U.S. Fabricated Metal Product Manufacturing (NAICS 332): An Investor Primer

A Histometrics rollup primer for a general investing audience — public-market and private investors alike. The North American Industry Classification System (NAICS) is the U.S. government's standard code set for industries; a three-digit code like this one is a "subsector." NAICS 332 sits inside the manufacturing sector (codes 31–33) and rolls up nine four-digit "industry groups." This page synthesizes the nine child primers plus our ground-truth federal statistics for the whole subsector. Figures are reported facts with citations; statements about the future are labeled as expectations or judgments.

1. Overview

Fabricated Metal Product Manufacturing is the layer of U.S. industry that takes mill metal — steel, aluminum, brass, and specialty alloys bought as coil, plate, bar, wire, and extrusion — and turns it into finished parts and structures. It sits one step downstream of the steel and aluminum mills (NAICS subsector 331) and one step upstream of the machinery, vehicle, building, and equipment makers who assemble the finished product. Almost nothing mechanical or built exists without it: the frame of a building, the valve on a pipeline, the bolt in a jet engine, the can in a vending machine, the hinge on a door, the spring in a mattress, the coating on a guardrail.

It is one of the largest and most fragmented corners of American manufacturing. Our federal ground-truth extract puts it at about $466 billion of shipments, 1.44 million workers, 53,790 plants, and 49,631 firms — and yet the four largest firms hold just 5% of revenue, and even the 50 largest hold only about 21%.[1][3] There is no pure-play stock for the subsector, no index or exchange-traded fund (ETF) that captures it, and no single company that is "fabricated metal." Public-market investors reach it as segments buried inside larger, diversified companies; private investors have the more direct — and often better-matched — routes, because the overwhelming majority of the base is family-owned shops, employee-owned companies, and private-equity (PE) platforms.

The one idea to carry through this page: NAICS 332 is not a market you can buy — it is a container for nine very different businesses, and its distinctive value to an investor is the contrast across them. Three of the nine children make up roughly 70% of revenue and 72% of jobs; the other six range from mid-sized to tiny. They serve end markets that barely overlap — construction, aerospace, defense, energy, water, autos, packaging, housing — which is exactly why the subsector as a whole rarely moves in unison, and why the interesting analysis is child-by-child.

Three findings from the rebuilt child research change the level view, and each is developed below. First, fiscal 2025 was a margin reset that ran across the subsector, not through one corner of it — selling prices reset faster than metal costs, and every child that carries disclosed proxies shows the same mechanism, with the damage falling almost entirely on the commodity end (Sections 5 and 8).[19][22][24] Second, reported dollars here carry a large price component: producer prices have outrun volumes in several children, and the one federal real-output series anywhere in the subsector fell while nominal output rose, so revenue growth needs a volume test (Sections 3 and 9).[19] Third, the federal statistics stop being uniform below this level — several children could not obtain or could no longer confirm current revenue, firm-count or concentration figures for their own children, and one has withdrawn them outright (Sections 2 and 3).[25][26][27]

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

All nine share the same physical economics (buy metal, add engineering and fabrication, sell for more) and the same master constraints (metal-cost pass-through, capacity utilization, freight-limited geography, certification moats). What separates them is what they make, who buys it, how concentrated it is, and how an investor can actually own it. The table below is the heart of this rollup; it is ordered by share of the subsector's 2022 receipts. Tickers are held back to Sections 4 and 10 by design.

CR4 = combined revenue share of the four largest firms. HHI = Herfindahl-Hirschman Index, a concentration score from near 0 (perfect fragmentation) to 10,000 (a monopoly); U.S. antitrust agencies treat below 1,500 as "unconcentrated." CR4/HHI are 2022 Economic Census; revenue share is 2022; jobs share is 2023 County Business Patterns.[2][3] † The hardware child's rebuilt research declines to publish a current receipts total or any concentration ratio, having been unable to confirm one from a published federal source; its revenue share and CR4/HHI here rest on our level extract alone and are no longer corroborated from below.[27]

Six contrasts do the real work of this level:

1. Three children carry it. Structural/architectural metals (3323), the valves-and-all-other box (3329), and machine shops/fasteners (3327) together are ~70% of revenue and ~72% of jobs; the other six are small-to-tiny. The subsector's fortunes are mostly these three, whose end markets (construction, energy/defense, and general machining) rarely peak together.

2. Revenue and jobs disagree — and that is informative. Machine shops (3327) hold a bigger share of jobs (~25%) than of revenue (~17%) because they are labor-intensive job shops with low revenue per worker (~$223,000); the can/tank/boiler child (3324) is the opposite — high revenue per worker (~$493,000) because beverage-can revenue carries a big aluminum pass-through, and finishing services (3328) show low revenue per worker (~$262,000) because those shops bill for labor on metal they don't own.[21][22][23] These ratios divide 2022 receipts by 2023 headcount, so read them as directional; judge each child by the currency it's big in.

3. Concentration is a per-child fact, not a subsector fact — and the level HHI is not the floor. The children run from atomistic machine shops and fasteners (HHI 6.4, and ~3.5 in the machine-shop half alone) to moderately concentrated kitchen/handtool niches (296) and the can oligopoly inside 3324.[21][22][25] The subsector's own HHI of 13.7 is lower than seven of the nine children — but not lower than 3327, which comes in at 6.4 for exactly the same reason the level does: it too pools unrelated markets. Correct the reading rather than the ranking. Meanwhile the genuine oligopolies are one or two levels down: small-arms ammunition at CR4 75.4% and HHI 1,766, large-caliber ammunition at CR4 87.1%, faucets at CR4 63.9%, cans at CR4 51.4% and HHI 844, heat treating at CR4 41.8% and HHI 517.[20][22][23] A low number at this level tells you nothing about the market you would actually be buying.

4. Investability doesn't track size. The coherent valve industry has no pure-play stock; the messier grab-bag (3329) hands you clean bearing and firearm pure-plays. Cans (a slice of 3324) hold nearly all the clean listed packaging exposure. In the whole of 3323, the cleanest definitional match to the federal product definition — metal ceiling-suspension grid — sits in the group's most fragmented, least investable corner.[19] Everywhere else, listed capital is a diluted segment inside a bigger company, and the deepest ownership is private.

5. Regulation stretches across two worlds. Most of the subsector answers to plant safety, environmental, product-standard, and trade rules. One corner — the ammunition and ordnance industries inside 3329 — adds an entire weapons regime (licensing, arms-export control, foreign-investment review). The rollup runs from drinking-water lead limits to gun-control politics.

