Chemical Manufacturing (U.S.) — NAICS 325

A Histometrics rollup primer for public-market and private investors. NAICS (North American Industry Classification System) 2022 code 325 is a subsector — the three-digit level — that gathers seven child industry groups (four-digit codes): basic chemicals (3251), resins/rubber/fibers (3252), agricultural chemicals (3253), pharmaceuticals (3254), paints/coatings/adhesives (3255), soaps/cleaning/beauty (3256), and "other" chemical products (3259). Reported figures are federal facts with citations; forward-looking statements are labeled as judgments. This page synthesizes the seven child primers and our ingested ground-truth federal statistics for the subsector.

1. Overview

Chemical manufacturing is where the U.S. economy turns oil, gas, minerals, air, and biomass into almost everything else: the plastics in packaging, the medicines in the cabinet, the fertilizer on the field, the paint on the wall, the shampoo in the shower, and the ink, glue, and explosives that build and print the physical world. It is one of the largest manufacturing subsectors in the country — about $901 billion of shipments in 2022 from roughly 14,500 plants employing 911,000 people [1][2].

But the single most useful thing to understand about NAICS 325 is that it is not one industry — it is seven, and underneath the seven sit three completely different economic engines wearing one code. Roughly 57% of the subsector's dollars are commodity molecule-makers — capital-heavy plants that buy a feedstock, convert it, and earn a volatile spread times volume (basic chemicals, resins/rubber/fibers, most of the "other" bin, and the fertilizer half of ag chemicals). About 28% are intellectual-property businesses that earn premium margins on patented, government-approved products until the patent runs out (pharmaceuticals, and the pesticide half of ag chemicals). The remaining 15% are brand-and-formulation businesses where the physical product is cheap and the value lives in a brand or a hard-to-switch specification (soap and beauty; paint, coatings, and adhesives) [1][3][4][5][6][7][8][9].

Those three engines have opposite cyclicality, opposite margins, opposite owners, and opposite ways to invest. A deep-cyclical polyethylene cracker and a defensive, recession-proof drug plant sit in the same subsector and have almost nothing in common as investments. The revised child research adds a second-order lesson that only shows up at this level: the defensiveness of the "safe" engines belongs to their installed base, not to their growth pipelines — the subsector's most contract-protected business (industrial gas) wrote the single largest project loss of the cycle, and the brand children's margins bent under 2025 tariffs. The distinctive value of a subsector-level view is the contrast across the seven children — which is where this primer begins.

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

The table below ranks the children by size (share of subsector receipts) and lines up the axes that matter to an investor. Tickers are held to Sections 4 and 10.

(CR4 = the four-firm concentration ratio, the share of receipts held by the largest four firms; HHI = the Herfindahl-Hirschman Index, a 0–10,000 concentration gauge where below 1,500 is "unconcentrated." Shares are child receipts measured against the subsector ground truth on the 2022 Economic Census basis; establishment and employment subtotals reconcile exactly to the subsector total, and receipts to the dollar — see §3.)

Five contrasts do most of the work:

• Money and jobs don't line up — the tell of the three engines. Basic chemicals is the biggest child by revenue (~33%) but only ~18% of jobs, because commodity plants are extremely automated (~$1.8 million of revenue per worker) [3]. Pharmaceuticals is the mirror image: only ~26% of revenue but ~35% of all jobs — the single largest employer in the subsector — because regulated, skilled drug-making is labor-heavy and high-wage (~$0.73 million of revenue per worker) [6]. Capital-intensity versus labor-intensity is the first fault line.
• Cyclical versus defensive — with the defensive side no longer flawless. The commodity children (3251, 3252, 3259, and fertilizer) live and die on the feedstock spread and the operating rate, and most are near or working through a 2023–25 trough [3][4][9]. The IP and brand children (3254 pharma, 3256 beauty, plus industrial gas inside 3251) remain defensive — demand for medicine, cosmetics, and contracted gases barely blinks in a recession [6][8] — but the revision shows the qualification clearly: an industrial-gas major booked ~$3.6 billion of project-exit costs on clean-energy projects in fiscal 2025, and the listed mass-beauty growth story attributed its fiscal-2026 gross-margin decline primarily to tariffs [3][8]. Defensiveness attaches to contracted volumes and established brands, not to whatever those businesses are building next.
• Investability ranges from "clean listed pure-play" to "off the public market entirely" — and it has loosened slightly. Pharma, ag chemicals, paint, and beauty offer focused U.S.-listed pure-plays; cleaning, resins, and adhesives are reachable mainly through diversified proxies. None of 3259's three children has a U.S.-listed pure-play, but both explosives majors now sponsor U.S. depositary receipts, so the trade is no longer strictly offshore [9]. Large private, foreign-state, and cooperative slices persist inside every other child [3][4][5][7][8].
• Listed purity is not the same as economic quality — and the children now prove it three ways. In fibers, the closest listed pure-play ran a negative 5.8% gross margin in its Americas segment while the fiber business embedded in a diversified major earned a 27% EBIT margin on $1.05 billion of sales [4]. In cleaning, the sole listed chemical pure-play carries roughly $1.1 billion of market value against a $45.8 billion child [8]. In "other chemicals," there is no pure-play anywhere, yet the richest segment margin in the subsector's specialty tier — 53.8% segment EBITDA on activated carbon — sits inside a diversified name [9]. Across 325, the best economics are frequently a segment or a private asset, not a ticker that matches the code.
• They don't even share a stock-market sector. In standard market classification, pharma is Health Care, beauty and much of cleaning are Consumer Staples/Discretionary, and the rest is Materials — so "owning Chemical Manufacturing" is not one index trade but three (§10).

3. How big it is (the subsector rollup)

Our ingested federal ground-truth figures for NAICS 325:

Read together: a ~$901 billion shipments business run by ~911,000 workers across ~14,500 plants — about $1.0 million of revenue per employee at well-above-average pay (~$97,700), the signature of a capital- and knowledge-intensive economy where long-lived plant and skilled, regulated labor do the work, not cheap headcount [1][2].

