U.S. Nonferrous Metal (except Copper and Aluminum) Rolling, Drawing, Extruding, and Alloying — NAICS 33149

1. Overview

North American Industry Classification System (NAICS) code 33149 groups two ways of turning nonferrous metals — everything except copper and aluminum, so nickel, titanium, zinc, lead, magnesium, tin, and precious and refractory metals — into usable industrial forms.[1] One half shapes metal: it rolls, draws, and extrudes it into plate, sheet, strip, bar, rod, wire, and tube. The other half recovers and alloys metal: it smelts scrap and blends purchased metal into ingot, billet, bar, and wire. The industry sits between raw-metal production and demanding end markets — aerospace, defense, medical devices, energy, electronics, and the vehicle battery system.

For investors, the single most useful fact about this level is that it is not a coherent market you can buy. No U.S.-listed company is a clean NAICS 33149 pure play. Public-market investors get partial, diluted exposure through specialty-materials companies, diversified industrials, global miners, and holding companies; private investors buy individual mills, refiners, and collection networks directly. The two child industries also behave differently enough that lumping them together hides more than it reveals — which is why this primer leads with the contrast between them.

2. What's inside — the two child industries and how they differ

NAICS 33149 contains exactly two child industries:

• 331491 — Nonferrous Metal (except Copper and Aluminum) Rolling, Drawing and Extruding. Rolls, draws, or extrudes purchased nonferrous metal into mill shapes — and also covers plants that recover metal from scrap and then form it into finished mill shapes in the same establishment. Value comes from metallurgy, tight tolerances, and hard-won customer qualifications.[2]
• 331492 — Secondary Smelting, Refining, and Alloying of Nonferrous Metal (except Copper and Aluminum). Recovers metal from scrap and alloys purchased metal, stopping at primary forms — ingot, billet, bar, wire. Value comes from feedstock access, environmental permits, and recovery yield.[3]

The dividing line is therefore not "new metal versus scrap." Both children can start from scrap. What separates them is whether the plant goes on to roll, draw, or extrude the metal it recovers.[2][3] Classification is establishment-based, so two plants owned by the same company can sit in different codes.[3]

They are adjacent on the factory floor but different as businesses. The first is a conversion business selling engineering; the second is a recovery business selling a metal spread and a permit.

Contrast table

The practical takeaway: 331491 rewards engineering and program qualification; 331492 rewards permits, feedstock, and metal-accounting discipline. An investor who likes one does not automatically like the other.

3. How big it is

The best single-year federal footprint comes from the U.S. Census Bureau's 2023 County Business Patterns (CBP), which counts establishments with paid employees.[4] Revenue, firm count, and concentration come from the 2022 Economic Census.[5] Both are drawn from our ingested ground-truth statistics for this level.

Payroll per employee is calculated from the reported totals and excludes benefits. Establishments are operating locations, not companies — the 386 establishments are run by roughly 265 firms, so many operators run more than one plant.[4][5]

The children reconcile exactly on employment, and not at all on revenue. The two child pages' 2023 CBP counts add back to the level figures above without residual: 199 + 187 establishments, 12,115 + 9,490 employees, $254.7 million + $185.3 million of first-quarter payroll.[4] Pay levels differ modestly — 331491's implied annual payroll per employee is roughly $83,800, and $751.3 million across 9,490 employees implies roughly $79,200 for 331492, against ≈$81,800 for the level.[4] But neither child page could obtain Economic Census receipts, firm counts, or concentration ratios at six digits. The $15.19 billion of receipts, the 265 firms, and every concentration measure in the next section exist only at this five-digit level and cannot be split between shaping and recovery.

Two partial revenue-side markers exist beneath this level, and they are not additive and not comparable. For 331491, the Occupational Safety and Health Administration's 2024 Hazard Communication economic analysis listed 226 firms, 254 establishments, and 16,187 employees, with an estimated $8.805 billion of industry revenue and $183.3 million of profit in 2022 dollars — figures extrapolated from 2017 business statistics and benchmarked to payroll, not reported shipments, and best read as an order-of-magnitude regulatory baseline.[6] For 331492, the U.S. Geological Survey valued U.S. secondary lead production at $2.4 billion in 2025, covering one commodity within a child that also handles zinc, precious metals, and specialty alloys.[7] Neither figure should be netted against, or added to, the $15.19 billion.