6. Federal coverage is no longer uniform across the children — and that is new. Four digits is still solid ground; below it, the picture degrades unevenly. Hardware (3325) has withdrawn its receipts and concentration figures entirely.[27] Cutlery and handtool (3322) could obtain no current six-digit revenue, firm count or concentration for either of its children.[25] Spring and wire (3326) has no defensible per-child revenue split — the level is now more informative than the page beneath it.[26] Architectural and structural metals (3323) can publish a current six-digit revenue figure for only one of its six industries.[19] Inside 3324, the heavy-gauge tank child has withdrawn its receipts, firm count and concentration as unconfirmable.[22] Forging and stamping (3321) has revenue for four of five children — but from three different federal programs on three vintages, which cannot be added.[24] The practical rule this forces on an investor is simple: pick one federal program and stay inside it, and treat any market share built on a six-digit denominator in this subsector as fragile.

3. Size — the subsector rolled up

Our ground-truth federal statistics for NAICS 332 come from two U.S. Census Bureau programs, covering different years, so they are complementary rather than a single snapshot: the 2022 Economic Census (receipts, firms, concentration) and the 2023 County Business Patterns (CBP; establishments, employment, payroll). Every figure below is a reported federal value; nothing is estimated and no suppressed cell is filled in.[1][2][3]

How the children add up — and where they now stop. The employment side of the rollup is still perfect: the nine children's 2023 establishments sum exactly to 53,790, their employment exactly to 1,441,471, and their annual payroll to ~$96.77 billion against the level's $96.764 billion.[2][19][20][21][22][23][24][25][26][27] The 2022 receipts no longer close. Eight of the nine children still publish a receipts figure and those remain consistent with the level total; the hardware child has withdrawn both its receipts total and its concentration ratios.[27] We do not fill that cell by subtraction — the residual would be an inference from the parent total, not a published figure, and quoting it as hardware's revenue would manufacture a statistic. Firm counts still do not sum, and now cannot even be fully tested: the eight children that publish one already total ~49,757 against the level's 49,631, before hardware, which no longer publishes one. A company that manufactures in more than one industry group is counted once at the subsector level but appears in each group it operates in. Firm counts never sum across levels; establishment and employment counts do.

What the rollup shows. Pay averages ~$67,100 per worker across the subsector, and the spread by child is wide and meaningful — from ~$59,200 in the for-the-trade finishing and thermal-processing shops of 3328, through the low-$60,000s in labor-intensive machining and wire work, to ~$76,900 in the capital-intensive boiler, tank and can businesses of 3324.[2][22][23] Inside individual children the same gradient reappears and tracks qualification rather than process: forging and stamping runs from ~$60,000 in powder metallurgy to ~$87,000 in aerospace-approved nonferrous forging.[24] With 49,631 firms across 53,790 establishments, the overwhelming majority of companies run a single plant. And the concentration figures are extraordinary: an HHI of 13.7 is near the floor of the federal data. This is, at the company level, one of the most fragmented subsectors in U.S. manufacturing — the entire backdrop for the investment case, because it is why no pure-play stock exists and why PE can build platforms by rolling up regional shops.

Undercount and caveats. These are employer figures — establishments and firms with paid employees. They understate the true economic footprint in three ways, all running toward understatement:

• Owner-only (nonemployer) shops are excluded. The small/individual-ownership undercount bites hardest where tiny shops cluster — machine shops (3327), the misc-metal tail of 3329, small ornamental and architectural fabrication (3323), and stamping, where 821 of 1,094 plants have fewer than 50 workers — so the true operator count in those corners runs above the figures here.[2][24] It bites least in the capital-heavy corners, and 3329 is a genuinely employer-based, factory-payroll child with essentially no informal tail.[20]
• Captive fabrication is classified elsewhere. Forging, stamping, machining, structural work, or finishing done in-house inside a vehicle, aircraft, machinery, appliance, or shipbuilding plant is booked under that finished product's industry, not under 332. In heat treating, industry sources have long estimated commercial (for-hire) shops perform only ~10% of all U.S. work — the rest is captive.[23] Captive pipe fabrication sits in construction rather than manufacturing, and one private industrial contractor alone runs more than 450,000 square feet of pipe-fabrication shop entirely outside the merchant figures.[20] This omission is large and systemic.
• Imports supply much of U.S. consumption in the consumer-facing and commodity corners — cookware, cutlery, hardware, faucets, fasteners, bearings, ammunition — so domestic factory output materially understates what the U.S. market actually spends on these products. Two of these are now sized: the U.S. ball-bearing market runs near $8 billion against ~$6.0 billion of domestic output, and the ammunition market near $7.7 billion against $5.83 billion of factory shipments.[20][25][27]

A fourth caveat now belongs alongside the others: reported dollars carry a large price component. Producer prices for plate work rose 68.2% between December 2020 and December 2025 and for coating services 35.6% from December 2019 to June 2026, while the only federal real-output series anywhere in the subsector — for ornamental and architectural metalwork — fell from 125.663 in 2017 to 97.239 in 2021 even as that industry's nominal output rose.[19][23] Much of the nominal growth in this subsector since 2019 is selling-price inflation on pass-through inputs, not more tons through the press, the kettle or the furnace. Any revenue comparison across years, companies or children here needs a volume test.

No industry-wide margin or return-on-capital series exists at this level, and no suppressed value is stated anywhere on this page. As a directional operating gauge, the Federal Reserve's capacity-utilization reading for the whole NAICS 332 group was about 76.9% in June 2026, below its long-run average of ~78.5%, even as sector output ran about 2.1% above the prior year — parent context, not a per-child measure.[5][24]

4. Investable universe — where value concentrates across the children

There is no clean listed pure-play for NAICS 332, and clean pure-plays exist for only a handful of the sub-industries beneath it. Public exposure is a set of partial windows inside larger companies, unevenly distributed across the nine children. Tickers below denote where the exposure lives, not a recommendation, and are reserved for this section and Section 10.