The rollup is unusually clean, and the revision left it that way. The seven children's establishments sum to 14,543 and employees to 911,245 — exact to the unit — and receipts sum to $900.97 billion to the dollar and payroll to within rounding [1][2][3][4][5][6][7][8][9]. Only the firm count doesn't add cleanly: the children total 11,205 companies versus 10,806 at the subsector, a gap of ~399 firms that operate in more than one child (a Dow in petrochemicals and resins, a Procter & Gamble in cleaning and beauty) and are counted once at the subsector level. That is a normal feature of how the Census aggregates, not an error. One vintage caveat worth naming: some children now cite newer survey figures alongside the census basis — the industrial-gas industry inside 3251, for example, leads with $17.3 billion of 2023 receipts from the Annual Integrated Economic Survey rather than the ~$16.4 billion 2022 figure that reconciles upward [3]. All arithmetic on this page stays on the 2022 Economic Census basis; mixing vintages would break the tie-out.

Concentration — and why the number is almost meaningless here. The subsector looks astonishingly unconcentrated: a CR4 of just 13% and an HHI of 76.8 [1], a small fraction of the 1,500 mark antitrust agencies treat as the "unconcentrated" ceiling. That number is an artifact of addition, not a read on market power. The subsector's HHI sits below every child whose HHI is published except 3259 — the residual "everything else" bin, itself a bundle of unrelated trades — because pooling seven product families whose leading firms are largely different companies in non-competing markets mechanically dilutes any one firm's share toward zero (3256's is suppressed in the federal data, so it cannot be compared). The revision shows this dilution is fractal. Nearly every child now documents the same effect inside itself: 3251 is less concentrated than four of its five children; 3252's HHI of 338.8 sits below both of its children (371.8 and 712.4), and 32521's in turn sits below both of its children; 3253's 530.5 sits below both children (948.4 and 1,088.7); 3254's 313.7 sits below its dominant finished-drug child's 557; and 3259's 59.4 sits below all three of its children [3][4][5][6][9]. The one exception is 3255, whose 401.4 sits between paint (785) and adhesive (359.6) rather than below both [7]. The real market structures live one and two levels down, and many are tight oligopolies: petrochemical crackers (CR4 ~74%), the industrial-gas duopoly, titanium dioxide, nitrogen and phosphate fertilizer (CR4 ~77.5% and ~86.4%), the plasma majors (an estimated 70–80% of U.S. collection and fractionation), and explosives (CR4 ~53%) [3][5][6][9]. An antitrust regulator sees a fragmented subsector; an investor sees a stack of concentrated markets. Do not read the 76.8 as competition.

Undercount caveat — and it runs toward understatement. This is not a hidden cash economy or a field of tiny sole proprietors; the great bulk of Sector 325 is large, permitted, well-counted plants, so there is no missing small-business tail inflating the picture. The distortions run the other way, and three matter:

1. Factory-gate, not consumption. The $901 billion is what U.S. plants ship, not what Americans use. Because enormous volumes arrive as imports, consumption is far larger: U.S. prescription-drug spending alone was about $805.9 billion in 2024 — more than triple pharma's entire domestic factory output — with the FDA reporting that roughly 53% of branded and 69% of generic finished drugs distributed here are made abroad and only 11% of active-ingredient manufacturers are U.S.-based [6]. Beauty shows the same gap and a genuine measurement dispute: one market-research estimate puts U.S. beauty retail near $130 billion for 2025, while Circana's measured-retail series reports $36.0 billion of prestige plus $72.7 billion of mass-market sales — differently scoped series the child declines to reconcile or combine — against the $36.5 billion shipped from domestic plants [8].
2. Off-code jobs and split value chains. The 911,000 factory workers exclude the research staff, headquarters, and sales forces filed under other codes — the biopharmaceutical industry alone directly employs over 1 million Americans against pharma's 321,000 factory workers here [6]. Integrated complexes book revenue across many codes (a Gulf Coast site reports under refineries, resins, and basic chemicals at once), so no single code captures the whole chain [3][4]. The starkest case is explosives, where most U.S. military energetics are made at government-owned Army ammunition plants run by prime contractors and so barely touch the $3.16 billion federal line, while the ATF counts just under 3.2 billion pounds of explosives used in 2022 and the industry association sizes the broader commercial value chain at more than 60,000 jobs and over $19 billion — expressly a value-chain estimate, not code revenue [9].
3. The soft edges are the small and captive tails. Where small, individual, family, or captive ownership dominates — hobbyist and artisan soap makers, indie/direct-to-consumer beauty brands that outsource all production, small compounders and blenders, in-house captive ink and resin volumes — the count is a floor, not a ceiling [8][9]. Compost is the sharpest illustration: the census counts 88 firms and $756 million, while industry surveys count close to 5,000 U.S. composting facilities, most run by municipalities or waste companies filed elsewhere [5]. A minor offsetting distortion: U.S. subsidiaries of foreign multinationals are counted as domestic "firms," which understates how concentrated global ownership really is.

Finally, the $901 billion is a 2022 figure — a high-price year at the tail of the post-pandemic boom; commodity-chemical prices have since fallen with the down-cycle, so current-dollar sales in the cyclical children are lower even where volumes held [3][4].

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

There is no way to own NAICS 325 as a unit — no single fund tracks the subsector, and (as noted) its names don't even share one stock-market sector. Exposure is assembled child by child, and how reachable the value is differs sharply. Tickers appear here and in §10.

