NAICS 3312 — Steel Product Manufacturing from Purchased Steel

An investor's rollup primer. NAICS is the North American Industry Classification System, the U.S. government's standard code for industries. This is an "industry group" (4-digit) that sits inside Primary Metal Manufacturing. It contains two child industries: 33121 (Iron and Steel Pipe and Tube Manufacturing from Purchased Steel) and 33122 (Rolling and Drawing of Purchased Steel). This page synthesizes the two child primers and adds the federal figures that exist only at this combined level.

---

1. Overview

Everything in NAICS 3312 begins with steel that somebody else already made. The firms here do not melt iron ore or run blast furnaces — that is the business of the big integrated mills one branch over (Nucor, Steel Dynamics, Cleveland-Cliffs, U.S. Steel, coded in NAICS 331110). Instead, companies in 3312 buy semi-finished steel — hot-rolled coil, plate, skelp, bar, wire rod — and add one more shaping step: they form and weld it into pipe and tube, cold-roll and coat it to tighter tolerances, or draw it down into wire and wire products.[3][4][5]

That single fact drives the economics of the entire group. Because the expensive raw material is purchased at market prices, the value these firms add is a thin conversion margin stacked on top of a large, volatile steel-cost pass-through. The central investment question is therefore not simply "will steel demand rise?" but "can a given plant hold its spread between selling price and steel input cost, keep its lines full, manage inventory, and stay diversified across cyclical end markets?"[4][5]

The revised children now put numbers on how thin that layer is, and they reach it from opposite ends of the group. Worthington Steel, the largest listed flat-rolled processor, earned an adjusted EBITDA (earnings before interest, taxes, depreciation, and amortization) margin of 7.0% on fiscal 2025 net sales of $3.093 billion.[7] Insteel, the wire pure-play, earned a 14.4% gross margin on fiscal 2025 sales of $647.7 million.[8] And U.S. Steel's tubular segment watched gross margin fall from 40% to 14% between 2023 and 2024 on almost unchanged shipments, as its average realized price dropped from $3,137 to $1,905 per ton.[9] Those are different income-statement lines at three companies with different product mixes and are not comparable to one another — but they make the same point three times: at this level, several points of margin move year to year on spread and inventory timing rather than on volume.

For scale, do not confuse 3312 with the mills that supply it. U.S. raw-steel production ran about 82 million metric tons in 2025 with sales value near $149 billion[6], against roughly $46 billion of receipts across this entire group.[2] For investors, 3312 is a small, deeply industrial, fragmented corner of heavy manufacturing — a late-cycle bet on U.S. construction, energy, autos, appliances, agriculture, and infrastructure, wrapped around a leveraged view on the steel price. There is no clean, sizable public basket for it; exposure comes through a handful of diversified public straddlers, one small pure-play, and a deep private and private-equity base.

---

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

The two children both buy steel and reshape it, but they make different products, serve different customers, and behave differently as businesses. 33121 turns purchased steel into hollow shapes — pipe and tube. 33122 re-rolls and draws purchased steel into flat, bar, and wire forms. They do not compete with each other; pooling them into one group is what makes 3312 look even less concentrated than either child alone (see Section 8).

Contrast table — size figures from our ingested federal statistics: County Business Patterns (CBP, the Census Bureau's annual establishment tally) 2023 for size, and the 2022 Economic Census for revenue and concentration; company and deal facts from the child primers.[1][2][4][5] Acronyms: HSS = hollow structural sections; OCTG = oil-country tubular goods (well casing and tubing); PC strand = prestressed-concrete strand (the high-strength cable inside bridge girders); WWR = welded wire reinforcement (the steel mesh in concrete slabs); HHI = Herfindahl-Hirschman Index (an antitrust concentration gauge, 0–10,000); CR4 = combined revenue share of the four largest firms.

