Wholesale Trade (U.S.) — NAICS 42

A Histometrics rollup primer for public-market and private investors. This is a sector — the two-digit top level of the North American Industry Classification System (NAICS, the U.S. government's standard scheme for grouping businesses). Sector 42 sits above the three-digit subsectors and gathers just three children: 423 (durable-goods merchant wholesalers), 424 (nondurable-goods merchant wholesalers), and 425 (wholesale agents and brokers). This page synthesizes the three child primers plus our ground-truth federal statistics for NAICS 42; its distinctive value is the contrast across the three — how the business models, size, growth, ownership, and investability differ. For the deep detail on any one, follow the child link.

Note: the size, count, and concentration figures below are reported federal facts; statements about "direction of travel" and outlook are forward-looking judgments and are flagged as such. Everything that has to add up is on the 2022 Economic Census / 2023 County Business Patterns basis — the only vintage published consistently for all three children and the sector. Two children now also carry newer federal figures; those are reported alongside and never summed.

1. Overview

Wholesale trade is the distribution layer of the physical economy — the tier that stands between the producers who make goods (factories, farms, refineries, mills, mines, importers) and the businesses that use or resell them (stores, restaurants, hospitals, contractors, factories, gas stations, pharmacies). It is the plumbing through which almost every physical product passes on its way from where it is made to where it is used. At roughly $11.9 trillion of receipts across 265,472 firms [2], NAICS 42 is one of the largest slices of the U.S. economy by gross throughput — but "gross throughput" is the operative phrase: most of that $11.9 trillion is product cost flowing through, not value the sector keeps. This is a thin-margin, high-turnover, picks-and-shovels complex that earns a small spread on an enormous flow of goods.

The single most important thing to understand at this level is that "wholesale trade" bundles two fundamentally different business models under one roof, split by a single legal test — does the firm take title to the goods? [1]

• Merchant wholesalers (subsectors 423 and 424) take title. They buy goods on their own account, own the inventory, extend the credit, and carry the price-and-spoilage risk, then resell in smaller lots at a markup. Together they are ~93% of the sector's receipts and ~96% of its employees. This is a spread-and-turns business: buy at a landed cost, mark up, resell, and repeat as inventory cycles. [3][4]
• Agents and brokers (subsector 425) never take title. They only arrange a sale for a commission or fee — a manufacturers' representative, a food broker, an auto-auction marketplace, a hospital group-purchasing organization. This is the capital-light "toll booth" of business-to-business (B2B) commerce: no inventory to finance, high margin on incremental volume, but only a thin slice of the goods that flow through, and permanently exposed to being cut out. [5]

For an investor, that split is the whole story of the sector's shape. The title-takers are working-capital-heavy, cyclical or defensive depending on what they carry, and judged like distributors (inventory turns, cash conversion, return on invested capital). The agents and brokers are asset-light, fee-based, and judged like marketplaces (take rate, transaction volume, disintermediation risk). Sections 2 and 4 draw those distinctions out; the rest covers the sector as a whole.

Three things have changed across the children since the last pass, and each cuts against the headline rather than with it. First, a large and now-measured share of the merchant subsectors' revenue is not contestable independent wholesaling at all — manufacturers' own sales branches are counted inside these codes, which the previous version of this page got wrong (Sections 2 and 3). Second, the listed way into the sector is narrowing in all three children simultaneously: 423 lost more distributors to strategics and sponsors in 2025–26 than it gained, 424's paper, grocery, and miscellaneous lanes each lost a listed handle, and 425's largest publicly traded group-purchasing organization was taken private (Sections 4 and 10). Third, both merchant children independently establish that gross margins across these businesses diverge enormously while operating margins converge into the low single digits, which changes what a margin comparison can tell you (Section 5). None of the three reduces the sector's economic importance; all three narrow what an investor can actually buy or safely compare. [3][4][5]

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

Sector 42 splits into three subsectors. The first cut is title vs. no title (merchant wholesaler vs. agent/broker); the second cut, within the merchant wholesalers, is durable vs. nondurable — long-lived goods kept for years (423) versus short-lived goods consumed and reordered (424). The three are not competitors; they are three different links in the same distribution chain, filed under one federal heading. The table below is the heart of this page. Shares are of the sector's ~$11.9 trillion in 2022 receipts — but read the 425 row's caveat carefully (its dollar figure is not economically comparable to the other two).

(PE = private equity; ESOP = employee-stock-ownership plan; GPO = group purchasing organization; ETF = exchange-traded fund. Tickers and multiples are held for Sections 4 and 10.)

The federal numbers behind the contrast (concentration measured by CR4 — the combined receipts share of the four largest firms — and HHI — the Herfindahl-Hirschman Index, the antitrust 0–10,000 concentration gauge where below 1,500 is "unconcentrated"):

* For agents and brokers the Census collects two numbers: the gross value of the goods whose sale they arrange (~$786B) and the commissions they actually keep (a fraction of it) — the Census's own worked example converts $200,000 of commissions earned at a 5% rate into $4 million of gross selling value. The dollar figure above is transaction value, so both it and the derived revenue-per-employee overstate the group's real economics; the sanity check is that $786B of "receipts" sits against just $17.5B of annual payroll and ~244,000 workers. Our federal file does not carry a separate national commission total, and this page does not state one. The 425 child also now reports a newer, differently-scoped figure — $856.4B of 2023 sales from the Census Annual Integrated Economic Survey — which is a different survey and year, not a contradiction; it is not used in any total on this page. ** 425's HHI is suppressed in our data. Child figures from the three child primers [3][4][5]; sector federal statistics [2].