• Cleanly investable through listed pure-plays. Four children offer focused U.S.-listed access: pharma (the finished-drug majors — household names, directly or as American Depositary Receipts, ADRs, U.S.-traded certificates for foreign shares — hold nearly all the public liquidity) [6]; ag chemicals (nitrogen's CF Industries, phosphate's Mosaic, crop-protection's Corteva and FMC) [5]; paint (Sherwin-Williams, PPG, RPM International, Axalta) [7]; and beauty (Estée Lauder, Coty, e.l.f. Beauty, Inter Parfums) [8]. Basic chemicals adds the subsector's quality anchor — the industrial-gas duopoly Linde and Air Products, contracted, toll-road compounders — plus diversified majors and a bench of pure-cycle plays [3]. Two of these maps are about to change: Corteva plans a Q4 2026 separation that would create a larger listed crop-protection pure-play than exists today, and Axalta should not be modelled as an unchanged standalone, with a shareholder vote on the AkzoNobel combination set for August 5, 2026 [5][7].
• Reachable mainly through diversified proxies. The cleaning half of 3256 has no focused U.S. consumer stock — you own it as a slice of staples giants (Procter & Gamble, Colgate-Palmolive, Church & Dwight, Clorox), with lone surfactant pure-play Stepan — roughly $1.1 billion of market value against a $45.8 billion child — at the chemistry end [8]. Resins and rubber (3252) run through the same diversified chemical majors (Dow, LyondellBasell, Westlake, Celanese, Eastman) [4]; adhesives through one pure-play (H.B. Fuller, ~$3.47 billion of FY2025 revenue) and otherwise inside diversified industrials or foreign leaders — Henkel's Adhesive Technologies unit alone booked ~€11.0 billion of 2024 sales, dwarfing any U.S.-listed adhesive name [7].
• Off the U.S. public market — with one route now opened. All three children of 3259 still have no U.S.-listed pure-play, but explosives is now tradeable domestically through depositary receipts on Orica (OCLDY) and the newly demerged Dyno Nobel (DNLZY), with LSB Industries a straightforward U.S.-listed way to sit upstream in ammonium nitrate; ink remains a Japanese-parent or private game (DIC, Sakata INX, artience), and the ~86% "all other" bin is reachable only as segments (Avient, Ingevity, Eastman Kodak) [9]. A striking share of capacity across every child is private, foreign-state, or cooperative-owned: Chevron Phillips, INEOS, Formosa, and SABIC in petrochemicals; Invista, Ascend, and Lycra in fibers; Koch and Simplot in fertilizer; Chinese-state Syngenta in pesticides; family firms like S.C. Johnson in cleaning; Austin Powder in explosives; and — as of a $9.7 billion buyout completed in January 2026 — the largest domestic chlor-alkali asset, OxyChem, now inside Berkshire Hathaway [3][4][5][8][9][12].

Where the value concentrates: by dollars, basic chemicals and pharma together are ~59% of the subsector, and both are broadly investable — pharma the more defensive, basic chemicals the more cyclical. But the recurring lesson of the revised children is that where the listed ticker is and where the economics are often diverge: the purest assets in almost every child are frequently private (contract manufacturers and CDMOs, plasma fractionators, farmer co-ops, family formulators, compounders and blenders), and the best margins are frequently a segment inside a diversified firm rather than a matching pure-play. That is the recurring opening for private capital. Full company rosters live in the seven child primers.

5. How the money works

The three engines named in §1 are the key to the whole subsector, because they get paid in three different ways:

• Commodity molecule-makers (≈57% of receipts): spread × utilization. Owners earn the gap between a volatile input (ethane, naphtha, natural gas, electricity, phosphate rock, corn) and a volatile product price, per ton or per gallon, times volume through a high-fixed-cost plant. Because fixed costs are high, earnings swing far more than revenue: a world-scale plant is very profitable near 90–100% utilization and loss-making in the 60s–70s. The revised children make the 2025 trough concrete rather than rhetorical — integrated polyethylene profit fell from ~$750 per ton in 2024 to ~$580 in 2025 against a ~$830 historical average; one TiO₂ producer's utilization fell from 96% to 77% and absorbed ~$111 million of unabsorbed fixed cost; a chlor-alkali segment's income fell from $296.4 million to $181.1 million even as sales rose; another producer took a $727 million goodwill impairment plus $393 million of closure costs; and ExxonMobil's Chemical Products earnings fell from $2.6 billion to $800 million on margins it called "deeply bottom-of-cycle" [3][4][10]. Two structural themes run through all of them — the U.S. shale-gas cost advantage that makes American petrochemicals, methanol, chlor-alkali, and fertilizer globally competitive exporters, and the ever-present threat of Chinese and Middle-East oversupply, with one industry estimate putting global chemical overcapacity at 222 million tonnes in 2024, the highest since 1978 [3][4][5][9][11]. The shale edge is also not uniform: light-ethane cracking yields far less butadiene than naphtha (0.0178 versus 0.0476 tonnes per tonne of feed), so the same feedstock shift that advantages U.S. resin starves U.S. tire-rubber makers of their backbone monomer [4]. The one genuine exception is industrial gas, which sells much of its output on 10-to-20-year take-or-pay contracts (the customer pays for a minimum volume whether it uses it or not) — one major reports roughly $59 billion of future minimum-purchase and plant-sale consideration plus a ~$7.1 billion project backlog — utility-like, recurring, and the subsector's most defensive commodity economics [3].
• IP businesses (≈28%): premium margin over a patent cycle. A patented drug or a newly registered pesticide commands premium pricing (70–90% gross margins on a near-costless pill) for its protected years, then generics or biosimilars flood in and price collapses toward cash cost. Value is an off-balance-sheet asset — the patent plus the government approval (an FDA drug approval, an EPA pesticide registration) — that takes years and fortunes to build and can be lost overnight to a patent cliff, a court vacatur, or a failed inspection [5][6]. The protection is not evenly granted: under Medicare negotiation, biologics get about 13 years of market pricing before negotiation bites versus 9 years for small-molecule pills, a gap that steers capital toward large molecules [6]. Genuinely new chemistry is scarce on the crop side too — one producer describes its newest herbicide as the first new mode of action in more than three decades, which is precisely why the premium is worth so much [5]. Owners get paid through dividends and buybacks (mature firms), cash burn toward a binary readout (clinical-stage biotech), or a slice of future sales (drug-royalty investors).
• Brand and formulation businesses (≈15%): brand equity and spec-in stickiness. The physical product is cheap; buyers pay for trust, scent, shelf presence, or a qualified specification. Prestige beauty routinely runs 70–80% gross margins, and even a mass-market name reported 70.7% — though the same company's selling, general and administrative expense reached 63% of sales, the reminder that these are brand-economics businesses wearing a manufacturer's code [8]. Paint and adhesive earn a formulation spread made durable by "spec-in" — once a coating is qualified on a car line or a glue designed into a diaper, switching is costly, so revenue repeats — and the mix premium is now measurable: one adhesive pure-play's Engineering Adhesives segment ran a 22.2% adjusted EBITDA margin against 15.6% in Building Adhesive Solutions [7].

