Wood Product Manufacturing (United States) — NAICS 321

A Histometrics rollup primer for public-market and private investors. NAICS = North American Industry Classification System, the U.S. government's standard scheme for sorting businesses by what they make; codes get more detailed as they get longer. Code 321 is a three-digit subsector that rolls up three four-digit industry groups: 3211, 3212, and 3219. Figures are marked to U.S. federal statistics; forward-looking statements are judgments, not facts.

---

1. Overview

NAICS 321 is the "turn a tree into a product" subsector — everything between the forest and the building site or shipping dock, but not the forest itself (growing and cutting trees is Forestry and Logging, NAICS 113) and not paper (that is NAICS 322). It is a big, quiet, deeply fragmented, overwhelmingly private slice of American manufacturing: about $172 billion of shipments, 432,000 workers, and 13,900 plants in the 2022–23 federal data.[1]

The single most useful way to read the subsector is as a rough value chain, from log to finished good:

• 3211 — Sawmills and Wood Preservation (cut the log): breaks logs down into lumber and, for the share headed outdoors or underground, forces preservative into it.
• 3212 — Veneer, Plywood, and Engineered Wood Product Manufacturing (press the panel): glues and presses veneers, chips, and fibers into panels and load-bearing members.
• 3219 — Other Wood Product Manufacturing (make the finished thing): converts that lumber and those panels into windows, doors, flooring, pallets, trusses, and manufactured homes.

Everything in 321 shares three traits: all of it buys wood as its main input; all of it is a conversion (spread-plus-throughput) manufacturer rather than a store, a landlord, or a fee business; and all of it rises and falls with the construction and industrial cycle, gated by interest rates. That common spine is why 321 hangs together as one subsector.

Two structural notes shape how deep you have to drill. First, the children are not equally deep: 3211 and 3212 each contain exactly one five-digit industry (32111 and 32121), so at the four-digit level they are pass-throughs and their real internal variety appears only at six digits — sawmills versus treating plants inside 3211, and four distinct panel and member markets inside 3212. Only 3219 fans out at the five-digit level, into millwork, pallets, and "all other wood."[2][3][4] Second, and more important for investors, the subsector's famous fragmentation dissolves as you descend: the revised child data now show real product markets inside 321 running from near-perfect fragmentation to genuine oligopoly.

The reason to study the rollup rather than a single number is the contrast across the three children — they differ on size, capital intensity, concentration, which macro pulse they ride, who owns them, and where (if anywhere) you can buy a public stock. The blended $172 billion / 76-HHI average describes no market anyone actually competes in. This page leads with that contrast, then treats the subsector as a whole.

---

2. What's inside — the three children and how they differ

All three convert wood; as businesses they pull apart on almost every axis. Shares are computed from federal receipts (2022 Economic Census); plant and job shares and concentration are from the same federal sources. CR4 is the combined revenue share of the four largest firms; HHI (Herfindahl-Hirschman Index) is a standard 0–10,000 concentration gauge where anything under 1,500 is "unconcentrated."[1]

How to read the contrast (the five things that matter):

• Center of gravity is downstream — and so is the pay cut. 3219 alone is ~46% of the subsector's revenue, ~59% of its jobs, and ~68% of its plants — because it is where thousands of small millwork, pallet, and craft shops live. Move left-to-right along the chain (log → panel → finished good) and revenue per worker falls hard (~$565K → ~$485K → ~$311K), with average pay following the same slope in reverse order of capital intensity: ~$63,000 in panels, ~$61,300 in sawmilling and treating, ~$54,000 downstream.[1][2][3][4]
• The scale peaks are one level below where the parent table shows them. At four digits, 3212 runs the biggest plants (~$28M each). But the children reveal the true extremes sit at six digits and straddle two children: softwood plywood mills average roughly $83 million of shipments each across just 80 sites, and wood-preservation plants about $29 million across 362 — while engineered-member plants (mostly small local truss shops) average about $20 million and sawmills about $15 million. Capital intensity in 321 is a product-market property, not a child-level one.[2][3]
• "Unconcentrated" is an artifact that dies at depth. On paper every child is unconcentrated (HHI 167 / 247 / 99) and the subsector is the loosest of all (HHI 76). One level down, real markets range from HHI 88 (pallets — nobody can dominate it) to ~1,036 (softwood plywood), with wood preservation at 927, reconstituted wood at 883, wood windows and doors at 718, and manufactured homes where three firms build 86.5% of the units. The subsector HHI, and largely the child HHIs, describe no market anyone competes in.[1][2][3][4][19]
• Which macro pulse matters is the sharpest divide, and it is now observable. Most of 321 rides housing — new starts plus repair-and-remodel, gated by mortgage rates. Two corners carry ballast against a housing slump: wood preservation (~$10.5B inside 3211) tracks infrastructure replacement — aging utility poles and railroad ties swapped on schedule regardless of housing — and pallets (~$18.3B inside 3219) track freight and industrial production. The two cycles visibly ran out of phase: pallet output rose ~34.9% in 2021 and ~33.5% in 2022, then fell ~13.9% in 2023, while homeowner remodeling spending kept grinding higher.[2][4]
• The chain is partly its own customer. The single biggest buyer of 3211's output sits inside 3219: pallet manufacturing consumes roughly 38% of all sawn lumber used in the United States. A freight-driven destocking cycle therefore reaches back up the chain into sawmills, and a housing recovery is not the only thing that fills a mill's order book.[4][22]
• Public investability is uneven, indirect, and thinning. There is no listed pure-play for the subsector, and none for two of its three children (nor for pallets inside the third). The cleanest listed bets sit inside 3219 (manufactured-home makers) and on the sawmill side of 3211 (timber REITs). The roster is also shrinking: PotlatchDeltic disappeared into Rayonier when their merger closed in January 2026, Masonite and PGT Innovations were both taken private in 2024, and Enviva's 2024 bankruptcy removed the listed wood-pellet route. Section 4 maps what is left.[3][4][6][16]

The one-sentence contrast: 3211 is the cyclical, commodity log-breaker with a defensive, genuinely concentrated infrastructure sidecar; 3212 is the capital-intensive panel oligopoly wearing an unconcentrated label; 3219 is the huge, fragmented, mostly private finished-goods tail with two real oligopolies buried in it — one vertical chain that rarely fires on all cylinders at once.