Where listed value actually sits (roughly in order of how directly a stock reads onto the work):

• Cans (inside 3324) hold nearly all the clean listed packaging exposure — the near-direct names are Ball, Crown Holdings, Silgan Holdings, Ardagh Metal Packaging, and Sonoco Products, with captive exposure inside Anheuser-Busch InBev. Two cautions travel with that menu: listed is not independent (Ardagh Metal Packaging was roughly 76% controlled as of late 2025), and segment revenue is not NAICS revenue.[22]
• Bearings and firearms (inside 3329) remain the only sub-industries in the whole subsector with U.S.-listed pure-plays that trade on their own product — Timken and RBC Bearings; Sturm Ruger and Smith & Wesson, both currently earning trough margins. The foreign-listed door opened wider in January 2026 when Czechoslovak Group (CSG N.V.), owner of the number-two U.S. ammunition producer, began trading on Euronext Amsterdam, alongside Prague-listed Colt CZ; the last small U.S.-listed ammunition maker exited manufacturing in April 2025.[20]
• Structural steel fabrication (3323) is reached through integrated steelmakers and now through more direct vehicles than this page previously carried: Nucor (NYSE: NUE) is the only listed name with disclosed activity in all six of that child's six-digit industries, Steel Dynamics (Nasdaq: STLD) offers the cleanest single disclosure (fabrication net sales $1.42 billion at a 28.7% operating margin), Commercial Metals (NYSE: CMC) runs 53 North American fabrication facilities, Valmont (NYSE: VMI) and Arcosa (NYSE: ACA) cover utility and infrastructure structures, INNOVATE (NYSE: VATE) gives the most concentrated custom-structural exposure through DBM Global, and BWX Technologies (NYSE: BWXT) and Mayville Engineering (NYSE: MEC) cover plate weldments. The largest dedicated prefab-building player, Cornerstone Building Brands, is private but SEC-reporting on its debt — a credit route where no equity route exists.[19]
• The building envelope (3323) concentrates in the metal window-and-door line — Janus International (NYSE: JBI), Apogee (Nasdaq: APOG), Griffon (NYSE: GFF), JELD-WEN (NYSE: JELD) — with sheet-metal and ornamental exposure genuinely indirect through Gibraltar (Nasdaq: ROCK), Carlisle, Atkore, Armstrong World and Worthington.[19]
• Coating and finishing (3328) has one large, reasonably pure U.S. operator — AZZ (NYSE: AZZ), North America's largest hot-dip galvanizer with 42 galvanizing plants; heat treating's global leader, Bodycote, is listed in London; plating has no listed shop at all and is reached only through its chemistry-and-equipment oligopoly (Element Solutions, MKS/Atotech), with Aalberts, Curtiss-Wright and Valmont spanning two or three trades inside larger companies.[23]
• Hardware (3325) offers relatively focused makers — Allegion (NYSE: ALLE), CompX (NYSE American: CIX), The Eastern Company (Nasdaq: EML) — alongside diversified Fortune Brands Innovations (NYSE: FBIN), global leaders listed abroad (ASSA ABLOY, dormakaba), and adjacent Hillman Solutions (Nasdaq: HLMN) in fasteners and key duplication.[27]
• Valves and fluid control (inside 3329) never arrive as a pure-play — always a segment of Emerson (NYSE: EMR), Flowserve (NYSE: FLS), Parker Hannifin (NYSE: PH), Crane (NYSE: CR), Masco (NYSE: MAS), Fortune Brands (FBIN), Zurn Elkay (NYSE: ZWS), Mueller Industries (NYSE: MLI), Mueller Water (NYSE: MWA) or Watts Water (NYSE: WTS), with Core & Main (NYSE: CNM) the distribution layer.[20]
• Fasteners (inside 3327) now have one segment large and profitable enough to analyze on its own — Howmet (NYSE: HWM) Fastening Systems, $1.745 billion at a 30.4% adjusted EBITDA margin — and the aerospace end consolidated hard in 2026, with Howmet buying Stanley Black & Decker's aerospace-fastener business (~$1.8 billion) and TriMas selling its aerospace segment to PennAero (~$1.45 billion). Neither seller retains the exposure. Elsewhere: Illinois Tool Works, Nucor Fastener, Berkshire's SPS, micro-cap Chicago Rivet, and distributors Fastenal and W.W. Grainger as volume proxies.[21]
• Machining, forging/stamping, cutlery/handtool and spring/wire are the thinnest. Machining is reached through marketplaces Xometry and Proto Labs and component makers NN, Inc., Helios and Kennametal; forging's listed value sits in aerospace-grade nonferrous work inside Howmet, ATI and Berkshire's Precision Castparts, with the closest listed specialist (SIFCO) at $84.8 million of sales — under a hundredth of Howmet's; cutlery/handtool is diversified tool majors and housewares brands (Stanley Black & Decker, Snap-on, Acme United, Griffon, Techtronic; Lifetime Brands, Helen of Troy, Newell, Groupe SEB); and spring/wire offers Insteel as the one near-direct U.S. listing, with springs foreign or diversified (Rosebank Industries after its May 2026 purchase of MW Components, Beijer Alma, NHK Spring, Leggett & Platt).[21][24][25][26]

One transaction now spans children: Baker Hughes completed its acquisition of Chart Industries on 16 July 2026, bringing boiler-side heat exchangers and cryogenic tanks under one roof and removing a listed plate-adjacent fabricator from the market — a single diversified buyer is now a meaningful counterparty on both sides of 3324 and adjacent to 3323.[19][22]

The dominant ownership reality across all nine children is private, and the deal record now carries priced anchors. Below the listed names sit tens of thousands of family firms, employee-owned companies, and PE platforms — and the deepest pools of acquirable independents are exactly where public exposure is thinnest: structural and ornamental fabrication, machine shops and fasteners, finishing shops, tanks, springs, and the misc-metal tail. What the private field pays now reads as a ladder that tracks certification rather than process: roughly 4× trailing EBITDA for ordinary tank fabrication, 7.2× for a regional duct maker serving data-center and healthcare work, 7–11× for quality aerospace/defense machining shops, ~10× for a multi-facility precision-components platform, and ~18× for elite aerospace fasteners.[19][21][22][26] For most of the subsector, the private field is richer and more directly investable than the thin public surface suggests.

