Canada’s merchandise trade surplus widened to $4.2 billion in August as shipments to the United States surged 8.1 per cent ahead of Washington’s 50 per cent Section 338 duties. Economists warn the windfall was borrowed from the months that follow.
OTTAWA, Oct. 8, 2026
Canadian exporters and their American customers spent August racing the clock, and the scoreboard released this week shows exactly how hard they ran. Statistics Canada reported on Tuesday that Canada’s merchandise trade surplus widened to $4.2 billion in August from $787 million in July, the sixth consecutive monthly surplus and the largest in more than four years. The surplus with the United States alone climbed to $11.2 billion from $6.1 billion, a one-month swing that the agency and several bank economists described as the largest positive change ever recorded in the bilateral balance.
The number blew past expectations. Market consensus, as tracked by InvestingLive, had pointed to a surplus near $1.55 billion. The actual figure came in at nearly triple that. Yet almost no one in Ottawa or on Bay Street treated the release as good news about Canadian competitiveness. Statistics Canada itself flagged the central caveat in its commentary, noting that new United States tariffs may have prompted businesses to bring shipments forward, so part of the export strength reflects timing rather than underlying demand.
That caveat is the story. August 2026 was the month in which an additional 50 per cent American duty on roughly US$20 billion of Canadian goods took effect, imposed under Section 338 of the Tariff Act of 1930 and applied on top of existing duties, taxes and fees. The measure was signed in a set of proclamations dated July 20, originally scheduled to bite on Aug. 19, then suspended three days and brought into force at 12:01 a.m. Eastern on Aug. 22. American importers had a month of warning and a hard deadline. They used it.
The numbers behind the headline
Exports rose 2.5 per cent month over month in August, rebounding from a 2.6 per cent decline in July. Unusually for a month in which commodity prices moved, the gain was not a price illusion. Export volumes also rose 2.5 per cent, meaning the increase was real goods crossing real borders rather than the same tonnage at higher prices.
Eight of eleven major product groups posted gains. Energy products rose 4.7 per cent, their first increase since April, carried by crude oil, diesel and nuclear fuel shipments. Industrial machinery, equipment and parts climbed 10.1 per cent. Electronic and electrical equipment rose 11.0 per cent. Consumer goods gained 6.6 per cent. The miscellaneous goods and supplies category, a smaller line that can move violently, jumped 43.3 per cent, according to TD Economics.
Imports went the other way, falling 2.0 per cent after an 8.3 per cent surge in July. It was the first monthly decline in seven months. Import volumes slipped 1.1 per cent. Motor vehicles and parts led the retreat, down 8.8 per cent as the category pulled back from a record high set in July. TD Economics also pointed to a 15.5 per cent drop in metal ores and non metallic minerals, which it attributed largely to the volatile movement of unwrought gold, silver and platinum. Other summaries of the release recorded a 7.0 per cent fall in metal and non metallic mineral products, a related but distinct series.
The currency added a wrinkle. The Canadian dollar averaged roughly 1.1 United States cents higher in August than in July. Measured in American dollars rather than Canadian ones, exports rose 4.0 per cent and imports edged down 0.6 per cent, a reminder that the headline Canadian dollar figures understate the real movement of goods priced in the United States currency.
The geographic split was the most revealing part of the release. Exports to the United States rose 8.1 per cent while imports from the United States fell 2.5 per cent. Exports to all other destinations fell 8.5 per cent, giving back a record set in July. In other words, in the month Canada’s largest customer raised the cost of entry, Canadian goods flowed toward that customer at a faster rate than at any point in the recent record, while the diversification trade that Ottawa has spent two years promoting went into reverse.
What front running actually is
Front running, sometimes called pull forward or tariff arbitrage, is one of the most predictable behaviours in trade policy. When a duty has a known effective date, the rational response from an importer is to land as much inventory as possible before that date, because every unit cleared beforehand carries the old duty rate for the life of that inventory.
Mechanically, the surge shows up on the Canadian side as an export boom and on the American side as an import boom, and in neither case does it signal demand. It is a shifting of transactions across the calendar. The inventory that arrived in American warehouses in August is inventory that will not be ordered in September, October or November. The term economists use, borrowing from the language of fiscal policy, is intertemporal substitution.
Marc Ercolao of TD Economics put the point directly in the bank’s note on the release, writing that some of August’s strength borrows from future activity. TD expects a payback in September and beyond as the tariffs, Canadian countermeasures and new American import restrictions all take hold. With two months of third quarter data in hand, the bank judged that net trade was tracking toward a slight boost to growth, while warning that the trade outlook is clouded and that risks are tilted to the downside. TD also wrote that it considers a major negotiating breakthrough before year end unlikely.
There is a second, less obvious distortion embedded in the August data. Because the Section 338 duty applies to the full entered value of covered goods and stacks on top of antidumping, countervailing and other applicable duties, the financial incentive to beat the deadline was unusually large. A shipment worth $1 million that cleared on Aug. 21 carried no additional charge. The same shipment on Aug. 22 carried $500,000 in extra duty. Few trade measures create so sharp a cliff, and few produce so clean a natural experiment in importer behaviour.
