Industrial Machinery Manufacturing (U.S.) — NAICS 33324

A Histometrics rollup primer for public-market and private investors. NAICS = North American Industry Classification System, the standard code the U.S. government uses to group businesses. This is a 5-digit "industry" that contains four more detailed 6-digit industries.

1. Overview

This is the "make the machines that make things" corner of the U.S. economy — one rung below the machinery sector as a whole, one rung above four very different equipment businesses. NAICS 33324 groups together the companies that build food-processing machinery (333241), semiconductor-making machinery (333242), sawmill, woodworking, and paper machinery (333243), and an "all other" catch-all of plastics, printing, textile, chemical, glass, and 3D-printing machinery (333248).[1]

Why an investor should care: every one of these four is a "picks-and-shovels" capital-goods business. The owners don't sell chips, bread, lumber, or plastic parts — they sell the equipment (plus the spare parts, consumables, service, and software) that the makers of those things must buy to build and modernize a plant. That gives the whole group a shared DNA: demand is derived from customers' capital-spending cycles, revenue is lumpy and backlog-driven, and the durable profit hides in a recurring aftermarket annuity. But it also hides enormous internal contrast — the four children differ by a factor of ten in how concentrated they are, differ sharply in who owns them, and are traveling in different directions at once.

The single most useful thing this rollup can tell you is how the four differ, because "industrial machinery" as a label averages away the very distinctions that decide where money is made and how you can reach it. Section 2 leads with that contrast; the rest covers the level as a whole.

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

The four are one family (capital goods, aftermarket-led, cyclical) but they are not interchangeable. The table below is the heart of this primer. Federal shares are computed from our ground-truth 2022 receipts; the child figures reconcile almost exactly to the level total (see §3).[1][3][4][5][6]

The three contrasts that matter most:

1. Value is concentrated; firm counts are not. Semiconductor machinery is only 144 firms — 6% of the level's companies — yet books a third of its revenue, at roughly $95 million of receipts per firm. The "all-other" bucket is the mirror image: two-thirds of the firms but $9.5 million of receipts per firm, a long tail of niche shops in which about 96.7% of companies fall below the SBA's 750-employee small-business threshold.[6][29] So the level's overall look ("fragmented," HHI 152.7, see §3) is an average of opposites, not a description of any one child.

2. Concentration ranges across the whole spectrum inside one code — and the federal numbers understate it everywhere. At the leading edge, semiconductor tools are effectively single-supplier monopolies: ASML holds essentially the entire market for extreme-ultraviolet lithography, Lam roughly 45% of etch, KLA around 52% of process control, and the top five suppliers about 65% of the global market.[4] The other three children run federal HHIs of 66–274.[3][5][6] But two of the four child primers now make the same point explicitly: the domestic HHI understates true market structure, because the global champions build much of their equipment offshore and barely register in the U.S. establishment count. The sawmill/paper child reads HHI 255 domestically while the high-speed paper-machine market is a global oligopoly of Voith, Valmet, and Andritz, and HOMAG claims an estimated global woodworking-machinery share above 30%.[5][20][21] Read concentration by product market, not by NAICS code — at either level.

3. The pay gap is the tell. Semiconductor machinery pays ~$151k per worker versus ~$79–82k for the other three — a signal of how much more R&D-intensive and precision-driven chip tools are than a dough mixer, a debarker, or an injection-molding press. Applied Materials spent $3.57 billion on R&D in fiscal 2025 (about 12.6% of revenue) and Lam $2.10 billion (11.4%).[8] It is why that one child, not the other three, behaves like a technology stock.

Even the government treats them as different animals: the SBA's small-business employee ceiling runs 500 for food machinery, 550 for sawmill/paper, 750 for all-other, and 1,500 for semiconductor.[2][3][4][5][6]

What they share, and why they sit together: all four are engineered-to-order capital goods sold on backlog; all four run a razor-and-blade model where the installed base drives recurring parts-and-service revenue (§5); all four have demand derived from a customer industry's capex (§6); and all four are systematically understated by U.S. production statistics because so much equipment is imported or built offshore (§3).

3. How big it is (the rollup figures)

Our ground-truth federal figures for NAICS 33324 as a whole:

CBP = County Business Patterns, an annual Census dataset. HHI is a standard concentration gauge; higher means more concentrated, and antitrust agencies treat ~1,500 as the "moderately concentrated" line.

A few readings. Average pay across the level is roughly $98,000 per worker ($10.95B ÷ 111,497) — above the U.S. manufacturing average, and pulled up by the semiconductor child. The level looks fragmented on paper (HHI 152.7, CR4 20.8%), but as §2 stressed, that is a statistical artifact of stapling four separate product markets together: no single firm can dominate "industrial machinery" when the machines in question are as unrelated as a wafer etcher and a paper-making line. Read concentration at the child level, not here.

A useful validity check: the four children's establishments (2,836), employment (111,497), and payroll ($10.95B) sum exactly to the level, and receipts sum to ~$40.24B against the level's $40.23B — confirming the ground-truth compilation is internally consistent.[1][3][4][5][6]

Two measurement caveats the revised children now force onto this page.

1. The employment figure is frame-dependent. The 111,497 above is a CBP count. For the all-other child alone, BLS payroll data reported 66,200 jobs in April 2026 against CBP's 51,199 for 2023 — different survey frames and reference periods, not a correction of one by the other.[6][28] Treat the level employment number as one defensible frame, not the only one.
2. Half this level's history is not splice-able. Two of the four children are 2022 NAICS constructs: 333243 merged the formerly separate sawmill/woodworking and paper-machinery industries, and 333248 was created by combining former 333244 (printing machinery) with former 333249 (other industrial machinery), the latter representing roughly 87–89% of the combined predecessors' receipts, firms, and employees.[5][6][29] Any long time series stitched across the 2022 revision at this level is unreliable.