Read across the three engines and a margin ladder emerges that no single child page can show: the thinnest returns sit in commodity conversion and low-value formulation (an ink-heavy segment near 5.7% operating margin), the middle in concentrated industrial conversion (explosives at low-twenties EBITDA margins), and the richest in specified-in, registered, or patented product (activated carbon at 53.8% segment EBITDA; prestige beauty and patented drugs at 70–90% gross margins) [6][8][9]. These are company and segment disclosures on different scopes, not like-for-like industry margins — but the ranking is consistent, and it is the same ranking as the three engines. The unifying investor lens differs accordingly: spread and operating rate for the commodity children, innovation pipeline and registration/patent strength for the IP children, and brand pricing power and spec-in retention for the brand children. Owning one is not a substitute for owning another.

6. What drives demand

Most of Sector 325 sells derived demand — its molecules are inputs to other industries — so the commodity and specialty children track the industrial cycle (autos, construction, packaging, agriculture, electronics) with amplification, and are broadly exposed to foreign demand and competitors' new capacity because so much output is exported [3][4][7][9]. The IP and brand children break from that cycle: medicine demand rests on an aging, chronically ill population — a 2025 CDC study found 76% of U.S. adults report at least one chronic condition — and is essentially non-cyclical [6], while beauty and cleaning track households and premiumization (buyers trading up faster than volumes grow), with U.S. prestige beauty retail up 4% and mass-market beauty up 5% in 2025 [8].

On top of the cyclical base sit the subsector's real growth pockets, which are increasingly secular and policy-driven:

• Decarbonization and electronics — blue/green hydrogen and carbon capture on a base the Department of Energy puts at roughly 10 million metric tons of U.S. hydrogen a year, plus ultra-high-purity gases and inorganics for the semiconductor fab build-out [3].
• Biologics and the obesity wave — GLP-1 (glucagon-like peptide-1) diabetes/weight-loss drugs drove roughly 29% of all 2024 U.S. drug-spending growth, and the shift toward biologics continues — though the sources genuinely disagree on how far it has gone: commonly put at ~37–38% of U.S. prescription spending, while the FDA said in March 2026 that biologics account for 51% of drug spending on just 5% of prescriptions. Take the direction, not the decimal [6].
• Electrification and lightweighting — heavier electric vehicles wearing tires faster (rubber), structural adhesives replacing welds and bolts, and engineered compounds for wiring, where an EV carries 2.5–4.0 km of wire against 1.0–1.8 km in a combustion car [4][7][9].
• Defense and critical minerals — a funded rearmament cycle lifting explosives and energetics (new federal capacity includes a ~$435 million Army TNT plant in Kentucky targeted for about 2028 and a ~$635 million Iowa artillery complex around 2029, though that output sits largely outside the private code), plus deeper mining for lithium and copper as ore grades fall [9].
• Regulation as a demand creator — recycled-content and low-emission mandates, state organics-diversion laws driving compost, and PFAS (per- and polyfluoroalkyl substances, "forever chemicals") drinking-water limits that name granular activated carbon as best-available treatment, with compliance potentially running to 2031 [5][8][9].

Two of these pockets carry an explicit brake the children now disclose. Explosives' mining engine is partly offset by structural thermal-coal decline — Powder River Basin coal was 14% of one major's Americas revenue in FY2024 and falling [9]. And ethanol set a record at 16.49 billion gallons in 2025, yet EPA projects conventional ethanol consumption of only ~14.2–14.3 billion gallons in 2026–2027, so growth has to come from higher blends, exports, and prospectively sustainable aviation fuel rather than the domestic base [3]. The near-term swing factor for the commodity children remains the inventory cycle — customers overshoot in both directions, and the 2023–25 downturn was largely a de-stocking event [4][5].

7. Regulation

Every child is a permit-driven, heavily regulated business, and regulation is both a permanent cost and a competitive moat that favors scale and disadvantages higher-emitting foreign supply. Two federal agencies dominate, splitting the subsector along the same lines as its economics:

• The EPA (Environmental Protection Agency) governs the commodity and specialty children. The Clean Air Act's hazardous-air-pollutant standards — including the April 2024 "HON" (Hazardous Organic NESHAP, National Emission Standards for Hazardous Air Pollutants) rule, which covers roughly 220 plants, mandates fenceline monitoring, and is projected to cut more than 6,200 tons of hazardous air pollutants a year — hit petrochemicals, resins, and ink [3][4][11]. The Toxic Substances Control Act (TSCA) governs chemical risk across the board, and its PFAS reporting obligation is the nearest shared hard deadline in the subsector, with most submissions due October 13, 2026 [7][9]. Product- and process-specific regimes reach FIFRA (the Federal Insecticide, Fungicide, and Rodenticide Act) pesticide and sanitizer registration (3253, 3256), the ATF (Bureau of Alcohol, Tobacco, Firearms and Explosives) Federal Explosives License for explosives (3259 — 9,185 active licenses and permits in FY2024, including 2,036 manufacturer licenses), the 2024 asbestos rule requiring six of eight remaining U.S. chlor-alkali plants to convert diaphragms within five years, and two determinations that cut across children: an unreasonable-risk finding on 1,4-dioxane in November 2024 that expressly covers byproduct exposure from ethoxylation — the surfactant chemistry linking 3251 and 3256 — and an unreasonable-risk conclusion on formaldehyde that reaches directly into adhesive formulations [3][5][7][8][9].
• The FDA (Food and Drug Administration) governs the IP and consumer children. It is the common overlord of medicine-making — current Good Manufacturing Practice, inspections, and approval pathways for drugs, biologics, and diagnostics, with 82 biosimilars approved as of March 2026 [6] — and, since MoCRA (the Modernization of Cosmetics Regulation Act of 2022, enforced from July 2024), of beauty as well; it also reaches medical gases (regulated as drugs) and food-contact materials, including printing inks and food-grade compounds [3][8][9][11].