The core divergence. 33121 (pipe & tube) is the bigger-plant, higher-pay, fewer-locations child: 294 plants averaging ~93 workers each, paying ~$87,000 a head, and carrying slightly more payroll than its sibling.[1] Its story is a cautiously constructive volume outlook (water, energy, and domestic-content-driven construction) set against uncertain margins. 33122 (rolling & drawing) is the more-plants, smaller-plant child — 442 locations averaging ~64 workers — and it splits internally: cold-finished and coated shapes are a reshoring and electrification growth story, while wire drawing is a steady-demand, fewer-and-larger-plants story.[1][5] By revenue the two children are close to even, with 33122 modestly ahead on the shipments measure; by payroll 33121 is ahead.

What changed this pass — where consolidation actually lives. This parent previously framed the group's clearest consolidation story as the wire-drawing roll-up. The revised children reverse that emphasis. Wire drawing (331222) is still shedding plants, but its deals are small bolt-ons — Insteel's $67.0 million purchase of most of Engineered Wire Products' assets and a $5.1 million purchase of selected O'Brien Wire Products inventory and equipment, both in 2024.[8] The two largest transactions anywhere in 3312 are now in rolled shapes and its service-center orbit: Worthington Steel's majority interest in Klöckner & Co and Ryerson's merger with Olympic Steel, both closing in 2026.[13][14] Rising plant counts and shrinking owner counts are not contradictory — that is what an industry looks like when capacity grows while the buyers get bigger.

One structural lever exists in only part of the group. Inside 33122, the rolled-shape child can run toll processing, where the customer keeps title to the steel and pays only a conversion fee, and plants can shift between toll and direct work as demand changes.[5] Wire drawers have no comparable escape hatch — they buy the rod, finance it, and carry its price risk — and the 33121 primer describes no toll model in pipe and tube either. So 331221 is the one place in 3312 with a genuine dial for turning commodity exposure up or down, and the one place where reported revenue per ton is not comparable across operators.

Measurement note: 33121's revenue is Economic Census receipts and 33122's is value of shipments — slightly different accounting concepts, so the ~46%/~54% split is directional, not exact.

---

3. Size (this level's rollup figures)

From our ingested ground-truth federal statistics for NAICS 3312:[1][2]

So the whole industry group is roughly a $46-billion-revenue, 56,000-job business run out of about 736 plants and 457 companies. Revenue per worker is about $824,000 — an extraordinarily high figure that, as at both children, is the clearest signal that purchased steel, not labor, dominates the cost base. Annual payroll is under 10% of receipts; steel and other materials are the overwhelming rest. Read that per-worker number as a steel-price artifact rather than a productivity measure — and, in the rolled-shape part of the group, partly an accounting artifact as well, since direct-sale revenue includes the full cost of the metal while toll revenue includes only the conversion fee for the same tonnage.[5]

Undercount and coverage caveats.

• Nonemployers omitted. CBP counts only employer establishments — those with payroll — and leaves out self-employed people with no employees.[1] For a capital-intensive steel-processing group this gap is trivial: mills and drawing lines need substantial equipment, working capital, and labor. Both children reach the same conclusion independently.[4][5]
• Firm counts don't add up cleanly. The children's Economic Census firm counts (162 for 33121, 305 for 33122) sum to 467, but the group reports 457 firms.[2] The difference is expected: a single company that operates in both children is counted once at the group level but once in each child below, so child counts overstate the group total. It also means multi-plant, multi-line operators are common — 457 firms run 736 establishments.
• Per-child sales are approximate. As noted, one child reports receipts and the other reports shipments; the group's $45.998B is the single clean revenue figure for the whole level.[2]
• Data quality is asymmetric below this level, and it got worse on this pass. 33121 is clean: the level equals its one child, and the Economic Census reports an unsuppressed $21.4 billion of receipts across 162 firms even though narrower sales figures are suppressed at 331210.[4] 33122 is not: neither 331221 nor 331222 could establish a current, defensible revenue figure or concentration ratio from public sources, so our ingested combined 33122 statistics are the only consistent size frame there.[5]
• An independent source disagrees on wire drawing. The American Wire Producers Association's preliminary December 2023 compilation, built from Bureau of Labor Statistics data, counts 260 facilities, 10,518 employees, and $716.9 million in annual wages for wire drawing, against CBP's 198 establishments, 14,260 employees, and ~$0.92 billion in payroll for the same child.[15][1] That is a definitional gap rather than an error in either source, and the child primer reports both. We use CBP at this level because only CBP reconciles to the 3312 totals above — but an investor sizing wire drawing should hold a range, not a point.

