Canadian exports of lumber and other sawmill products have climbed 13.7% from their March low, and price and volume are rising together. The 21.4% half-year fall still being quoted is arithmetically right and describes a first quarter the market has spent the second quarter climbing out of.
Every figure here comes from two Statistics Canada tables that anyone can open, and both are linked at the end. We pulled them through the agency's own data service rather than retyping them from a summary. Where we could not check something, we have left it out and say so at the close.
The year opened badly and hasn't stayed that way
Measured on export value against the same quarter a year earlier, the first quarter of 2026 came in 32.8% down. The second came in 9.0% down. June on its own was 1.7% below June 2025.
That is a different market from the one the half-year headline describes, and the gap between the two matters commercially. An operator who set a budget assumption in the spring, when the only reading available covered the worst quarter in the series, is carrying a view of demand that the last three months have overtaken. The half-year number isn't wrong. It's just old by the time it's published, and it's the one that gets quoted.
The recovery isn't tidy, and it would be easy to oversell. March was worse than February. May was fractionally worse than April, and seasonally adjusted value dipped that month. What holds is the shape across a quarter, not a clean climb month by month.
Price and volume are rising together, which is what makes this demand
Value on its own can't tell you what happened, because value is price multiplied by volume and a fall in either produces a fall in the total. The agency publishes the split as two indexes for the same product class, so this is arithmetic rather than inference.
From the March low to June, the export price index rose 8.4% and the export volume index rose 5.0%. Running the same pair backwards, from a June 2024 base to that March low, price fell 14.3% and volume fell 10.7%.
Both directions point the same way. If the binding constraint were fibre access or closed capacity, volume would fall while price rose, because less wood chasing the same buyers is a supply story. Instead the two moved down together and have turned up together, which is what a demand cycle looks like.
For anyone selling into this, the distinction is practical rather than academic. Rising volume with falling price is share bought with discount, and it flatters the order book while emptying the margin. Rising volume with rising price means buyers came back. Only one of those justifies putting a shift back on.
The tariff effect isn't in these numbers yet
Duties on Canadian softwood remain in force, and it would be convenient to read the recovery as evidence they aren't biting. The data won't carry that claim, and it's worth being precise about why.
From the exporting side, a duty and weak demand look identical. A duty raises what the buyer pays and lowers what the mill receives, so the exporting country records a lower price and a lower shipped volume. Softer demand produces exactly the same two movements. Nothing in a Canadian export series can separate them.
What the last quarter does say is narrower. Netback price has been rising since March, and that isn't the shape of a burden getting heavier. It isn't evidence that the burden is absent, and this series can't size the level of it at all. The figure that would settle the question is the gap between what the buyer pays and what the mill receives, because that gap is the duty. Neither side of it appears in this table.
This series does move on policy, which is why it is worth watching
July 2025 sits well above the months either side of it. The volume index reaches 114.9 while price barely moves, and the following month volume drops to 88.1. Flat price with a volume spike and an immediate hole behind it is shipments being pulled forward to beat a date, then a lull while the pipeline refills.
We know how easy that is to misread, because we misread it first. Measuring from July to August gives a 24% drop in a single month and a tidy story about a step change. The step isn't there. Measured from June, either side of the spike, the move is about 11%.
The useful part is what it proves about the instrument. A trade policy date shows up in this data inside a month, in shipment timing before it shows up in price. So when the next change lands, expect the same shape: a pull-forward month, a hole behind it, then the new level. The first two readings after any policy date are the least informative in the series and they are the two that get written about. The third is the one worth waiting for.
The cadence supports that. These are monthly figures running about five weeks behind, with June available in the first week of August. A quarterly review will miss a turn by a full quarter. Monthly is the right frequency for a series that moves on announcements.
What we build for this
The reason this took an afternoon rather than a week is that the split between price and volume already existed in the source. Most operators don't have that. A sales line arrives as one number per month, and when it moves, the argument about why gets settled by whoever is most confident in the meeting.
The pattern we build is a revenue line carrying price and volume as separate measures from the start, at the grain the business actually trades on, so a fall can be attributed instead of described. The outcome that matters isn't a chart. It is that the conversation moves from whether the drop is real to which grades and which customers it sits in, and that question has an answer. You can see how the model is assembled in how it works.
Sources
- Statistics Canada, table 12-10-0163, international merchandise trade by commodity, monthly. Export value, customs basis, lumber and other sawmill products, unadjusted and seasonally adjusted, to June 2026. Canadian dollars, all destinations.
- Statistics Canada, table 12-10-0168, trade price and volume indexes, monthly. Export price and volume indexes for the same product class, seasonally adjusted, to June 2026.
- Lesprom, 6 August 2026. The half-year softwood lumber figures now in circulation, quoted here as the reported position rather than as a source we have verified.
Left out on purpose. This product class is broader than softwood lumber, it is measured in money rather than cubic metres, and it counts every destination, so it is not the same series as the softwood-to-one-market figures in the trade press. We could not build that narrower comparison, because the public tables carry commodity detail and destination detail in different places and never together, which needs the customs database at tariff-line level. We have not verified the half-year softwood figures themselves, and the source for them does not state which currency its values are in. We have not separated the duty from weaker demand, for the reason given above. And there is no customer figure here, because none of ours was cleared for it.