U.S.-Canada metal tariffs will choke EV and defense supply chains

The metal tariff that looks cheap but isn’t

The United States slapped a 50% tariff on a wide range of Canadian goods. Canada fired back on September 8 with $20 billion in retaliatory tariffs on more than 700 American products.

Markets reacted exactly as you would expect. Steel and materials stocks jumped on Monday — Nucor, Steel Dynamics, Cleveland-Cliffs, Century Aluminum all climbed after trade talks collapsed. The VanEck Steel ETF rose 1.6% that single day. The State Street Materials ETF hit an all-time high.

Then the week ended, and the gains evaporated. XLB closed down for the five-day stretch. SLX flatlined.

Both funds are up significantly year-to-date — SLX over 28%, XLB over 18% — so this was always going to be a bumpy ride regardless. The point is that the headline reaction told you nothing about what actually happens next.

What happened next is where the damage lives.

The border is not a line on a map

Here is what most coverage of this trade war is missing: North American steel and aluminum do not move in one direction. They move back and forth across the U.S.-Canada border multiple times during production.

Canadian primary aluminum goes into U.S. extruders. U.S.-melted steel travels north for finishing. Both come back south inside finished vehicles and appliances. A tariff does not tax that shipment once. It compounds every time the metal re-crosses.

That is the single most important fact in this story, and it is why the auto sector has no winners here.

“For the auto sector, our view is that the sector is so integrated that the tariffs just don’t impact the country you are tariffing but your own country,” said Atsi Sheth, chief credit officer at Moody’s Ratings.

Parts cross the border repeatedly during assembly. Every crossing now carries a tax. That raises the cost of building a car in Michigan, an Ontario plant, or anywhere in between. The tariff on Canadian inputs becomes a tariff on American manufacturing.

Kyle Mohrbach at o9 Solutions put it plainly. The greatest exposure sits in components that are “Canadian-sourced, single-sourced, hard to substitute, or required to keep an assembly line moving.” That includes steel, stampings, powertrain components, braking systems, electronics, specialized subassemblies.

These are not things you replace on a spreadsheet. They are things you have qualified, tested, and embedded in production lines for years.

Defenses and data centers have the same problem

The auto sector gets the attention, but the defense and semiconductor supply chains face the same trap — and they have even less room to maneuver.

Canada supplies a dominant share of the primary aluminum the United States consumes. It also produces significant quantities of critical minerals used in defense applications and electronics. You cannot tariff your dependence away overnight because smelters are enormously capital- and energy-intensive, and new capacity takes years rather than months to build.

Scott Beaulier, dean of the College of Business at the University of Wyoming, said it directly: “The United States remains heavily import-dependent, and Canada has supplied an extraordinary share of U.S. primary aluminum.”

The same logic applies to defense procurement. Military aircraft, ships, and missile systems require aluminum and specialty metals at scales and timelines that do not bend to tariff policy. The Pentagon cannot simply reroute procurement to domestic sources when those sources do not yet exist at the required volume.

Chip manufacturing faces a parallel constraint. The supply chain for semiconductor-grade materials is even more concentrated than auto, and Canada plays a nontrivial role in the North American distribution network. Disrupting it does not create domestic capacity. It creates delays.

The illusion of durable winners

Moody’s Sheth acknowledged that U.S. steel companies have a structural edge simply because the American market is larger. But she also warned that this advantage is narrow and temporary.

“The new tariffs create a meaningful but manageable headwind,” said Angelo Kourkafas, senior global investment strategist at Edward Jones. “A headwind that cuts both ways, as higher steel and aluminum costs also start working through U.S. manufacturers, autos, and construction on this side of the border.”

Dan Luttner, managing partner at NEOS by Argon & Company, pushed further. He argued that the market pop in metals stocks was a headline reflex — mills reprice to replacement cost the moment a 50% wall goes up. Of course the stocks jumped. That does not mean the businesses are durable winners.

“The interesting question is who controls their feedstock inside the wall versus who’s still exposed to it?” Luttner asked.

Nucor runs electric arc furnaces with integrated capacity that never touches Canadian ore or slab. It keeps the price umbrella structurally. Cleveland-Cliffs also fits that description but is in negative territory in 2026 due to ongoing balance sheet stress — so the theoretical advantage has not translated into market performance. Century Aluminum is different again: U.S. primary aluminum capacity is thin, and a lot of what feeds it crosses the border as alumina or semi-finished product. The upside is real, but it is “not immune to the same friction it’s supposed to be protected from.”

And then there are holdings like Freeport-McMoRan, the third-largest position in the XLB at 6.5%. That is a copper and critical-minerals play driven by AI demand, not a pure metals tariff trade. It belongs to a different story entirely.

The timeline no one is talking about

Luttner drew a distinction between the Strait of Hormuz chokepoint and this trade war. Hormuz is geography — oil has nowhere else to go. This is policy, and the volume can reroute, reshore, or get absorbed into price. The catch is the timeline: it takes 12 to 24 months of capital and requalification to do any of those things.

“It’s a slow-motion reallocation,” Luttner said. “So the real story isn’t which stock popped. It’s which manufacturers had already de-risked their supply chain before this week, and which ones are only now finding out how many times their product crosses that border.”

That timeline is critical for two reasons.

First, the 2028 election cycle creates a political window where the costs of this trade war will be fully visible but the remedies will be politically impossible. Automotive plants, defense contractors, and semiconductor facilities are spread across swing states. The pain will be geographic and concentrated, even as the benefits are diffuse and delayed.

Second, companies are not waiting around. Melissa Irmen, director of advocacy for the National Association of Foreign-Trade Zones, said members are already adjusting sourcing decisions and reshaping supply chains. Some warehousing may shift permanently to Canada. Others will restructure origin rules to avoid duties. But these are multi-year decisions made under conditions of extreme uncertainty.

“Companies are not able to make the fast decisions required for the tariff changes,” Irmen said. “Supply chains don’t work that way. We tell our members things will not go back to the way they were pre-2025. Try to look as long term as you can.”

Who wins, who loses, and what happens next

The short list of clear winners is small: domestic steel producers with self-contained feedstock and balance sheets that can absorb the disruption while competitors stumble. Even that list is contested — Cleveland-Cliffs is proof enough.

The losers are far more numerous. Automotive manufacturers on both sides of the border. Defense contractors dependent on Canadian-origin metals. Semiconductor and electronics firms navigating a supply chain that now taxes every cross-border movement. Construction companies already squeezed by material costs. Canadian exporters hit by $20 billion in U.S. tariffs on top of the reciprocal hit to their own economy.

Moody’s Sheth said the rating agency will watch heavy manufacturing, steel, and aluminum closely. Larger companies can usually withstand shocks better, but uncertainty causes the most damage as firms reconfigure long-term options.

“Companies won’t sit on their hands and wait,” Sheth said.

The market priced this trade war in a single afternoon. The real pricing — in raised costs, delayed deliveries, requalified suppliers, and reshored capacity that may or may not materialize — will unfold over the next two years. By then, the political calculus will have shifted, and the industries that absorbed the cost will be counting losses that the original tariff math never accounted for.