The number worth holding onto in a schedule that spans 15% to 50% is the 35 points between the ends [9], because a range that wide taxes how an invoice is written more than it taxes the metal itself: on a $1,000 cross-border shipment the two ends are $150 and $500, and that $350 gap is a fixed cost that no amount of mill output can absorb [12]. The top of the schedule is more than three times the bottom [10].
That geometry is also why an exemption for critical minerals reads better in a summary than on a purchase order. The source has minerals exempt from some specific tariffs while the metals sector remains significantly affected [4], and those two statements only sit together if relief attaches upstream, near ore and concentrate, while the duty attaches to the rolled, extruded and drawn shapes a fabricator actually orders. Exempting what goes into a smelter does nothing for the buyer of what comes out.
Worth saying plainly what this material does not contain. It is an aggregator account relaying CNBC's finding that the duties could disrupt cross-border flows and raise production costs [2], and it names no product lines and no effective dates [11], which means the volume-weighted rate could sit near the bottom of the range or near the top and nothing here tells you which.
The hedge being reached for is gold [5], and prediction-market pricing has drifted slightly toward higher gold by end-December 2026 [6]. "Slightly" is not a number. A contract resolving that far out prices persistence rather than the next monthly print, which is a different bet from the one a purchasing manager is making this quarter, and the same account describes positioning on inflation as cautious [7].
This is probably wrong, but the first-order corporate response to a 35-point spread is customs work rather than capacity: reclassification, re-routing and redesigning a bill of materials are cheap relative to a new line, and every dollar spent on tariff engineering is a dollar not spent adding tonnage in either country. A counter-thesis worth weighing is that Canadian sellers facing a 50% wall on some lines concede price to keep volume, in which case US input costs move far less than the headline rate implies and the pass-through lands on the seller's margin instead. What would settle it is boring and specific: the line-by-line schedule, and whether landed prices for mill products move by anything close to the duty.
