Metal and Mineral (except Petroleum) Merchant Wholesalers (U.S.) — NAICS 4235

An investor's primer (industry-group rollup). NAICS (North American Industry Classification System) is the coding scheme the U.S. government uses to sort businesses into industries; codes get more specific as they get longer. This page covers the four-digit industry group 4235, which bundles the merchant wholesalers — firms that take title to (buy and own) the material they resell — that move metals, coal, and ores from producers to the factories, builders, and power plants that consume them. It rolls up two five-digit industries: 42351 (metal service centers) and 42352 (coal, coke, ores, and minerals). This is a synthesis page: the substance lives in the two child primers, and this level's job is to show how the two halves differ and to give the group's own official numbers.

---

1. Overview

Industry group 4235 is the dry-bulk industrial-materials middle layer: the merchants who sit between the mills and mines that make raw metal and mineral feedstock and the manufacturers, contractors, and utilities that use it. They are not producers and not end users — they buy in producer-scale volumes, hold inventory, add handling or processing value, and resell in the sizes and grades customers actually order.

For an investor, the useful thing about 4235 is that it packages two very different businesses under one label, and the contrast is the whole story. One child — metal service centers (42351) — is a large, asset-heavy distribution and light-processing industry tied to construction, autos, and aerospace. The other — coal and ore wholesalers (42352) — is a small, asset-light commodity-trading industry tied to power and steel. They share a legal definition ("merchant wholesaler, takes title") and almost nothing else: different capital intensity, different margins, different customers, different concentration, different direction of travel, and completely different ways to get exposure. Metal service centers are ~93% of the group by sales and ~99% of its employment [1][2]; everything below is mostly about them, with coal-and-ore as the high-velocity, structurally shrinking sidecar.

Three contrasts are worth fixing in mind before the detail. First, throughput per person differs by more than five times: roughly $2 million of sales per employee in metal against roughly $11 million in coal-and-ore [1][2][3]. Second, the two halves price their business in opposite ways — the metal child is overwhelmingly transactional (Reliance describes its segment as operating primarily in the spot market, and Ryerson reports no long-term fixed-price metal purchase contracts [4][5]), while the coal child is overwhelmingly contracted (87% of Peabody's 2025 mining sales volume was committed under long-term agreements, and most utility coal moves under contracts negotiated directly between mine and power plant [6][7]). Third, and unusually for two industries in the same group, they nonetheless share a formal size marker: the Small Business Administration sets the same 200-employee small-business threshold for both [8].

---

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

At the five-digit level, NAICS splits 4235 into exactly two industries. Both happen to contain a single six-digit child, so 42351 = 423510 and 42352 = 423520 — the leaf primers carry the company-by-company detail.

HHI (Herfindahl-Hirschman Index) is the standard concentration gauge — the sum of each firm's squared market share; U.S. antitrust reviewers treat anything under 1,500 as "unconcentrated." The two children sit on opposite sides of that idea in spirit: metal distribution is a genuinely fragmented, roll-up-in-progress industry, while coal-and-ore trading is a relationship-and-balance-sheet business where a handful of desks dominate — top-heavy, though at an HHI of 777 still below the line regulators treat as moderately concentrated [2].

Scope note. Both children are merchant wholesalers — they own the goods they resell. That excludes commission agents and brokers who never take title (those sit in NAICS 425, specifically 425120), the mills and mines themselves (NAICS 331/2121/2122), scrap dealers (423930), fabricators (332), petroleum (4247), precious metals and gems (423940), and construction sand and gravel (423320) [9]. On the metal side the Census definition explicitly folds in custom sawing, shearing, bending, leveling, cleaning, and edging performed as part of a sale — so "processing" is inside this code, not adjacent to it [9].

---

3. How big the group is

These are the group's own ground-truth federal figures for NAICS 4235.