---

3. How big it is (the rollup, federal figures)

Our ground-truth federal statistics for NAICS 321 as a whole:[1]

The children add up cleanly. The three children sum to the subsector almost to the unit on the hard counts — receipts ($51.8B + $41.16B + $79.50B ≈ $172.5B), establishments (3,053 + 1,456 + 9,391 = 13,900, exact), and employment (91,690 + 84,946 + 255,619 = 432,255, exact) — and payroll matches within rounding ($5.62B + $5.31B + $13.88B ≈ $24.82B).[1][2][3][4] Firm counts sum slightly higher than the subsector total (2,677 + 1,026 + 8,457 = 12,160 vs. 12,029) because a company active in more than one child is counted once at the subsector level but in each child it operates in — a small, expected overlap of ~130 firms (Weyerhaeuser and UFP Industries are the textbook examples), not an error. Note also that the derived subsector average of ~$12.4 million per plant matches no child: they run ~$17M, ~$28M, and ~$8.5M.

Concentration — and why the whole is less concentrated than its parts, at every level. At the subsector level the four largest firms hold just 12.7% of receipts (CR4), the top eight 19.7%, the top 20 32.7%, the top 50 46.1%, and the HHI is 76.2 — extraordinarily low.[1] That is lower than any individual child (HHI 167 / 247 / 99), and the revised children now show the same pattern repeating one rung down: 3212's HHI of 247 is lower than every one of its four products (341 / 579 / 883 / 1,036), and 3219's HHI of 99 is lower than two of its three children.[3][4] This is a classic aggregation artifact compounding at each level — combining separate industries dilutes any one firm's share of the enlarged whole. A company that dominates OSB panels is a rounding error across all of pallets, windows, and manufactured homes; a company that dominates utility-pole treating (where CR4 is 54.9%) vanishes inside a sawmill industry ten times its size. The subsector HHI describes no real market, and for freight-limited products it is worse than useless — panels, pallets, and trusses compete inside a delivery radius (~150–250 miles), not nationally.[3][4] This is the central reason to read the rollup child-by-child, and then product-by-product, rather than off the headline number.

Undercount and scope caveats — the factory dollars are solid; the edges and the surrounding economy are not. Because these are real factory industries with fixed plants and paid employees, the ~$172 billion of shipments is reasonably reliable — this is not a case like logging or many services where informal, sole-proprietor activity distorts the totals. No suppressed values are relied on here. But five real gaps apply:

• Material scope only. NAICS classifies by material, so 321 captures only the wood version of each product. Plastic pallets, vinyl and aluminum windows (vinyl is today the most common U.S. residential window frame), and metal containers all sit outside the subsector — even when made by the same companies on adjacent production lines.[3][4]
• Imports are excluded — and in one corner they are the market. Federal statistics count domestic production at the factory gate. On the hardwood side that omission dominates: a U.S. International Trade Commission analysis of 2021 apparent U.S. hardwood-plywood consumption put imports at 92.7% of volume and 84.0% of value, and U.S. hardwood-plywood imports ran roughly $490 million in Q1 2025 alone (+18% year over year) against that product's ~$3.5 billion of domestic shipments. Read 321's dollars as domestic making, not domestic consumption.[3][23]
• Small and individual owners are undercounted — concentrated in 3219 (and a fringe of 3211). Pallets (an informal repair-and-recycle economy of small yards, owner-operators, and cash "core"-buying) and the artisan "all other wood" tail are the segments most prone to mild undercount, along with a fringe of very small "portable" band-sawmills; the smallest self-employed producers fall below payroll-based counts. They dominate the count of firms but are tiny in dollars and do not move the subsector's scale.[2][4]
• Vertical integration and classification lines understate conversion. A firm that planes its own logs is booked under sawmills, not millwork; and a large share of trusses and wall panels is fabricated captively inside builder- or dealer-owned component plants classified under construction or distribution codes rather than manufacturing. The engineered-member and cut-stock figures are the "merchant" slice, not the total.[3][4]
• Downstream value chains excluded — and the gap is large. The data measure factories, not the surrounding economy: the pallet pooling/rental business (CHEP, PECO — booked in leasing), the captive lending behind manufactured homes, timberland ownership, and all retail and installation sit elsewhere. One company makes the scale visible — Berkshire reported roughly $12.9 billion of Clayton Homes revenue in 2025 across factories, retail, and lending, nearly double the entire census receipts of the manufactured-home industry Clayton leads (~$7.1B).[4][18]
• These are 2022 receipts, and they now understate the subsector. Building-products prices and manufactured-home values generally rose after 2022; Champion and Cavco alone reported combined revenue above $4.5 billion in their fiscal-2026 years. Different federal series also disagree in level — for pallets, Census shipments were $18.27 billion (2022) while BLS sectoral output was $15.4 billion in 2023; read direction, not level, when mixing them.[4][17]

---

4. The investable universe — where value concentrates across the children

Three facts dominate the public-market map. First, there is no pure-play for the subsector, and no pure-play for two of the three children — most listed exposure is a segment inside a diversified forest-products or building-products firm, so a "wood-products stock" is really a basket of several sub-industries at once. Second, a handful of companies stitch the chain together: UFP Industries is the only public name touching all three children (treated lumber and millwork, PalletOne pallets, trusses and components), while Weyerhaeuser, West Fraser, and now Rayonier each span sawmilling (3211) and panels or engineered members (3212) atop timberland — Rayonier picked up PotlatchDeltic's softwood-plywood mill when the merger closed in January 2026. Berkshire Hathaway is quietly the largest owner inside 3219, holding both Clayton Homes and hardwood-flooring maker Shaw.[2][3][4][6][14][18] Third, the leader in most product markets is private or foreign — Sierra Pacific and Georgia-Pacific in sawmilling, Columbia Forest Products in hardwood panels, Andersen in windows, Clayton in manufactured homes, 48forty and CHEP in pallets. For a private investor, the whole subsector is a target-rich, fragmented, mostly private M&A field.