Undercount caveat. CBP excludes businesses without paid employees, but that omission is modest here because both children are capital-intensive.[8] The larger measurement gaps are (a) captive operations — a battery manufacturer that recycles its own lead, or an integrated mill, may be recorded under a different primary activity — and (b) the fact that diversified public companies do not report results on a six-digit NAICS basis, so their contribution to this level is invisible in their financial statements.[8] The inverse error is just as common: adding up the revenue of ATI, Carpenter, Precision Castparts, and Haynes does not produce industry size, because those groups carry foreign operations, stainless steel, powders, forgings, castings, machined components, and distribution that sit outside the code.[2] Because a meaningful share of this level is owned by private, family, and small independent operators, treat listed names as windows onto the industry, not measures of it.

4. Investable universe — where value concentrates across the children

The 2022 Economic Census reports this level as fragmented, not concentrated. The four largest firms account for 29.7% of receipts, the top eight 41.9%, the top twenty 59.7%, and the top fifty 81.8%; the Herfindahl-Hirschman Index (HHI) is 337.6.[5] For reference, U.S. antitrust guidelines treat an HHI below 1,500 as unconcentrated — this level is well below that.

That average conceals a genuinely concentrated pocket. The Environmental Protection Agency (EPA) counted only 11 operating U.S. secondary-lead smelters in 2025 — five fewer than at its 2012 rulemaking — with six companies owning 10 of the 11 plants, among them East Penn Manufacturing, Gopher Resource, Ecobat subsidiaries, Clarios, Sanders Lead, and Doe Run's Buick facility.[9] Because the federal record publishes concentration only for the combined five-digit level, that plant-level structure cannot be extrapolated to either child as a whole; it does show that "unconcentrated on average" is compatible with an oligopoly inside one metal stream. Size is also relative: the Small Business Administration applies a 900-employee standard to 331491, so a technically "small" qualifying manufacturer can still be substantial.[10]

Because the federal data does not split receipts or concentration between the two children, the map below is organized by child. All tickers and company names belong to the investable universe, not to the NAICS definition itself.

331491 — shaping metal (partial exposure through specialty materials)

Notable private operators include Ulbrich Stainless Steels & Special Metals (precision strip/foil/wire), Perryman (integrated titanium), Fort Wayne Metals (medical wire, owned by Axel Johnson), Elgiloy (nickel/cobalt/titanium strip), and G.O. Carlson / Electralloy (specialty plate).[18] Route to market is mixed: ATI reported that about 40% of its nickel-alloy, titanium, and specialty-alloy plate volume moved through independent service centers in fiscal 2025, so distributor inventory behavior sits between the mills and true end demand.[11]

331492 — recovering metal (diluted exposure through miners, recyclers, and holding companies)

Foreign-listed Johnson Matthey and Umicore add diversified global precious-metal refining with U.S. operations. Johnson Matthey is also the clearest illustration of why headline revenue misleads in this child: fiscal 2025/26 revenue of £12.6 billion, but sales excluding precious metals of only £2.6 billion.[25] The largest U.S. recovery assets remain privately controlled: Ecobat's lead-recycling network, Gopher Resource (Energy Capital Partners), Element Resources (Atlas Holdings), family-owned East Penn, and Doe Run (Renco Group).[26]

Bottom line for the level: public markets offer no direct way in. The cleanest exposure to the shaping half is a basket of listed specialty-materials companies; the cleanest exposure to the recovery half is private ownership of a permitted plant, since the listed proxies bundle U.S. recycling with mining, overseas operations, or downstream manufacturing.

5. How the money works

The two children share a structural feature — reported revenue is a poor guide to profit — but for opposite reasons.