How the rollup reconciles. Establishments, employment, and payroll tie out exactly to the two children added together (8,608 + 284 = 8,892 establishments; 145,207 + 2,151 = 147,358 employees) [1]. Firm counts do not: 6,006 + 238 = 6,244 against the group's 6,239 [2]. That five-firm gap is what you would expect when a handful of companies operate in both industries — they are counted in each child but only once at the group. And one number deliberately does not roll up: the metal child now leads with a newer figure, ~$296.4 billion of 2023 wholesale sales from the Annual Integrated Economic Survey [3], which comes from a different survey with a different collection frame. Do not add it to the coal child's 2022 Census $24.0 billion to manufacture a fresher group total; the $329.3 billion above is the internally consistent one.

Three things to read off these numbers:

1. The group looks very fragmented — but that is metal talking. The rollup HHI of 108.7 is even lower than metal service centers' own 122, because adding the small coal-and-ore segment enlarges the denominator and shrinks every big metal firm's share of the combined pie. The headline "unconcentrated" reading is real for 93% of the group by sales, but it masks a genuinely concentrated second segment: coal-and-ore on its own runs an HHI of 776.6 with the top four firms at 47.8% and the top fifty at 96.6% — against 55.4% for the metal child's top fifty [2]. Rollup concentration averages away that difference; don't mistake it for a description of both halves.
2. Revenue is a price snapshot, not a volume gauge — and there is no single price on either side. The ~$329 billion reflects 2022 selling prices, historically high after the 2021–22 metals and energy spike. Both children's dollar sales move with commodity prices, so this NAICS can post large year-to-year revenue swings driven by price, not tons moved. The children now show how coarse "the commodity price" is as a concept here: in coal, the 2024 average U.S. mine sales price was $52.47 per short ton, but thermal coal averaged $37.85 and metallurgical coal $180.02 [10]; in metal, inventory accounting adds its own noise — Reliance recorded $113.7 million of LIFO expense in 2025 after $144.4 million of LIFO income in 2024 [4].
3. Both halves leak business outside the code — but only one leak is measured. On the metal side the boundary quirk is now quantified: of the ~$296.4 billion of 2023 wholesale sales, $82.5 billion (about 28%) flowed through producer-owned sales branches classified as manufacturing rather than through independent merchant wholesalers [3]. Otherwise metal is well-measured. Coal-and-ore is materially under-counted and the shortfall is not quantified at all: big miners run in-house marketing and export desks booked under mining (the largest U.S. coal exporter, Xcoal, is one such marketer) [11]; most utility coal moves under direct mine-to-buyer contracts that skip the merchant layer [7]; and global houses (Glencore, Trafigura, Vitol, Mercuria) book U.S. coal and iron-ore volumes through affiliates classified elsewhere. For scale, the commodity U.S. coal market alone was roughly $68 billion in 2024 [12] against only ~$24 billion of merchant-wholesale receipts here [2].

---

4. Investable universe — where the value sits

Almost all the listed value in 4235 lives in the metal-service-center child — and in 2026 that listed slice got shorter even as the companies got bigger. Two deals closed: Ryerson completed its merger with Olympic Steel in February 2026, retiring the ZEUS listing (Ryerson now trades as NYSE: RYZ) [13], and Worthington Steel completed its acquisition of a controlling interest in Germany's Klöckner & Co in June 2026 [14]. What remains is a short list: Reliance ($14.29 billion of 2025 net sales across 311 locations — the North American scale leader) [15][4], Ryerson (~$4.6 billion of standalone 2025 revenue before a full year of Olympic Steel) [5], Worthington Steel (~$9.5 billion combined pro forma) [14], and Canada-listed Russel Metals (~C$4.3 billion of 2024 revenue, with large U.S. operations) [16]. Even here the industry is overwhelmingly private: 6,006 firms operate 8,608 establishments [1][2], so the typical firm runs one or a few locations — family-owned groups (O'Neal Industries, Alro Steel, Central Steel & Wire, Kenwal), U.S. arms of foreign trading houses (Samuel Son & Co., thyssenkrupp Materials NA, Marubeni-Itochu Steel, Toyota Tsusho America), and private-equity roll-ups [17].