Where the listed exposure actually sits, by child (tickers reserved for this section):

There is no ETF (exchange-traded fund) dedicated to the subsector; broad timber/forestry funds (WOOD, CUT) give only incidental exposure.[2] Private-market routes are where most of 321 actually lives — timberland and family mills plus a live treater roll-up (3211), employee-owned and foreign-parented panel plants and hundreds of small truss fabricators (3212), and the vast fragmented tail of millwork, pallet, and craft shops plus manufactured-home communities and their captive loan books (3219). Realistic private participation runs from buying or lending to a regional mill/treater/pallet yard, to building-products and PE roll-ups, to timberland via private funds and TIMOs (timberland investment management organizations) or the REITs.

---

5. How the money works

Owners across all three children make money the way cyclical conversion manufacturers do — not the way stores, landlords, or fee businesses do. Four levers run through every child:[2][3][4]

• The conversion spread. Buy wood (logs, OSB, hardwood, resin, energy), add machining and finishing, sell the finished piece for more. The wood share is enormous — lumber and plywood were 41.6% of UFP Industries' net sales in 2025 (roughly $1.7 billion of lumber purchased) — so the spread compresses fast when input prices spike ahead of selling prices, or when inventory is bought high into a falling market. The spread is thinnest where the product is nearly a commodity (framing lumber, plain OSB, plain pallets, where wood can be 50–70% of unit cost) and richest where branding, engineering, or customization add value (branded windows, architectural millwork, engineered members, engineered siding).[3][4][15]
• Capacity utilization. Plants carry high fixed costs (buildings, kilns, presses, saws, finishing lines), so profitability swings hard with how full the lines run. The 2023–25 downturn shows the operating leverage in both children that report it cleanly: Boise Cascade's Wood Products segment earned $337 million of operating income in 2023, $232 million in 2024, and just $5.8 million in 2025; JELD-WEN's North American revenue fell 20.5% to $2.15 billion while its North American adjusted EBITDA fell 60.9% to $99.5 million (segment margin 9.4% → 4.6%). A fifth off the top line took three-fifths off earnings.[3][4][12][16]
• Price-cost lag. Every child prices ahead of buying or drying its wood, so cost moves land late — in millwork, green-lumber price changes historically take roughly five to six months to reach earnings because the wood must be kiln-dried first. It is the most under-appreciated margin risk in the subsector.[4]
• Product mix, and freight as a moat. In every child, commodity work competes on price, is import-exposed, and earns little, while custom, branded, and engineered work earns real margin and resists imports; in soft markets buyers trade down, hurting mix even when volume holds. And because panels, mouldings, pallets, and trusses are bulky and cheap per pound, plants serve a limited radius — which makes these regional markets and is why the subsector stays fragmented no matter how much capital tries to roll it up.[3][4]

Beyond the shared levers, each child adds a distinct profit engine — the reason to study them separately. On the sawmill side of 3211 the swing factor is the lumber-to-log spread, a violently cyclical commodity margin (framing lumber cleared $1,500 per thousand board feet in the 2021 boom and sat around $903 by October 2025); its wood-preservation sidecar earns a steadier treating spread on engineered, contract products, with roughly 75% of Koppers' North American railroad-products sales under long-term contracts.[2][21] In 3212, the most durable profits come from vertical integration and value-add — owning timber → veneer → panel → distribution and escaping the raw commodity (Louisiana-Pacific's engineered-wood siding runs near a 25% adjusted-EBITDA margin against commodity OSB).[3][13] In 3219, millwork's edge is brand and dealer networks; pallets' best economics are in recycling (~95% of wood pallets are reused or recycled) and, outside the code, in pooling — CHEP Americas earned a 20%-plus operating margin on $3.67 billion of FY2025 revenue; and manufactured homes run a captive-finance flywheel, with Clayton's 21st Mortgage and Vanderbilt holding roughly $29.5 billion of net loan balances at the end of 2025.[4][18][20]

One figure the children disagree on, and it matters. The parent previously stated that roughly 70% of manufactured-home buyers use chattel loans (personal-property loans secured by the home, not a land mortgage). Industry and FHFA-adjacent sources do cite ~70%, but a CFPB analysis of Home Mortgage Disclosure Act data found only ~42% of manufactured-home purchase loans were chattel. The gap is definitional — all placements versus reported mortgage-disclosure loans — and the honest read is a range, not a point.[4][27]

The resulting margin ranking is the subsector's cleanest financial contrast: the finance-attached home builders report the fattest gross margins (Champion 26.4% consolidated and Cavco 22.1% on factory-built in fiscal 2026), while pure conversion businesses sit in the mid-teens (UFP Industries 16.8%, JELD-WEN 16.0% company-wide in 2025). These are company-level figures spanning more than any one code — read the gap, not the decimals.[4][15][16][17] The headline numbers to watch across 321 are conversion/gross margin, capacity utilization, inventory turns, yield, price-cost lag, and product mix — not the same-store, occupancy, rate-base, or funds-from-operations metrics used in retail, property, utility, or REIT analysis. (The one nuance: the timber-REIT owners on the sawmill side do layer real-estate and biological-growth economics on top of manufacturing — but they pay income out as cyclical, not bond-like, dividends.)[2]

---

6. What drives demand

Every part of 321 answers to overlapping but importantly different demand engines — the single most useful thing to understand about the subsector:

• New residential construction — the master switch. Every new home needs framing lumber, sheathing panels, engineered members, windows, floors, and a roof structure. Total U.S. starts were 1.36 million in 2025 (down 0.6%), with single-family at 943,000 (down 6.9%), and 2026 projected near 1.3 million; the softening carried into 2026, with a seasonally adjusted annual rate of 1.177 million in May 2026 (882,000 of it single-family). One study quantifies the dependence: a 2026 U.S. Forest Service-supported analysis put the housing-start elasticity of softwood-lumber demand at 0.59 against an income elasticity of just 0.14.[2][3][4][25]
• Repair, remodel & replacement (R&R) — the biggest lever for millwork. There are far more existing homes than new ones, so replacement dominates windows, doors, flooring, and trim (roughly two-thirds of residential window demand). Harvard's Joint Center put owner-occupied improvement and repair spending at $509 billion in 2025, tracking toward roughly $518–524 billion by end-2026 — near a record, but with growth slowing to about 1.6%.[3][4][26]
• Interest and mortgage rates — the variable behind almost all of it. Rates gate new construction, the appetite to finance big remodels, and (at chattel rates commonly running 7–13%) manufactured-home affordability. Rate relief is the clearest shared upside catalyst for the whole subsector.[4][27]
• Infrastructure replacement — the ballast inside 3211. Wood preservation is driven mostly by replacement of an aging wood-pole grid — roughly 150 million poles in service across North America, with analysts estimating 8–12% annual increases in replacement need from grid hardening, wildfire resilience, and electrification — plus a railroad-crosstie maintenance cycle anchored by the Class I railroads. Only a minority slice (decks, fences) tracks housing.[2]
• Housing affordability — the driver for manufactured homes. A new manufactured home averaged roughly $134,500 in December 2025 against ~$424,000 for a new site-built house — about a third of the price. As site-built prices and rents climb, demand shifts toward the factory-built option, a partly counter-cyclical pull.[4]
• Freight and industrial production — the subsector's one non-housing pulse. Pallet demand is derived from how much the economy makes and moves — industrial output, food-and-beverage volumes, e-commerce, and the restock/destock cycle. The swing is violent and out of phase with housing: output rose 34.9% in 2021 and 33.5% in 2022, then fell 13.9% in 2023, with the pallet-products producer price index sliding from 245.6 in April 2022 to 184.0 by September 2024.[4]
• Intra-wood substitution — the winners inside the subsector. Not all substitution runs against wood. OSB keeps taking sheathing share from plywood and now holds roughly 75% of the North American structural-panel market; engineered members gain from prefabrication and from displacing dimensional lumber, steel, and concrete; and mass timber, still small, consumed 179 million board feet in 2024. Prefabrication itself is slower than the logic suggests — only ~3% of single-family completions in 2024 used modular or panelized methods, below the 7% share of 1998.[3][4]
• Substitution away from wood — a structural headwind on the wood share. Vinyl and fiberglass have taken window-frame share; luxury vinyl tile has taken flooring share (wood is now ~12.5% of U.S. flooring sold, with 2025 hardwood sales down ~3.8% and volume down ~6.8%); plastic nibbles at pallets, though wood still holds 90%+. This caps the long-run growth of the wood portion even in a strong building cycle.[3][4]

---

7. Regulation

Because all of 321 buys wood and runs woodworking plants, it shares a regulatory spine — then diverges sharply on top of it, which is itself a defining feature:

• Trade duties and tariffs — the most active lever, and it cuts both ways. Combined U.S. antidumping and countervailing duties on Canadian softwood lumber roughly doubled to ~35% in 2025, and Section 232 national-security tariffs added 10% on imported softwood timber and lumber and 25% on kitchen cabinets and vanities effective October 14, 2025, with a December 31, 2025 action delaying scheduled increases and holding the 25% rate through 2026.[2][4][7][8] Two refinements the children add: plywood under HTS 4412 was excluded from Section 232, so the panel child took the input-cost hit without the import protection[3]; and the children do not agree on the all-in burden — 3211 reads combined duties plus the Section 232 tariff at about 34.8%, while 3219 stacks the ~35% AD/CVD on top of the 10% tariff to reach ~45% on some Canadian lumber. Treat it as a range. Higher duties help domestic producers who sell lumber (much of 3211) but hurt the shops that buy wood as feedstock (much of 3212 and 3219) — a genuine cross-current inside the same subsector.
• A live trade case on hardwood plywood — new since the last revision. On July 16, 2026 the Commerce Department issued final affirmative determinations on hardwood and decorative plywood from China, Indonesia, and Vietnam, with exporter-specific dumping margins of 15.40%–187.27% and subsidy rates of 4.22%–165.39%. The USITC's final injury vote is not until August 19, 2026, so orders are not yet assured — the single clearest binary policy catalyst anywhere in the subsector. Separately, duties on Chinese engineered wood flooring remain in force after the 2023 sunset review (roughly 31.6% antidumping plus ~20.8% countervailing for most exporters, atop a 25% Section 301 tariff).[3][4][10]
• Formaldehyde, air permitting, and product rules across the board. The EPA's formaldehyde-emission rule under Title VI of the Toxic Substances Control Act (TSCA) bites composite panels — hardwood plywood, MDF, and particleboard — hardest, while structural plywood, OSB, structural composite lumber, and I-joists are explicitly exempt, making the rule a burden for some children and a compliance moat against non-compliant imports for others. EPA's June 2026 amendments to the Plywood and Composite Wood Products NESHAP reach 33 major-source softwood-plywood facilities and 52 reconstituted-wood facilities. The Lacey Act requires importers to declare wood species and legal harvest (Phase VII declarations began December 2024), and OSHA governs woodworking machinery and combustible wood dust everywhere.[3][4][23]
• Wood-preservation is a pesticide business — distinctive to 3211. Treating is regulated by the EPA as a pesticide use: chromated copper arsenate was cancelled for residential use in 2003 and pentachlorophenol ("penta") in 2022, with manufacture and sale ending after February 2024 and treating plants given three further years to burn off stocks — pushing treaters toward costlier chemistries such as DCOI and copper naphthenate and leaving legacy contamination as a real remediation liability. Federal-land timber policy is the sawmill side's parallel constraint: Endangered Species Act and National Forest Management Act restrictions cut Pacific Northwest federal timber output by roughly 85% from the late 1980s and pushed the industry South.[2][9]
• Border phytosanitary rules — distinctive to pallets. Any wooden pallet or crate crossing a border must meet ISPM 15 (debarked and either heat-treated to a 56 °C core for 30 minutes or fumigated, then stamped), enforced on U.S. imports by USDA APHIS since 2006. No other corner of 321 carries this regime.[4][24]
• Federal building preemption — distinctive to manufactured homes. Manufactured homes are built to a single national HUD Code (in force since 1976 and given its most comprehensive update in roughly three decades in 2024 — 87 changes, including multi-unit HUD-code buildings), which preempts local codes; financing policy is shaped by FHFA's "Duty to Serve" mandate, barely applied to chattel loans so far, and DOE has delayed manufactured-housing energy-standard compliance deadlines. Site-built structural products instead meet state/local building codes and grade-stamping standards (PS 1/PS 2, APA stamps, ANSI/TPI 1 for trusses, and the tall mass-timber provisions of the 2021/2024 International Building Code).[3][4][24][27]
• One credit expired. The Section 25C Energy Efficient Home Improvement Credit (up to $600/year for qualifying windows) was terminated for property placed in service after December 31, 2025, removing a modest support for replacement demand in 2026.[4]
• No price or entry regulation anywhere. There is no rate base, license regime, or output control across 321 — barriers are capital, freight logistics, environmental permitting, and building codes, not price regulation.