Two policy fronts are reshaping the biggest child specifically. The Inflation Reduction Act's Medicare drug-price negotiation is a structural headwind to branded pricing — the first 10 prices take effect January 2026, at least 38% below 2023 list. And Section 232 trade tariffs, in an April 2026 action on certain patented pharmaceuticals and ingredients, are tiered rather than flat: a 100% default rate, cut to 20% for firms with approved onshoring plans, 15% for EU, Japanese, South Korean, and Swiss product, and 0% for firms signing pricing-and-onshoring agreements, phasing in from July 31, 2026, aimed at an active-ingredient base of which only about 15% of patented volume is made domestically [6]. Trade policy moves the commodity children too — and it points in opposite directions for different products, and reverses fast. Anti-dumping duties protect domestic fiber and fertilizer while export-facing resin fears retaliation; a 15% reciprocal tariff on nitrogen imports imposed in April 2025 was lifted for Trinidadian product by executive order that November; and on June 29, 2026 the White House authorized an eight-month suspension of certain countervailing duties on Moroccan phosphate fertilizer, duties that had been in place since 2021 [4][5]. Neither child's protection should be underwritten as durable.

8. Consolidation

Every child is consolidating, and the barriers that drive it are common across the subsector — multi-billion-dollar plants, feedstock and energy access, permits, patents, registrations, brand equity, and slow customer qualification — which is why the field keeps getting fewer and larger. The marquee deals span all three engines:

• Commodity children. Industrial gas consolidated into a global oligopoly (the ~$70-billion-plus Praxair–Linde merger, Air Liquide–Airgas); Berkshire Hathaway completed its $9.7 billion purchase of OxyChem in January 2026, with Occidental retaining specified legacy environmental liabilities; Middle-East national oil companies are moving downstream (ADNOC/OMV merging into Borouge and buying Nova Chemicals for ~$13.4 billion, ADNOC acquiring Covestro for ~$16.3 billion); Koch completed a ~$3.6 billion purchase of OCI's Wever, Iowa nitrogen complex in August 2024; and resin, rubber, and fiber producers are cutting hard — Dow halved its dividend in July 2025 and moved to cut ~4,500 jobs, LyondellBasell is selling four European plants while buying the rest of a Louisiana cracker, Westlake closed a ~1.0-billion-pound PVC plant in December 2025, Goodyear sold most of its polymer-chemicals business to private equity for $650 million, and fibers ran restructuring through bankruptcy court (Ascend filed April 2025 and emerged that December; Lycra used a prepackaged Chapter 11 to shed ~$1.2 billion of debt) [3][4][12].
• IP children. Pharma runs on M&A as patent-cliff insurance (2025 pharma deal value ~$240 billion, up ~81% year over year), alongside a contract-manufacturing roll-up — the $16.5 billion Catalent take-private, plus roughly $24.9 billion of new CDMO capacity announced in 2025, about 74% of it flowing to the United States — and a tariff-driven reshoring capex cycle in which drugmakers have pledged on the order of $500 billion of new U.S. manufacturing [6]. Crop protection consolidated in one dramatic 2015–2018 wave — Dow–DuPont spinning out Corteva, Bayer buying Monsanto for about $66 billion (with roughly $9 billion of DOJ-required divestitures), ChemChina buying Syngenta for about $43 billion [5].
• Brand children. Paint produced headline mergers (Sherwin-Williams–Valspar at ~$9.5 billion; the AkzoNobel–Axalta all-stock combination announced in November 2025 at roughly $25 billion of enterprise value, with an Axalta shareholder vote set for August 5, 2026); adhesives runs a continuous drip of tuck-ins and carve-outs measured in the hundreds of millions; beauty runs a steady drip of strategics buying indie brands and private equity buying the contract manufacturers; and cleaning is pruning thin-margin volume — Henkel sold its North American private-label detergent business (~€500 million) to First Quality in 2025, and Reckitt sold a majority of Essential Home to Advent at the end of 2025, retaining 30% [7][8].

The revision surfaces a genuinely new pattern at this level: the corporate action now splits as much as it combines, and the two run on opposite ends of the subsector. The commodity end keeps concentrating (Berkshire/OxyChem, Koch/OCI, Orica buying out its U.S. explosives joint venture in June 2026), while at the IP and specialty end the merger wave is finished and structures are being refined: Corteva plans a Q4 2026 separation into two public companies, Incitec Pivot's 2025 demerger created the listed pure-play Dyno Nobel, and imaging split Carestream in two at the end of 2025 [5][9]. The crucial point for reading the concentration data is unchanged: consolidation happens within product markets, not across them, so it tightens the real oligopolies one and two levels down while leaving the subsector-level HHI stuck near zero (§3). The aggregate fragmentation and the on-the-ground market power are both true at once.

9. Risks

The risks rhyme across the subsector but weight differently by engine:

• Cyclicality and operating leverage (commodity children). The fixed-cost structure that magnifies profits in an up-cycle magnifies losses when utilization falls — the 2023–25 troughs in petrochemicals, resins, TiO₂, chlor-alkali, and fertilizer all showed it, and the 2025 write-downs are now visible ($727 million of goodwill impairment plus $393 million of closure costs at one chlorovinyls producer, a 50% dividend cut at another major, and Chapter 11 for two fiber makers) [3][4].
• Feedstock, energy, and a narrowing U.S. edge. Ethane, naphtha, natural gas, electricity, corn, and mined inputs all swing, and margins can invert even when volumes hold — one producer reported U.S. natural-gas and ethane costs up 51% and 33% in 2025 while its own selling prices fell 4% and volumes fell 6%. As LNG and ethane exports pull up domestic gas prices, the shale-cost advantage slowly erodes [3][4].
• Global oversupply, led by China. Excess Chinese and Middle-East capacity can flood commodity grades and crush spreads regardless of U.S. demand, with global chemical overcapacity estimated at 222 million tonnes in 2024 [4][9][10][11].
• Patent/exclusivity cliffs and policy pricing (IP children). A top drug can lose 80–90% of revenue within a year of generic or biosimilar entry, and 2026–2030 is an exceptionally heavy wave — more than $230 billion of revenue across ~190 drugs, about 69 of them blockbusters. IRA negotiation and pesticide registration vacaturs add policy risk that can strand inventory or reset prices [5][6].
• Environmental, safety, and litigation liability (all). Ethylene-oxide and PFAS exposure, opioid and lead-paint suits, legacy Superfund and phosphogypsum sites, chlorine and explosives incidents, and tightening emissions rules all carry real remediation and legal tails. Roundup (glyphosate) is the largest single case: roughly $10 billion-plus already paid, with a $7.25 billion class settlement sought in 2026 to cap future claims [3][5][6][7].
• Import dependence and tariff whiplash — pointing opposite ways. The subsector depends on imported active pharmaceutical ingredients and pesticide active ingredient (largely from China) — U.S. pharmaceutical imports rose from $73 billion in 2014 to $215 billion in 2024 — yet is also a major exporter of resins and basic chemicals, so the same tariff shock can help an import-competing child and hurt an export-facing one [4][5][6]. Tariffs now bite the brand children directly too: 2025 Section 301 duties, including 25% on Chinese titanium dioxide, lifted pigment prices roughly 10–15% and pushed ink makers into open price increases, and the listed mass-beauty growth story attributed its fiscal-2026 margin decline primarily to tariffs [8][9].
• Big-project and stranded-asset risk — including inside the defensive engine. Multi-billion-dollar, hard-to-shut plants punish mistiming the cycle; and the subsector's most defensive child wrote its largest single loss on growth, roughly $3.6 billion of project-exit costs at one industrial-gas major in fiscal 2025 [3].
• Single-plant fragility and supply concentration. Concentrated single-site supply can halt customers nationwide — the world's film supply runs through Rochester; a September 2024 pool-chemical plant fire in Conyers, Georgia forced roughly 17,000 residents to evacuate, required removal of nearly 14 million pounds of reactive chemicals, and ended with the owner deciding not to rebuild manufacturing there; and an upstream ammonium-nitrate force majeure disrupted a major explosives supplier into a tight 2025 market [3][9].

10. How to invest, and the outlook

There is no clean way to own the subsector — no NAICS 325 fund exists, and the names span three stock-market sectors, so even the index route fractures: a broad materials fund captures the commodity and specialty children (Dow, Linde, Sherwin-Williams, and peers) but not pharma, which lives in health care funds, nor beauty and cleaning, which sit in consumer staples/discretionary. Exposure is built engine by engine, matched to a view.

Public routes, by engine:

• Defensive anchors — pharma finished-drug majors and biotech/sector ETFs (exchange-traded funds) [6]; the industrial-gas duopoly Linde (LIN) and Air Products (APD), remembering that the contracted base and the project pipeline are not the same asset [3]; and defensive consumer names in cleaning and beauty (Procter & Gamble (PG), Colgate-Palmolive (CL), Church & Dwight (CHD); beauty pure-plays Estée Lauder (EL), e.l.f. (ELF)) [8]. Owned for stability, priced at premium multiples.
• Cyclical value — diversified chemical majors Dow (DOW), LyondellBasell (LYB), Westlake (WLK) for petrochemicals and resins [3][4]; fertilizer's CF Industries (CF) and Mosaic (MOS) [5]; and the pure-cycle plays in TiO₂, chlor-alkali, and ethanol — best bought near a trough, with earnings that compress hard in downturns and dividends that can be cut, as 2025 showed [3][4].
• Specialty and IP tilts — crop protection's Corteva (CTVA) and FMC (FMC), with Corteva's planned Q4 2026 split the single event most likely to change how that child is owned [5]; paint's Sherwin-Williams, PPG, and Axalta (the last pending its August 2026 shareholder vote) and adhesive's H.B. Fuller (FUL) [7]; and diversified specialty names carrying an "other-chemical" segment (Avient (AVNT), Ingevity (NGVT)), plus LSB Industries (LXU) and the explosives ADRs Orica (OCLDY) and Dyno Nobel (DNLZY) [9].

In every case, share prices, dividend yields, and valuation multiples belong to each company's own disclosures — judge commodity names on spread and cycle timing, IP names on pipeline and registration/patent strength, and brand names on pricing power.

Private routes. A large share of the subsector's capacity is off-market, and this is where much of the purest exposure sits: energy/infrastructure private equity and midstream in petrochemicals; distressed and restructured debt in fibers, where the Ascend and Lycra Chapter 11s converted lenders into owners of real, cash-generating plants; farmer cooperatives, mineral royalties, and compost platforms in fertilizer; venture and PE in pesticide biologicals; contract manufacturers and CDMOs (contract development and manufacturing organizations) across pharma, beauty, and cleaning; and buy-and-build roll-ups of the fragmented ink, compound, and blender tail of 3259 [3][4][5][8][9]. All of NAICS 3259's pure-play ownership and large slices of every other child are reachable only privately or through foreign listings and depositary receipts.