The coal-and-ore child adds no public pure-play at all — every genuine 42352 wholesaler is private: independent traders Xcoal, Javelin Global Commodities, Gerald Group, the U.S. desks of Trafigura, Mercuria, and Vitol, plus Oxbow in petroleum coke and industrial carbon and DTE Vantage in coal blending and transshipment [11][18][19][20]. Public-market investors can only approximate that theme through coal producers that also market coal (Peabody, Core Natural Resources, Alpha Metallurgical, Warrior Met, Ramaco, Hallador), coke maker SunCoke, railroads such as CSX for transport exposure, and the global trader-miner Glencore — names carrying far more mining and production risk than the asset-light trading this code describes [child 42352].

One consequence of that ownership split matters at the group level: the only place merchant-trading economics are visible at scale is in the disclosures of the global houses — Glencore's Marketing arm produced $2.9 billion of adjusted earnings before interest and tax in 2025, inside its stated long-run $2.3–3.5 billion range [21], and Trafigura's Metals and Minerals segment turned $73.3 billion of fiscal-2025 revenue into $2.0 billion of operating profit before depreciation and amortization [22]. Neither is U.S.-only and neither maps onto NAICS 42352, but together they show the shape of a business the federal statistics can only imply. Net: if you want to own 4235 through the stock market, you are almost entirely buying metal service centers; the coal-and-ore exposure is a private-market and producer-proxy story.

---

5. How the money works

The two children monetize the same middleman position in opposite ways. At the group level, Census's own gross-margin table puts merchant wholesalers here at $230.5 billion of sales against $173.8 billion of cost of goods sold — a gross margin of $55.5 billion, or 24.2% of own-account sales [23]. Treat that as a midpoint, not a norm: the dispersion around it is enormous.

Metal service centers earn on spread plus processing. The metal spread — buy from the mill, resell higher — is most of the revenue but a thin slice of profit; because a center holds weeks or months of inventory, that inventory is effectively a market position. Falling prices squeeze spreads as higher-cost stock is liquidated, but rising prices are not automatically good either: margins compress when mill costs rise faster than a center can pass them through [5]. The better money is in value-added processing (cutting, slitting, leveling, laser and plate work), which carries far higher margins and makes customers stickier [24] — Reliance says roughly half its orders now include value-added processing against closer to 40% a decade ago, and Ryerson reports nearly 80% of 2025 products were processed to customer requirements, though the two define processing differently [4][5]. How much that matters shows in the results: in 2025 Reliance earned a 28.7% gross margin, $4.1 billion of gross profit, and $741.6 million of net income on $14.29 billion of sales, while Ryerson earned a 17.1% gross margin on $4.57 billion of revenue and posted a $56.4 million net loss [4][5]. "Steel distribution" implies no standard margin. A useful quirk: because inventory is the biggest asset, a downturn releases cash as the center runs down stock — so well-run distributors throw off free cash flow precisely when revenue falls, a working-capital reversal rather than sustainable earnings. The metrics that matter are tons sold, gross profit per ton, and inventory turns.

Coal-and-ore wholesalers earn on trading and logistics. Here margin per ton is thinner still and the value is in quality and geographic arbitrage (blending coals to a buyer's spec, moving tons from weak to strong markets), logistics (controlling rail cars, barges, terminals, and ocean freight), financing smaller counterparties, and hedging on futures exchanges. Freight is a first-order cost, not overhead: transporting coal to the U.S. electric-power sector averaged $17.55 per ton in 2024, down from $19.23 in 2023 [25] — against a thermal mine price of $37.85 per short ton [10], the freight bill is roughly half the value of the coal itself. Working capital is the balance sheet: merchants pay suppliers and freight providers well before collecting, finance routine trading through self-liquidating trade-finance facilities, and hold corporate credit lines against derivative margin calls [22]. The metrics are trading-desk metrics — tons handled, gross margin per ton, working-capital turns, counterparty exposure, and the shape of the forward price curve. The punchline is that volatility is the product: dislocations create the arbitrage, but the same swings can produce mark-to-market losses [child 42352].