---

8. Consolidation

The subsector is unconcentrated to an extreme (HHI 76), but that headline hides very different competitive structures and one common direction: consolidation at the top of each niche, over a long fragmented tail — scale is bought, not born. The revised children add a pointed corollary: in 321, consolidation mostly removes public options rather than creating them.[1]

• 3211 — a completed timber-REIT merger and a treater roll-up. Sawmilling is migrating to the U.S. South (roughly 25% of North American softwood capacity in 2009 to about 38% by mid-2024), with heavy rationalization elsewhere — more than 5 billion board feet idled or closed since 2023. The landmark deal is the Rayonier–PotlatchDeltic $8.2 billion all-stock merger, announced October 2025 and closed in January 2026, forming the second-largest U.S. timber REIT and retiring a listed name. Wood preservation consolidates through a decades-long roll-up of regional treaters by Stella-Jones, Koppers, and Doman, protected by high environmental-permitting barriers and long customer qualification cycles — recent examples include Koppers' ~$100 million purchase of Brown Wood Preserving's utility-pole assets and Doman's ~$255 million acquisition of C.M. Tucker.[2][5][6][21]
• 3212 — panel oligopolies, upstream integration, and capacity discipline. The capital-intensive OSB and plywood businesses are the subsector's most concentrated real markets (softwood plywood HHI ≈1,036; nine companies make all North American OSB), consolidated by forest-products majors integrating from timber to panel — West Fraser's ~C$4.0 billion acquisition of Norbord made it the largest North American OSB producer, and Boise Cascade paid ~$512 million for Coastal Plywood. The current cycle is being managed by subtraction: West Fraser indefinitely curtailed its High Level, Alberta OSB mill; Georgia-Pacific closed softwood plywood at Emporia, Virginia; and Roseburg exited hardwood plywood outright in September 2025.[3][12][13]
• 3219 — consolidation in opposite directions internally. Millwork is consolidating fastest and shedding public names (Owens Corning bought Masonite for ~$3.9 billion; MITER Brands bought and delisted PGT Innovations for ~$3.1 billion). Pallets roll up on two tracks — manufacturing (UFP/PalletOne, which agreed in May 2026 to pay roughly $48 million for three John Rock plants representing $82 million of annual sales) and recycling (48forty, 270+ facilities on ~$709 million of revenue) — beneath a CHEP pooling oligopoly (~90% of North American pooling). Manufactured homes is already concentrated and still consolidating: Clayton ~46.8%, Champion ~22.1%, Cavco ~17.7% — together 86.5% of 2025 units (102,962 homes across 149 production lines). The artisan craft tail barely consolidates at all, though off a very low base (its four-firm share rose from 10% in 2017 to 21% in 2022).[4][16][19][20]

Across all three children, a slow structural force — material substitution (vinyl/fiberglass windows, LVT flooring, plastic pallets, steel/concrete structure) — caps the wood segment and pushes survivors toward premium, custom, and engineered products.[2][3][4]

---

9. Risks

• Deep cyclicality and rate sensitivity — the dominant, shared risk. Every child tracks housing, construction, or industrial freight with little natural recurring-revenue cushion; a rate-driven downturn hits volumes and margins together, and high fixed costs magnify the earnings swing (Boise Cascade's Wood Products operating income fell from $337 million in 2023 to $5.8 million in 2025; JELD-WEN's North American adjusted EBITDA fell ~61% on a ~20% revenue decline). Manufactured homes is the extreme case in the subsector's history — annual shipments fell from ~373,000 units in 1998 to ~50,000 in 2009, recovering only to ~103,000 in 2024–2025.[3][4][12][16][19]
• Input-cost and tariff whipsaw. Framing lumber ranged roughly $600 to $900+ per thousand board feet across 2025 (about $903 in October), and glass, resin, and energy prices move sharply and are now tariff-inflated; thin-margin commodity producers (dense in sawmilling, pallets, and the misc. tail) have the least room to pass costs through, and trade policy can be raised, delayed, or trimmed by courts. Raw-material risk is species- and grade-specific, not reducible to one lumber benchmark.[2][3][4][7][8]
• Trade policy cuts both ways — and can deliver nothing. The same duties that protect domestic lumber sellers raise costs for the many shops that buy wood; plywood's exclusion from Section 232 left one child with the costs and none of the shelter; and the hardwood-plywood AD/CVD case still hinges on an ITC injury vote that could go either way.[2][3][4][10]
• Overcapacity where the capital is heaviest. West Fraser has described North American OSB as oversupplied following greenfield and restart announcements for 2025–27 — the classic failure mode of the subsector's most capital-intensive child.[3][13]
• Structural substitution. Continued wood-share loss to vinyl/fiberglass, LVT, and plastic is secular, not cyclical.[3][4]
• Environmental and legal liability — concentrated in wood preservation. Legacy creosote, penta, and arsenic sites carry real remediation exposure (Koppers alone held $10.2 million of environmental reserves at the end of 2025 and guided to about $13 million of 2026 environmental capital spending), and regulatory chemistry transitions raise costs.[2][9][21]
• Customer and channel concentration. Wood preservation sells to a handful of utilities and Class I railroads; millwork sells through two retailers (Home Depot and Lowe's were ~30% of JELD-WEN's 2025 revenue, with the top ten customers ~48%); the pallet economy is framed by a pooling oligopoly; and manufactured-home credit is concentrated too, with the top five lenders accounting for more than 40% of purchase lending and nearly 75% of chattel lending.[2][4][16][27]
• Financing fragility — concentrated in manufactured homes. The industry's worst crash began with loose chattel underwriting; reliance on high-rate personal-property credit remains its structural weak point.[4]
• Thin and uneven public liquidity, and thinning further. There is no subsector pure-play, no pure-play for panels or pallets, the only near-pure-play in millwork (JELD-WEN) is a volatile small-cap spanning four frame materials and two continents, and the biggest manufactured-home operator (Clayton) is unbuyable directly — while the listed roster keeps shrinking through mergers, take-privates, and one bankruptcy. Cleanly owning "321" on a U.S. exchange is impossible.[4][6][16]
• Labor and safety. An aging, hard-to-hire manual workforce plus combustible-dust and machinery hazards across every child; the pallet industry posted a 2024 recordable injury rate of 3.9 cases per 100 full-time workers versus 2.3 for all private industry.[3][4]