The undercount caveat — read this carefully. These figures are honest, but they measure the wrong thing for gauging market size. Two gaps:

1. Imports and offshore production. The statistics count equipment produced by U.S.-located establishments, not equipment bought by U.S. factories. The children now put numbers on this. In plastics, 2024 imports of equipment equaled 68.8% of domestic shipments, and mold imports equaled 93%.[26] The largest paper machines and integrated sawmill lines installed in America are built by Valmet, Andritz, Voith, and Dürr/HOMAG abroad; vendors put the global paper-making-machine market near $19 billion in 2025 against the whole child's ~$3.97B of domestic shipments (an indicative third-party estimate on a much broader scope).[5] And in semiconductor the U.S. champions build much of their gear offshore (Singapore, Malaysia), so the domestic figure understates even the U.S. firms' true scale: Applied Materials alone booked $28.4 billion of global revenue in fiscal 2025 against the entire child's ~$13.75B of domestic shipments, while global equipment billings ran about $135 billion in 2025.[4][7][8] Treat $40B as "domestic factory output," not "the size of the U.S. market."
2. Private ownership limits disclosure. This is not a gig-economy or sole-proprietor industry — these are real factories with 40 workers each on average — so the usual small-operator undercount does not apply. But a large share of the industry sits inside private-equity platforms and family-owned firms (§4), so public financial disclosure captures only a slice of the economics.

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

The dominant fact for an investor is that your route in changes completely depending on which child you want — and the child with the most public value (semiconductor) is the one where private access is nearly impossible, while the children full of private value (all-other, sawmill/paper) offer the thinnest public menu. Tickers appear here and in §10 only.

Semiconductor machinery (the deepest public menu). This is a stock-picker's industry: the meaningful players are nearly all listed. The U.S. leaders are Applied Materials (Nasdaq: AMAT) (~$28.4B FY2025 revenue), Lam Research (LRCX) (~$18.44B), and KLA (KLAC) (~$12.16B) — large, liquid, dividend-paying, cyclical — trailed by focused specialists such as Axcelis (ACLS), Onto Innovation (ONTO), MKS Instruments (MKSI), and Kulicke & Soffa (KLIC), plus Veeco (VECO), now being acquired by Axcelis.[8][10] Foreign leaders ASML (the sole maker of extreme-ultraviolet lithography tools; €32.7B of 2025 sales) and Tokyo Electron round out the tools U.S. fabs actually buy.[4][9] Private access is essentially limited to the subsystem, component, and startup layer beneath the majors.

Food-product machinery (two thin pure-plays, then foreign or private). The closest U.S.-listed pure-plays are recent creations: JBT Marel (NYSE: JBTM), formed when JBT bought Iceland's Marel in January 2025 (~$3.80B of 2025 revenue, ~50% recurring), and Midera Food Processing (Nasdaq: MFP), spun out of Middleby with the separation completed July 6, 2026 (~$853M of 2025 net sales, of which $341M — 40% — is aftermarket parts and service).[13][14] Broad but partial exposure comes through foreign groups (Germany's GEA and Krones, Sweden's Alfa Laval); much of the industry is private, being rolled up by private equity (KKR's Fortifi, Warburg Pincus's Duravant, Leonard Green's ProMach) or held by family specialists (Heat and Control, Reiser, Urschel, Bühler).[3][18]

Sawmill, woodworking & paper machinery (no U.S. pure-play, but one more listed route than this page previously showed). The nearest U.S.-listed proxy is Kadant (NYSE: KAI) — a diversified process-industries supplier, ~71% parts and consumables, with wood processing only about 22% of 2025 consolidated revenue.[19] The genuine paper-machine leaders are foreign-listed — Valmet (Helsinki; €5.20B of 2025 net sales) and Andritz (Vienna; Pulp & Paper backlog €2.83B at end-2025) — or private (Germany's Voith).[20] For the woodworking side specifically there is now a clear listed platform: Dürr/HOMAG (XETRA: DUE), with HOMAG at ~€1.4B of sales and a claimed global woodworking-machinery share above 30%.[21] The dominant U.S. sawmill-systems makers — USNR (private-equity-backed) and Wood-Mizer (private) — have no public shares.[5]

All-other industrial machinery (mostly private/foreign; a speculative public fringe). There is no large U.S.-listed champion. The listed pure-plays are small, loss-making 3D-printing companies — Stratasys (SSYS) (~$551M 2025 revenue), 3D Systems (DDD) (~$387M, 33.9% gross margin, $(45)M adjusted EBITDA), Nano Dimension (NNDM), Velo3D (VLD) (which has flagged going-concern doubt).[27] Steadier partial exposure comes from diversified industrials that own machine brands — Hillenbrand (NYSE: HI), whose Advanced Process Solutions segment ran ~$2.07B of FY2025 revenue at 34.0% gross margin, and Nordson (NDSN) at ~$2.7B and 55.2%.[24] The classic plastics, printing, and extrusion capacity sits with PE (Bain now holds ~51% of Milacron with Hillenbrand at 49%; Gamut's Davis-Standard) or foreign makers (Koenig & Bauer, Bobst).[6][25]

The pattern: public-market value in this rollup is heavily skewed toward one child (semiconductor), which by itself dwarfs the market capitalization of the other three combined. Just as important, the public menu outside semiconductor is not only thin but compromised — the largest food pure-play recorded a $49.7 million loss from continuing operations in 2025 despite $3.8B of revenue, the closest sawmill/paper proxy derives only ~22% of revenue from wood, and the all-other pure-plays lose money.[13][19][27] If you want diversified, liquid exposure to "industrial machinery," you are mostly buying chip-tool makers. Reaching food, wood/paper, or the plastics-and-printing world in size means foreign listings or private equity.

5. How the money works

All four children run the same capital-goods-plus-aftermarket model, and the split is the whole game.

