First-Ever Section 338 Tariffs (July 2026): 50% on $20bn of Canadian Goods, USMCA Preference Overridden — Why the Loonie Only Fell 0.6%
The 2026 USMCA review has produced its first hard consequence. On 20 July 2026 the White House issued three proclamations invoking Section 338 of the Tariff Act of 1930 — an authority unused for roughly three-quarters of a century — imposing an additional 50% duty on nearly $20bn of Canadian goods from 12:01 a.m. Eastern time on 19 August. Critically, the duties apply whether or not a shipment qualifies for USMCA preference, which is why Prime Minister Mark Carney called them a direct violation of the agreement. And yet the Canadian dollar fell only about 0.6%, from 1.4014 on 17 July to 1.4095 on 21 July. That gap between a 50% headline and a 0.6% currency move is the whole lesson: tariffs reach a currency through fundamental channels, and this one landed on a narrow slice of them.
This is a textbook case of why a fundamental currency-strength read beats a price-only one. A chart shows a barely-changed loonie and tells you the market shrugged. It cannot tell you why: that the measure covers roughly 5% of Canada's exports to the US, that energy — the loonie's commodity channel — was explicitly carved out, and that the currency's dominant driver right now is a 125–150 basis point carry gap that a tariff on 5% of trade cannot overturn. Trade policy doesn't move a currency directly; it moves it through channels, and the size of the move depends on which channels it touches. Here is the map.
- Three proclamations on 20 July 2026 invoked Section 338 of the Tariff Act of 1930 — its first use as a tariff instrument — for an extra 50% duty on Canadian dairy, alcoholic beverage and motor vehicle goods plus a long tail of consumer items, effective 19 August.
- Exposure is about $20bn, roughly 5.2% of the $382bn of goods the US imported from Canada in 2025. Energy, potash, critical minerals and fish are excluded.
- USMCA preference does not exempt affected goods — the first measure of the 2026 review dispute to override the agreement rather than sit beside it.
- The loonie's response was small: USD/CAD 1.4014 (17 July) → 1.4054 (20 July) → 1.4095 (21 July), a one-week low and about a 0.6% CAD decline on Bank of Canada daily averages.
- Reason: the rate factor still dominates. A Bank of Canada rate of 2.25% against a Fed range of 3.50–3.75% outweighs a narrow-based tariff — and the excluded categories are exactly the ones that feed the commodity factor.
- No federal retaliation yet; Carney says he and President Trump agreed to "intensify discussions" before the 19 August start date.
- See how the growth, rates, risk and commodity factors are scoring CAD right now on the live meter.
What actually happened: the first-ever Section 338 tariffs
Section 338 lets the President "offset any burden or disadvantage placed on the commerce of the United States by an unequal imposition or discrimination by a foreign country by specifying and declaring additional duties." It has sat dormant for decades. On 20 July 2026 it was activated against a single trading partner, with the White House citing discriminatory treatment of US alcohol, automobile and dairy products, alongside a US goods trade deficit with Canada of $46.4bn in 2025.
The headline categories are dairy, alcoholic beverages and motor vehicle products, but the annexed lists run much wider — wine, cement, plywood, furniture, clothing, wigs, seeds, fishing rods, swimming pools and hockey sticks among them. The duty is an additional 50%, stacking on existing rates, and it bites on goods entered for consumption or withdrawn from warehouse on or after 12:01 a.m. ET on 19 August 2026. Excluded: oil and gas, potash, critical minerals, fish, and products already carrying sectoral tariffs. Contemporaneous coverage of the announcement and Canada's reaction is available from Al Jazeera and NPR; the process background sits with the Office of the US Trade Representative.
This is the escalation the 1 July review left room for. At that mandatory six-year review under Article 34.7, the US declined to confirm a 16-year extension of USMCA while Canada and Mexico backed it, which moved the pact into an annual review cycle running to 2036 rather than terminating it. The agreement stayed in force — and three weeks later a measure arrived that runs straight through its central promise.
Why "USMCA preference doesn't help" is the important sentence
For a currency, the tariff rate is less informative than the architecture. Two features of this action matter more than the 50%.
First, the duties apply regardless of USMCA origin qualification. Until now, the working assumption for exporters and for investors was that compliant goods travel duty-free and the disputes happen at the margins — steel, aluminium, autos, lumber. A measure that reaches qualifying goods changes what the agreement is worth as insurance, and insurance value is precisely what a risk premium prices.
Second, the legal route is novel. Reaching for a 1930 authority that has no modern precedent means the boundaries of its use are untested, which is a different kind of uncertainty from a known instrument applied at a known rate. Markets can price a tariff. They struggle to price the size of a toolbox.