Both, then, are commodity-linked and cyclical — but one is an inventory-and-fabrication business that rewards asset ownership, and the other is a thin-margin trading book that rewards speed, balance sheet, and logistics control.

---

6. Demand drivers

The two halves answer to different end markets.

Metal service centers are a leveraged read on the industrial economy. Demand tracks construction (~33%), automotive (~27%), aerospace (~16%), with machinery and other rounding out the rest [26]. Nonresidential construction and infrastructure are the single largest and most interest-rate-sensitive driver — Reliance identifies nonresidential construction as its largest end market by tons and reported particularly strong demand there in 2025 [4] — and reshoring mega-projects (chip fabs, data centers, EV and battery plants) are a forward tailwind. Orders are short-cycle and often just in time, so volumes move fast. Sitting on top of it all is the price of metal itself, which moves dollar revenue independent of volume. The Metals Service Center Institute (MSCI) Metals Activity Report is the industry's real-time demand pulse; its late-2025 readings showed U.S. steel shipments roughly flat-to-down and aluminum modestly up — a soft-but-stabilizing volume backdrop [27].

Coal-and-ore demand runs on power, steel, and — newly — critical-mineral sourcing. Thermal (steam) coal feeds electricity, now only ~15–17% of U.S. power from a fleet down 43% since 2010 [28]: U.S. coal consumption fell 3.5% to 410.9 million short tons in 2024, of which electricity generation took 373.3 million, the lowest in the federal primary-energy series back to 1949 [10]. The decline is not smooth — first-quarter 2025 consumption ran 18% above the prior year as natural-gas prices rose, and AI data-center load has since kept aging plants running [29]. Metallurgical coal and coke feed blast-furnace steelmaking — export-heavy and higher-margin, with China and India the marginal buyers. Because so much high-margin volume is export-oriented, global steel and seaborne trade matter as much as U.S. demand: of 502.2 million short tons of total 2024 U.S. coal distribution, domestic distribution fell 13.5% to 393.9 million while foreign distribution rose 8.1% to 108.3 million (56.6 million metallurgical, 51.7 million thermal), before exports fell to about 93 million tons in 2025 as Chinese tariffs cut U.S.-to-China shipments by roughly 92% [10][30]. This is physical work, not clerical: in 2024 rail carried 71.4% of U.S. coal shipments, river transport 11.0%, trucks 7.8%, and tramways or conveyors 9.8%, with electric utilities and independent power producers taking 90.5% of shipments [31]. The "other mineral and ore" half has its own engine: the 2025 U.S. critical-minerals list runs to 60 commodities, with the United States wholly net-import-reliant for 13 and at least half import-reliant for 20 more, and China the primary import source for 8 of those 33 categories [32] — sourcing diversification and inventory buffering are merchant work.

The common thread is that both children are downstream of heavy industry and both are price-takers — but metal leans domestic and construction-led, while coal-and-ore leans global, export-led, and structurally shrinking at home.

---

7. Regulation

Neither child is licensed or rate-regulated as a business; both are shaped by rules aimed at their commodities and their customers, and both live with policy whipsaw. The useful distinction at this level is where the rule lands: for metal it lands on the goods (tariffs on what the distributor buys), for coal-and-ore it lands on the customer (emissions rules on what the buyer burns) — and only in the coal child does a compliance regime land squarely on the merchant's own profit and loss.

For metal service centers, the story that moves the numbers is trade policy. In June 2025 the U.S. raised Section 232 tariffs (national-security tariffs on imports) on steel and aluminum to 50% from 25% [33], and 2026 proclamations introduced full-value, derivative-product, and product-specific treatment [34][35]. Tariffs prop up domestic prices — supporting inventory values and margins, and steering demand toward the domestic mills service centers buy from — but raise costs for the manufacturers those centers serve, and can leave a distributor exposed if policy shifts while it holds expensive stock. Antidumping and countervailing duties (AD/CVD, extra duties on unfairly cheap or subsidized imports) add another layer, alongside ordinary workplace, transport, and environmental rules.