The outlook (forward-looking judgment). Read Sector 325 as a barbell across its three engines. The commodity children — over half the subsector — are near or working through a hard cyclical trough, with recovery gated by Chinese oversupply, the industrial inventory cycle, a narrowing U.S. feedstock edge, and new Gulf Coast capacity arriving in 2026 into a still-soft market; one integrated major called its 2025 chemical margins "deeply bottom-of-cycle," and the down-cycle may run toward the end of the decade. Their durable winners are the lowest-cost producers with a feedstock or energy advantage [3][4][10][11]. The IP children stay defensive and profitable but face narrowing branded pricing power under IRA negotiation and a heavy 2026–2030 patent-cliff wave, offset by biologics and a real, capital-intensive but incomplete reshoring of the API base [6]. The brand children compound steadily on premiumization and spec-in stickiness, with beauty the growthier half and cleaning the more defensive — though the revised research walks back one old claim: paint and adhesive grow at a similar mid-single-digit value rate over the long run, and adhesive's advantage is a less rate-exposed end-market mix and a richer specialty tilt, not faster growth [7][8]. The genuine growth, across all three engines, is increasingly secular and policy-shaped — decarbonization and electronics, biologics and GLP-1, EV lightweighting, defense energetics, and PFAS-driven treatment demand. The practical conclusion of this rollup: don't buy "chemicals" — decide which of the three engines you want (spread, patent, or brand), then choose the child and the name whose economics you actually want to own, and check whether the listed ticker or the private asset is the one that actually holds those economics. For the full argument on any one child, read its dedicated primer.