---

10. How to invest & the outlook

Public-market routes are all indirect — think in slices matched to a thesis, not whole companies:

• Broadest single names: UFP Industries (UFPI) (~$6.3B revenue, 2025) touches all three children; Weyerhaeuser (WY), West Fraser (WFG), and Rayonier (RYN) each span sawmilling and panels atop timberland.[2][3][11][13][14]
• Housing-cycle torque (upstream): sawmill-side timber REITs WY, RYN and lumber producers WFG, IFP, CFP (3211); OSB/panel names BCC, LPX and mass-timber optionality MERC (3212).[2][3][11][12][13]
• Defensive corners: wood-preservation infrastructure names KOP, SJ, DBM (3211); manufactured-home makers SKY, CVCO, LEGH — the subsector's cleanest pure-plays — with leader Clayton only via BRK.B, and steadier rent-driven exposure through community REITs SUI, ELS (3219).[2][4][17][18][21]
• Millwork: thin near-pure-play JELD, plus OC / FBIN for doors and MHK for flooring.[4][16]
• Pallets: no U.S. pure-play — UFPI or foreign-listed pool BXB / BXBLY.[4][20]
• Channel/demand proxies: distributor BLDR and retailers Home Depot (HD), Lowe's (LOW), Floor & Decor (FND) — a liquid way to bet on the demand rather than the making. Expect all of these to trade on interest rates and housing/freight data more than on wood specifically; earnings are cyclical, so trailing multiples look cheap at cycle peaks and expensive at troughs.[3][4]

Private-market routes — where most of the subsector's ~12,000 firms actually change hands: timberland, family sawmills, and a live buy-and-build treater roll-up (3211); employee-owned and foreign-parented panel plants plus hundreds of small truss fabricators (3212); and the vast fragmented tail of branded and PE-platform millwork, a lower-middle-market pallet making-and-recycling roll-up under the pooling oligopoly, and manufactured-home communities plus freight-radius truss shops (3219). All are asset-based, locally moated, and sensitive to the same cycles as the stocks. Underwriting should separate manufacturing earnings from retail, installation, and lending earnings, and examine yield by species and grade, plant utilization, delivered freight radius, price-cost lag, customer concentration, backlog quality, warranty reserves, and maintenance capital.[2][3][4]

Outlook (forward-looking judgment, not fact). NAICS 321 is best understood as a cyclical wood value chain that rarely fires on all cylinders at once — which is exactly why it rewards a child-by-child (and then product-by-product) read rather than the blended $172 billion, HHI-76 average that describes no real market. Heading into 2026 the whole subsector is cyclically soft but stabilizing: total starts of 1.36 million in 2025 with single-family down 6.9%, 2026 projected near 1.3 million, a May 2026 SAAR of 1.177 million, remodeling growth slowing to ~1.6%, and ~6–7% mortgages capping the upside across the board.[2][3][4][25][26] The structural cases diverge by child. The clearest ballast sits in the two corners least tied to housing — wood preservation (aging-grid replacement, wildfire hardening, and electrification pointing to mid-to-high-single-digit utility-pole demand growth) and pallets (low-growth but cash-generative, tracking freight, and a consolidation story rather than a growth one) — plus manufactured housing, where a real long-run affordability tailwind is hostage to financing and to unresolved energy-standard rules. The most cyclical torque sits in sawmilling and panels, muted while rates keep housing flat, but leveraged to a widely cited ~3.7–4 million-home shortage, the South's cost advantage, elevated Canadian duties, disciplined OSB supply, and emerging mass-timber demand — with one live binary in hardwood plywood, where the ITC's August 19, 2026 injury vote either delivers pricing power to a structurally challenged product or delivers nothing. The single clearest shared upside catalyst is rate relief, which would revive new construction, remodeling, and freight together — and, as with all cyclicals, the best historical entry points have come when construction pessimism is deepest.