5. How the money works

Under one shared model — buy metal, add fabrication, sell for more — the whole subsector runs on the same handful of levers, even though the end products could not be more different:

• The metal spread is the margin engine, and timing is the risk — 2025 showed exactly how it fails. Profit is the contracted price less the cost of steel, aluminum, brass, alloys, energy, and labor. Purchased metal is usually the single largest variable cost — Steel Dynamics puts purchased steel at about two-thirds of its fabrication manufacturing cost — and it is typically a pass-through, not a margin. The danger is not a higher metal price by itself but a timing mismatch. At Steel Dynamics in fiscal 2025, fabrication selling prices fell 13% and volumes 8% while consumed steel cost only 7% less, so the metal spread contracted 17% and segment operating income fell 39%; the same mechanism compressed margins at five other disclosed proxies on both sides of 3323 alone.[19] That is a live risk into 2026, with producer prices for steel-mill products up 16.9% and aluminum-mill shapes up 52.4% in the year to June 2026.[6][7]
• Pass-through inflates the top line, so separate price of input from price of work. This is the discipline the revised children insist on, and it applies everywhere: Ardagh puts variable costs, which include metal, at roughly 75% of cost of sales; Element Solutions disclosed that pass-through metals pricing added $64.4 million to 2025 electronics sales; Bodycote's 2024 organic revenue decline was mostly lower energy surcharges, with underlying revenue down only 0.8%.[22][23] A year, a company or a child cannot be compared until the pass-through is stripped out.
• Capacity utilization is the master operating lever. Presses, furnaces, mills, welding cells, machining centers, plating tanks, and skilled crews are fixed costs whether busy or idle, so incremental volume drops through at high margins and lost volume hurts fast. The cost of slack is measurable: SIFCO booked $992,000 of idle-capacity cost in fiscal 2025, and Nucor's broader Steel Products facilities ran at 61% in the final quarter of 2025.[19][24] Operating leverage cuts both ways, which is why a modest cyclical dip can cause a much larger earnings dip.
• Certification and qualification are the moat. Once a die, part, or process is tooled, tested, and approved — under aerospace, defense, food-contact, pressure-vessel, drinking-water, or automotive standards — customers are slow and costly to re-source. It is not absolute (customers often own the dies and can move them), and approvals are frequently facility- or line-specific rather than portable, which makes their transferability a live diligence item in any acquisition.[23][24]
• Value migrates up the value-add ladder. Heat treatment, machining, nondestructive testing, coating, and assembly capture more revenue per part; basic cutting and catalog commodity items compete on price and are easily replaced.
• Freight keeps it regional, and the radius is now measured. Bulky, low-value-density products do not ship far economically: Silgan's metal-container plants generally serve customers within a 300-mile radius, and Valmont puts a normal galvanizing service radius at roughly 300–500 miles.[22][23] This is the structural reason the subsector stays fragmented and local, and it caps winner-take-all dynamics even where consolidation is active.

The margin ladder is the same in every child, and it is set by qualification — not by process, product, or size. Disclosed 2025 segment results run from the high twenties and low thirties at the certified end (Howmet Fastening Systems at a 30.4% adjusted EBITDA margin; AZZ's Metal Coatings at roughly 31%; Allegion's Americas segment at 27.9% operating; RBC Bearings at 22.5% operating), through the mid-teens in diversified industrial and building-products work (Gibraltar's residential segment at 16.6%; Bodycote's Precision Heat Treatment at 16.0% adjusted operating; GKN Powder Metallurgy at 9.1% adjusted operating), down to low single digits in commodity fixed-price fabrication and consumer-cyclical work (Matrix Service's storage segment at a 4.0% gross margin; Olin's Winchester at 3.9%; Park-Ohio's Engineered Products at 1.4% operating; NHK Spring's automotive suspension springs at 0.3%).[19][20][21][22][23][24][26] Do not average these and do not rank them. They are different measures — segment operating, adjusted segment EBITDA, gross, and company adjusted EBITDA margins — on different scopes, most of them global rather than U.S. and none of them a NAICS margin. Read the ordering, which holds across all nine children; read the levels only against a like measure.

The specific economic engine also varies by child — project/engineered-to-order (boilers, tanks, structural, valves), high-volume manufacturing (cans, fasteners, wire products), for-the-trade service (finishing), consumable razor-and-blades (ammunition, replacement saw blades), and contract-and-backlog defense work where a government-owned plant earns the contractor a capped management fee rather than a product margin.[20][22][23] Demand visibility spans nearly the full available range within a single child: Silgan has roughly 90% of projected 2026 metal-container sales under multi-year arrangements while Greif reports that many industrial-packaging customers order weekly for same-week delivery.[22] Useful cross-child gauges: orders, book-to-bill and backlog margin (project businesses); unit shipments and price/mix excluding metal pass-through, line utilization, and scrap (volume businesses); aftermarket/replacement share; and cash conversion.

6. Demand drivers

Because the nine children serve end markets that barely overlap, the subsector is genuinely diversified and rarely moves as one — that diversification is a feature of the level as a whole. The major pulls:

• Nonresidential & institutional construction — warehouses, factories, data centers, hospitals, schools (structural, architectural, hardware, roll forming). The near-term signal is negative: total U.S. construction spending in May 2026 ran at a $2.21 trillion seasonally adjusted annual rate, down 1.5% year over year, with the first five months of 2026 running 2.7% below 2025 and manufacturing construction down 21.9%.[9] The American Institute of Architects' July 2026 consensus panel forecast a 0.3% decline in nonresidential spending for the year and noted architecture billings have been falling since early 2023 — a series that typically leads by nine to twelve months.[10]
• Infrastructure, grid & data centers — bridges, transmission structures, substations, and rising electricity load (structural, tanks, wire products, pipe fabrication, galvanizing). The Department of Energy put data centers at 4.4% of national electricity in 2023, projected at 6.7%–12% by 2028.[11] The pull is uneven in how it converts: developers planned 86 gigawatts of U.S. generating additions for 2026, but only 6.3 gigawatts is natural gas — solar and batteries dominate and use little steam equipment.[22]
• Aerospace & defense — long qualifications and multiyear programs (forging, fasteners, machine shops, bearings, large-caliber ammunition). This is where the growth currently is: ATI reported 92% of High Performance Materials & Components revenue from aerospace and defense in 2025, with those sales up 14%, commercial jet engines up 21% and defense up 24%; Howmet's fastening sales rose 11%.[21][24]
• Automotive & heavy vehicles — high operating leverage, sharp cyclicality (forging, powder metallurgy, fasteners, springs, machining). Currently the soft side: Park-Ohio's forged-and-machined revenue declined three years running, Mayville's 2025 sales fell 6.0%, and North American metal-powder shipments rose only 0.6%.[24]
• Energy — oil, gas, LNG, nuclear — capital-spending cycles (boilers, tanks, valves, pipe fabrication, machining). North American LNG export capacity is forecast to rise from 11.4 billion cubic feet per day in 2023 to 24.4 billion by 2028.[12]
• Water infrastructure — federally funded renewal and lead-service-line replacement (tanks, waterworks valves and fittings). The 2021 infrastructure law directs more than $50 billion through the EPA, including $15 billion specifically for lead service lines, against an EPA-assessed $625 billion of twenty-year drinking-water need.[13][20]
• Consumer & household — housing repair-and-remodel, cooking-at-home, food and beverage packaging (faucets, cookware, cutlery, cans). Still the most defensive demand in the subsector, but not uniform: beverage and pet-food cans are growing while U.S. human-food can shipments fell from 26.3 billion units in 2022 to 23.7 billion in 2024, and drums and pails are in an acknowledged industrial contraction.[22]