In 331491, revenue is metal value plus conversion value. Alloy surcharges, metal indices, and hedging pass metal cost through with a timing lag: Carpenter reported $531.0 million of surcharge revenue in fiscal 2025 and notes that surcharges protect dollar profit but dilute percentage margins, while roughly 40% of its sales sit under firm-price arrangements that retain raw-material risk.[12] Rising prices also inflate working capital — Carpenter carried $793.8 million of inventory at June 30, 2025, including $454.9 million of work-in-process, against $2.877 billion of sales.[12] What actually moves profit is mix, utilization, and yield, and mix can overwhelm volume: Carpenter's fiscal 2025 gross margin reached 26.7% against 21.2% a year earlier, and its Specialty Alloys Operations segment earned a 23.0% operating margin on $2.564 billion of sales even as shipment volume fell 11%.[12] ATI's Advanced Alloys & Solutions segment posted a 16.3% EBITDA margin on $2.146 billion of external sales, up from 15.4%, on exotic-alloy pricing and richer nickel-alloy and titanium mix; its more downstream segment earned 23.6%.[11] These are segment margins, not industry margins.

In 331492, the useful measure is margin per ton excluding metal-price pass-through. High metal prices inflate sales without improving profit, and customer-owned metal processed under tolling (processing someone else's material for a fee) may not appear as revenue at all.[3][21] The four recovery business models — own-account refining, toll refining, closed-loop integration, and byproduct recovery — convert the same feedstock into cash differently, and the swing factor is feedstock, not price. Befesa's mature steel-dust operation recorded €799.7 million of revenue and €212.1 million of adjusted EBITDA in 2025, a 26.5% margin against 20.6% in 2024, on 1.215 million tonnes of furnace dust at roughly 70% utilization — a global segment broader than U.S. 331492.[22] Sibanye's PGM recycling shows the downside: in the first half of 2023, feed volumes fell 52% as collection networks contracted, adjusted EBITDA fell 49% to $20 million — a 5% margin — while the business advanced about $2 million per day to suppliers and carried roughly $178 million of recycling advances.[27] Policy can dominate the result outright: Sibanye recognized $126 million of cumulative Section 45X credits in the first half of 2025 against recycling EBITDA excluding those credits of just $4 million.[27] At the frontier, ramp economics are brutally negative — American Battery Technology reported $4.3 million of fiscal 2025 revenue against $14.9 million of cost of goods sold, a $10.6 million gross loss, plus a $2.9 million inventory write-down.[23]

The contrast worth carrying away: in shaping, margin is set by what you make and can rise while tonnage falls; in recovery, margin is set by what you can buy and can collapse while metal prices are merely soft — with tax credits and ramp costs capable of swamping the operating result entirely.[12][27] For both, the honest operating metrics are the physical ones: throughput, capacity utilization, furnace uptime, recovery yield, payable metal, inventory days, maintenance spending, and environmental provisions. Note that no valid six-digit capacity-utilization series is published for either child — the Federal Reserve reports only broader primary-metals utilization — so utilization must be assessed plant by plant.[29]

6. Demand drivers

Shaping (331491) is pulled above all by aerospace and defense: ATI derived 68% and Carpenter 62% of fiscal 2025 sales from that market, which describes the leading listed producers rather than every establishment in the code.[11][12] Newer airframes carry more titanium and newer engines more nickel- and titanium-based alloys, and demand tracks build rates, engine shop visits, defense procurement, and supply-chain inventories.[11] Nuclear and energy want corrosion-resistant, traceable alloys; medical wants titanium implants and nickel-titanium wire; electronics, automotive, and industrial equipment fill out the rest.[11][12] A structural risk sits upstream: the United States produced no titanium sponge in 2025 and relied entirely on imports — about 44,000 metric tons with a customs value near $460 million, with Japan supplying 77% of sponge imports during 2021–24 — while net import reliance for cobalt was 79%, and 51% of U.S. cobalt consumption went into superalloys.[30] Domestic rolling capacity does not remove supply risk.