---

Sources

1. U.S. Census Bureau, 2022 Economic Census (receipts, firm counts, concentration ratios CR4/CR8/CR20/CR50 and HHI) and 2023 County Business Patterns (establishments, employment, annual payroll) — NAICS 321 and its children 3211/3212/3219. Histometrics federal-statistics compilation. https://www.census.gov/programs-surveys/economic-census.html; https://www.census.gov/programs-surveys/cbp.html
2. Sawmills and Wood Preservation child primer (NAICS 3211 / 32111) and its sources — single-child pass-through structure; six-digit split (321113 sawmills ~$41.3B, 2,691 plants, ~$15M each; 321114 wood preservation ~$10.5B, 362 plants, ~$29M each); sawmill HHI 191 / CR4 22.2% vs. preservation HHI 926.6 / CR4 54.9%; average pay ~$61,300; lumber ~$903/MBF (Oct. 2025); housing starts 1.36M (2025) and 1.177M SAAR (May 2026); ~3.7–4M home shortage; 150M utility poles and 8–12% replacement growth; PNW federal timber −85%; ~34.8% effective Canadian-lumber burden; South migration 25%→38% and >5 Bbf idled; WOOD/CUT ETFs; ~250 private treaters. https://www.census.gov/naics/
3. Veneer, Plywood, and Engineered Wood Product Manufacturing child primer (NAICS 3212 / 32121) and its sources — single-child pass-through structure; six-digit shares and plants (321215 ~48%/$19.8B/983; 321219 ~27%/$11.2B/192; 321212 ~16%/$6.6B/80; 321211 ~9%/$3.5B/201); child HHIs 341 / 579 / 883 / 1,036; ~$20M and ~$83M shipments per plant; average pay ~$63,000; freight as a regional moat; plywood excluded from Section 232; NESHAP and formaldehyde incidence; capacity curtailments and closures. https://www.census.gov/naics/
4. Other Wood Product Manufacturing child primer (NAICS 3219) and its sources — three five-digit children and their shares/concentration (32191 ~47%/$37.3B, CR4 27%, HHI 263, inner range 225–718; 32192 ~23%/$18.3B, CR4 16.2%, HHI 88.1; 32199 ~30%/$24.0B, CR4 28.9%, HHI 252.9, manufactured homes CR4 66%); average pay ~$54,000; pallet output and PPI swings; Census vs. BLS output; five-to-six-month price-cost lag; ~95% pallet reuse/recycle; wood ~12.5% of U.S. flooring; modular ~3% of completions; manufactured-home prices and shipment history; framing lumber $600–900+/MBF; Section 25C termination; engineered-wood-flooring duties; Lacey Act Phase VII; pallet injury rate 3.9 vs. 2.3; Andersen ~$3.6B; USFS housing-start elasticity 0.59. https://www.census.gov/naics/
5. Forisk, "Top 10 North American and U.S. Lumber Producers in 2024," 2024 (47.9 billion board feet of 2024 U.S. softwood capacity; top ten ~50%). https://forisk.com/top-10-north-american-and-u-s-lumber-producers-in-2024/
6. Rayonier Investor Relations, "Rayonier and PotlatchDeltic Announce Closing of Merger of Equals," 2026 ($8.2 billion all-stock merger closed January 2026; ~4.2 million acres; seven wood-products facilities). https://ir.rayonier.com/news/news-details/2026/Rayonier-and-PotlatchDeltic-Announce-Closing-of-Merger-of-Equals/default.aspx
7. Congressional Research Service, U.S.–Canada Softwood Lumber Trade: Current Issues for Congress (R48781), 2025; NAHB, "Canadian Lumber Duties Hit 35% — And May Go Higher Soon," 2025. https://www.congress.gov/crs-product/R48781; https://www.nahb.org/blog/2025/08/canadian-lumber-cvd-rates
8. The White House / Federal Register / U.S. Customs and Border Protection, Adjusting Imports of Timber, Lumber, and Their Derivative Products (Section 232) — 10% on softwood timber and lumber, 25% on kitchen cabinets and vanities effective Oct. 14, 2025; Dec. 31, 2025 delay holding the 25% rate through 2026; plywood under HTS 4412 excluded. https://www.federalregister.gov/documents/2025/10/06/2025-19482/adjusting-imports-of-timber-lumber-and-their-derivative-products-into-the-united-states; https://www.whitehouse.gov/presidential-actions/2025/09/adjusting-imports-of-timber-lumber-and-their-derivative-products-into-the-united-states/
9. U.S. Environmental Protection Agency, EPA Requires Cancellation of Pentachlorophenol, 2022 (manufacture/sale ending after February 2024; three-year stock burn-off; shift to DCOI and copper naphthenate). https://www.epa.gov/pesticides/epa-requires-cancellation-pentachlorophenol-protect-human-health
10. U.S. Department of Commerce, Final Affirmative Determination — Antidumping and Countervailing Duty Investigations on Hardwood and Decorative Plywood from China, Indonesia, and Vietnam (dumping 15.40%–187.27%; subsidies 4.22%–165.39%), July 2026; U.S. International Trade Commission, Case Calendar (final injury vote August 19, 2026). https://www.trade.gov/final-affirmative-determination-antidumping-and-countervailing-duty-investigations-hardwood-and; https://ids.usitc.gov/case/8288/investigation/8781
11. Weyerhaeuser Company, Form 10-K FY2025 (Wood Products ~$4.96B; engineered solid section, I-joists, softwood plywood), 2026. https://www.sec.gov/Archives/edgar/data/106535/000119312526051422/wy-20251231.htm
12. Boise Cascade Company, Form 10-K FY2025 (Wood Products segment operating income $5.8M vs. $231.5M prior year; ~40% LVL and ~37% I-joist share; Coastal Plywood acquisition ~$512M), 2026; Mercer International / Mercer Mass Timber, Acquisition of Structurlam (~$81M), 2023. https://www.sec.gov/Archives/edgar/data/1328581/000132858126000006/bcc-20251231.htm; https://www.bc.com/boise-cascade-reaches-agreement-to-acquire-coastal-plywood-operations/
13. West Fraser Timber Co., 2025 Annual Report (High Level, Alberta OSB curtailment; North American OSB oversupply) and Completes Acquisition of Norbord (~C$4.0B); Louisiana-Pacific Corporation, Form 10-K FY2025 (siding ~25% adjusted-EBITDA margin); Forisk Consulting, Top 2025 OSB Producers in North America (9 companies; 26.5B sq ft, 3/8-inch basis). https://www.westfraser.com/sites/default/files/2026-03/West%20Fraser%20Annual%20Report%202025.pdf; https://www.sec.gov/Archives/edgar/data/60519/000006051926000012/lpx-20251231.htm; https://forisk.com/top-2025-osb-producers-in-north-america/