For coal-and-ore, the master switch is environmental policy on the customer side: Environmental Protection Agency (EPA) power-plant rules set thermal demand, and 2025 executive actions push to keep coal plants open, roll back EPA limits, and expand federal coal leasing [36]. Trade policy moves the export book overnight [30]; the Surface Transportation Board, Federal Maritime Commission, and Mine Safety and Health Administration touch the physical chain; and the Commodity Futures Trading Commission (CFTC) oversees the futures and swaps traders use to hedge. The children add one exposure the parent previously omitted, and it belongs here because it falls on the merchant itself: cross-border trading routinely involves state-owned producers, opaque intermediaries, and sanctioned jurisdictions — the Office of Foreign Assets Control restricts parts of the Russian metals and mining sector [37], and in 2024 Trafigura pleaded guilty to a Foreign Corrupt Practices Act conspiracy and agreed to pay more than $126 million [38], a sum that can overwhelm years of thin per-ton spreads.

---

8. Consolidation

Both children are consolidating, from opposite starting points.

Metal service centers are a fragmented industry with a consolidating top — thousands of small players, no dominant giant (HHI 122, top four 16.8%) [2], and lots of room to roll up. That is exactly what the leaders are doing, and in 2026 the marquee deals stopped being announcements and became facts: the Ryerson–Olympic Steel merger closed in February 2026 into a group positioned as the second-largest North American service center, targeting ~$120 million of annual synergies [13], and Worthington Steel–Klöckner closed in June 2026 at ~$9.5 billion combined [14]. Reliance built its position through 70-plus acquisitions [15], and Russel Metals bought seven U.S. Kloeckner locations for ~$119 million in 2025 [16][17]. Cutting the other way are mills moving downstream into fabrication and distribution, and digital price transparency — Ryerson explicitly names online competitors and quoting transparency as a source of margin pressure [5].

Coal-and-ore starts already top-heavy (top four 47.8%, top fifty 96.6% of revenue) [2] and is concentrating further along two lines: producers are absorbing the merchant function through in-house marketing desks, long-term commitments, and joint ventures — reinforced by the January 2025 CONSOL–Arch merger into Core Natural Resources [39] and visible in how tightly producers lock up their own volume [6] — and global trading houses are pushing into metals and ore, with Vitol hiring iron-ore traders away from Glencore and Trafigura [18]. Small brokers survive only on a logistics niche or a regional relationship.

So the group as a whole is consolidating, but the metal half is a roll-up of many into fewer, while the coal half is a squeeze on an already-narrow field.

---

9. Risks

The two children share a cyclical, commodity-linked risk base but diverge on the big secular question.

• Shared: commodity-price and inventory risk on thin margins; cyclicality (a manufacturing or power-demand downturn hits volume directly); trade-and-tariff shocks that can reverse fast; working-capital swings; and freight and energy costs.
• Metal-specific: end-market concentration in construction and autos; mill disintermediation (producers selling direct); upstream supplier concentration (even a many-branch distributor depends on a small set of mills, so outages, quotas, or long lead times impair availability); borrowing-base contraction when metal prices fall; labor and safety exposure around cranes, saws, and cutting equipment; material substitution (composites and lightweighting); and environmental liabilities, including legacy remediation imported through acquisitions [4][5].
• Coal-and-ore-specific: the secular decline of thermal coal (the domestic coal fleet is scheduled to fall from ~172 gigawatts in mid-2025 toward ~145 by end-2028) [40]; basis and quality risk — a benchmark hedge may not track the purchased grade, location, or delivery period, and disputes over moisture, calorific value, sulfur, ash, or ore assay can change settlement values or cause outright rejection; counterparty and credit risk (a history of producer bankruptcies); capital and insurance pullback from coal on environmental-and-governance grounds; and sanctions and anti-corruption exposure, which the Trafigura case shows can dwarf a trading book's annual margin [38].