---

4. Investable universe — where the value sits across the children

There is no large, clean, listed vehicle for 3312. Value concentrates in a short list of public companies that each straddle this level (mixing it with other businesses) plus a long tail of private and private-equity-owned firms. Where value sits differs sharply between the two children — and this pass moved the two rosters in opposite directions: 33121's widened, while 33122's narrowed as two of its listed names merged into one.

Inside 33121 (pipe & tube) the public and large-private exposure is real but partial — no name is a clean NAICS pure-play:

• Public / diversified: Nucor (NUE, its tubular group), Tenaris (TS), Atkore (ATKR), Maruichi, and now Vallourec (French-listed premium seamless and OCTG) and Nippon Steel, which gained U.S. Steel's Tubular Products business when that acquisition closed on June 18, 2025.[12]
• Public / narrower direct plays: NWPX Infrastructure (NWPX, large-diameter water pipe) and Friedman Industries (FRD).
• Large private owners: Zekelman Industries, Bull Moose / Caparo, Stupp Bros., plus Lerman Enterprises (Lock Joint Tube), Benteler International, and Grupo Villacero (Tex-Tube). Full company-by-company detail is in the 33121 primer and its 331210 child.[4]

Inside 33122 (rolling & drawing) the standout remains the one genuinely clean listed read on the level:

• Pure-play: Insteel Industries (IIIN) — draws wire and makes PC strand and welded wire reinforcement, with fiscal 2025 sales of $647.7 million, and has grown by acquiring smaller wire-products makers.[8] Small-cap, steel-price-sensitive, and a read on one of 33122's two children.
• Adjacent / partial: Worthington Steel (WS), whose majority acquisition of Klöckner & Co adds a broad international service-center business and makes the post-deal company a thinner read on cold rolling[13]; Ryerson (RYZ), which absorbed Olympic Steel on February 13, 2026 and mixes distribution, aluminum, and stainless with carbon processing[14]; Reliance (RS, through Feralloy); Nucor (NUE, through its half-ownership of NuMit and Steel Technologies); Leggett & Platt (LEG, captive rod-and-wire buried inside Bedding Products); Bekaert (BEKB, Belgian-listed global wire transformation).[5]
• Private / PE: on the rolled side, Heidtman Steel and Samuel, Son & Co. anchor a base of regional coil processors and toll operations; on the wire side, Heico Companies (Davis Wire, National Standard), WireCo WorldGroup, and specialists including Fort Wayne Metals, Sumiden, Mid-South Wire, Tree Island, Kiswire, Tokusen, Wire Mesh Corporation, Oklahoma Steel & Wire, National Wire Products, and Concrete Reinforcements. No source establishes a defensible ranking of these firms by U.S. revenue — treat the list as major identifiable participants, not a league table.[5]

Reality check. Public exposure to 3312 is either niche-and-illiquid (Insteel, NWPX, Friedman) or heavily diluted by other businesses (Nucor, Tenaris, Vallourec, Atkore, Nippon Steel, Worthington Steel, Ryerson, Leggett & Platt). The children now quantify how wide that dilution runs: NWPX's water-transmission segment was 66.7% of its 2025 sales[10], while tubular products were just 9% of Friedman's fiscal 2025 sales.[11] Sizing the relevant slice is the first step of any position. Note also the direction of travel: consolidation in 33122's orbit has made the listed names bigger while making them thinner reads on actual conversion processing.[5] There is no dedicated public fund or index tracking this level, and broad steel and metals ETFs mainly own steelmakers and miners — economically different exposure. The private and PE tail is where much of the tonnage, and much of the transactable opportunity, actually lives.

