Algoma's Q3 shows the cost of a conversion year — and why the second EAF, not the price, is the number that matters

Algoma Steel's third quarter is best read not as a steel-market result but as an industrial-construction result. Shipments of 144,527 tons are roughly a third of the 419,173 tons moved in the same quarter a year earlier, revenue fell 52.4% year on year to C$249.4 million, and the net loss widened sequentially to C$129.6 million from C$96.0 million. The headline that matters, however, came after the quarter closed: first steel from Electric Arc Furnace Unit Two on 9 October 2026, roughly fifteen months after Unit One's first heat.

What the numbers actually say

The loss shrank sharply year on year (from C$485.1 million in Q3 2025), but that comparison is misleading — the prior-year quarter carried what were evidently heavy impairment-type charges during the blast furnace wind-down. The sequential comparison is the informative one, and it points the wrong way.

Source: Algoma Q3 2026 figures as reported by SteelOrbis. Percentage changes are as reported; the implied per-ton revenue figures below are my own calculations.

Price realisation improved substantially — C$1,510 per net ton, up 33.7% year on year — on higher transaction prices and a richer plate mix. Algoma is Canada's only discrete plate producer, and plate has been the strongest part of the North American flat-rolled complex, reinforced by defence-linked demand from its Roshel and Hanwha arrangements. Yet a better price per ton coincided with a worse loss. Calculated on shipped tons, the loss was about C$897/t in Q3 against roughly C$529/t in Q2 — an increase of around 70%. That is the signature of fixed-cost under-absorption and ramp-related expense, not of weak pricing.

The revenue-per-shipped-ton arithmetic (C$249.4m ÷ 144,527 t ≈ C$1,725/t) exceeds the reported C$1,510/t net figure by roughly 12%; the gap will sit in freight and revenue deductions. The C$1,510 is the number management wants read, and it is directionally consistent with the mix story, but it cannot be reconstructed from the two headline figures alone.

Three causes, only one of which is temporary

Algoma attributed the shortfall to the EAF-only ramp-up following the halt of all blast furnace operations on 18 January 2026, US Section 232 tariffs on its export sales, and an unplanned outage at the Lake Superior Power generating facility, which had already forced the company to suspend EAF production in August and restart under interim power arrangements in early September.

These three causes are not equivalent in nature, and conflating them understates the risk profile:

• The ramp-up is transitional. Two EAFs plus the existing rolling complex should support a return to a multiple of Q3's annualised ~578,000-ton run-rate. The company has not yet demonstrated, in the figures available to me, that it can do so reliably.
• The power outage is structural, not one-off. An electric steelmaking route converts a fuel-and-ore risk into an electricity-availability risk. Algoma's own decarbonisation case rests on Ontario's low-emissions grid — a genuine advantage on carbon intensity, but one that makes the melt shop hostage to generator reliability in a way a blast furnace complex never was. The LSP outage cost a full quarter of output; the mitigation (pulling forward maintenance) reduces future downtime but does not reduce the exposure.
• Section 232 is a policy variable with a long half-life. For a mill on the Great Lakes with a historically substantial US book, tariffs do not merely shave margin — they force re-routing into a smaller domestic market that Algoma partly monopolises (plate) and partly competes in against Stelco and imports.

The balance-sheet question behind the conversion

A company posting sequential losses of this size through a multi-year capital programme is a liquidity story as much as an operating one. The C$500 million of government liquidity support secured in late 2025 — including C$400 million from the Canada Enterprise Emergency Funding Corporation — is what makes the 2026-27 ramp feasible at all. That backing is also a constraint: public lenders will want visible progress on volume, cost per ton and EBITDA inflection, which places unusual weight on the next two quarters.

What to watch

Three indicators will determine whether Q3 was the trough:

1. Unit Two's learning curve. Unit One took roughly a year to progress from first heat to meaningful contribution; Unit Two benefits from a trained workforce and shared infrastructure, so a faster ramp is plausible — but it is an expectation, not a reported fact.
2. Power reliability at LSP. Any repeat of the August outage would be far more damaging now that no blast furnace provides fallback tonnage.
3. Plate realisation and tariff treatment. With shipments at a fraction of capacity, every dollar of plate premium matters more than the headline steel price. Watch whether the improved C$1,510/t net realisation holds as hot-rolled share returns.

The strategic logic of Algoma's conversion — lower carbon intensity, scrap-based flexibility, a defensible plate franchise — remains intact. The Q3 result simply prices the execution risk that sits between that logic and its realisation. The second EAF is the pivot; whether output and unit costs follow it is the question the next two quarters must answer.