Recovery (331492) is pulled by lead batteries (67% of apparent U.S. lead consumption in 2025, with steady replacement demand plus data-center and telecom backup), steel activity (zinc from electric-arc-furnace dust, supported by steel decarbonization), precious metals (U.S. autocatalyst recycling recovered roughly 50,000 kilograms of palladium and 8,600 kilograms of platinum in 2025), and refinery operations (spent catalysts).[7][22] Lithium-ion is a real but early stream: the Department of Energy reported U.S. capacity to reclaim 35,500 tons of battery materials in 2023, with another 76,000 tons announced.[31] Replacement batteries are relatively defensive; the other streams track production cycles, and electrification cuts both ways — Johnson Matthey expected automotive use of both platinum and palladium to fall 5% in 2025 as battery-electric powertrains gained share, eroding future autocatalyst scrap.[25] Policy is a mild tailwind — lead joined the federal critical-minerals list in 2025 — but incentives do not fix poor feedstock economics.[32]

The two halves meet in the metals themselves. Secondary nickel from purchased scrap supplied roughly 130,000 metric tons, about 60% of apparent U.S. consumption in 2025, while nickel and cobalt superalloys are the highest-value output of the shaping half.[7][30] The shared theme is supply security: both children benefit from a national preference for domestic sourcing of strategic metals, whether qualified titanium and nickel alloys or recycled lead.

7. Regulation

Environmental and worker-safety rules are the dominant regulatory burden across the level, and they are heaviest on the recovery half.

• Effluent and air. The Environmental Protection Agency (EPA) regulates wastewater from rolling, drawing, extrusion, heat treatment, and finishing under 40 CFR Part 471, and process and fugitive emissions from secondary lead under Clean Air Act National Emission Standards for Hazardous Air Pollutants (NESHAP); a 2025 proposal covering 11 secondary-lead facilities remained proposed, not final, as of 2026. Confusing the forming and smelting regimes is a common classification error.[33][34]
• The cost frontier is visible. EPA calculated that requiring wet electrostatic precipitators at the eight secondary-lead facilities then lacking them would cost $621 million upfront and $73 million annually for about 3.8 tons per year of additional lead-emission reduction — roughly $19 million per ton — and did not propose it. Later rulemaking or litigation could revisit that judgment, which is the single largest identified contingent capital item in the recovery half.[9]
• Hazardous waste. The Resource Conservation and Recovery Act (RCRA) governs hazardous residuals, precious-metal reclamation, and spent lead-acid batteries — central to 331492.[35]
• Worker exposure. Occupational Safety and Health Administration (OSHA) rules set a lead permissible exposure limit of 50 µg/m³ (action level 30) and a beryllium action level of 0.1 µg/m³. OSHA identifies secondary-lead smelter workers as experiencing some of the highest occupational lead exposures in any lead-related industry.[36]
• Defense sourcing. Defense Federal Acquisition Regulation Supplement (DFARS) clause 252.225-7009 generally requires covered nickel, cobalt, titanium, and zirconium specialty metals to originate domestically or in a qualifying country — a competitive advantage for qualified U.S. mills, at the cost of documentation and traceability.[37]
• Trade and tax policy. In January 2026 the Administration directed negotiations over imports of processed critical minerals and derivative products following a Section 232 national-security finding, leaving future treatment unsettled — helpful to domestic mills, potentially inflationary given import dependence on sponge and alloying inputs.[38] On the recovery side, the Section 45X credit is generally 10% of production costs for qualifying critical minerals and begins phasing down in 2031, terminating after 2033.[28]

For both children a strong permit is a double-edged asset: it is a barrier to entry that protects incumbents, and a standing liability if emissions, exposure, or legacy contamination trigger shutdowns and remediation.

8. Consolidation

The level is fragmented (HHI 337.6; top-four share under 30%), but the two children are consolidating by different mechanisms.[5]

In shaping, consolidation happens by acquisition. Entry barriers are expensive melt-and-roll equipment, environmental permits, skilled labor, and long product-validation cycles; once a metal is specified for an engine or implant, switching suppliers is slow and costly. The qualified field is already narrow — Carpenter reports fewer than 10 major competitors in its most demanding applications, and ATI names Precision Castparts subsidiaries, Howmet, and Carpenter in high-performance products, with Haynes and VDM in flat-rolled alloys.[11][12] Acerinox's acquisition of Haynes International, combined with VDM Metals, built a larger high-performance-alloys platform backed by roughly $200 million of planned U.S. investment over four years; family firms nonetheless retain defensible niches in precision strip, foil, titanium, and medical wire.[17]