---

5. How the money works

Across both children this is a conversion-spread business, and the model is the same one you would apply to any purchased-input reprocessor. A firm buys steel at whatever the market charges, runs it through forming, welding, rolling, or drawing equipment, and sells it at that steel cost plus a conversion margin — or, in the rolled-shape child, processes metal the customer owns for a fee. Four consequences follow:

1. Steel price dominates revenue. Annual payroll of $4.493 billion against $45.998 billion of receipts is under 10%[1][2]; purchased steel and other materials are the overwhelming rest, so reported sales rise and fall with steel prices even when tonnage is flat.
2. Inventory and contract timing make or break a quarter. When steel prices fall, firms holding higher-cost stock take inventory holding losses; when prices rise faster than customer contracts reset, the spread compresses. Insteel quantifies the exposure from the wire side: a 10% increase in wire-rod cost would have cut annual pretax earnings by $37.5 million had selling prices not moved.[8] Margins are therefore lumpier than the underlying demand.
3. Utilization is the controllable lever. Payroll is a modest share of sales, but forming, rolling, and drawing lines are energy-hungry and capital-intensive, so keeping the lines full is the key operating metric and swings in it move profit disproportionately. There is no published capacity-utilization series for either child specifically; the Federal Reserve's broader iron-and-steel-products measure averaged 72.2% in 2025.[16]
4. The spread between product lines is wide enough that "the group's margin" is not a useful number. The children bookend it. Energy tubulars are the volatile end — U.S. Steel's tubular gross margin fell from 40% to 14% and EBITDA from $638 million to $135 million between 2023 and 2024, on price and spread rather than volume.[9] Project-based water pipe is the steady end — NWPX's water-transmission gross margin was 19.1% in 2025 against 18.5% in 2024, even as its average steel purchase cost moved from $994 per ton in 2023 to $914 in 2024 and $967 in 2025, with monthly 2025 averages ranging from $838 to $1,093.[10]

The mitigations are structural rather than clever, and both children describe the same ones: match customer and mill contract indices, keep inventory turns high, use firm-price contracts, and — where the model allows — shift toward toll work so the customer finances the metal.[5][7] Useful gauges are also common: tons shipped, price per ton, the steel-to-product conversion spread, capacity utilization, backlog and book-to-bill, and inventory days. Because purchased steel drives the P&L (profit-and-loss statement), the framings used for other sectors do not apply here — there is no regulated-utility rate base, no real-estate-investment-trust funds-from-operations, and no mining all-in sustaining cost. Watch the spread, the utilization, and the inventory position.

---

6. Demand drivers

The two children lean on overlapping but distinct end markets, which is exactly why product and customer diversity is valuable at the group level — these markets do not move in lockstep.

• Energy (mainly 33121): drilling, well completions, and pipelines drive OCTG and line pipe. The energy transition cuts both ways — long-run substitution away from fossil fuels threatens conventional OCTG and hydrocarbon line pipe, while carbon dioxide transport, hydrogen, geothermal, and carbon-storage wells call for specialized tubulars.[4]
• Construction and infrastructure (both): HSS, piling, and sprinkler pipe from 33121; PC strand, welded wire reinforcement, fencing, mesh, and nails from 33122. Insteel estimated 85% of its fiscal 2025 sales related to nonresidential construction and 15% to residential.[8] Federal spending under the IIJA (Infrastructure Investment and Jobs Act) is a multi-year tailwind, reinforced by domestic-content rules (Section 7).
• Municipal water (33121): large-diameter transmission pipe, plus recurring replacement demand from aging systems.[10]
• Autos, appliances, and machinery (mainly 331221): cold-finished bar and cold-rolled strip feed precision parts, motors, and stampings; reshoring supports domestic cold-finishing. Demand follows vehicle production and platform awards rather than vehicle sales alone, so strikes and model changeovers can remove local volume abruptly.[5]
• Electrification (a 331221-only growth pocket, new to this rollup): non-grain-oriented electrical steel for motors and grain-oriented steel for transformers. The Department of Energy points to data centers, electric vehicles, charging stations, and renewable generation as transformer-demand drivers, and estimates that roughly 55% of in-service distribution transformers are more than 33 years old.[22] This favors processors with stamping, annealing, and core-assembly capability over commodity slit-coil operations.
• Agriculture and general industry (both): mechanical tube, fencing, springs, fasteners, welding consumables, tire bead wire, and rope.