• New equipment (the "razor"). Large, engineered-to-order systems — a poultry line, a wafer etcher, a paper machine, an injection-molding press. These range from millions to hundreds of millions of dollars (advanced EUV lithography systems reportedly run well over $100M each and the newest High-NA tools around $380M apiece; Valmet booked a single Arauco pulp-mill order worth over €1 billion), are sold on long lead times against detailed specs, and carry only modest gross margins. Revenue here is backlog-driven: the order book, not the current quarter, is the leading indicator of the next 12–24 months.[4][5][20] Backlog quality matters as much as its size — cancellation rights, deposits, escalation clauses, and remaining engineering content determine what it is actually worth.[6]
• Aftermarket (the "blades"). Once a machine is installed it runs for 10–20 years and needs a steady stream of spare parts, consumables (cutting blades, belts, paper-machine clothing, chamber parts), rebuilds, service, and increasingly software (yield optimization, predictive maintenance). This revenue is higher-margin, stickier, and far less cyclical — an annuity on the installed base. Disclosed shares now span a wide band: Kadant ~71% parts and consumables (up from 66% in 2024), JBT Marel ~50% recurring, Midera 40%, ASML's service roughly a quarter of revenue, and Lam's customer-support line at $6.94B against $11.49B of systems.[8][9][13][14][19] Lifting that share is the explicit strategy everywhere.
• The metrics owners and investors actually watch are common across all four: order backlog and book-to-bill (new orders ÷ shipments; above 1.0 means demand is building), recurring-revenue mix (the higher, the more resilient), gross/EBITDA margin (project work runs thin, parts and service run fat, so mix drives margin more than volume), and customers' capacity utilization (full plants order machines; idle plants freeze capex first).

EBITDA = earnings before interest, taxes, depreciation and amortization, the profitability measure most often quoted here.

Margin: a direction, not a level. The previous version of this page implied a clean split — chip tools rich, mechanical machinery thin. The revised children do not support a single benchmark. Semiconductor gross margins do cluster high (Applied 48.7%, Lam 48.7%, ASML 52.8%, KLA 60.9%), but inside the all-other child Nordson runs a 55.2% company gross margin on proprietary components and recurring parts while Hillenbrand's process-solutions segment runs 34.0% and 3D Systems 33.9%; Kadant's consolidated gross margin was 45.2% in 2025, JBT Marel's 35.1%, Midera's 36.2%.[8][9][13][14][19][24][27] What travels reliably across all four is the direction: parts and service are richer than projects, and mix moves margin. Kadant's gross margin rose from 44.3% to 45.2% in 2025 precisely because capital-equipment revenue fell 16% while parts and consumables grew 11%.[19] Conversely, Hillenbrand's segment margin fell from 35.5% to 34.0% on lower volume and unfavorable mix, and Midera's from 39.6% to 36.2%.[14][24]

Where the children disagree: working capital. The food primer argues that deposits and progress payments on custom projects make this business less working-capital-intensive than many capital-goods categories; the all-other primer describes working capital as heavy, with cash tied up in long-lead work-in-progress and customer advances flattering cash generation in an order upcycle and reversing when bookings slow.[3][6] Both are defensible — the answer is set by contract terms, not by the NAICS code — so check billing structure company by company rather than assuming a level-wide trait. What is common to all four is high operating leverage: fixed engineering and plant overhead means margins expand fast in an upcycle and compress hard in a downturn.

6. What drives demand

Because demand is derived, it rises and falls with customers' willingness to invest. Three drivers are common to all four; the differences are in which end market each one rides.

Common tailwinds:

1. Automation to offset labor. The dominant structural driver across every child. PMMI found 95% of surveyed consumer-packaged-goods companies struggling to hire skilled operators and technicians, and BLS reports that only 6.6% of food-manufacturing plants used robots in 2022 — a low base against projected ~22% output growth for food and beverage manufacturing from 2024 through 2034.[15][3] The same logic drives optimization controls in sawmills and automation across plastics and printing.[5][6]
2. Reshoring and new-factory construction. One tracker logs roughly $1.66 trillion of announced U.S. manufacturing investment since January 2025; separately, Commerce said semiconductor and electronics companies had announced nearly $450 billion of U.S. private investment as of January 2025.[12][30] Both children stress the same caveat: announcements are not completed spending, and construction schedules determine when equipment orders actually land.[4][6] Add lumber reshoring for the wood child — U.S. mills now supply roughly 75% of domestic softwood consumption, having added ~8.7 billion board feet of capacity since 2016 — and CHIPS-funded fabs for the semiconductor child.[12][23]
3. The capital-spending cycle itself (the swing factor). When customers have cash and confidence they green-light big lines; when they retrench, they defer. This is the single biggest short-term mover and the source of the group's cyclicality. The current cycle is visibly split: Lam's revenue fell from $17.43B (FY2023) to $14.91B (FY2024) before recovering to $18.44B (FY2025), while over roughly the same window Kadant's capital-equipment revenue fell 16%, Hillenbrand's process-solutions backlog fell 9%, and BLS payrolls in the all-other child fell 4.2% year over year.[8][19][24][28]

Where the children part ways — the end market each is chained to:

• Semiconductor machinery → chip capex, itself driven by AI data centers, memory, and each new transistor node. Industry chip capex is projected near $200 billion in 2026, with TSMC alone more than a quarter of it; in 2025 wafer-processing equipment sales rose 12%, test billings 55%, and assembly and packaging 21%.[7][12] This is the strongest demand backdrop of the four — and the most geographically concentrated, with China, Taiwan, and Korea together about 79% of global equipment spending.[7]
• Food machinery → food-processor capex, a defensive base (people eat in all weathers) with a soft near-term top: more than half of food processors surveyed planned to cut 2026 capital budgets, with the aggregate rising only ~2.9% and spending skewed toward brownfield upgrades over new plants.[16] Within the end market, meat and poultry is the largest slice (~29.2% of equipment demand), then prepared foods (~14.0%) and dairy (~12.4%).[15]
• Sawmill/paper → housing and packaging. The wood side rides mortgage rates and homebuilding — U.S. housing starts fell ~4% in 2024 and a further ~2.3% in 2025, and a U.S. Forest Service study puts softwood-lumber demand elasticity to starts at 0.59.[23] The paper side is tilting hard from graphic paper toward packaging, tissue, and recycling: 2025 U.S. paper and paperboard production fell 3.7% to 66.3 million tons, but containerboard ran 36.1 million tons at a 91.9% operating rate and packaging paper grew 1.7%, while printing-and-writing capacity fell 13.9% to 7.7 million tons.[22]
• All-other → broad manufacturing capex, the most diffuse and the most internally dispersed. In Q3 2025 North American injection-molding-machine shipments rose 30% sequentially and 4.2% year over year while single-screw extruders fell 24.2% and twin-screw extruders 38.6% — three equipment types in three different cycles inside one child.[26] The whole child is sensitive to interest rates, since customers finance equipment and higher rates lengthen payback.[6]

