Nucor (NUE)

Nucor vs ArcelorMittal: One Comparison

Nucor is worth more than ArcelorMittal on less than half the tonnage, and that gap is the cleanest single illustration of what the market pays for one steelmaking route over the other. Nucor melts recycled scrap in electric arc furnaces (EAFs). ArcelorMittal and Cleveland-Cliffs mostly smelt iron ore with coke in a blast furnace and refine it in a basic oxygen furnace, the BF-BOF route, which carries far more fixed cost. On 9 June 2026 Nucor closed at a market capitalisation of about $57.9 billion; ArcelorMittal, three quarters blast-furnace, closed at about $50.4 billion. ArcelorMittal shipped 54.0 million tonnes of steel in FY2025. Nucor's external mill shipments, the basis comparable to ArcelorMittal's reported group figure, were 19.8 million tons.

Mind the units. Nucor and Cleveland-Cliffs report short tons of 2,000 lb; ArcelorMittal reports metric tonnes, which are about 10% heavier. Put Nucor's $210 per short ton onto a metric basis and it becomes roughly $232 per tonne against ArcelorMittal's $121. The unit correction widens the gap rather than closing it, which is why each figure should stay in the unit its company reports until the moment you line two of them up.

Enterprise value per shipped ton widens the gap further. Nucor's roughly $62.3 billion of enterprise value ($57.9 billion of equity plus ~$4.4 billion of net debt) works out to about $3,100 per external short ton shipped, or about $3,460 per tonne; ArcelorMittal's ~$58.3 billion over 54.0 Mt is roughly $1,080 per tonne. On FY2025 reported EBITDA the multiples are about 14.9x against 8.9x. Part of the premium is US listing and earnings quality; part is simply what the market pays for a franchise at this point in the cycle. The through-cycle EV/EBITDA guide covers how to judge whether the premium is earned; the short answer is that Nucor has spent two decades earning it, but the entry multiple still matters.

The Model: Scrap In, Margin Out

The FY2025 10-K states that Nucor's steelmaking operations use EAFs for 100% of their production. The route needs no blast furnaces, coke ovens, or captive iron ore mines. The furnaces are fed by roughly 20 million gross tons of recycled scrap a year, and Nucor owns the supply chain behind that number: the David J. Joseph Company, its wholly owned scrap processing and brokerage business, plus two direct reduced iron plants (Louisiana and Trinidad) that sent about 3.3 million metric tonnes of DRI to the mills in FY2025. When scrap tightens, Nucor is on both sides of the trade.

The earnings engine is the steel spread: the hot-rolled coil price minus the scrap cost, with volume as the multiplier. Our steel spread guide works through the arithmetic. What makes the EAF route structurally different is that the biggest cost input, scrap, falls when steel prices fall. The cost base flexes with the cycle instead of grinding against it.

FY2025 showed exactly what that means. Nucor earned $4,174 million of EBITDA on $32.49 billion of net sales, about $210 per external ton shipped. Cleveland-Cliffs, an integrated BF-BOF producer selling into the same US market in the same year, earned roughly $2 per ton of adjusted EBITDA and booked a $1.4 billion net loss. Both faced the same domestic demand and the same tariff shelter. The route explains most of the gap, with automotive at 28% of Cliffs' steel revenue doing much of the rest. It is why Nucor's market capitalisation is eight times Cliffs' on shipments only a fifth larger.

Which Tons? The Per-Ton Denominator Trap

Nucor discloses two shipment figures and they produce two very different per-ton numbers. Total steel mill shipments were 25.271 million tons in FY2025, but 5.4 million of those went to Nucor's own downstream products businesses. External shipments were 19.848 million tons. Divide the $4,174 million of EBITDA by each and you get ~$165/ton on the total or ~$210/ton on external shipments. Neither is wrong; they answer different questions. The external basis is the one comparable to ArcelorMittal's reported group shipments, so cross-producer tables should use it, and any per-ton figure you quote should say which denominator it sits on. The EBITDA per tonne guide treats this, along with the short-ton versus metric-tonne trap, as the two checks to run before trusting any comparison table.

What to Watch in the Financials

The spread against planning. Our planning spread, from the bulk deck's $900 HRC and $375 scrap, is $525/st. When spot HRC and scrap diverge from those marks, profitability swings with the spread, not the absolute level alone. Spot HRC near $1,193/st against scrap near $427/st puts the June 2026 spread around $766, roughly $240 above planning, and a spread that wide is not something to capitalise at a full multiple. The question for any Nucor model is what EBITDA per ton looks like at $525.

Section 232. The 50% tariff on primary steel imports, in force since June 2025 and charged on a product's full customs value rather than its metal content since April 2026, is what holds US HRC roughly $208/st above what Asian coil costs to land in the US once the duty is paid, and that premium was already narrowing through 2026. Nucor is the biggest beneficiary of the regime, which makes it the most exposed to a reversal.

The balance sheet. Net debt of ~$4.4 billion at 31 December 2025 is about 1.1x FY2025 EBITDA, inside the 1.0-2.5x band we treat as normal for a steelmaker. Watch whether buybacks and the growth capex programme hold it there as the spread normalises.

Shipments and utilisation. Falling utilisation precedes falling spreads, so the volume line is the early-warning indicator. Total mill shipments of 25.3 million tons in FY2025 are the baseline.

Key Risks

The multiple is the risk. At roughly 14.9x FY2025 EBITDA, Nucor trades far above the ~4-6x through-cycle band we use for commodity steelmakers on normalised earnings. The premium rests on the EAF cost structure and scrap integration, plus Nucor's position in the US sheet market. But FY2025's $210 per external ton was struck above the $175 our mid-cycle build gives an EAF producer, so 14.9x is a high multiple on earnings that were themselves above mid-cycle. Pay it and the downside maths is unforgiving if the spread reverts to $525 and the multiple compresses at the same time.

Tariff dependence. A meaningful slice of the current spread is policy, not economics. China exported HRC at $496-504/mt FOB in early June 2026, about $450-457 per short ton, while US domestic material traded above $1,100/st. Any carve-out, exclusion regime, or post-2027 reset that narrows the gap flows straight through to Nucor's realised prices.

Scrap-side compression. The spread can close from the input side too. US EAF capacity keeps being added off an already dominant base: EAFs made 71.3% of US crude steel in CY2025. More furnaces chasing the same pool of prime scrap, the clean, low-residual grades the better product mixes need, raises the metallics bill. DJJ and the DRI plants are a partial hedge, not an exemption.

Spread cyclicality still bites. EAF economics dampen the cycle; they don't repeal it. Nucor's EBITDA per ton still swings hundreds of dollars between peaks and troughs, and a US construction or automotive downturn would take volumes and spread down together. Cleveland-Cliffs' FY2025 is the live reminder of what the bottom of this cycle does to steel earnings.