Both children are late-cycle, and both are more volatile than their end markets: customers build stock when they expect price increases and destock when prices fall, so shipments swing harder than underlying consumption.[5]

---

7. Regulation

The regulatory picture is common to both children and centers on trade policy, which is the single biggest lever — and a double-edged one.

• Tariffs. Section 232 of the Trade Expansion Act of 1962 lets the President impose national-security tariffs on steel; the additional rate has generally been 50% since June 4, 2025,[17] and 2026 proclamations retained that rate for core metal articles while introducing different treatment and further product-specific exceptions for several derivative categories.[18] This helps domestic pipe makers, re-rollers, and wire drawers by keeping cheap imported finished product out — but it also raises their own input costs. That is not theoretical: Insteel sourced approximately 27% of fiscal 2025 rod purchases from imports, up from 15% the prior year.[8] The useful question is not "do tariffs help steel" but what each company's feedstock and finished-goods import exposure, pass-through ability, and inventory position look like when the policy moves.
• Antidumping and countervailing duties. Targeted cases on specific imports benefit domestic producers more cleanly than blanket tariffs. In the OCTG investigation initiated in April 2026, alleged dumping margins run 43.64%–55.16% for Austria, 73.68%–75.31% for Taiwan, and 124.15%–126.08% for the United Arab Emirates — petition allegations rather than final duties, but a live source of price support or disruption depending on where a mill sits.[20]
• Domestic-content rules. The Build America, Buy America Act (BABA) and long-standing Buy America requirements generally mandate U.S.-made iron and steel through all manufacturing processes — from initial melting through coating — on federally funded infrastructure, subject to waivers.[19] That is a direct advantage for domestic pipe, wire, strand, and mesh, and a more indirect one for rolled shapes, arriving where purchased coil is further rolled, coated, or cut for compliant products.[5]
• Safety and environment. Plants answer to the Occupational Safety and Health Administration (OSHA, machine-guarding and lockout/tagout), the Environmental Protection Agency (EPA, pickling, coating, and wastewater permits), and — for pipeline products — the Pipeline and Hazardous Materials Safety Administration (PHMSA). The hazard set is real: the rolled-shape child runs a recordable injury and illness rate above the private-industry average.[5] Because no one here runs a blast furnace, the environmental footprint and scrutiny are far lighter than at the integrated mills upstream, and carbon regulation arrives indirectly through electricity, fuel, and purchased-steel prices.
• Customer-imposed regulation. Specifications, quality audits, traceability, destructive and nondestructive testing, and product-liability exposure are a genuine compliance burden in both children — and a barrier to entry that protects incumbents.[4][5]

---

8. Consolidation

At the national level the whole group is unconcentrated — and notably more fragmented than either child on its own. Our ingested 2022 concentration figures for 3312 show the top four firms at 21.8% of revenue, the top eight at 31.5%, the top twenty at 48.6%, the top fifty at 69%, and an HHI of just 196[2] — well under the 1,500 mark that U.S. antitrust agencies treat as the edge of "unconcentrated." Compare 33121, where the same distribution runs 30.3% / 45.5% / 68.5% / 87.5% with an HHI of 388.8, and 33122 at 25.3% / 36.4% / 56.1% with an HHI of 249.8.[4][5] Pooling two non-competing sub-markets mechanically dilutes measured concentration at every point of the curve, so the group looks more competitive than the businesses inside it actually are — a reminder to read concentration at the level where firms genuinely compete. Note too that neither of 33122's six-digit children could establish a concentration figure of its own, so the combined 33122 numbers are the finest resolution available on that side.[5]