7. Regulation

None of these industries is economically regulated — there is no price setting, rate base, or reimbursement regime like a utility, REIT, or hospital. Regulation instead shapes product design, creates replacement cycles, and — in one child — directly gates revenue. The intensity ranges from light to unusually heavy across the four:

• Machine and worker safety (all four, light-touch but universal). Equipment must meet U.S. Occupational Safety and Health Administration (OSHA) machine-guarding rules — 29 CFR 1910.212 generally, with hazardous-energy control (lockout/tagout) under 1910.147 for servicing — plus voluntary consensus standards (ANSI, ISO, UL); European sales need CE marking. For the wood child, OSHA's woodworking and sawmill standards (1910.213 and 1910.265) directly drive demand for guarded, retrofitted equipment.[31]
• Sector-specific hygiene and materials standards (food child). Food-contact machinery must meet 3-A Sanitary Standards and support the customer's obligations under the FDA Food Safety Modernization Act (FSMA) preventive-controls regime and USDA FSIS sanitation performance standards. Certification is a commercial requirement that both favors established makers and forces a periodic replacement cycle — regulation as a moat and a demand engine, not just a cost.[32]
• Environmental rules (mostly on the customer, sometimes on the builder). Emissions, effluent, and energy requirements on pulp and paper mills push spending toward more efficient and recycling-oriented equipment, and Kadant warns that environmental requirements can force abatement investment, redesign, or temporary operating restrictions; on the builder side, EPA's area-source air-toxics standards cover industrial machinery and equipment finishing operations.[5][6]
• Trade policy (all four, the shared swing variable) — now with numbers. Tariffs on imported machinery can protect domestic builders, while tariffs on imported steel, castings, and components raise their input costs. Numerous steel and aluminum categories face 50% Section 232 tariffs, imported Chinese pulp-and-paper machinery carries a 25% Section 301 tariff, and combined softwood-lumber duties on Canada reached roughly 45% on most producers by late 2025 — the last of which tilts production toward U.S. mills and indirectly lifts sawmill-equipment demand.[19][23] Against that, plastics-equipment imports at 68.8% of domestic shipments and mold imports at 93% show how exposed the all-other child is to the same measures on the input side.[26]
• Export controls and industrial policy (the semiconductor exception — uniquely heavy). This is what sets one child apart, and the timeline has extended. U.S. Bureau of Industry and Security (BIS) rules in October 2022, October 2023, April 2024, and December 2024 progressively restricted advanced chip-tool sales to China; in August 2025 BIS removed license-free treatment previously available to certain foreign-owned fabs in China. Allied governments followed: the Netherlands added certain metrology and inspection systems to its control list in January 2025, and the EU incorporated Dutch national controls in November 2025.[11] At the same time the CHIPS and Science Act ($52.7B, ~$33.7B in direct awards finalized by early 2025, plus a 25% equipment tax credit) is pulling domestic fab construction — and tool orders — forward, though several marquee builds have slipped (Intel's Ohio fab is now targeted around 2030).[12] Dual-use additive-manufacturing gear in the all-other child faces lighter EAR/ITAR export rules.[6]

Net: for three children regulation is a background cost and a mild demand tailwind; for the semiconductor child it is a front-page revenue lever moved by government policy.

8. Consolidation

All four children are consolidating, but through different vehicles — a useful tell about who holds the value:

• Semiconductor: public M&A among the mid-caps; giants stay put. The defining recent deal is Axcelis's ~$4.4B acquisition of Veeco (announced October 2025), which would create the #4 U.S. wafer-fab-equipment supplier at combined pro-forma 2024 revenue near $1.7 billion — still a fraction of the big three. The giants are too large to merge without antitrust objection, so scale-building happens below them.[10]
• Food: a public mega-merger plus PE roll-ups plus a spin-off — all at once. JBT + Marel (2025) created a ~$3.8B pure-play; KKR's Fortifi (which closed Provisur in 2025, alongside Bettcher and Frontmatec), Warburg Pincus's Duravant, and Leonard Green's ProMach stitch private specialists into end-to-end platforms; and Middleby un-bundled its food unit into Midera (separation completed July 2026) to close a valuation gap. Every consolidation strategy is on display in one child.[13][14][18]
• Sawmill/paper: a domestic roll-up under a foreign oligopoly. USNR has rolled up much of U.S. sawmill systems (Timber Automation, Norwood, Burton Mill Solutions), Kadant does serial bolt-ons, and Valmet earlier merged with automation firm Neles — while the global high-speed paper-machine market remains an oligopoly of Voith, Valmet, and Andritz and HOMAG holds a claimed 30%-plus share of global woodworking machinery.[5][20][21]
• All-other: PE consolidation and a brutal 3D-printing shakeout. Private equity is the consolidator in mature niches — Bain Capital completed a ~51% majority investment in Milacron in March 2025, with Hillenbrand retaining 49%; Gamut owns Davis-Standard.[25] In additive manufacturing, the 2020–21 bubble deflated violently: Nano Dimension bought Desktop Metal and Markforged in 2025, Desktop Metal filed Chapter 11 within months and its assets were sold off, and Markforged was resold to Stratasys for $42.5 million against the $116 million Nano Dimension had just paid.[27]

The common strategic logic across all four: buyers increasingly want fewer suppliers, integrated lines, and one throat to choke on service — so installed base, aftermarket reach, and a broad process portfolio beat single-machine excellence. Scale is being bought more than built.

9. Risks

The group shares a cyclical spine; the child-specific risks are where the real differentiation sits.