Two structural forces cut across the cycle. Reshoring/onshoring of supply chains is a broad tailwind for domestic fabrication capacity and the reason a qualified second domestic supplier is prized — the Reshoring Initiative recorded 244,000 announced U.S. manufacturing jobs from reshoring and foreign direct investment in 2024, and U.S. metalworking machinery orders reached $5.74 billion in 2025, 22.5% above 2024 after three years of decline.[17][18] The electric-vehicle (EV) transition is a mixed signal — and here the children genuinely disagree, which is worth naming rather than averaging. The forging child argues battery-electric drivetrains remove crankshafts, connecting rods and much conventional transmission content but still need high-torque gears, shafts, wheel-end, suspension and safety-critical parts, with hybrids preserving most of it: a negative mix shift for forging, a sharper structural headwind for powder metallurgy.[24] Inside the finishing child, the heat-treating research argues EVs preserve or increase demand for heat-treated drivetrain and structural parts, while its plating research declines to assert any net effect for want of defensible sources.[23] The safe reading is that electrification changes mix more than total, with the clearest structural loser being powder metallurgy inside 3321 and no settled sign elsewhere. In mid-2026 the overall picture was uneven: aerospace/defense, grid/data-center, water, pipe fabrication and defensive packaging demand were firm while heavy-truck, automotive, general-industrial, and rate-sensitive commercial and residential construction were softer.[5][9][19][21][24]

7. Regulation

None of the subsector is rate-regulated; regulation is about worker safety, environment, product quality, and trade — so regulated-utility rate base, real-estate funds-from-operations, and mining all-in-sustaining-cost frameworks do not apply here. The shared threads:

• Worker safety. The Occupational Safety and Health Administration (OSHA) governs forging machines, mechanical power presses, machine guarding, welding and hot work, and control of hazardous energy across fabrication shops; powder plants add combustible-dust risk, and field steel erection falls under a separate construction standard with a dedicated section for systems-engineered metal buildings.[19][24] The hazard shows up in the data and in the penalties: 2024 total recordable rates ran 5.7–5.8 cases per 100 full-time workers in nonferrous forging and custom roll forming against 3.4 in fabricated structural metal and metal window-and-door work, and OSHA proposed $1.326 million against one heat treater and $338,094 against one powder coater.[19][23][24]
• Environment. The Environmental Protection Agency (EPA) covers air emissions, metal-finishing effluent, and hazardous waste; finishing and plating remain the most heavily regulated corners. Two live items have hardened since this page was last written: California is phasing out hexavalent chromium — decorative plating by 2027 with a pathway to 2030, functional hard chrome by 2039 — and a first-ever EPA per- and polyfluoroalkyl-substances (PFAS) wastewater rule is expected around 2026, following a 2023 survey of roughly 2,000 chrome-finishing facilities.[23] The state layer has also moved beyond disclosure on the consumer side: Minnesota prohibited intentionally added PFAS in cookware from January 1, 2025, where California's rule was disclosure only.[25] Older plating and coating sites carry legacy soil-and-groundwater liabilities that are make-or-break in diligence.
• Product standards and certifications — frequently mandatory, and functioning as barriers to entry: American Institute of Steel Construction (AISC) certification and IAS AC472 for structural and metal-building work; American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code for boilers and tanks, with ASME B31 and Section IX welder qualification for pipe fabrication (a welder's qualification lapses after six months of not using the process); Nadcap, AS9100 and AMS 2750H/CQI-9 for aerospace and thermal processes; Food and Drug Administration (FDA) food-contact rules for cans, closures and cookware; NSF/ANSI lead-free certification for plumbing valves and faucets; ENERGY STAR and National Fenestration Rating Council certification for windows; and building, fire, and accessibility codes (including the Americans with Disabilities Act) for hardware and architectural metal.[19][20][22][23] The Fastener Quality Act also applies to certain threaded fasteners — but its practical reach is narrower than commonly assumed, with numerous proprietary, aerospace-approved, and quality-system-produced categories exempt.[21]
• Trade policy — the universal, double-edged lever, and it is no longer a single number. Section 232 national-security tariffs on steel and aluminum went to 50% in June 2025, a separate 50% tariff on semi-finished copper and copper-intensive products followed in August 2025, and the regime was restructured again in June 2026.[8][20] The children do not agree on how derivatives are now treated. One reading has covered articles and certain derivatives at 50% with exceptions; another has upstream steel at 50%, derivatives at 25%, and certain industrial or grid equipment at a temporary 15% through 2027; another describes a 25% full-value charge on defined lists; and one child reports steel-spring tariff lines sitting in the 50% full-value annex as of April 2026.[19][20][22][25][26] We do not reconcile them, because the operative rate depends on a product's Harmonized Tariff Schedule classification, metal content and origin — not on its NAICS code. The rollup-level consequence is what matters: there is no headline rate to model, coverage is a per-part legal question, Commerce stopped accepting new product-exclusion requests in February 2025, and 80.3% of manufacturers surveyed by the National Association of Manufacturers in the fourth quarter of 2025 said they had paid tariffs on imported inputs during the year.[16][19] Tariff-classification diligence with customer pass-through clauses is now a real underwriting item. Antidumping and countervailing duties add product-specific protection, and Build America, Buy America (BABA) domestic-content rules steer federal infrastructure dollars to qualified U.S. makers.
• One corner reaches further. The ammunition and ordnance industries inside 3329 add an entire weapons regime the rest of the subsector never touches — licensing through the Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF); a 10%/11% federal excise tax; arms-export control (International Traffic in Arms Regulations, ITAR); and national-security review of foreign acquirers by the Committee on Foreign Investment in the United States (CFIUS), which cleared the sale of the number-two U.S. ammunition producer to a Czech buyer. One targeted change landed this cycle: the National Firearms Act's $200 transfer tax on suppressors and short-barreled firearms fell to $0 effective January 1, 2026, with registration retained.[20]

The recurring theme: compliance is also commercial. A safety, quality, food-contact, or certification failure can cost more than a year's profit, and certification is often the very thing that keeps low-cost entrants out of the best work.