Direction of travel differs by child — and the emphasis has shifted. Incumbents in 33121 continue to grow by acquisition (Nucor assembling its tubular group, Tenaris acquiring IPSCO, Zekelman's multiple deals, and Nippon Steel's completed purchase of U.S. Steel in June 2025[12]), yet the industry stays fragmented and antitrust still bites — a Department of Justice investigation led Tenaris and Benteler to abandon a proposed U.S. seamless-pipe combination.[21] Within 33122, the plant-count picture is unchanged: wire drawing (331222) fell from 252 establishments to 198 between 2017 and 2023 while rolled shapes (331221) rose from 204 to 244 on reshoring and tariff cover.[5] But the corporate consolidation story now runs the other way from what this page previously implied. The group's two largest transactions are both on the rolled-shape side — Worthington Steel securing approximately 62% of Klöckner & Co as of June 3, 2026[13], and Ryerson closing its merger with Olympic Steel on February 13, 2026, with former Olympic holders taking roughly 37% of the combined company[14] — while wire consolidation remains a series of small bolt-ons.[8] Freight costs on heavy steel and bulky wire also create pockets of local concentration; a mesh, nail, or water-pipe plant can dominate its shipping radius even where the national picture looks fragmented.

One measure puts the whole group's scale in perspective: the U.S. Small Business Administration size threshold for both of 33122's children is 1,000 employees, and with average plants employing 58 to 72 people there, essentially the entire sub-industry qualifies as "small business."[5] The handful of listed names sit well outside the typical participant — at either child.

---

9. Risks

The risk set is shared across both children and compounds at the group level:

• Steel-price mismatch — the dominant risk: inputs move before customer prices reset, swinging both revenue and, via inventory timing, margins.[8][9]
• Broad cyclicality — energy, construction, autos, appliances, and capital spending can all turn down together, and customer destocking amplifies the swing.
• Captive and upstream competition — the integrated mills also make pipe, cold-rolled steel, and wire, competing directly with independents that must buy steel from those same mills. Nucor is simultaneously a supplier, a named competitor, and an owner of processing and tubular capacity.[5]
• Trade-policy whipsaw — tariffs protect the output but inflate the input, and the regime can flip with administrations or courts, creating the opposite inventory shock.[17][18]
• Excess capacity and discounting — incumbent capacity additions can outrun demand, capping pricing power; the broad iron-and-steel-products utilization measure averaged 72.2% in 2025.[16]
• Thin margins and capital intensity — lines cost money whether or not they run, and there is little cushion when utilization drops.
• Customer concentration and project risk — large OEMs, distributors, and contractors hold bargaining power (one listed processor's three largest customers were about 32% of fiscal 2024 net sales), and water and line-pipe projects can be delayed or cancelled.[5]
• Supplier consolidation and specialty inputs — fewer upstream mills reduces purchasing leverage, and inputs such as zinc have a limited supplier base.[7]
• Substitution, in both directions — plastic, concrete, or ductile iron can displace steel pipe; aluminum and advanced high-strength grades reduce steel content per vehicle; synthetic rope competes with steel rope. Against that, engineered welded mesh is taking share from hand-tied rebar, and higher-strength grades raise the value of precision processing.[4][5]
• Labor and safety — nationwide manufacturing labor shortages have affected recruitment of skilled and unskilled workers, and injury rates run above the private-sector average in at least one child.[4][5]
• Energy transition — a genuine long-run threat to conventional OCTG and hydrocarbon line pipe, only partly offset by low-carbon tubular applications.[4]
• Acquisition integration — the 2026 mergers introduce execution risk at exactly the companies investors would use for exposure.[13][14]
• Investability and data risk — limited pure-play public options, an illiquid private tail, and, as Section 3 shows, no reliable current per-child revenue or concentration data below 33122 to underwrite against.[5][15]