How Canada got here
The legal architecture of the current dispute is unusual, and it matters for how the next few months play out.
For most of 2025, American tariffs on Canada rested on emergency economic powers. That foundation collapsed on Feb. 20, 2026, when the United States Supreme Court struck down the tariffs imposed under the International Emergency Economic Powers Act in Learning Resources, Inc. v. Trump, invalidating the 2025 duties on Canada, Mexico and global imports. Tariffs grounded in the Trade Expansion Act of 1962, covering steel, aluminum, copper, autos and lumber, survived the ruling.
What followed was a steady substitution of one legal authority for another. On Feb. 24 the United States imposed a temporary global import surcharge of 10 per cent under Section 122 of the Trade Act of 1974, limited to 150 days without congressional approval and excluding goods compliant with the Canada United States Mexico Agreement. On April 6 metal tariffs were restructured to apply to the full value of steel, aluminum and copper articles and derivatives rather than to metal content alone, at rates ranging from 10 to 50 per cent. On June 8 Washington adjusted the metal regime again, cutting agricultural machinery and some heating and cooling equipment to 15 per cent and setting CUSMA compliant goods at 25 per cent on non American content with a total minimum of 15 per cent, a structure legislated to remain in force until Dec. 31, 2027.
On July 1 the CUSMA joint review concluded without the sixteen year extension that Canadian negotiators had sought. The agreement now runs to 2036 subject to annual reviews, a far shorter horizon for investment planning than the one Canadian manufacturers had assumed.
Then came Section 338, a provision of the Smoot Hawley Tariff Act of 1930 that had sat essentially unused for most of a century. It permits duties of up to 50 per cent where a trading partner is found to discriminate against American commerce, and it permits outright exclusion of that partner’s products if the discrimination persists. Three proclamations issued July 20 targeted dairy, alcoholic beverages and motor vehicles, with findings that cited Canada’s cheese tariff rate quota practices and provincial restrictions on American alcohol. In practice the scope ran well past those headline categories, reaching wine, hockey sticks and cement, among other goods.
Canada responded on Sept. 8 with counter tariffs of 15, 25 and 50 per cent on roughly $27.6 billion of American goods across 629 tariff lines, covering steel, dairy, appliances and a long list of consumer items. Washington answered three weeks later. On Sept. 29 import bans took effect on roughly US$1 billion of Canadian alcoholic beverages, motorcycles, molasses, whey and dairy, enforced not through duties but through outright refusal of entry. Covered goods can no longer be admitted to a foreign trade zone, entered into a bonded warehouse, moved in bond or entered for consumption.
Hanging over all of it is the threat of a 50 per cent tariff on Canadian cars, trucks and auto parts beginning Jan. 1, 2027.
Reaction
The response in Ottawa has been to treat the August figures as a warning rather than a win. Statistics Canada’s own commentary was cautious, attaching the front loading caveat directly to the release rather than leaving it to analysts.
Private sector economists were blunter. TD’s note framed the result as borrowed activity and projected a reversal. The bank argued that the direct macroeconomic impact of the recent measures should be modest given their targeted scope, while cautioning that prolonged uncertainty could delay investment and hiring and raise costs for businesses and consumers. That second effect, the uncertainty channel, is the one the Bank of Canada has leaned on most heavily in its own analysis.
In its July 2026 Monetary Policy Report, the central bank put the average American tariff rate on Canadian goods at about 5.0 per cent, down marginally from 5.1 per cent in April, and Canada’s average rate on American goods at 1.5 per cent after remissions. Those averages look small, and they are, because the exemption for compliant goods still shields the bulk of bilateral trade. The bank’s April analysis made the asymmetry explicit. Industries facing sectoral American tariffs account for roughly 1 per cent of Canadian output and employment but around 15 per cent of Canadian exports. The damage is narrow and deep rather than broad and shallow.
The bank also assumed that American trade policy would be one reason potential output growth eases to 1.1 per cent in 2026 before recovering to 1.3 per cent in 2027 and 1.5 per cent in 2028. That is the quiet cost of the dispute. It is not a recession. It is a slower ceiling.
Economic impact analysis
Three conclusions follow from the August release, and none of them are comfortable.
The first is that the headline surplus is close to meaningless as a signal of Canadian export health. A trade balance that widens because a customer is stockpiling before a price increase is not evidence of competitiveness. The honest reading of the month is that Canadian producers and American buyers jointly pulled forward perhaps several hundred million dollars of activity to avoid a duty, and that the bill for doing so arrives in the autumn data.