8. Consolidation

The fragmentation quantified in Section 3 (HHI 13.7, CR4 5%) is precisely what makes NAICS 332 a consolidation hunting ground, and the pattern is carve-outs and regional roll-ups, not industry-wide mega-mergers. The drivers are the same everywhere: an aging population of owners facing succession with no family successor; rising compliance and capital intensity (environmental, certification, and equipment capex) that pushes sub-scale shops to sell rather than reinvest; and abundant unspent private-equity capital hunting platform and bolt-on deals. Value creation is usually capability- or channel-led — specifications, engineering, certifications, distribution density, metal-purchasing scale, and the ability to balance work across plants — rather than simply adding fabrication capacity or leverage. The recurring integration risk is the departure of the local estimators, engineers, and project managers whose relationships and know-how are the real asset.

Concentration is a per-child fact, and the subsector HHI hides it — but read the artifact correctly. The blended 13.7 is lower than seven of the nine children because bundling nine unrelated markets means the leader in one (say, structural steel) makes none of the leader's product in another (cans, or valves, or ammunition), so no firm looms large over the combined base. The exception is instructive: machine shops and fasteners (3327) reads lower still at 6.4, for exactly the same pooling reason one level down, with its machine-shop half at roughly 3.5.[21] Read both as "separate markets pooled together," never as a single free-for-all. The real structure ranges from atomistic job shops through moderately concentrated niches to genuine oligopolies inside a child — the beverage-can business, where regulators forced eight plant divestitures in the last major merger and five companies make substantially all North American aluminum beverage cans; small-arms ammunition at an HHI of 1,766; faucets at a CR4 of 63.9%; heat treating at 41.8%.[20][22][23]

Two things about the deal record are new at this level. First, the multiple tracks certification, not process — the ladder in Section 4 runs from roughly 4× trailing EBITDA for plain tank fabrication to ~18× for elite aerospace fasteners, which is the same ordering as the margin ladder in Section 5.[19][21][22][26] Second, and more sobering for a roll-up thesis, all this activity has barely moved measured share. Over two decades the other-metal-valve industry went from 238 companies at a CR4 of 19.8% and an HHI of 217.4 to 175 firms at 21.1% and 230.5, while fabricated pipe went the other way entirely, its CR4 falling from 23.2% to 14.7%.[20] Roll-ups here consolidate ownership faster than they consolidate share — because freight economics, differing estimating cultures, project-specific relationships and non-portable approvals all cap winner-take-all dynamics. The one structural advantage that does travel is vertical integration: Cornerstone states plainly that competitors owned by steel producers may receive more favorable raw-material pricing or delivery priority, which is the clearest single explanation for why integrated steelmakers appear in this subsector's investable universe at all.[19] Deal activity spans the whole subsector — take-privates and carve-outs at platform scale (Cornerstone at ~$5.8 billion, Barnes Group at ~$3.6 billion, Oldcastle BuildingEnvelope at ~$3.45 billion, MITER–PGT at ~$3.1 billion), strategic consolidation of certified niches (Howmet's aerospace-fastener purchase, RBC–Dodge, Baker Hughes–Chart, Flowserve–Trillium, Aalberts–Paulo), foreign acquisition of the ammunition base under CFIUS review, and PE platform-building in finishing, machining, fasteners, springs, tanks and fabrication.[19][20][21][22][23][26]

9. Risks

Shared across the subsector, with emphasis varying by child:

• Cyclicality and operating leverage. Fixed-cost plants mean low utilization compresses margins fast; downturns in construction, autos, energy, or general industrial output hit quickly, and the leading construction indicators are currently negative.[9][10]
• Metal price-cost timing. Fixed-price contracts and backlog lose money when procurement lags or escalation protection is weak; falling metal can trigger customer repricing before high-cost inventory clears. Higher metal is not automatically good — it helps an integrated mill-fabricator while immediately hurting an independent shop with unlocked material.[19]
• Nominal growth mistaken for real growth. Producer prices have led output in several children, and the subsector's only real-output series fell from 2017 to 2021 while nominal output rose.[19][23]
• Measurement risk — new, and material. Current six-digit revenue, firm-count and concentration data do not exist in publishable federal form across much of this subsector, and even where two federal programs both report, they disagree by wide margins.[19][24][25][26][27] Two errors follow predictably: treating a commercial "global market" headline as this industry, and dividing a conglomerate's global segment revenue by a partial NAICS denominator to claim market share. Neither survives contact with the data.
• Customer and channel concentration. Losing a qualified program can strand specialized, single-purpose equipment, and the disclosed figures show how much bargaining power large retail and OEM channels hold: JELD-WEN's ten largest customers were about 48% of 2025 revenue (Home Depot 17%, Lowe's 13%), Mayville's top ten were 62.3%, and Hillman's two largest home centers were 22.5% and 20.9%.[19][27]
• Quality and certification failures. Recalls, lost approvals, or a failed audit can outweigh years of profit and shut a shop out of its most profitable work — and because most of the finishing and heat-treating base holds someone else's metal, a bad furnace cycle or contaminated bath triggers replacement cost far exceeding the processing invoice.[23]
• Skilled-labor scarcity. The Bureau of Labor Statistics counted 457,300 welders, cutters, solderers and brazers in 2024 and projects about 45,600 openings a year through 2034, mostly replacement; machinists and tool-and-die makers show roughly 34,200 annual openings against flat-to-declining employment, with tool-and-die jobs projected down 11%.[14][15] The American Welding Society projects a need for 320,500 new welding professionals by 2029 on a different basis. These are retention and training problems more than growth-driven shortages — but 72.1% of manufacturers looking to hire cited skilled production workers as the binding gap.[16][19]
• Thin qualified capacity, and thin supply behind it. Two single-site events showed how little slack exists: a 2025 fire at Precision Castparts' Jenkintown facility affected more than 700 sole-sourced parts, and a fire at SPS Technologies reportedly took out roughly 15% of U.S. aerospace-fastener supply.[21][24] Upstream is thinner still — only two major iron-powder suppliers remain in North America, and the U.S. had a single active titanium-sponge producer in 2023 against China holding an estimated 65.8% of potential world capacity.[24] A buyer of shop capacity is not buying input security, and qualifying an alternative component supplier can take nearly two years.[20]
• Trade-policy whipsaw, now distinct from tariff level. The schedule moved materially three times in roughly twelve months and coverage is now determined product by product through customs classification — a quoting, sourcing and inventory risk on top of the cost itself.[8][19][20]
• Substitution — broader than this page previously allowed. Concrete tilt-up, engineered timber, masonry and modular systems compete in buildings; glass-fiber-reinforced-polymer rebar and fiberglass grating compete where corrosion resistance justifies the cost; plastics (PVC, CPVC, PEX) are the principal substitute for metal plumbing fittings; composites threaten steel leaf and coil springs and synthetic rope competes with steel rope; castings and forgings can replace multi-piece weldments at volume. Additive manufacturing is more complementary than substitutive at the finishing end — printed parts still need stress relief, aging or hot isostatic pressing, and usually machining of critical surfaces.[19][20][23][26]
• Ownership structure. Listed names bury the exposure inside larger businesses (dilution); private roll-ups carry acquisition debt into cyclical downturns (leverage and integration risk).