14. UFP Industries, Inc., Form 10-K, Fiscal 2025 (~$6.3B revenue; gross margin 16.8%; lumber and plywood 41.6% of net sales, ~$1.7B purchased; Packaging $1.60B with PalletOne $522M; ~8–10% of the U.S. new machine-built pallet market; John Rock acquisition), 2026. https://www.sec.gov/Archives/edgar/data/912767/000110465926019567/ufpi-20251227x10k.htm
15. JELD-WEN Holding, Inc., 2025 Form 10-K and FY2025 results (~$3.2B revenue; North America −20.5% to $2.15B; North American adjusted EBITDA −60.9% to $99.5M, margin 9.4%→4.6%; gross margin 18.2%→16.0%; top ten customers ~48%, Home Depot ~17%, Lowe's ~13%); Owens Corning, $3.9B Acquisition of Masonite and 2025 Form 10-K (~$10.1B); MITER Brands, Completes Acquisition of PGT Innovations (~$3.1B; delisted), 2024. https://investors.jeld-wen.com/; https://investor.owenscorning.com/; https://www.businesswire.com/news/home/20240327387754/en/
16. Champion Homes, Inc., Form 10-K (FY2026) (revenue $2.664B; gross margin 26.4%; 46 plants); Cavco Industries, Inc., Form 10-K (FY2026) (factory-built revenue $2.157B; factory-built gross margin 22.1%; CountryPlace Mortgage and Standard Casualty). https://www.sec.gov/Archives/edgar/data/90896/000119312526239333/sky-20260328.htm; https://www.sec.gov/Archives/edgar/data/278166/000162828026037782/cvco-20260328.htm
17. Berkshire Hathaway, Inc., 2025 Annual Report (Clayton Homes revenue ~$12.9B and pretax earnings ~$1.9B; 21st Mortgage and Vanderbilt net loan balances ~$29.5B); Shaw Industries Group company profile (~$6B sales). https://www.berkshirehathaway.com/2025ar/2025ar.pdf; https://shawinc.com/company-profile
18. Manufactured Housing Institute, Q4 2025 Market Share Report (Clayton 46.75%, Champion 22.10%, Cavco 17.65%; top three 86.5%; 102,962 homes on 149 production lines in 2025). https://www.manufacturedhousing.org/wp-content/uploads/2026/04/Q4-2025-Market-Share-Report2.pdf
19. Brambles Limited (CHEP), FY2025 Annual Report (CHEP Americas revenue $3.67B, operating profit $737M); 48forty Solutions / Summit Partners (~$709M revenue, 270+ facilities). https://www.chep.com/files/download/Brambles-FY25-Annual-Report.pdf; https://www.summitpartners.com/companies/48forty-solutions
20. Koppers Holdings Inc., 2025 Form 10-K (~75% of North American railroad-products sales under long-term contracts; $10.2M environmental reserves; ~$13M 2026 environmental capital spending) and Brown Wood Preserving acquisition (~$100M), 2024–2026; Stella-Jones Inc., 2025 Annual Report (poles and ties ~76% of 2025 sales); Doman Building Materials Group Ltd., C.M. Tucker acquisition (~$255M), 2024. https://www.sec.gov/Archives/edgar/data/1315257/000131525726000012/kop-20251231.htm; https://www.stella-jones.com/sites/default/files/reports/SJ-AR-2025-EN_0.pdf; https://cdn.domanbm.com/files/10-01-24-DBMGL-CM-Tucker-Final-PR.pdf
21. Fastmarkets, "US Pallet Market: Trends and Future Projections" (~38% of U.S. sawn-lumber consumption; wood 90%+ of the pallet market). https://www.fastmarkets.com/insights/annual-pallet-estimates-viewpoint/
22. U.S. International Trade Commission, Hardwood Plywood from China — Five-Year Review, 2023 (2021 apparent U.S. consumption: imports 92.7% of volume, 84.0% of value); IndexBox / Woodworking Network, U.S. Hardwood Plywood Imports — Q1 2025 (~$490M, +18% year over year), 2025. https://www.usitc.gov/sites/default/files/publications/701_731/pub5426_0.pdf; https://www.woodworkingnetwork.com/news/woodworking-industry-news/imports-us-hardwood-plywood-surge-58
23. U.S. Environmental Protection Agency, Formaldehyde Emission Standards for Composite Wood Products (TSCA Title VI, 40 CFR Part 770), 2016/updated, and Plywood and Composite Wood Products NESHAP — Economic Impact Analysis (33 affected softwood-plywood facilities; 52 reconstituted-wood facilities), June 2026; APA — The Engineered Wood Association, structural plywood/OSB/I-joist exemptions and PS 1 / PS 2 grade-stamping. https://www.epa.gov/formaldehyde/formaldehyde-emission-standards-composite-wood-products; https://www.epa.gov/system/files/documents/2026-06/plywood_eia_neshap_final_20260629.pdf; https://www.apawood.org/structural-plywood-osb-exempt-from-new-formaldehyde-ruling
24. USDA Animal and Plant Health Inspection Service (APHIS), Wood Packaging Material / ISPM 15 (in force in the U.S. since 2006); U.S. Department of Housing and Urban Development, HUD Code (Manufactured Home Construction and Safety Standards; 2024 update with 87 changes); Truss Plate Institute / SBCA, ANSI/TPI 1. https://www.aphis.usda.gov/plant-imports/wood-packaging-material; https://www.hud.gov/hud-partners/manufactured-home-resources; https://www.sbcacomponents.com/media/tpi-announces-the-release-of-ansitpi-1-2022
25. National Association of Home Builders / Census-HUD, Overall Housing Starts Inch Lower in 2025 (total 1.36M, −0.6%; single-family 943K, −6.9%); Forisk, U.S. Housing Starts Outlook — Q1 2026 Update (2026 near 1.3M). https://www.nahb.org/news-and-economics/press-releases/2026/02/overall-housing-starts-inch-lower-in-2025; https://forisk.com/u-s-housing-starts-outlook-q1-2026-update/
26. Joint Center for Housing Studies of Harvard University (JCHS), Leading Indicator of Remodeling Activity (LIRA) — $509B of owner-occupied improvement and repair spending in 2025, ~$518–524B by end-2026, growth slowing to ~1.6%. https://www.jchs.harvard.edu/research-areas/remodeling/lira
27. Consumer Financial Protection Bureau, Manufactured Housing Finance: New Insights from the Home Mortgage Disclosure Act Data (~42% of manufactured-home purchase loans chattel; top five lenders >40% of purchase lending and ~75% of chattel lending); Federal Housing Finance Agency, Duty to Serve; HousingWire, "FHFA pushes GSEs to embrace chattel loans," 2026 (~70% industry estimate; chattel rates commonly 7–13%). https://www.consumerfinance.gov/archive/newsroom/manufactured-housing-loan-borrowers-face-higher-interest-rates-risks-and-barriers-to-credit/; https://www.housingwire.com/articles/duty-to-serve-chattel-loans-fhfa/