In recovery, consolidation happens substantially by attrition. Legacy environmental liabilities cap what buyers will pay, so a functioning, permitted brownfield plant can be worth more than a larger unpermitted project — and the secondary-lead count has fallen to 11 operating smelters, five fewer than at EPA's 2012 rulemaking, with six companies now owning 10 of them.[9] Lead favors vertical integration because battery makers control core-return networks; precious-metal refining competes instead on assay integrity, security, and settlement speed. Private sponsors have assembled platforms (Gopher Resource, Element Resources) while Sibanye-Stillwater expanded into U.S. precious-metal recycling through Reldan and Metallix.[20][26] Because official concentration is published only for the combined level, the visible tightness in lead should not be read as the concentration of the child as a whole.

9. Risks

Risks common to both children:

• Cyclicality and destocking can leave high-fixed-cost capacity underused; USGS has documented repeated titanium expansions followed by excess capacity when aircraft programs slipped.[30]
• Metal-price and working-capital swings move inventory values and financing needs faster than physical volume.
• Operational failure — furnace outages, poor yield, contamination, or failed certification — can halt an entire customer program at a capital-intensive asset.
• Asset concentration. Both halves run few substitutable plants; Carpenter specifically identifies Reading and Latrobe, Pennsylvania, and Athens, Alabama, as concentrations of specialized equipment.[12]
• Skilled labor. Metallurgists, furnace operators, maintenance technicians, and inspectors are scarce; Carpenter cites hiring and retention, labor disputes, and work stoppages as material risks.[12]
• Environmental and worker liabilities, including inherited contamination, can turn a cheap private acquisition expensive.
• Substitution — additive manufacturing and composites in shaping; changing battery chemistries and vehicle electrification in recovery, where lithium-iron-phosphate cells reduce recoverable nickel and cobalt value.

Child-specific: shaping carries upstream supply concentration (titanium sponge, cobalt, several alloying elements), customer/program concentration, and tariff exposure on imported inputs; recovery carries feedstock scarcity (exports, captive networks, and collectors withholding material when prices fall), assay risk in precious metals, ramp risk on new plants, and policy risk as Section 45X credits phase down from 2031 and terminate after 2033.[28][30][38] For public investors, the overarching risk is dilution — every listed proxy bundles this activity with something larger.

10. How to invest & outlook

Public investors have no pure play, so the task is to size the relevant slice inside a larger company and judge its quality. For shaping names, compare qualified aerospace and defense exposure, backlog quality, conversion margins, surcharge lag, the share of sales under firm-price versus indexed terms, mill utilization, inventory age, and customer concentration — and separate accounting earnings from cash generation during a ramp. For recovery names, first pick the metal exposure you want, then examine segment throughput, utilization, recovery yields, treatment fees, hedging, supplier advances, tax credits, and environmental provisions — consolidated revenue and spot metal prices are weak guides to recycling profit, as Johnson Matthey's £12.6 billion of revenue against £2.6 billion of sales excluding precious metals makes plain.[25]

Private investors should diligence what the federal data cannot show. For a shaping mill: ownership of customer qualifications, change-of-control approvals, furnace and mill bottlenecks, maintenance capital, scrap accounting, inventory valuation, environmental history, and skilled-worker retention. For a recovery plant: permits and inherited contamination, feedstock contracts, collection density, settlement records, realistic sustaining capital, and downside-tested metal spreads. With no defensible current industry multiple published for either child, transactions must be valued plant by plant, and expansion is safest when backed by contracted customer demand or secure feed.

Outlook — cautiously constructive, for structural reasons. Both children benefit from a durable preference for domestic supply of strategic metals: the strongest shaping prospects are in qualified aerospace, defense, nuclear, medical, and high-temperature power alloys plus strategic domestic titanium capacity, while recovery is underpinned by the fact that the United States had no primary refined-lead production in 2025 and secondary output of roughly 1.0 million metric tons supplied about 70% of apparent domestic consumption.[7][30] Near-term earnings will stay uneven in both — aircraft schedules, outages, metal-cost timing, and feedstock availability can outweigh underlying demand in any given quarter. Returns are more likely to come from qualification advantages, secure feedstock, reliable operations, and disciplined metal-risk management than from simple exposure to metal prices. Existing, permitted, qualified assets look better positioned than speculative greenfield capacity.