For a group-level investor the practical point is that the small child carries the structural-decline and compliance-tail risk while the large child carries the cyclical-industrial and inventory risk — so a 4235 "index" would be dominated by the metal cycle, with a small, higher-beta coal tail.

---

10. How to invest and the outlook

Public. In practice, buying 4235 through the stock market means buying metal service centers: the listed leaders — Reliance (NYSE: RS), Ryerson (NYSE: RYZ) now combined with Olympic Steel, Worthington Steel (NYSE: WS) scaled up via Klöckner, and Russel Metals (TSX: RUS) — plus producer-with-distribution names such as Steel Dynamics (NASDAQ: STLD) and Nucor (NYSE: NUE) for mill-style economics, though those give exposure to metal production and pricing rather than distribution. These trade at cyclically modest earnings multiples, so where you are in the cycle matters more than the headline number; they can look statistically cheap at peaks and expensive at troughs. The coal-and-ore child has no listed pure-play; the closest public proxies are the trader-miner Glencore — the truest read on the trading model [21] — U.S. coal producers with marketing arms, coke maker SunCoke, and railroads such as CSX, all deep-cyclical equities carrying far more production risk than the trading business the code describes.

Private. Both children are mostly private, and that is where direct exposure lives. In metal, thousands of family-owned regional centers make direct ownership, private-equity roll-ups, and adding processing capacity all viable — prized for counter-cyclical, asset-backed cash generation, with the usual traps being slow-moving stock, earnings inflated by a favorable price period, underinvested equipment, and reliance on an asset-based revolver. In coal-and-ore, exposure comes through trade finance and private credit to traders and producers, minority stakes or joint ventures in marketing platforms, or ownership of the logistics assets (export terminals, barge fleets, rail access) every trader must rent [11][18][19][20]; the underwriting question there is gross spread after logistics, inventory turns, collateral control, hedge basis, assay procedures, customer concentration, and sanctions controls — not headline sales.

Outlook. The two halves point in different directions and should be held for different reasons. Metal distribution offers steady, consolidation-driven exposure to U.S. industrial activity, with elevated Section 232 tariffs supporting domestic prices [33] and reshoring feeding the project pipeline, against a mixed near-term setup: late-2025 shipments were only stabilizing [27] and rate-sensitive construction is uneven. Coal-and-ore offers cyclical support over structural decline — AI and data-center power growth plus supportive 2025 policy are delaying coal-plant retirements [29][36][40], and premium coking coal has been volatile but firm into early 2026 [41] — while the long-run domestic trajectory stays downward [28][40] and value migrates toward export and steel-grade markets that reward scale, balance sheet, and global reach, with thermal exports hostage to seaborne prices and China trade policy [30]. The structural through-line across both, whatever the cycle, is consolidation. For the complete treatment, read the two child primers (42351 and 42352).

