There Is No Longer One Aluminum Price
For most of the past two decades, a US foundry could check the London Metal Exchange (LME) aluminum price, add a modest regional premium, and know roughly what the next truckload of metal would cost. That relationship broke down over the course of 2025, and as of 2026 it hasn't come back.
The United States enacted a large Section 232 tariff increase on steel and aluminum imports in March 2025, and the aluminum rate has climbed steadily since: from 10% in early 2025, to 25%, and up to 50% by mid-2025 — where it remains as of April 2026. Layered on top of that, the "Midwest premium" — the extra amount US buyers pay above the LME benchmark for physical delivery — hit a record of roughly $2,182 per tonne in 2026. The combined effect: US foundries now pay an all-in price of approximately $5,340 per tonne, compared with $3,400–3,800 per tonne on the LME. That's a US premium of roughly 40–50% over the global reference price.
Since June 28, 2025, aluminum imports without verifiable smelting or casting origin face a 200% tariff — not a typo. And the exclusion process (which lets specific products request exemptions) has already resulted in over 400 products approved for the 50% tariffs, with the Department of Defense listing castings and forgings as a national security priority area, suggesting further inclusions are likely rather than a near-term rollback.
Why US Aluminum Decoupled From the Global Price
Section 232 tariffs work by making imported metal significantly more expensive for US buyers relative to buyers everywhere else. That constrains the effective supply available to the US market at the global price, and buyers who need metal on a production schedule — a foundry can't simply wait out a pricing dispute — end up bidding up the domestic premium to secure delivery. The Midwest premium is the visible result: it's the market's way of pricing the scarcity that tariffs create.
The practical consequence for foundries is a three-tier cost structure that didn't really exist five years ago: US-only casters paying the full domestic premium, international casters selling into other markets at the global price, and vertically integrated producers who control more of their own supply chain and are partially insulated either way. Which tier a given foundry sits in now matters more for its cost competitiveness than almost any operational decision it makes internally.
The Three Cost Levers Foundries Actually Control
Tariff policy and LME pricing sit entirely outside any individual foundry's control. But industry analysis points to three levers that are fully actionable at the plant level, and all three are seeing renewed attention in 2026:
- Recycled aluminum content. Remelting scrap consumes a fraction of the energy required to process primary metal, which cuts direct energy cost. In markets with carbon border adjustment mechanisms, higher recycled content also carries compliance value beyond the raw cost savings.
- Alloy and casting route selection. Choosing the alloy grade and casting process appropriate to each application — rather than defaulting to a familiar standard — reduces scrap generation and downstream machining requirements per unit produced.
- Defect and inclusion reduction. Every casting rejected for porosity, inclusions, or other defects represents metal that has to be re-melted or scrapped entirely. This is the lever most directly tied to filtration quality at the pour — and it's the one where the tariff environment has quietly changed the math.
Why Defect Reduction Pays for Itself Faster Now
Casting filters exist to reduce inclusion defects — oxide films, dross, and other contaminants that get carried into the mold and compromise the finished part. That function hasn't changed. What has changed is what a unit of avoided scrap is now worth.
If better filtration prevents, say, 500 kg of aluminum per month from being scrapped to inclusion defects, that scrap was worth roughly $1,700–1,900 at the pre-2025 LME-plus-modest-premium pricing many foundries budgeted around. At today's all-in US price of approximately $5,340/tonne, the same 500 kg of avoided scrap is worth roughly $2,670 — a jump of 40-50% in dollar value for identical physical performance. The filtration equipment didn't get better. The metal got more expensive.
This is the part of the tariff story that's easy to miss if you're only tracking headline metal prices: the ROI calculation on any process improvement that reduces scrap — better filters, tighter process control, improved gating design — has shifted in the improvement's favor by roughly the same percentage as the tariff-driven price increase itself. A payback period that used to run 14 months now plausibly runs 9–10 months, without the equipment changing at all.
Outlook: Policy Is the Swing Factor, Not Demand
Industry analysis is fairly consistent on one point heading into 2027: the three-tier cost structure between US-only casters, international casters, and vertically integrated producers is likely to persist as long as the current tariff regime holds, and the policy variable — not underlying demand for castings — is the primary factor that will determine US foundry costs through the remainder of this year and into next. Cautious optimism about the broader $52 billion US foundry industry, driven by tax incentives, infrastructure spending, and reshoring trends, doesn't change the fact that metal cost per tonne is now a function of trade policy as much as commodity markets.
For foundries evaluating capital investment in filtration or process control this year, the practical takeaway is straightforward: the business case for reducing scrap was already there before 2025. It's simply stronger now, and it will likely stay stronger for as long as the tariff structure remains in place.