• Cyclicality (all four). Big-ticket orders are the first thing customers defer in a downturn; revenue and margins swing hard on operating leverage. Lam's $17.43B → $14.91B → $18.44B fiscal-year arc is the sharpest illustration; Kadant's 2025 capital-equipment revenue fell 16%, Hillenbrand's segment backlog 9%, Midera's organic sales 4.4%.[8][14][19][24] The semiconductor child swings the most (memory capex is famously boom-and-bust); the food child the least (defensive end market).[3][4]
• The aftermarket cushions but does not cancel the cycle. In the same 2025 that Kadant's capital equipment fell 16%, its parts and consumables grew 11% — the razor-and-blade divergence working as designed, and the single best reason to weight recurring mix over headline sales at every one of the four.[19]
• Customer concentration (all four, differently expressed). Applied reported two customers at approximately 19% and 15% of fiscal 2025 revenue; food demand is tied to a handful of large processors' capital plans; sawmill and paper buyers are a small universe in which one delayed project moves a builder's year.[3][5][8]
• The undercount cuts both ways as a risk. Because so much value is private or offshore (§3), public investors see only a slice, and the diversified/foreign names carry currency and non-core-business dilution.
• Import competition and FX (all four). Formidable German, Swiss, Italian, Japanese, Finnish, Austrian, Dutch, and increasingly Chinese makers set the bar; plastics-equipment imports at 68.8% of domestic shipments and molds at 93% quantify the exposure in one child.[26] A strong dollar or trade retaliation whipsaws competitiveness.[5][6]
• Tariff and input-cost whipsaw (all four). Steel and aluminum are the dominant bill of materials; 50% Section 232 metal tariffs simultaneously compress builders' margins and raise customers' project costs, delaying orders — a double hit made worse by fixed-price backlog that embeds yesterday's steel, component, and freight assumptions.[3][6][19]
• Labor constrains the builders too, not just their customers (all four). Automation demand exists because operators are scarce — but the OEMs themselves need controls engineers, welders, field technicians, and commissioning specialists, and service growth stalls without technician density near the installed base.[3][6] In semiconductor, an SIA/Oxford Economics study projected 67,000 expected new U.S. industry jobs — 58% of the total — could go unfilled by 2030 at then-current degree-completion rates.[4]
• Concentration and single-child dominance (rollup-specific). A public-market bet on "industrial machinery" is really a bet on semiconductor capex plus a handful of diversified industrials; the group is not as diversified as the four-industry label implies.
• Child-specific tail risks: China/export-control exposure, ~79% demand concentration in China/Taiwan/Korea, and node-transition technology risk in semiconductor;[4][7][11] avian influenza and protein-price shocks in food, where even the largest pure-play posted a $49.7M loss from continuing operations in 2025;[3][13] structural decline of graphic paper (printing-and-writing capacity −13.9% in a single year) and housing-rate sensitivity in sawmill/paper;[22][23] and outright going-concern and cash-burn risk in the listed 3D-printing pure-plays of the all-other child.[27]

10. How to invest, and the outlook

Matching route to child (tickers and multiples belong here).

• Broad, liquid public exposure ≈ semiconductor. The cleanest large-cap ways to own this rollup are the chip-tool makers — AMAT, LRCX, KLAC — plus specialists (ACLS, ONTO, MKSI, KLIC) and foreign leaders (ASML via ADR, Tokyo Electron). Broad semiconductor ETFs (SOXX, SMH) hold them heavily but blend in chipmakers. These are volatile cyclicals; entry valuation matters as much as company quality.[4][8][9]
• Focused public pure-plays are scarce elsewhere. Food offers JBTM (global, protein-heavy, ~50% recurring) and the newly spun MFP (40% aftermarket); sawmill/paper offers the diversified proxy KAI (~71% parts and consumables, 2026 revenue guided to roughly $1.18–1.20B), foreign Valmet/Andritz, and — for woodworking specifically — Dürr/HOMAG; all-other offers speculative small-cap 3D printers (SSYS, DDD, NNDM, VLD), diversified industrials (HI, NDSN), or foreign press builders (Koenig & Bauer, Bobst). Track the same metrics everywhere — backlog, book-to-bill, recurring-revenue mix, margins — plus each child's leading demand signal (chip capex, food-processor capex intentions, housing starts, broad manufacturing capex).[13][14][19][21][24][27]
• Downstream proxies are an honest second-best. Because demand is derived, the customers' own fortunes lead the machine orders: lumber and packaging producers for the wood/paper child, chipmakers for semiconductor, food processors for food. When their margins are strong, machinery orders follow.[5]
• Private routes hold much of the value the public market can't reach. In food, sawmill/paper, and all-other, private equity is the dominant owner and consolidator (KKR/Fortifi, Warburg/Duravant, Leonard Green/ProMach, Bain/Milacron, Gamut/Davis-Standard, USNR), alongside family firms reachable only through M&A; in semiconductor the private layer is the subsystem, component, and venture-stage tool suppliers beneath the majors.[3][4][5][6][18][25] This is illiquid and largely closed to public investors, but it is where a great deal of the value creation is happening.
• Diligence questions that travel across all four children. The revised child primers converge on the same short list: organic orders rather than acquisition-driven revenue; backlog age, cancellation rights, deposits, and expected project margin; equipment-versus-aftermarket mix and parts capture; service-technician capacity near the installed base; customer and end-market concentration; and pricing lag against metals and components. A nominally high-margin OEM can turn out to be a project contractor with fragile backlog, while a slow-growth machine builder with strong parts capture behaves like an industrial service franchise.[3][5][6]

Outlook (a forward-looking judgment, not a reported fact). The four are not moving together, which is the whole point of a rollup view:

• Semiconductor has the strongest backdrop — SEMI projects record global equipment sales of ~$139B in 2026 and ~$156B in 2027, up from ~$135B in 2025 — but the oldest risk (cyclicality, China policy, whether AI merely front-loads the cycle) is live.[4][7]
• Food is steady, not spectacular: an automation-versus-labor tailwind and a regulation-driven replacement cycle against a cautious 2026 capex year. PMMI and FPSA put 2025 U.S. food and beverage processing-machinery shipments at $6.2B, up 3.2%, reaching ~$6.7B by 2027; independent forecasts see ~3.3% annual growth into the early 2030s — the two lenses differ in scope but agree on the order of magnitude.[15][16][17]
• Sawmill/paper is soft near-term (weak housing, mill closures, graphic-paper capacity down 13.9% in 2025) with a medium-term case resting on lumber reshoring under tariff protection, automation retrofits, and the graphic-to-packaging machine-conversion cycle; the aftermarket annuity is the anchor through it.[19][22][23]
• All-other is a cyclical, reshoring-levered bet currently in the trough — payrolls down 4.2% year over year, extruder shipments down sharply even as injection molding recovers — whose additive-manufacturing subset is still consolidating and whose steadier exposure is diversified or private.[6][26][27][28]

The through-line across all four: expect value to keep shifting from selling machines to selling the parts, service, and software that keep them running — Kadant's move from 66% to 71% recurring in a single soft year is this trend in miniature — and expect consolidation to keep favoring scale and installed base over single-machine excellence. For most investors, the practical takeaway is that "industrial machinery" is not one bet — it is four, dominated in the public market by one (semiconductor), with the other three best reached through foreign listings or private equity.

---

Sources

Figures for this level are drawn from our ground-truth federal compilation; company-, market-, and policy-level facts are carried up from the four child primers (333241, 333242, 333243, 333248).