The second is that the sectoral pattern inside the aggregate matters more than the aggregate. Energy, which is largely exempt from the Section 338 regime and which moves on pipeline capacity and global pricing rather than tariff schedules, contributed a 4.7 per cent gain. Machinery and electrical equipment, categories with substantial CUSMA compliant content, posted double digit increases. Those are different stories with different durability. The Bank of Canada’s April assessment documented how uneven the damage has been across tariffed sectors, with steel exports roughly halved, softwood lumber running about 20 per cent below 2024 averages as of February, aluminum falling 50 per cent below 2024 levels by July 2025 before recovering more than half of the loss, and copper running about 40 per cent above its 2024 average.
The third is that the diversification story took a bad month. Exports to destinations other than the United States fell 8.5 per cent in August after reaching a record in July. One month is not a trend, and the July record is itself evidence that the diversification push has produced real volume. But the August reversal illustrates the gravitational problem that Canadian exporters have described to the central bank in consultations: transportation costs and proximity make it hard to redirect goods away from the American market, even when the American market is charging a premium for entry.
For third quarter gross domestic product, the arithmetic is straightforward and temporarily flattering. Net trade is tracking toward a modest positive contribution. For the fourth quarter, the same arithmetic runs in reverse, with the added drag of counter tariffs raising input costs for Canadian importers and the September import bans removing certain shipments from the trade accounts entirely.
What it means for importers and exporters
For Canadian exporters, the practical lesson of August is that the tariff calendar is now a planning variable on par with demand forecasting. The Jan. 1, 2027 threat on autos, trucks and parts is the next hard date on the board, and the same incentive structure that produced the August surge will produce another one in the fourth quarter if that measure proceeds. Firms in the automotive supply chain should be modelling a pull forward of orders into November and December and a corresponding hole in the first quarter of 2027.
Exporters of goods already covered by Section 338 face a different calculation. Drawback remains available on the additional duties, which is a meaningful cash recovery for goods that are subsequently exported from the United States. Goods entered under Chapter 98 of the American tariff schedule are generally outside the duty, with listed exceptions where it applies only to the value of repairs, processing or assembly. Goods already subject to Trade Expansion Act tariffs are excluded from the additional 50 per cent charge, so classification work that establishes which regime governs a given product is worth real money. Energy products, potash, certain civil aircraft, fish and critical minerals sit outside the measure.
CUSMA origin, it bears repeating, does not exempt a good from the Section 338 duty or from the import bans. That is a reversal of the intuition that carried Canadian exporters through 2025, when compliance with the agreement was the reliable shield. Compliance still matters enormously for the Trade Expansion Act tariffs and the Section 122 surcharge. It does not help under Section 338.
For Canadian importers, the counter tariff regime that took effect Sept. 8 is now the binding constraint on landed cost. Rates of 15, 25 and 50 per cent across 629 tariff lines reach deep into industrial inputs as well as consumer goods, and the remission process is narrow. Importers who have not reviewed their American sourcing against the counter tariff schedule since September are almost certainly paying duty they could in some cases avoid through substitution or through a remission application.
For businesses on both sides of the border, the September import bans introduce a risk that duty planning does not address. A banned good is not an expensive good. It is a good that cannot enter. Goods imported before 12:01 a.m. on Sept. 29 may still be entered, and goods already sitting in a bonded warehouse or foreign trade zone may be withdrawn at the 50 per cent rate, but new shipments are refused. Customs and Border Protection has been rejecting non compliant entries and cancelling unreleased filings in its automated system. Any Canadian exporter in the covered categories needs to confirm the status of in transit inventory before it reaches a port.
Outlook
The most useful thing the August data provides is a timestamp. It marks the last clean month before the full architecture of the current dispute, American duties, Canadian counter tariffs and American import bans, was simultaneously in force. September will be the first month that shows all three operating together, and it will also carry the payback from August’s pull forward. The two effects point the same direction.
Statistics Canada will publish the September figures in early November. Economists expect them to be ugly, and the market will have to separate the genuine deterioration from the mechanical reversal of August’s surge. The honest answer is that a single month will not permit that separation cleanly. The fourth quarter will.
What the August release does establish is that the trade relationship is still functioning, still enormous and still overwhelmingly oriented toward a single customer. Canadian exporters moved more goods into the United States in August than in any comparable month of the dispute, because the price of waiting was 50 per cent. That is not resilience. It is a stampede ahead of a closing gate, and the pasture on the other side is smaller than it was.
The small exporter problem
Lost inside a national trade surplus is the fact that the capacity to front run a tariff is not evenly distributed. Pulling shipments forward requires a customer willing to carry inventory, warehouse space to put it in, and working capital to finance the gap between production and payment. Large exporters with established American distribution had all three in August. Small and medium sized firms frequently had none of them.
The Canadian Federation of Independent Business has documented the asymmetry through this dispute, reporting that a substantial share of small exporters expected serious revenue damage from the 50 per cent measures, with a significant minority anticipating revenue declines of half or more. For those firms, August was not a windfall month. It was the month their larger competitors cleared inventory at the old rate while they continued shipping at the new one.
That distributional point is invisible in the aggregate data and central to the policy response. Remission, retraining and liquidity support are the instruments that reach the firms the surplus did not.