Wood product manufacturing

The constraint is fibre, and it is getting worse rather than better. Canfor wound down its Polar sawmill north of Prince George explicitly for 'a shortage of economically available fibre', removing about 140 million board feet over six months, and West Fraser lost $937M on sales that fell to $5.462B from $6.174B under softwood duties, oversupply and soft demand. A small entrant buys fibre in the same market as those two and sells into the same prices, without their tenure. The value-added niches — mass timber components, remanufacturing — are real and they sit downstream of a supply problem no operator can solve alone.

Two businesses share this code and they have opposite years. Wood preservation is the attractive one: Stella-Jones sold C$3,492M of treated poles, ties and lumber in 2025 at an 18.9% EBITDA margin, with utility products alone at C$1,822M and rising [A]. But its customers are electric utilities and Class I railways buying on multi-year supply programmes, its moat is a continental network of treating plants plus pole-quality fibre under contract, and it spent C$259M on acquisitions in 2025 absorbing what adjoins it [A]. A new treating plant is a permitted, capital-heavy site with two or three possible customers per region, each already supplied. Sawmilling is the reachable-looking one and the numbers are worse. The Value-Added Wood Manufacturing record cuts small sawmilling on fibre supply; this screen adds what happens at the border. Interfor lost C$344.4M on C$2,805.9M of sales in 2025, booked a C$69.1M impairment on its Eastern Canadian mills, and has US$663.5M of cumulative duties on deposit; the combined countervailing and anti-dumping rate is 35.16%, with a further 10% Section 232 tariff on top [A]. That is cash paid before the customer pays, on a commodity averaging C$599 per thousand board feet in the fourth quarter. The size bands say this is a plant industry — 199 of 726 establishments employ fifty or more [A] — and a mill is sunk capital whose working capital is then taxed at the border. The small custom and portable mills that make up the bottom of the count sell locally and avoid the duty, and are the subject of the other record.

This is the part of the wood sector with a growth story: mass timber is displacing some concrete and steel in mid-rise buildings, codes have moved to allow it, and engineered members carry more value per cubic metre than lumber. The pre-screen said plant capital, and two anchors show what that means in practice at both ends of the group. At the commodity end, Louisiana-Pacific's OSB segment fell from US$1,184M of sales and US$298M of adjusted EBITDA in 2024 to US$832M and US$7M in 2025 — commodity OSB prices down 26% — and the company guides the segment to breakeven for 2026 [A]. A panel mill is a continuous press line that cannot be run at half speed, and the segment's margin went from a quarter of sales to nothing in one year without the owner doing anything wrong. At the growth end, Structurlam — the best-known Canadian mass-timber producer — built a 280,000 sq ft plant in Conway, Arkansas in 2021 and was sold out of bankruptcy court in 2023 for US$81.1M, the price covering Conway and three British Columbia facilities together [A]. The buyer was Mercer, a pulp company with the balance sheet to wait. The mechanism is the same in both: the plant is sunk years before the order book is known, in OSB because price is set by housing starts, in mass timber because each sale is a building project that can be delayed or cancelled. The Value-Added Wood Manufacturing record treats fibre; this one is about the press. The size bands agree — 237 of 414 establishments employ twenty or more [A]. The reachable exception is the roof-truss plant (321215), which is regional and order-driven; it was not examined here.

A residual group, and the pre-screen was right that it is owner-operator territory: 1,284 of 2,168 establishments — 59% — have fewer than ten employees [A]. The two reachable niches are the pallet and crate shop and the millwork shop. Millwork sells through the same builder and designer relationships as the Custom Cabinet & Millwork Shop record and is cut there on distribution; this screen takes the pallet shop, which is the easier entry — saws, nailers, a yard, and local shippers who reorder every month. The cut is that nothing in a pallet belongs to the maker. The product is a published specification, the customer keeps two or three suppliers and rebids, recycled pallets undercut new ones, and the main input is lumber bought at a commodity price. The one listed company with a pallet business shows the result at scale: UFP Industries' Packaging segment — which owns PalletOne — had US$1,604M of sales in 2025, down 2.0%, with PalletOne's organic unit volume down 8% in the fourth quarter and company-wide selling prices down 2%; UFP as a whole earned an 8.9% adjusted EBITDA margin [A]. That is a company buying lumber by the trainload and acquiring pallet makers to hold volume. A new shop buys the same lumber dearer and sells the same specification to the same shippers. This is a soft cut, not a clean kill: buying an established shop with its customer list is a different proposition from starting one, and a full study would have to test what a local pallet shop actually earns — no such figure was found. Wood windows and doors, prefabricated buildings and manufactured homes are also in this code and were not examined.