---

Sources

1. U.S. Census Bureau, County Business Patterns 2023, NAICS 4235 and children (establishments, employment, payroll). https://www.census.gov/programs-surveys/cbp.html
2. U.S. Census Bureau, 2022 Economic Census — Comparative/Concentration Statistics, NAICS 4235 and children (firms, receipts, CR4/CR8/CR20/CR50, HHI). https://www.census.gov/programs-surveys/economic-census.html
3. U.S. Census Bureau, 2023 Annual Integrated Economic Survey, NAICS 42351 (total sales; merchant wholesalers vs. manufacturers' branches). https://data.census.gov/table?codeset=naics~42351&g=010XX00US
4. Reliance, Inc., 2025 Form 10-K (locations, spot-market model, processing mix, gross margin, LIFO, end markets, risks). https://www.sec.gov/Archives/edgar/data/861884/000110465926020651/rs-20251231x10k.htm
5. Ryerson Holding Corporation, 2025 Form 10-K (revenue, gross margin, net loss, LIFO, processing mix, digital competition). https://www.sec.gov/Archives/edgar/data/1481582/000119312526062397/ryi-20251231.htm
6. Peabody Energy, 2025 Form 10-K (87% of 2025 mining sales volume under long-term agreements), 2025. https://www.sec.gov/Archives/edgar/data/1064728/000106472826000006/btu-20251231.htm
7. U.S. Energy Information Administration, Coal explained: prices and outlook (long-term contracts vs. spot; benchmark grades), 2025. https://www.eia.gov/energyexplained/coal/prices-and-outlook.php
8. U.S. Small Business Administration, Table of Size Standards, 2023 (NAICS 423510 and 423520 — 200 employees). https://www.sba.gov/document/support-table-size-standards
9. U.S. Census Bureau, 2022 NAICS Manual — definitions, scope, and cross-references for 423510 and 423520, 2022. https://www.census.gov/naics/reference_files_tools/2022_NAICS_Manual.pdf
10. U.S. Energy Information Administration, Annual Coal Report 2024 (consumption 410.9M short tons; mine prices; distribution and export breakdown), 2025. https://www.eia.gov/coal/annual/pdf/acr.pdf
11. Global Energy Monitor / BNamericas, Xcoal Energy & Resources — privately held; largest exporter of U.S.-origin coal, 2024. https://www.gem.wiki/Xcoal_Energy_%26_Resources
12. IndexBox, United States' Coal Market Overview 2024 (commodity market ~$67.8B), 2024. https://www.indexbox.io/blog/coal-united-states-market-overview-2024/
13. Ryerson Holding Corporation, Ryerson and Olympic Steel Announce Successful Closing of Merger, Feb 2026 (~$120M synergies; RYZ listing). https://ir.ryerson.com/news/news-details/2026/Ryerson-and-Olympic-Steel-Announce-Successful-Closing-of-Merger/default.aspx
14. Worthington Steel, Inc., Worthington Steel Completes Acquisition of Klöckner & Co, June 2026 (~$9.5B combined). https://www.sec.gov/Archives/edgar/data/1968487/000119312526254547/d31520dex991.htm
15. Reliance, Inc., Fourth Quarter and Full Year 2025 Financial Results, 2026 (net sales $14.29B; acquisition history). https://www.globenewswire.com/news-release/2026/02/18/3240630/0/en/Reliance-Inc-Reports-Fourth-Quarter-and-Full-Year-2025-Financial-Results.html
16. Russel Metals Inc., 2024 Annual & Fourth Quarter Results, 2025 (C$4.3B revenue; Kloeckner U.S. service-center acquisition ~$119M). https://www.prnewswire.com/news-releases/russel-metals-announces-2024-annual--fourth-quarter-results-302375306.html
17. Metal Center News, 2024 Top 50 Metals Service Centers, 2025 (industry structure; leading public and private companies). https://ftp.metalcenternews.com/downloads/2509_Top50-Companies.pdf
18. Mining.com, Vitol hires Glencore, Trafigura iron-ore traders in metals push, 2024. https://www.mining.com/web/vitol-hires-glencore-trafigura-iron-ore-traders-in-metals-push/
19. Oxbow, Solutions — petroleum coke, industrial carbon, terminals, bulk logistics, 2024. https://www.oxbow.com/Solutions
20. DTE Vantage / DTE Energy, Coal blending and transshipment services, 2024. https://dtevantage.com/company/
21. Glencore, Preliminary Results 2025 (Marketing adjusted EBIT $2.9B; long-run $2.3–3.5B range), 2025. https://www.glencore.com/media-and-insights/news/preliminary-results-2025