---

10. How to invest & outlook

Public-market investors face a choice between narrow-and-illiquid and diluted. The cleanest direct reads are small: Insteel (IIIN) for wire and reinforcement inside 33122 — and only for one of that child's two halves — and NWPX Infrastructure (NWPX) or Friedman Industries (FRD) for narrower slices of 33121, where the relevant segment is 66.7% and 9% of company sales respectively.[8][10][11] Broader exposure comes through diversified names — Nucor (NUE), Tenaris (TS), Vallourec, Atkore (ATKR), Nippon Steel, Worthington Steel (WS), Ryerson (RYZ) — but each mixes 3312 activity with other businesses, and the 2026 deals made several of them thinner reads, not thicker ones.[13][14] No dedicated fund or index tracks this level. There is no dividend-yield or utility-style income thesis here; treat these as cyclical industrials tied to the steel spread, and value them on mid-cycle earnings rather than peak spreads — watching enterprise value to EBITDA, free-cash-flow yield, leverage, and replacement cost per ton.

Private-market investors get the larger opportunity set: regional coil processors, toll operations, and coating lines on the rolled side; commodity drawers, coating and heat-treatment specialists, and sub-scale wire, mesh, and reinforcement plants on the wire side. The wire roll-up thesis survives — the shrinking establishment count is real — but it should now be framed as plant rationalization at modest ticket sizes, because the headline corporate M&A in this group has moved to rolled shapes and service centers.[8][13][14] The children converge on one diligence checklist, worth restating as this level's own: separate distribution gross margin from real conversion earnings; direct tons from toll tons; spot from indexed contracts; a commodity inventory windfall from sustainable processing margin; maintenance from growth capital expenditure; and customer-approved capacity from merely installed capacity. Verify that a pipe or tube target actually manufactures rather than distributing or fabricating purchased product.[4][5] Steel-purchase pricing formulas, customer qualifications, line utilization, inventory turns, environmental history, and freight radius matter more than headline revenue.

Forward-looking judgment. The volume outlook is cautiously constructive — water infrastructure, domestic-content rules, energy, reshoring, electrification, and industrial construction support several end markets at once, and the two children's demand cycles don't fully overlap.[19][22] The margin outlook is less certain, because spare capacity (broad iron-and-steel-products utilization averaged 72.2% in 2025), volatile steel costs, and changeable trade policy can overwhelm demand growth — and the same tariff wall that protects the output inflates the input.[16][17][18] The through-line across 3312: the winners are operators that match purchase and sale indices, keep plants full, hold inventory turns tight, and sell qualified products where metallurgy, service, and reliability beat the lowest quote — not simply the largest nameplate capacity. Put concisely, 33121 is the diversified pipe-and-tube volume story with no pure-play; inside 33122, 331221 is the reshoring-and-electrification story with no pure-play either, and 331222 is the construction-demand, spread-and-consolidation story with exactly one — and all of it is, at one remove, a leveraged bet on the price of steel.

For the full company-by-company detail, economics, and diligence checklists, read the two child primers: 33121 and 33122.