---

Sources

1. U.S. Census Bureau. 2022 Economic Census — Concentration by Largest Firms and receipts, NAICS 325 (receipts $900,972,532 thousand; 10,806 firms; CR4 13.0%, CR8 20.0%, CR20 32.2%, CR50 46.5%; HHI 76.8). Histometrics ingested federal ground-truth statistics. https://www.census.gov/programs-surveys/economic-census.html
2. U.S. Census Bureau. County Business Patterns 2023 — NAICS 325 (14,543 establishments; 911,245 employees; ~$89.05B annual payroll; ~$25.18B Q1 payroll). Histometrics ingested federal ground-truth statistics. https://www.census.gov/programs-surveys/cbp.html
3. Histometrics child primer 3251 — Basic Chemical Manufacturing (receipts $297.9B; 1,374 firms; 2,621 establishments; 168,047 employees; ~$18.56B payroll; CR4 23.8%, HHI 221.7; group HHI below four of five children; petrochemicals CR4 74.3% and polyethylene profit ~$750/ton 2024 to ~$580/ton 2025 vs. ~$830 average; industrial gas CR4 66.8%, take-or-pay economics, ~$59B future minimum-purchase consideration and ~$7.1B backlog, ~$3.6B of FY2025 project-exit costs at one major; chlor-alkali segment income $296.4M→$181.1M and a $727M impairment plus $393M closure costs; TiO₂ utilization 96%→77% with ~$111M unabsorbed fixed cost; ethanol record 16.49B gallons 2025 against EPA's projected 14.2–14.3B conventional gallons for 2026–2027; DOE ~10 Mt/yr U.S. hydrogen; EPA 2024 asbestos rule for chlor-alkali diaphragms; 222 Mt global overcapacity in 2024; Berkshire–OxyChem completed January 2026; Praxair–Linde and Air Liquide–Airgas; ADNOC/OMV–Nova ~$13.4B and ADNOC–Covestro ~$16.3B; Koch, Chevron Phillips, INEOS, Formosa, SABIC, POET).
4. Histometrics child primer 3252 — Resin, Synthetic Rubber, and Fibers Manufacturing (receipts $129.7B; 1,291 firms; 1,567 establishments; 114,089 employees; $10.20B payroll; CR4 30.9%, HHI 338.8 — below both children's; resin/rubber ~95% of receipts vs. fibers ~5%; feedstock and energy ~70% of one major's cost of sales; 2025 segment margins ~4.1% and ~11.7% EBIT/EBITDA vs. ~10.9% and ~21.2% in 2024; ExxonMobil Chemical Products $2.6B→$800M; Westlake gas +51% / ethane +33% against prices −4% and volumes −6%; synthetic-rubber industrial production index 80.46 on 2017=100 in 2025; butadiene yields 0.0178 mt/mt ethane vs. 0.0476 naphtha; ~40% of polyethylene exported vs. fiber AD/CVD protection; Unifi Americas −5.8% gross margin vs. Eastman Fibers $1.05B at 27% EBIT margin, EBIT −38%; Dow 50% dividend cut and ~4,500 job cuts; Goodyear–Gemspring $650M; Ascend and Lycra Chapter 11s; Celanese Lanaken closure; 2026 Golden Triangle startup; HON rule ~220 facilities and 6,200+ tons/yr).
5. Histometrics child primer 3253 — Pesticide, Fertilizer, and Other Agricultural Chemical Manufacturing (receipts $48.81B; 766 firms; 1,033 establishments; 37,403 employees; $3.53B payroll; CR4 38.9%, HHI 530.5 — below both children's 948.4 and 1,088.7; fertilizer $30.19B ~62% vs. pesticide $18.6B ~38%; nitrogen CR4 77.5%, phosphate CR4 86.4% and HHI 2,402; compost census 88 firms / $756M vs. ~5,000 U.S. facilities; Bayer–Monsanto ~$66B with ~$9B of DOJ divestitures; ChemChina–Syngenta ~$43B; Koch–OCI Wever ~$3.6B, August 2024; Yara–Gulf Coast Ammonia $1.3B, July 2026; Corteva's planned Q4 2026 separation; Roundup ~$10B+ paid and a $7.25B class settlement sought in 2026; FIFRA registration review; April 2025 nitrogen tariff lifted for Trinidad in November 2025; June 29, 2026 Morocco phosphate duty suspension; phosphate rock as LFP battery feedstock).
6. Histometrics child primer 3254 — Pharmaceutical and Medicine Manufacturing (receipts $234.2B; 2,222 firms; 2,905 establishments; 321,068 employees; ~$36.5B payroll; CR4 28.4%, HHI 313.7 — below its dominant finished-drug child's 557; 76% of U.S. adults with ≥1 chronic condition; U.S. drug spending $805.9B in 2024 with GLP-1s ~29% of growth; biologics disagreement — ~37–38% of prescription spending commonly cited vs. FDA's March 2026 figure of 51% of spending on 5% of prescriptions; 82 biosimilars approved; IRA first 10 prices January 2026 at ≥38% below 2023 list, and 13 vs. 9 years of market pricing for biologics vs. small molecules; April 2026 Section 232 tiered tariffs — 100% default, 20%/15%/0% carve-outs, phasing in from July 31, 2026, against ~15% domestic patented API volume; 53% branded and 69% generic finished drugs made abroad, 11% of API manufacturers U.S.-based; imports $73B in 2014 to $215B in 2024; 2026–2030 cliff >$230B across ~190 drugs; 2025 M&A ~$240B, +81%; Catalent take-private $16.5B; ~$24.9B of 2025 CDMO capacity, ~74% to the U.S.; ~$500B of reshoring pledges; plasma oligopoly 70–80%; >1M direct biopharma jobs).
7. Histometrics child primer 3255 — Paint, Coating, and Adhesive Manufacturing (receipts $53.23B; 1,422 firms; 1,664 establishments; 77,815 employees; ~$5.98B payroll; CR4 31.3%, HHI 401.4 — between paint's 785 and adhesive's 359.6; paint ~62% / ~$33.2B and CR4 45.5%, adhesive ~38% / ~$20.05B and CR4 30.7%; ASC North American adhesive value growth 4.4%/yr to 2030, the same mid-single-digit band as paint's structural forecast; H.B. Fuller ~$3.47B FY2025 revenue, Engineering Adhesives $1.062B at 22.2% adjusted EBITDA vs. Building Adhesive Solutions $860.0M at 15.6%, raw materials ~75% of cost of sales; Henkel Adhesive Technologies ~€11.0B 2024; Sherwin-Williams 4,853 stores; AkzoNobel–Axalta ~$25B enterprise value with an August 5, 2026 shareholder vote; Sherwin-Williams–Valspar ~$9.5B; The Pittsburgh Paints Company; EPA formaldehyde unreasonable-risk finding; SCAQMD Rule 1168; PFAS reporting due October 13, 2026).
8. Histometrics child primer 3256 — Soap, Cleaning Compound, and Toilet Preparation Manufacturing (receipts $82.38B; 2,370 firms; 2,502 establishments; 106,155 employees; $7.51B payroll; CR4 38.0%, HHI suppressed; cleaning $45.83B with CR4 42.8% and HHI 651.8 vs. beauty $36.5B with CR4 46.5%; Stepan ~$1.1B market value on ~$2.2B of sales and a $220M Pasadena plant started April 2025; e.l.f. 70.7% gross margin and SG&A 63% of sales in fiscal 2026, with a tariff-driven margin decline and Target/Walmart/Amazon/Sephora at 18%/13%/11%/10% of sales; Coty ~81% internal manufacturing and $853M of future royalty payments; Circana 2025 — prestige beauty $36.0B +4%, mass $72.7B +5%, prestige fragrance +5%, mass fragrance +15% — against a ~$130B market-research retail estimate the child declines to reconcile; BLS beauty employment 58,200 vs. CBP 48,968; P&G Fabric & Home Care ~$29.6B worldwide; S.C. Johnson ~$11.8B; Henkel–First Quality ~€500M and Reckitt–Advent Essential Home; EPA 1,4-dioxane unreasonable-risk finding, November 2024; MoCRA enforced July 2024).
9. Histometrics child primer 3259 — Other Chemical Product and Preparation Manufacturing (receipts $54.75B; 1,760 firms; 2,251 establishments; 86,668 employees; $6.79B payroll; CR4 8.6%, HHI 59.4 — below all three children's; 32599 ~86% of receipts and ~81% of jobs, explosives $3.16B with CR4 53.3% and HHI ~1,014, ink $4.56B with CR4 47.2%; no U.S.-listed pure-play in any child, but ADRs for Orica (OCLDY) and Dyno Nobel (DNLZY), plus LSB Industries upstream; Dyno Nobel Americas FY2025 revenue US$976.5M at 22.1% EBITDA margin, Orica North America EBIT A$212.2M +15%; Powder River Basin thermal coal 14% of Dyno Nobel Americas revenue in FY2024; Ingevity Performance Materials $606.9M at 53.8% segment EBITDA margin; Eastman Kodak 2025 revenue $1.07B; DIC Packaging & Graphic ~5.7% operating margin and group net sales ~$6.9B; 2025 Section 301 duties incl. 25% on Chinese TiO₂ lifting pigment prices ~10–15%; ATF 9,185 licenses and permits in FY2024 incl. 2,036 manufacturer licenses, just under 3.2 billion pounds of explosives used in 2022, IME value-chain estimate >60,000 jobs and >$19B; EPA 2024 PFAS drinking-water limits naming granular activated carbon, compliance potentially to 2031; EV wiring 2.5–4.0 km vs. 1.0–1.8 km; ~$435M Kentucky TNT plant ~2028 and ~$635M Iowa artillery complex ~2029; Incitec Pivot demerger 2025 and Orica–Nelson Brothers June 2026; BioLab Conyers fire, September 2024).
10. C&EN (American Chemical Society) and Wood Mackenzie — North American petrochemical down-cycle and global ethylene capacity at risk of closure, 2025–2026 (via child primers 3251 and 3252). https://cen.acs.org/business/petrochemicals/party-over-North-American-petrochemical/104/web/2026/02; https://www.woodmac.com/press-releases/global-ethylene-closure/
11. U.S. Environmental Protection Agency and U.S. Food and Drug Administration — Clean Air Act "HON" rule (April 2024; ~220 plants, fenceline monitoring, 6,200+ tons/yr reduction) and MoCRA (enforced July 1, 2024), with TSCA, FIFRA, and Medicare/Section 232 policy context (via child primers 3251, 3252, 3253, 3254, 3255, 3256, 3259). https://www.epa.gov/hazardous-air-pollutants-ethylene-oxide/final-rule-strengthen-standards-synthetic-organic-chemical; https://www.fda.gov/cosmetics/cosmetics-laws-regulations/modernization-cosmetics-regulation-act-2022-mocra
12. Chemical & Engineering News (C&EN). "Berkshire Hathaway to buy OxyChem for $9.7 billion," 2025 (transaction completed January 2026; via child primer 3251). https://cen.acs.org/business/petrochemicals/Berkshire-Hathaway-buy-OxyChem-97/103/web/2025/10