Sources

1. U.S. Census Bureau, 2022 NAICS Definitions: Industry Group 3314, 2022.
2. U.S. Census Bureau, 2022 NAICS Sector 31–33 Definitions: NAICS 331491, 2022.
3. U.S. Census Bureau, North American Industry Classification System, 2022 (NAICS 331492), 2022.
4. U.S. Census Bureau, 2023 County Business Patterns, 2025 release.
5. U.S. Census Bureau, 2022 Economic Census — Concentration ratios, firm count, receipts, and HHI (NAICS 33149), 2025 release.
6. Occupational Safety and Health Administration, Hazard Communication Standard Final Rule: Economic Analysis, 2024.
7. U.S. Geological Survey, Mineral Commodity Summaries 2026: Lead, 2026.
8. U.S. Census Bureau, County Business Patterns: About the Program / Methodology, current.
9. U.S. Environmental Protection Agency, Secondary Lead Smelting NESHAP Proposed Rule, Federal Register, 2025.
10. U.S. Small Business Administration, Table of Size Standards, 2023.
11. ATI Inc., Form 10-K for Fiscal 2025, 2026.
12. Carpenter Technology Corporation, Form 10-K for Fiscal 2025, 2025.
13. Berkshire Hathaway Inc., Form 10-K for 2025, 2026.
14. Materion Corporation, Form 10-K for 2025, 2026.
15. Luxfer Holdings PLC, Form 10-K for 2025, 2026.
16. AMETEK Inc., Form 10-K for 2025, 2026.
17. Acerinox, Acerinox Completes the Acquisition of Haynes International, 2024.
18. Ulbrich Stainless Steels & Special Metals, Who We Are; Perryman Company, Who We Are; Fort Wayne Metals, Company News and Ownership Transition; Elgiloy Specialty Metals, About Us; G.O. Carlson, Company Overview, current–2026.
19. Brookfield Business Corporation, Fourth-Quarter 2025 Letter to Shareholders, 2026.
20. Sibanye-Stillwater Limited, Form 20-F for 2025, 2026.
21. AMG Critical Materials N.V., Annual Report 2025, 2026.
22. Befesa S.A., Annual Report 2025, 2026.
23. American Battery Technology Company, Form 10-K for Fiscal 2025, 2025.
24. United States Antimony Corporation, Form 10-K for 2025, 2026.
25. Johnson Matthey, Annual Report 2025/26, 2026; and PGM Market Report 2025, 2025.
26. Ecobat, Our Locations; Gopher Resource, Capital Raise; Atlas Holdings, Element Resources; East Penn Manufacturing, Sustainability in Design and Manufacturing; The Doe Run Company, Sustainability Report 2022, 2020–2026.
27. Sibanye-Stillwater Limited, Form 6-K, First Half 2023, 2023; and Form 6-K, First Half 2025, 2025.
28. Congressional Research Service, Section 45X Advanced Manufacturing Production Credit, 2025.
29. Federal Reserve Board, Industrial Production and Capacity Utilization, Table 2, 2026.
30. U.S. Geological Survey, Mineral Commodity Summaries 2026 (Titanium, Cobalt, Nickel), 2026.
31. U.S. Department of Energy, FOTW #1350: 2023 U.S. Battery Recycling Facilities, 2024.
32. U.S. Geological Survey, Interior Department Releases Final 2025 List of Critical Minerals, 2025.
33. U.S. Environmental Protection Agency, Nonferrous Metals Forming and Metal Powders Effluent Guidelines (40 CFR Part 471), 2026.
34. U.S. Environmental Protection Agency, Secondary Lead Smelting: NESHAP, updated 2026.
35. U.S. Environmental Protection Agency, Regulatory Exclusions and Alternative Standards for the Recycling of Materials, accessed 2026.
36. Occupational Safety and Health Administration, Lead and Beryllium FAQs, current.
37. Acquisition.gov, DFARS 252.225-7009 — Restriction on Acquisition of Certain Articles Containing Specialty Metals, 2026.
38. The White House, Adjusting Imports of Processed Critical Minerals and Their Derivative Products into the United States, 2026.