The loonie's answer: 0.6%, not 6%
The Canadian dollar's reaction was orderly. On Bank of Canada daily average rates, USD/CAD sat at 1.4014 on 17 July — a one-month high for the loonie — then moved to 1.4054 on announcement day, 1.4095 on 21 July, and 1.4093 by 24 July. That is a Canadian dollar decline of roughly 0.6% over the week and a one-week low, worth about 70.97 US cents at the trough. Official daily rates are published by the Bank of Canada.
| Date (2026) | USD/CAD (BoC daily avg) | Context |
|---|---|---|
| 17 July | 1.4014 | Loonie's best level in about a month |
| 20 July | 1.4054 | Section 338 proclamations signed |
| 21 July | 1.4095 | One-week CAD low as details land |
| 24 July | 1.4093 | Holds the level into month-end data |
Under a price-only lens this looks like indifference, or worse, like a market that missed the story. Under a factor lens it reconciles cleanly. The tariff hits the growth channel — but on a slice of trade small enough that Canada's Q2 growth tracking near the Bank of Canada's 2.5% annualised estimate is not meaningfully re-cut by it, with roughly four-fifths of exports still crossing duty-free. It leaves the commodity channel untouched by design, because energy and potash are exempt and Brent has been trading in the high-$90s. It does widen the risk channel. And it does nothing at all to the channel that is actually setting this exchange rate.
The channel that is in charge: rates
The Federal Reserve's target range is 3.50–3.75% with a hawkish tilt; the Bank of Canada's policy rate is 2.25%. That 125–150 basis point carry disadvantage is why USD/CAD has been pinned near 1.41 through a stretch of improving Canadian data, as we set out in the Canada GDP preview, and it echoes the dynamic behind the loonie's one-year low. A tariff announcement that touches 5% of exports arrives as a second-order input to a currency whose first-order input has not changed.
| Fundamental channel | Section 338 read for CAD | Direction |
|---|---|---|
| Interest rates (BoC vs Fed) | 2.25% vs 3.50–3.75%; unchanged by the tariff | Dominant headwind |
| Growth / trade outlook | ~5.2% of exports to the US exposed; four-fifths still duty-free | Modest headwind |
| Risk sentiment | USMCA preference overridden; novel legal authority | Persistent headwind |
| Commodities (oil) | Energy, potash and critical minerals excluded; Brent high-$90s | Neutral — carved out |
| Positioning | Speculative flows already cautious on the loonie | Context-dependent |
What Canada does next, and how each path scores
Because the duties do not bite until 19 August, the window between now and then is the fundamental variable — not the proclamation itself. Three broad paths, each landing on different factors.
- Negotiated settlement before 19 August. Carney said he had spoken with President Trump and that the two "agree to intensify discussions", with a comprehensive agreement the stated first objective. A deal that suspends or narrows the duties removes the risk-premium increment and leaves the rate gap as the sole story — mildly CAD-supportive, but only mildly, because it restores the prior status quo rather than improving it.
- Duties take effect, no retaliation. The growth drag is real but small and slow; the risk premium stays embedded. Broadly the current market pricing, which is why the move so far has been measured.
- Retaliation spiral. No federal counter-measures have been announced as of 26 July, though Carney has said he would consider all options and Ontario Premier Doug Ford has pushed for a dollar-for-dollar response. This is the scenario that would matter for the currency, because tit-for-tat escalation widens the risk premium, raises the odds that the Bank of Canada leans dovish to cushion growth, and therefore threatens to hit the rate channel too — the one that actually sets the price. Watch for whether the exempt categories stay exempt: energy entering the dispute is what would pull the commodity factor in.
The takeaway
The instinct on a "50% tariff" headline is to expect a currency crisis, and the instinct is wrong here for reasons a price chart cannot show you. The measure reaches about 5% of Canada's exports to the US, carves out the energy complex that drives the loonie's commodity factor, and leaves untouched the 125–150 basis point rate gap that has been the binding constraint on CAD all month. So the loonie fell 0.6%, not 6%.
What the action did change is the risk factor, and in a way that does not decay quickly: the 1 July review left the extension question open every year to 2036, and 20 July demonstrated that the window has teeth — including for goods that qualify under the agreement. That is the kind of driver a fundamental model carries as a standing premium rather than a one-day repricing. Read the channels separately and the apparent contradiction between a dramatic headline and a quiet currency stops being a puzzle: the headline was loud on a channel that was carved out, and silent on the channel in charge.
To learn how PIPTHEORY builds its fundamental currency-strength scores from five factors, see the methodology overview. For the case where crude was the channel driving CAD, see the oil-shock breakdown.
Educational macro context only — not investment advice.