1. U.S. Census Bureau, 2022 Economic Census — Concentration Ratios and Statistics for NAICS 33324 (Industrial Machinery Manufacturing) and County Business Patterns 2023 — receipts, firms, establishments, employment, payroll, concentration ratios, and HHI (Histometrics ground-truth compilation, stats-33324.md). https://www.census.gov/programs-surveys/economic-census.html; https://www.census.gov/programs-surveys/cbp.html
2. U.S. Small Business Administration, Table of Small Business Size Standards Matched to NAICS Codes (2023) — employee ceilings of 500 (333241), 1,500 (333242), 550 (333243), and 750 (333248). https://www.sba.gov/document/support-table-size-standards
3. Histometrics child primer, Food Product Machinery Manufacturing (NAICS 333241), and its cited federal figures — Economic Census 2022 / CBP 2023 (receipts ~$6.62B; 399 firms; 525 establishments; 19,033 employees; ~$1.57B payroll; CR4 25.3%; CR8 36.1%; CR20 56%; CR50 75.3%; HHI 273.6).
4. Histometrics child primer, Semiconductor Machinery Manufacturing (NAICS 333242), and its cited federal figures — Economic Census 2022 / CBP 2023 (receipts ~$13.75B; 144 firms; 170 establishments; 28,799 employees; ~$4.34B payroll; CR4 60.7%; CR8 78.2%; CR20 92.7%; HHI 1,153), plus segment shares and the SIA/Oxford Economics labor projection.
5. Histometrics child primer, Sawmill, Woodworking, and Paper Machinery Manufacturing (NAICS 333243), and its cited federal figures — Economic Census 2022 / CBP 2023 (receipts ~$3.97B; 316 firms; 391 establishments; 12,466 employees; ~$1.01B payroll; CR4 25.7%; CR8 38.7%; CR20 57.5%; CR50 77.4%; HHI 255), plus USNR, Wood-Mizer, and Voith ownership.
6. Histometrics child primer, All Other Industrial Machinery Manufacturing (NAICS 333248), and its cited federal figures — Economic Census 2022 / CBP 2023 (receipts ~$15.9B; 1,678 firms; 1,750 establishments; 51,199 employees; ~$4.03B payroll; CR4 ~11%; HHI 65.7).
7. SEMI, Global Semiconductor Equipment Billings Reached $135 Billion in 2025; Global Total Semiconductor Equipment Sales Forecast to Reach a Record $139 Billion in 2026; …$156 Billion in 2027 — 2025 segment growth (wafer processing +12%, test +55%, assembly and packaging +21%) and the ~79% China/Taiwan/Korea share. https://www.semi.org/en/SEMI-Reports-Global-Semiconductor-Equipment-Billings-Reached-135-Billion-in-2025; https://www.semi.org/en/semi-press-release/global-total-semiconductor-equipment-sales-forecast-to-reach-a-record-of-dollar-139-billion-in-2026-semi-reports; https://www.semi.org/en/semi-press-release/global-semiconductor-equipment-sales-projected-to-reach-a-record-of-156-billion-dollars-in-2027-semi-reports
8. Applied Materials, 2025 Form 10-K (FY2025 revenue $28.37B, gross margin 48.7%, R&D $3.57B, two customers ~19% and ~15% of revenue); Lam Research, 2025 Form 10-K (FY2025 revenue $18.44B, gross margin 48.7%, R&D $2.10B, systems $11.49B, customer support $6.94B, FY2023 $17.43B and FY2024 $14.91B); KLA, 2025 Annual Report (FY2025 revenue $12.16B, gross margin 60.9%). https://www.sec.gov/Archives/edgar/data/6951/000162828025056742/amat-20251026.htm; https://www.sec.gov/Archives/edgar/data/707549/000070754925000075/lrcx-20250629.htm; https://www.sec.gov/Archives/edgar/data/319201/000119312525213412/d912971dars.pdf
9. ASML, Q4 2025 Financial Results (2025 sales €32.7B, gross margin 52.8%, net income €9.6B; service roughly a quarter of revenue; High-NA system pricing ~$380M). https://www.asml.com/en/news/press-releases/2026/q4-2025-financial-results
10. Axcelis Technologies, Axcelis Technologies and Veeco Instruments to Combine (2025) — ~$4.4B transaction; combined pro-forma 2024 revenue ~$1.7B. https://investor.axcelis.com/news-releases/news-release-details/axcelis-technologies-and-veeco-instruments-combine-creating
11. U.S. Bureau of Industry and Security, Commerce Strengthens Export Controls (2024) and Department of Commerce Closes Export Controls Loophole for Foreign-Owned Semiconductor Fabs in China (August 2025); CSIS, The True Impact of Allied Export Controls… (2024) — 2022–2024 rules, Dutch January 2025 and EU November 2025 measures. https://www.bis.gov/press-release/commerce-strengthens-export-controls-restrict-chinas-capability-produce-advanced-semiconductors-military; https://www.bis.gov/press-release/department-commerce-closes-export-controls-loophole-foreign-owned-semiconductor-fabs-china; https://www.csis.org/analysis/true-impact-allied-export-controls-us-and-chinese-semiconductor-manufacturing-equipment
12. Semiconductor Industry Association and TSMC — CHIPS and Science Act ($52.7B; ~$33.7B awarded; 25% equipment tax credit; TSMC Arizona ~$165B; Intel Ohio timeline); U.S. Department of Commerce (January 2025) — nearly $450B of announced U.S. private semiconductor and electronics investment; FinancialContent/Token Ring — industry chip capex ~$200B in 2026 with TSMC over a quarter. https://www.semiconductors.org/chip-supply-chain-investments/; https://pr.tsmc.com/english/news/3122; https://www.commerce.gov/news/press-releases/2025/01/us-department-commerce-announces-chips-incentives-awards-corning; https://markets.financialcontent.com/wral/article/tokenring-2026-1-1-the-silicon-renaissance-us-chips-act-enters-production-era-as-intel-tsmc-and-samsung-hit-critical-milestones
13. JBT Marel Corporation, 2025 Form 10-K and Fourth Quarter and Full Year 2025 Results — revenue $3.798B, ~50% recurring, gross margin 35.1%, adjusted EBITDA margin 15.8%, $49.7M loss from continuing operations. https://www.sec.gov/Archives/edgar/data/1433660/000143366026000053/jbt-20251231.htm; https://ir.jbtmarel.com/news/press-releases/detail/530/jbt-marel-corporation-reports-fourth-quarter-and-full-year-2025-results-and-establishes-2026-guidance-with-continued-growth
14. Midera Food Processing, Registration Statement (Form 10), and Middleby Corporation, Middleby Completes Spin-Off of Midera Food Processing (July 2026) — net sales $853.2M, $341M (40%) aftermarket, gross margin 36.2% (39.6% in 2024), operating margin 11.9%, adjusted EBITDA margin 17.8%, organic sales −4.4%. https://www.sec.gov/Archives/edgar/data/2088281/000119312526241891/d14360dex991.htm; https://www.middleby.com/newsroom/middleby-completes-spin-off-of-midera-food-processing/
15. PMMI and the Food Production Solutions Association, 2026 Processing State of the Industry Report ($6.2B of 2025 U.S. shipments, +3.2%, ~$6.7B by 2027; meat and poultry 29.2%, prepared foods 14.0%, dairy 12.4%), and PMMI, Finding Skilled Workers Remains a Major Challenge (95% of surveyed CPG companies). https://www.pmmi.org/news/pmmi-and-fpsa-release-inaugural-2026-processing-state-of-the-industry-report-and-infographic; https://www.pmmi.org/news/finding-skilled-workers-remains-a-major-challenge-that-may-worsen
16. Food Processing, 2026 Capital Spending Outlook: Tightening the Belt — more than half of processors planning cuts; aggregate up ~2.9%. https://www.foodprocessing.com/business-of-food-beverage/capital-spending/article/55362973/2026-capital-spending-outlook-tightening-the-belt
17. Grand View Research, United States Food Processing Equipment Market Size & Outlook, 2033 (~3.3% CAGR) — via child primer 333241. https://www.grandviewresearch.com/horizon/outlook/food-processing-equipment-market/united-states