The clearest structural risk pockets sit in specific corners: powder metallurgy (EV drivetrain deletion) inside 3321; undifferentiated commodity job shops (imports and automation) inside 3327 and the misc-metal tail; the fabricated-wire side of 3326, where industry employment fell 62.3% between 2000 and 2024 on automation, imports and reclassification; and legacy environmental liabilities at older plating and coating sites inside 3328.[23][24][26] The weapons corner of 3329 carries political, legal, and ESG-screening risk that no other child bears — and is currently in a genuine margin trough rather than a simple volume normalization, with background checks down two years running.[20]

10. How to invest and outlook

Public investors should treat NAICS 332 as segment exposure, not a pure play. There is no ETF and no single proxy; each ticker is a company-specific thesis, so match the vehicle to the child you actually want:

• Structural/architectural (3323): Nucor (NUE) for the only listed name touching all six sub-industries, Steel Dynamics (STLD) for the cleanest fabrication disclosure, Commercial Metals (CMC), Valmont (VMI), Arcosa (ACA), INNOVATE (VATE) for the most concentrated custom-structural exposure, and Mayville (MEC) or BWX Technologies (BWXT) for plate weldments; Janus (JBI), Apogee (APOG), Griffon (GFF), JELD-WEN (JELD) and Gibraltar (ROCK) on the envelope side.
• Packaging cans (3324): Ball, Crown, Silgan, Ardagh, Sonoco — compared on normalized conversion earnings, not revenue inflated by metal pass-through, and noting Ardagh Metal Packaging's small float against ~76% control. Boilers via Babcock & Wilcox, BWX Technologies, Graham, Baker Hughes and Modine; tanks only through diversified, illiquid or adjacent contractor names.
• Bearings & firearms (3329): Timken and RBC Bearings; Sturm Ruger and Smith & Wesson — the subsector's cleanest pure-plays, with the firearms pair currently at trough margins; Olin for the one meaningful U.S. ammunition exposure, plus CSG N.V. (Euronext Amsterdam) and Colt CZ (Prague) for foreign-listed access.
• Valves (3329): Emerson (EMR), Flowserve (FLS), Parker Hannifin (PH), Crane (CR), Masco (MAS), Fortune Brands (FBIN), Zurn Elkay (ZWS), Mueller Industries (MLI), Mueller Water (MWA), Watts Water (WTS) — always a segment; Core & Main (CNM) for distribution economics rather than manufacturing.
• Hardware (3325): Allegion (ALLE), CompX (CIX), Eastern (EML), Fortune Brands (FBIN), with ASSA ABLOY and dormakaba abroad and Hillman (HLMN) adjacent.
• Finishing (3328): AZZ (coating), Bodycote (heat treating, London-listed), Element Solutions and MKS (plating chemistry), with Aalberts, Curtiss-Wright and Valmont as diluted spanners.
• Fasteners, machining, forging/stamping, cutlery/handtool, spring/wire (3327/3321/3322/3326): the thinnest — Howmet (HWM) is now the most direct large-cap fastener read after the 2026 aerospace consolidation, with ITW, Nucor, Chicago Rivet and distributors Fastenal and Grainger around it; Xometry, Proto Labs, NN, Helios and Kennametal for machining; Howmet, ATI, Park-Ohio and Mayville for forging/stamping; Stanley Black & Decker, Snap-on and the housewares brands for cutlery/handtool; Insteel and Rosebank for spring/wire.

In every case, value the actual segment exposure — its volumes, margins, backlog, end-market mix, and metal-price sensitivity — not the whole company, and separate real volume growth from metal-driven nominal growth. Compare on mid-cycle margins and cash conversion, and normalize enterprise value to earnings before interest, taxes, depreciation, and amortization (EV/EBITDA) for metal-price and capital-spending effects rather than peak-cycle earnings. The 2025 results across the subsector are the reason: a single strong year of metal spread is not a run rate.

Private investors get the most direct — and best-matched — access, precisely because 49,631 firms and an HHI of 13.7 leave an abundance of independent targets: owner-succession buyouts, regional add-ons, contract-manufacturing and finishing platforms, distribution-and-repair networks, equipment finance, and private credit. Diligence should reconstruct backlog contract-by-contract rather than accept a headline total, verify metal commitments and escalation clauses, test estimating accuracy and utilization, confirm that certifications actually transfer (many are facility- or line-specific), check environmental history (make-or-break for any plating or coating deal), probe customer and channel concentration and tariff classification, normalize EBITDA for owner compensation, deferred maintenance capex and one-time projects across a full cycle, and separate genuine manufacturing revenue from lower-margin distribution and field installation. Validate the NAICS classification at the establishment level rather than accepting a company-wide label — a plant producing engineered packages, ducts or weldments is manufacturing, while a dealer, erector, installer or distributor is construction or trade with entirely different economics.[19][21] Replacement cost is not value if the equipment lacks qualified work.