---

Sources

1. U.S. Census Bureau, 2023 County Business Patterns (CBP) — establishments, employment, and payroll for NAICS 3312 and its children (our ingested ground truth). https://www.census.gov/programs-surveys/cbp.html
2. U.S. Census Bureau, 2022 Economic Census: Concentration and Firm Size Statistics (EC2200) — receipts, firm count, and concentration (CR4/CR8/CR20/CR50, HHI) for NAICS 3312 and its children. https://api.census.gov/data/2022/ecnsize.html
3. U.S. Census Bureau / Office of Management and Budget, 2022 NAICS Manual — industry definitions for 3312, 33121, and 33122. https://www.census.gov/naics/reference_files_tools/2022_NAICS_Manual.pdf
4. Histometrics child primer, NAICS 33121 — Iron and Steel Pipe and Tube Manufacturing from Purchased Steel (and its 331210 detail) — company detail, economics, and regulation for pipe and tube. /primers-preview/33121
5. Histometrics child primer, NAICS 33122 — Rolling and Drawing of Purchased Steel (331221 Rolled Steel Shape; 331222 Steel Wire Drawing) — company detail, economics, and regulation for rolling and drawing. /primers-preview/33122
6. U.S. Geological Survey, Mineral Commodity Summaries 2026 — U.S. raw-steel production and sales value. https://pubs.usgs.gov/periodicals/mcs2026/mcs2026.pdf
7. Worthington Steel, 2025 Annual Report. https://www.sec.gov/Archives/edgar/data/1968487/000119312525181888/d879196dars.pdf
8. Insteel Industries, Fiscal 2025 Form 10-K. https://www.sec.gov/Archives/edgar/data/764401/000143774925031597/iiin20250927_10k.htm
9. United States Steel Corporation, 2024 Form 10-K — Tubular segment pricing and margin. https://www.sec.gov/Archives/edgar/data/1163302/000116330225000018/x-20241231.htm
10. NWPX Infrastructure, 2025 Form 10-K. https://www.sec.gov/Archives/edgar/data/1001385/000143774926005861/nwpx20251231_10k.htm
11. Friedman Industries, Fiscal 2025 Form 10-K. https://www.sec.gov/Archives/edgar/data/39092/000143774925020219/frd20250331_10k.htm
12. Nippon Steel, Completion of U.S. Steel Acquisition, June 18, 2025. https://www.nipponsteel.com/en/newsroom/news/2025/20250618_100.html
13. Worthington Steel, Klöckner Acquisition Form 8-K, 2026. https://www.sec.gov/Archives/edgar/data/1968487/000119312526254547/d31520dex991.htm
14. Ryerson Holding Corporation, Ryerson and Olympic Steel Announce Successful Closing of Merger, 2026. https://ir.ryerson.com/news/news-details/2026/Ryerson-and-Olympic-Steel-Announce-Successful-Closing-of-Merger/default.aspx
15. American Wire Producers Association, National Data 2023, 2024. https://www.awpa.org/wp-content/uploads/2024/08/National-Data-2023.pdf
16. Federal Reserve Bank of St. Louis and Federal Reserve Board, Capacity Utilization: Iron and Steel Products. https://fred.stlouisfed.org/series/CAPUTLG3311A2A
17. White House, Adjusting Imports of Aluminum and Steel into the United States, 2025 — Section 232 rate of 50% effective June 4, 2025. https://www.whitehouse.gov/presidential-actions/2025/06/adjusting-imports-of-aluminum-and-steel-into-the-united-states/
18. White House, Further Adjusting the Tariff Regimes for Imports of Aluminum, Steel, and Copper into the United States, 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/
19. Office of Management and Budget, Build America, Buy America Implementation Guidance, 2023. https://www.whitehouse.gov/wp-content/uploads/2023/10/M-24-02-Buy-America-Implementation-Guidance-Update.pdf
20. U.S. Department of Commerce, Commerce Initiates Antidumping and Countervailing Duty Investigations on Oil Country Tubular Goods, 2026. https://www.trade.gov/commerce-initiates-antidumping-duty-and-countervailing-duty-investigations-oil-country-tubular
21. U.S. Department of Justice, Benteler and Tenaris Abandon Merger After Investigation, 2023. https://www.justice.gov/archives/opa/pr/benteler-steel-tube-manufacturing-corp-abandons-merger-tenaris-sa-after-justice-department
22. U.S. Department of Energy, Distribution Transformer Types and Demand Drivers, 2024. https://www.energy.gov/oe/articles/energy-department-researches-distribution-transformer-types-and-demand-drivers

Ground-truth note: level figures (establishments, employment, payroll, firm count, receipts, and concentration) are drawn from our ingested federal statistics for NAICS 3312 [1][2]. Per-child size splits are from the child primers' ingested CBP 2023 data [1][4][5]. Company facts (tickers, deal history, private owners, segment shares, and margins) originate in the SEC filings and public sources cited in the child primers [4][5]; tickers are provided only for the investable-universe and how-to-invest sections per house style. Two confidence caveats carried up from the children: per-child receipts and concentration are not separately available below 33122, and independent trade-association data disagrees with CBP on the size of wire drawing [15]; and the revised 331222 research rests on a single independent brief whose intended cross-validation could not be completed, so wire-drawing figures carry less corroboration than the rolled-shape and pipe-and-tube figures.