leave the analyst to build the bridge.

The gap changes the margin arithmetic materially. At our planning price of $90/t (62% Fe CFR China), a C1-only reading suggests Vale earns roughly $69 on every tonne. Against the 2025 all-in cost the real number is closer to $36/t. Both are healthy margins, but one is nearly double the other, and the difference compounds across 314 Mt of sales. The all-in cost, not C1, is the figure to test against a long-term price assumption.

Freight is the structural reason Vale's bridge is wider than its Australian peers'. The Brazil-to-China haul is far longer than Pilbara-to-China, so more of Vale's delivered cost sits in shipping, a line item C1 never touches. Vale partially offsets this with ore quality, which brings us to the mix.

Product Mix: The Realisation Lever the Index Hides

The 314.4 Mt Vale sold in 2025 was not one product. It was 273.0 Mt of fines, 32.8 Mt of pellets and 8.5 Mt of run-of-mine ore. That split matters because each product realises a different price against the same 62% Fe benchmark: pellets earn a premium over fines, while ROM sells below them. A model that applies the headline index to every tonne misses a realisation lever that the income statement captures and the index does not.

The contrast with Fortescue makes the point from the other direction. Fortescue's lower-grade hematite realised just 88% of the Platts 62% index in the quarter to December 2025, so its cost advantage over Vale ($17.99/wmt against $21.3/t) is partly given back on the revenue line. Two producers with similar C1 costs can earn very different margins per tonne once grade and product mix set the realised price. Realisation against the index belongs next to C1 in any cross-miner comparison, not as an afterthought.

One more number worth holding onto: production of 336 Mt against sales of 314.4 Mt. A persistent gap between the two builds inventory and flatters unit costs, so the pair is worth tracking together quarter by quarter.

What to Watch in the Financials

Iron Ore Solutions EBITDA. The iron ore business generated adjusted EBITDA of $13.8 billion in 2025 within a group proforma figure of $15.9 billion. Iron ore is not a segment at Vale; it is the company, with base metals attached. Whatever happens to copper and nickel, the equity prices off the iron ore line. The fourth quarter of 2025 made the point in the crudest way: a $3.5 billion write-down of the Canadian nickel assets and a $2.8 billion deferred-tax write-off produced a $3.8 billion quarterly loss, and the iron ore cash flow never moved.

Leverage. Expanded net debt stood at $15.6 billion at the end of 2025, which is about 1.1x the Iron Ore Solutions EBITDA. That ratio is our derivation, not a Vale disclosure (the company's narrower net debt measure, including leases, was $11.2 billion), and the expanded definition includes commitments such as the dam-related provisions that conventional net debt omits. We screen 1.0-2.5x on mid-cycle earnings as normal territory. The caveat is the denominator: 2025 was a profitable year for iron ore, and the same debt against EBITDA built on our $90/t planning price reads less comfortably.

The C1 trajectory. Fines C1 came down from $21.8/t in 2024 to $21.3/t in 2025, then turned. In August 2026 Vale raised its 2026 guidance to $22.50-23.50/t of C1 and $58-62/t all-in, from $20.00-21.50 and $52-56, and put around 70% of the C1 increase on the exchange rate and the diesel price. A Brazilian cost base earning dollars carries a currency inside its unit costs, so work out what is driving a rising C1 before concluding the operation is slipping.

The share count, of all things. We put Vale's market cap at US$64.6 billion, the $15.14 ADR close on 9 June 2026 against 4.27 billion shares outstanding. Screens that use the issued count of 4.44 billion get about 4% more, because Vale still holds roughly 170 million shares in treasury after cancelling a tranche in February 2026. Small, but it moves every per-share number you build on top of it. Pick a basis and state it.

Peer Context: Vale vs Fortescue, and the Customers Who Compete

Fortescue is the natural pure-play comparison. It shipped 198.4 Mt in its financial year to June 2025 (mind the June year-end when lining up the numbers) at a hematite C1 of $17.99/wmt, the bottom of the big-4 curve, and carried net debt of just US$1.1 billion. Vale ships more volume, runs higher grades, and sells pellets; Fortescue sits at the bottom of the curve on C1 with a much cleaner balance sheet, though the grade discount gives back part of that edge. Neither dominates the other. They answer the same cost-curve question from opposite ends, which is what makes the pair useful.

The customer side of the chain is also worth a glance, because some steelmakers have opted out of being Vale's customers at all. ArcelorMittal produced 48.8 Mt of captive iron ore in 2025, covering 72% of its needs, and Cleveland-Cliffs internally sources the vast majority of its pellets. Vertical integration internalises exactly the margin Vale sells. When mill margins are tight, that choice, and the discounts and premiums mills will pay for different ore grades, feed straight back into Vale's realisations. The EBITDA per tonne guide shows how the steelmakers' side of this equation gets compared.

Key Risks

The iron ore price, which means China. China produced 960.8 Mt of the world's 1,849.4 Mt of crude steel in 2025, roughly 52%, and seaborne iron ore demand is largely a derivative of that number. Vale has no hedge against it. At our planning price of $90/t, the all-in margin compresses to roughly $36/t on 2025 costs and $28-32/t on the 2026 cost guidance. Still profitable, but the equity's sensitivity to the price assumption is the largest single input in any Vale model.

Brazil. Jurisdiction is the second axis of the Vale story. Royalties and licensing sit with federal and state governments whose terms can move, and the tailings dam failures at Mariana (2015) and Brumadinho (2019) left a long tail of remediation obligations, reparation payments and regulatory scrutiny; the expanded net debt figure exists partly to capture those commitments. An Australian producer with the same cost position would trade differently, and the discount is not irrational.

Freight. The long haul to China sits inside the all-in cost, so Vale's margin carries a freight-market exposure its Australian peers largely avoid. A sustained rise in bulk shipping rates widens the C1-to-all-in bridge with no offsetting revenue.

Realisation compression. The pellet premium and high-grade realisations that flatter Vale's revenue line are themselves cyclical. When steel mill margins are squeezed, mills shift toward cheaper feed, premiums on quality compress, and the discount on lower grades widens. The product-mix lever works in both directions, and it tends to turn against producers at exactly the point in the cycle when the headline price is also falling.