Outlook — selectively constructive over a full cycle, weaker near term than this page previously carried (a judgment, not a guarantee). The subsector's virtue is that it is diversified across cycles that rarely align, and that has not changed. What has changed is the near-term evidence: fabrication margins compressed across every disclosed proxy in fiscal 2025, capacity utilization sits at 76.9% against a 78.5% long-run average, total construction spending is running below year-ago with manufacturing construction down 21.9% and architecture billings falling since early 2023, and reported dollar growth carries a large price component.[5][9][10][19] Secular tailwinds still favor scarce, qualified, well-utilized capacity pointed at structural growth — aerospace and defense, grid expansion and electrification, data centers, LNG and pipe fabrication, bridge and water-system replacement, reshored advanced manufacturing, defensive packaging, and (in one corner) allied ammunition restocking, though that ramp is running well behind its stated goal. The offsets are real and shared: rate-sensitive commercial, warehouse and residential construction, weak heavy-vehicle and agriculture demand, scarce skilled labor, thin qualified and upstream capacity, and a metal-tariff regime that is no longer a single headline rate but a product-by-product schedule keyed to customs classification — still sheltering domestic sellers, still keeping input costs elevated, and now adding administrative and quoting risk on top.[8][19][20] Value therefore concentrates ever more tightly in the specialized, certified, backlog-rich, and branded operators, and in the same ordering across all nine children. The most exposed are undifferentiated fixed-price job shops with concentrated customers, weak utilization and unprotected metal, counting on a broad rebound. For the full case on any child — company tables, sub-industry economics, and diligence checklists — read its own primer.

Sources

1. Histometrics ground-truth federal statistics, NAICS 332 Fabricated Metal Product Manufacturing — 2022 Economic Census (receipts $465,673,224K; 49,631 firms; CR4 5.0%, CR8 7.9%, CR20 13.4%, CR50 20.9%; HHI 13.7) and 2023 County Business Patterns (53,790 establishments; 1,441,471 employees; annual payroll $96,763,880K; Q1 payroll $24,066,428K). U.S. Census Bureau.
2. U.S. Census Bureau, County Business Patterns: 2023 (establishments, employment, and payroll for NAICS 332 and children), 2025. https://www.census.gov/programs-surveys/cbp.html
3. U.S. Census Bureau, 2022 Economic Census — Concentration of Largest Firms for the U.S. (EC2200SIZECONCEN), NAICS 332 and children (receipts, firms, CR4/CR8/CR20/CR50, HHI), 2025. https://data.census.gov/table/ECNSIZE2022.EC2200SIZECONCEN
4. U.S. Census Bureau, 2022 NAICS Definitions — Subsector 332 and its nine industry groups (3321–3329), 2022. https://www.census.gov/naics/
5. Board of Governors of the Federal Reserve System, Industrial Production and Capacity Utilization (G.17), NAICS 332, June 2026. https://www.federalreserve.gov/releases/g17/current/
6. U.S. Bureau of Labor Statistics via Federal Reserve Bank of St. Louis, Producer Price Index: Steel Mill Products (WPU1017), 2026. https://fred.stlouisfed.org/series/WPU1017
7. U.S. Bureau of Labor Statistics, Producer Price Indexes — June 2026 (Table 2), 2026. https://www.bls.gov/news.release/ppi.t02.htm
8. The White House, Further Adjusting the Tariff Regimes for Imports of Aluminum, Steel, and Copper into the United States (Section 232), June 2026. https://www.whitehouse.gov/presidential-actions/2026/06/further-adjusting-the-tariff-regimes-for-imports-of-aluminum-steel-and-copper-into-the-united-states/
9. U.S. Census Bureau, Construction Spending — May 2026, 2026. https://www.census.gov/construction/c30/current/index.html
10. American Institute of Architects, July 2026 Consensus Construction Forecast, 2026. https://www.aia.org/resource-center/july-2026-consensus-construction-forecast
11. U.S. Department of Energy, Report Evaluating Increase in Electricity Demand from Data Centers, 2024. https://www.energy.gov/articles/doe-releases-new-report-evaluating-increase-electricity-demand-data-centers
12. U.S. Energy Information Administration, North American LNG Export Capacity Is on Track to More Than Double by 2028, 2025. https://www.eia.gov/todayinenergy/detail.php?id=64128
13. U.S. Environmental Protection Agency, EPA's 7th Drinking Water Infrastructure Needs Survey and Assessment, 2026. https://www.epa.gov/dwsrf/epas-7th-drinking-water-infrastructure-needs-survey-and-assessment
14. U.S. Bureau of Labor Statistics, Occupational Outlook Handbook: Welders, Cutters, Solderers, and Brazers, 2025. https://www.bls.gov/ooh/production/welders-cutters-solderers-and-brazers.htm
15. U.S. Bureau of Labor Statistics, Occupational Outlook Handbook: Machinists and Tool and Die Makers, 2026. https://www.bls.gov/ooh/production/machinists-and-tool-and-die-makers.htm
16. National Association of Manufacturers, Fourth Quarter 2025 Manufacturers' Outlook Survey, 2025. https://nam.org/2025-fourth-quarter-manufacturers-outlook-survey/
17. Reshoring Initiative, 2024 Reshoring Report, 2025. https://reshorenow.org/june-9-2025/
18. Association for Manufacturing Technology, Manufacturing Technology Orders Set Record in December 2025, 2026. https://www.amtonline.org/article/manufacturing-technology-orders-set-record-in-december-2025
19. Histometrics child primer, Architectural and Structural Metals Manufacturing (NAICS 3323), 2026 — company, deal, margin, construction-indicator and end-market detail with underlying SEC and Census citations.
20. Histometrics child primer, Other Fabricated Metal Product Manufacturing (NAICS 3329) — metal valves and all-other (bearings, ammunition, pipe fabrication, misc. metal), 2026.
21. Histometrics child primer, Machine Shops; Turned Product; and Screw, Nut, and Bolt Manufacturing (NAICS 3327), 2026.
22. Histometrics child primer, Boiler, Tank, and Shipping Container Manufacturing (NAICS 3324), 2026.
23. Histometrics child primer, Coating, Engraving, Heat Treating, and Allied Activities (NAICS 3328), 2026.
24. Histometrics child primer, Forging and Stamping (NAICS 3321), 2026.
25. Histometrics child primer, Cutlery and Handtool Manufacturing (NAICS 3322), 2026.
26. Histometrics child primer, Spring and Wire Product Manufacturing (NAICS 3326), 2026.
27. Histometrics child primer, Hardware Manufacturing (NAICS 3325), 2026.