18. Private-equity consolidation of food machinery — Food Engineering, Fortifi Completes Acquisition of Provisur Technologies (KKR platform incl. Bettcher, Frontmatec); Warburg Pincus / Duravant; ProMach under Leonard Green & Partners. https://www.foodengineeringmag.com/articles/103418-fortifi-completes-acquisition-of-provisur-technologies; https://warburgpincus.com/investments/duravant/; https://www.promachbuilt.com/press-releases/promach-announces-new-ownership-by-leonard-green-and-partners/
19. Kadant Inc., 2025 Form 10-K (parts and consumables 71% of 2025 revenue, up from 66% in 2024; wood processing ~22% of consolidated revenue; consolidated gross margin 45.2% vs 44.3%; capital-equipment revenue −16%; Section 301/232 tariff exposure) and company overview (2024 revenue ~$1.05B; ~$3.9B market capitalization; 2026 guidance $1.178–1.203B; $0.36 quarterly dividend, 13-year raise streak). https://www.sec.gov/Archives/edgar/data/886346/000088634626000018/kai-20260103.htm; https://www.stocktitan.net/overview/KAI/
20. Valmet Oyj, 2025 Annual Review (2025 net sales €5.197B; comparable EBITA margin 11.9%) and Financial Statements Review 2024 (order backlog ~€4.5B; ~€2.5B of orders including an Arauco pulp-mill order over €1B); ANDRITZ AG, 2025 Annual Report (Pulp & Paper backlog €2.833B; EBITA margin 10.3%). https://www.valmet.com/globalassets/investors/reports--presentations/annual-reports/2025/valmet-annual-review-2025.pdf; https://www.valmet.com/media/news/stock-exchange-releases/2025/valmets-financial-statements-review-january-1--december-31-2024orders-received-increased-but-net-sales-and-comparable-ebita-remained-steady-in-2024/; https://www.andritz.com/resource/blob/105246/6ad9400073c46323b95b1be977870245/andritz-annual-report-2025-data.pdf
21. HOMAG Group company profile (estimated global woodworking-machinery share above 30%; ~€1.4B sales) and 2025 results (order intake €1.380B); Dürr AG 2025 results (HOMAG operating margin 5.5%, up from 3.6%). https://www.homag.com/en/company/about-us/homag-group; https://www.homag.com/en/company/news/news/article/homag-group-with-significant-improvement-in-earnings-in-fiscal-year-2025; https://www.durr-group.com/en/media/news/news-detail/view/duerr-group-significantly-improves-profitability-114080
22. American Forest & Paper Association, 66th Annual Paper Industry Capacity and Fiber Consumption Survey — 2025 production 66.3M tons (−3.7%); containerboard 36.1M tons at 91.9% operating rate; printing-and-writing capacity −13.9% to 7.7M tons; tissue ~7.8M tons; packaging paper +1.7%. https://www.afandpa.org/news/2026/afpa-releases-66th-annual-paper-industry-capacity-and-fiber-consumption-survey
23. Housing, lumber capacity, and softwood duties — Farm Credit East / NAHB (U.S. housing starts −4% in 2024, −2.3% in 2025); U.S. Forest Service, Factors Affecting U.S. Softwood Lumber Demand (0.59 elasticity to starts); U.S. Lumber Coalition / NAHB (~75% domestic self-sufficiency; +8.7 billion board feet since 2016); Congressional Research Service, U.S.–Canada Softwood Lumber Trade (combined duties ~45% by late 2025). https://www.farmcrediteast.com/en/resources/Industry-Trends-and-Outlooks/Reports/2503KEP_Forestry-Outlook; https://research.fs.usda.gov/treesearch/80470; https://www.nahb.org/blog/2025/12/sawmill-lumber-production; https://www.congress.gov/crs-product/R48781
24. Hillenbrand, Inc., 2025 Form 10-K (Advanced Process Solutions FY2025 revenue $2.069B, gross margin 34.0% vs 35.5%, backlog $1.522B, −9%; Milacron 49% stake) and Nordson Corporation, 2025 Form 10-K (FY2025 revenue ~$2.7B; gross margin 55.2%). https://www.sec.gov/Archives/edgar/data/1417398/000162828025053207/hi-20250930.htm; https://www.sec.gov/Archives/edgar/data/72331/000007233125000144/ndsn-20251031.htm
25. Milacron / Hillenbrand, Bain Capital Completes Majority Investment in Milacron (March 2025; Bain ~51%, Hillenbrand 49%). https://www.milacron.com/press-release/bain-capital-completes-majority-investment-in-milacron-a-leading-global-provider-of-highly-engineered-plastic-processing-solutions/; https://hillenbrand.com/corporate-news/hillenbrand-completes-sale-of-majority-stake-in-milacron-injection-molding-and-extrusion-business/
26. Plastics Industry Association, Tariffs and the U.S. Plastics Industry Supply Chain (2024 imports: plastics equipment 68.8% of domestic shipments; molds 93%) and 2025 Q3 Committee on Equipment Statistics Report (injection molding +30% sequential and +4.2% year over year; single-screw extruders −24.2%; twin-screw −38.6%). https://www.plasticsindustry.org/blog/tariffs-and-the-u-s-plastics-industry-supply-chain-where-do-we-stand/; https://www.plasticsindustry.org/newsroom/plastics-industry-association-releases-2025-q3-committee-on-equipment-statistics-report/
27. Additive-manufacturing shakeout and results — 3D Printing Industry / VoxelMatters (Nano Dimension's purchases of Desktop Metal and Markforged, Desktop Metal's Chapter 11, and the $42.5M Markforged resale to Stratasys against $116M paid; Velo3D FY2025 guide $50–60M and going-concern doubt); 3D Systems FY2025 results (revenue $386.9M, gross margin 33.9%, $(45.4)M adjusted EBITDA); Stratasys FY2025 results (~$551M); CONTEXT/TCT on industrial-system shipment declines and an expected 2026 recovery. https://3dprintingindustry.com/news/nano-dimension-takes-substantial-loss-on-sale-of-markforged-to-stratasys-251862/; https://www.voxelmatters.com/velo3d-posts-13-6-million-in-q3-2025-revenue-following-nasdaq-uplisting/; https://www.sec.gov/Archives/edgar/data/910638/000162828026015798/a3d2025-12x31earningsrelea.htm; https://investors.stratasys.com/news-events/press-releases/detail/975/stratasys-releases-fourth-quarter-and-full-year-2025; https://www.tctmagazine.com/context-figures-q4-2024-tough-year-am-3d-printer-shipments-down/
28. U.S. Bureau of Labor Statistics, Table B-1b — Employees on Nonfarm Payrolls by Industry (NAICS 333248: 66,200 in April 2026 vs 69,100 in May 2025). https://www.bls.gov/ces/data/employment-and-earnings/2026/table1b_202605.htm
29. U.S. Small Business Administration / Federal Register, Small Business Size Standards: Revision of Monetary Based Size Standards (July 2022) — NAICS 333248 created by combining 333244 and 333249, with 333249 representing ~87–89% of the combined base; ~96.7% of firms below the 750-employee threshold. https://www.govinfo.gov/content/pkg/FR-2022-07-05/pdf/2022-13250.pdf
30. IoT Analytics, US Manufacturing Reshoring Boom: What the Data Says (~$1.66 trillion in announced private commitments since January 2025). https://iot-analytics.com/us-manufacturing-reshoring-boom-what-the-data-says/
31. U.S. Occupational Safety and Health Administration — 29 CFR 1910.212 (general machine guarding), 1910.147 (control of hazardous energy), 1910.213 (woodworking machinery), and 1910.265 (sawmills). https://www.osha.gov/laws-regs/regulations/standardnumber/1910/1910.212; https://www.osha.gov/laws-regs/regulations/standardnumber/1910/1910.147; https://www.osha.gov/laws-regs/regulations/standardnumber/1910/1910.213; https://www.osha.gov/laws-regs/regulations/standardnumber/1910/1910.265
32. Food-contact hygiene regime — 3-A Sanitary Standards, Inc.; U.S. Food and Drug Administration, FSMA Final Rule for Preventive Controls for Human Food; USDA Food Safety and Inspection Service, Sanitation Performance Standards Compliance Guide. https://www.3-a.org/documents/3-a-sanitary-standards-and-3-a-accepted-practices; https://www.fda.gov/food/food-safety-modernization-act-fsma/fsma-final-rule-preventive-controls-human-food; https://www.fsis.usda.gov/inspection/compliance-guidance/sanitation-performance-standards-compliance-guide