---
title: "Canada could send the US every tonne of aluminium it makes and still fall short"
description: "Canadian smelters produced nearly 1.6 million tonnes in the first half of 2026. American importers needed 1.68 million. The gap is why a tariff cut moves the price and not the supply."
category: "Markets"
category_url: https://boursel.com/category/markets
author: "Daniel Okonkwo"
published: 2026-08-23T10:37:10.000Z
updated: 2026-08-23T10:37:10.000Z
canonical: https://boursel.com/article/canada-could-send-the-us-every-tonne-of-aluminium-it-makes-and-still-fall-short
tags: ["aluminium", "commodities", "tariffs", "canada", "trade", "morgan stanley"]
---
# Canada could send the US every tonne of aluminium it makes and still fall short

Canadian smelters produced nearly 1.6 million tonnes in the first half of 2026. American importers needed 1.68 million. The gap is why a tariff cut moves the price and not the supply.

The case for cutting the tariff on Canadian aluminium usually rests on an implied premise: that Canada can cover what the United States needs, so the duty is a self-inflicted cost. The production figures do not support the premise, and the arithmetic is worth doing carefully because it changes what tariff relief can be expected to achieve.

In the first half of 2026 the United States imported [1.68 million tonnes of aluminium, about 280,000 tonnes a month](https://www.investing.com/news/economy-news/could-canada-fully-supply-the-us-with-aluminium-4872433). Canada shipped 1.16 million tonnes over the same period, averaging 192,000 tonnes a month, covering about 68 percent of American import needs. Canadian smelters produced nearly 1.6 million tonnes in total across that half year, which is roughly 95 percent of what American importers required.

That last number is the one to sit with. It describes the ceiling, not the current position. If Canada exported its entire output to the United States and supplied no domestic customer at all, it would still leave American importers about 5 percent short.

## And imports are not the whole demand

The shortfall is larger than it first appears, because imports are not the same thing as consumption. The United States relies on imports for [about 80 percent of the aluminium it consumes](https://www.investing.com/news/economy-news/could-canada-fully-supply-the-us-with-aluminium-4872433), the remainder coming from domestic smelters and, separately, from recycled metal.

Put those together. Canada's maximum possible contribution is around 95 percent of imports, and imports are around 80 percent of consumption, so the most Canada could supply is roughly three quarters of American aluminium demand. Self-sufficiency within North America is not on the table at current capacity, whatever the tariff is set at.

## What the tariff actually moves

The duty on Canadian aluminium currently stands at 50 percent, with a reduced 25 percent rate proposed. Because aluminium is a globally priced commodity, a tariff does not show up as a separate charge so much as inside the regional premium.

That premium is worth explaining, because it is where this argument is settled. The London Metal Exchange sets a global price for aluminium, but a buyer in Ohio needs metal delivered in Ohio, not a warehouse receipt. The Midwest premium is the extra amount paid for physical delivery in that region, and it absorbs freight, warehousing, financing and any duty owed on imported metal. When Washington raises a tariff, the premium is where the cost lands, and it is US buyers who pay it.

The premium has been running about 30 percent above what the tariff alone would justify. Morgan Stanley estimates that a cut to 25 percent could see it give up roughly half of that excess, on the order of 10 to 12 cents a pound. That is an analyst's estimate of a price effect, not a forecast to be relied on, and it says nothing about volumes.

Which is the point. A lower tariff would reduce what American manufacturers pay per pound. It would not conjure additional tonnes from smelters that are already running.

## Why the tonnes take years

New primary aluminium capacity is slow in a way that has little to do with permitting appetite. Smelting is an electrochemical process that runs continuously and consumes electricity on the scale of a small city, so a project is only viable where a producer can contract firm power for a decade or more at a price that works, and those contracts are negotiated long before the first pot is lit. The capital cost runs to billions, and a potline that is shut down cannot simply be switched back on: the electrolyte freezes and restarting is expensive enough that idled capacity often never returns.

Emirates Global Aluminium and Century Aluminum are planning a 750,000-tonne-a-year smelter in the United States. Set against a first-half import requirement of 1.68 million tonnes, or roughly 3.4 million tonnes annualized at that rate, a plant of that size would cover something in the region of a fifth of annual import demand once it is running. That is a material change to the balance and it is not self-sufficiency, and the timeline is measured in years rather than in negotiating rounds.

For anyone with aluminium in their cost base, the two questions therefore separate cleanly. The tariff decides the price this year. The smelters decide the supply, and they decide it much later.