22. Trafigura, 2025 Financial Review (Metals and Minerals segment $73.3B revenue, $2.0B operating profit; trade-finance and working-capital model), 2025. https://www.trafigura.com/news-and-insights/insights/2025-trafigura-annual-results/financial-review/
23. U.S. Census Bureau, 2022 Economic Census — Wholesale Gross Margin Table, NAICS 4235 (sales, cost of goods sold, gross margin for merchant wholesalers). https://data.census.gov/table/ECNGRMARGPROF2022.EC2242GRMARGPROF?q=EC2242GRMARGPROF
24. Mead Metals / The Fabricator, Value-added services a metal service center provides, 2023–2024. https://www.meadmetals.com/blog/what-value-add-services-does-a-metal-supplier-provide
25. U.S. Energy Information Administration, Coal transportation costs to the electric-power sector ($17.55/ton in 2024; $19.23 in 2023), 2025. https://www.eia.gov/todayinenergy/detail.php?id=67346
26. Global Growth Insights, Metal Service Centers Market — end-market breakdown, 2025 (third-party industry estimate). https://www.globalgrowthinsights.com/market-reports/metal-service-centers-market-105083
27. Metals Service Center Institute, Metals Activity Report (North American steel/aluminum shipments), 2025. https://www.msci.org/research_data/metals-activity-reports/
28. Energy Innovation, What is coal's future in the United States? (coal ~15% of generation; fleet down 43% from 2010), 2025. https://energyinnovation.org/expert-voice/what-is-coals-future-in-the-united-states/
29. E&E News (POLITICO), AI gives coal plants a lifeline as data-center demand rises, 2025. https://www.eenews.net/articles/ai-gives-coal-plants-a-lifeline-as-trump-makes-them-dirtier/
30. U.S. Energy Information Administration, U.S. coal exports decreased in 2025 after four years of growth (108 MMst in 2024 → 93 MMst in 2025; China tariff impact), 2026. https://www.eia.gov/todayinenergy/detail.php?id=67405
31. U.S. Energy Information Administration, Annual Coal Distribution Report (modal split; customer split), 2024. https://www.eia.gov/coal/distribution/annual/
32. U.S. Geological Survey, Mineral Commodity Summaries 2026 (critical-minerals list and import reliance), 2026. https://pubs.usgs.gov/periodicals/mcs2026/mcs2026.pdf
33. White & Case LLP, Trump administration increases steel and aluminum Section 232 tariffs to 50%, 2025 (effective June 4, 2025). https://www.whitecase.com/insight-alert/trump-administration-increases-steel-and-aluminum-section-232-tariffs-50-and-narrows
34. White House, Proclamation on Strengthening Actions on Aluminum, Steel, and Copper, April 2026. https://www.whitehouse.gov/presidential-actions/2026/04/strengthening-actions-taken-to-adjust-imports-of-aluminum-steel-and-copper-into-the-united-states/
35. White House, Proclamation Further Adjusting Tariff Regimes for Aluminum, Steel, and Copper, 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/
36. The White House, Reinvigorating America's Beautiful Clean Coal Industry (Executive Order), April 2025, 2025. https://www.whitehouse.gov/presidential-actions/2025/04/reinvigorating-americas-beautiful-clean-coal-industry-and-amending-executive-order-14241/
37. U.S. Treasury Department, Office of Foreign Assets Control, Metals and mining sanctions guidance (FAQ 1117), 2024. https://ofac.treasury.gov/faqs/1117
38. U.S. Department of Justice, Swiss commodities trading company pleads guilty to foreign bribery scheme (Trafigura, $126M), 2024. https://www.justice.gov/archives/opa/pr/swiss-commodities-trading-company-pleads-guilty-foreign-bribery-scheme
39. Core Natural Resources / industry reporting, CONSOL Energy–Arch Resources merger creating Core Natural Resources, January 2025, 2025. https://www.mining.com/
40. U.S. Energy Information Administration, U.S. coal-fired capacity retired in 2025 was the least in 15 years; fleet path to ~145 GW by 2028, 2026. https://www.eia.gov/todayinenergy/detail.php?id=67427
41. Argus Media, Australian premium coking coal price hits 17-month high, 2026. https://www.argusmedia.com/en/news-and-insights/latest-market-news/2778303-australian-premium-coking-coal-price-hits-17-month-high