A material share of these dollars is not contestable independent wholesaling — and in the merchant subsectors it is now measured. This is the biggest correction the revised children force, and it reverses part of what this page previously said. Federal merchant-wholesale statistics exclude commission agents (they sit in 425), but they include manufacturers' own sales branches and offices located away from the plant, and the Census Bureau's Annual Integrated Economic Survey now discloses that split code by code. In 423 the spread is enormous: $464.8 billion of $838.5 billion of 2023 whole-vehicle sales — more than half — came from manufacturers' sales branches; producer-owned branches were about 28% of metal service centers, 16–23% across machinery lines, and under 7% in electronic parts. [3] In 424 only three lines publish the split, and they show the same pattern: 46% of the paper subsector's jan/san child, 56% of "other grocery," and 46% of dairy. [4] These figures come from different survey scopes and vintages and must not be summed — both children say so explicitly, and no sector-wide branch share exists or should be constructed. But the directional point holds at this level and matters: the pool an independent distributor can actually contest is smaller than the headline, by an amount that runs from under a tenth to more than half depending on which lane you are in. Combined with 425's gross-transaction-value problem, the $11.9 trillion overstates the independent intermediary opportunity for two entirely separate reasons. [3][4][5]

Three contrasts worth internalizing:

1. Two title-takers dominate; the broker layer is a rounding fraction of the dollars but the most concentrated of the three. Merchant wholesalers (423 + 424) are ~93% of receipts and ~96% of employment; agents and brokers are the thin remainder. Yet 425 is by far the most concentrated child (top-four firms hold 25.4% of receipts, top eight 31.0%) while the two giant merchant subsectors look almost perfectly competitive (CR4 9.0% and 12.5%). Size and concentration run in opposite directions here. [2][3][4][5]

2. Dollar rank and headcount rank flip between the two merchant subsectors — and that flip is the tell. Nondurable (424) is the bigger by dollars (52%) but the smaller by jobs (38%); durable (423) is the reverse (41% of dollars, 58% of jobs). Nondurable wholesaling moves more dollars per worker because fuel and drugs are expensive per unit ($2.6M of sales per employee), while durable wholesaling is more handling- and headcount-intensive ($1.35M). Revenue measures the value of what flows through; employment measures how much handling it takes. Both child averages hide far wider internal ranges than the sector average suggests: inside 424, sales per employee runs from ~$1.0M in alcohol to ~$19M in petroleum — a 20-fold spread within one subsector — and inside 423 from ~$0.39M in car dismantling to ~$5.6M in whole-vehicle wholesaling. Pay follows a third ordering entirely, tracking how technical or licensed the work is rather than how expensive the box: roughly $68,000 in grocery against ~$160,000 in drug distribution and ~$164,000 in electronic-parts distribution. [3][4]

3. Sector-level concentration is a statistical mirage — and the mirage now demonstrably repeats at every rung below. For the whole of 42 the four largest firms hold just 6.7% of receipts, the top 50 hold 30.5%, and the HHI is a strikingly low 26.8 [2] — lower than both merchant children's HHIs (36.5 and 73.6; 425's is suppressed, though its CR4 of 25.4% is nearly four times the sector's). The top-50 comparison makes the same point: 30.5% at the sector against 30.7% in 423, 46.1% in 424, and 44.8% in 425. That is the aggregation effect at maximum strength — pooling three non-competing worlds, where the leader in fuel is not the leader in machinery is not the leader in auto auctions, mechanically dilutes every firm's share of the combined $11.9 trillion. What the revised children add is that the same arithmetic recurs one and two rungs down: 423's HHI of 36.5 sits below all nine of its children's, and that pattern repeats inside eight of those nine; 424's 73.6 sits below all nine of its children's (which run from ~79 to ~806), and seven of the nine report the same inversion against their own parts. Never read 42's HHI as the level of real competition; competition lives in the niches, where it is fierce — three firms handle 90%+ of U.S. prescription-drug distribution, a two-firm duopoly (roughly 50% and 35%) dominates auto salvage, roofing distribution runs a CR4 of 51.3%, a few merchant houses move ~80% of world cotton, and a beer distributor holds one brand per territory by law. [2][3][4][5]

3. Size (this level's rollup figures)

Our ingested ground-truth federal statistics for NAICS 42 as a whole [2]:

(CBP = County Business Patterns, the Census Bureau's annual employer-establishment count.) The rollup is a genuine aggregation. The three children reconcile into these totals almost exactly: establishments sum to 382,081 and employment to 6,236,357 — to the unit — while receipts (~$11,870B) tie to the sector total and annual and first-quarter payroll tie within rounding [2][3][4][5]. The one metric that sums above the sector is the firm count: the children total 266,754 versus the sector's 265,472, a gap of ~1,282, because a company that operates across more than one subsector (a firm that both wholesales durable goods and brokers deals, say) is counted in each child but only once at the sector level. That ~1,282-firm gap is the fingerprint of cross-subsector operators, not a data error — and it is a fractal fingerprint: 423 reports the same excess of ~1,752 firms against its own nine children, 424 an excess of ~716 against its nine, and each of 424's nine children reports the same small excess against its own sub-industries. Multi-line operators show up at every rung of the tree. [2][3][4]

That reconciliation holds only on one vintage — read the table accordingly. Several of 423's children and all of 425 now lead with, or report alongside, the Census Bureau's Annual Integrated Economic Survey for 2023, and 425's newer sales figure ($856.4B) sits ~9% above its 2022 Economic Census receipts line. The two bases can differ by tens of billions inside a single code and do not even move consistently in one direction. The 2022 Economic Census / 2023 CBP basis is the only vintage published consistently for all three children and the sector, and the only basis on which they reconcile. We keep it for anything that has to add up, and we do not assemble a fresher sector total out of the newer figures: different surveys with different collection frames, and adding them would manufacture a statistic. [3][5]

"~$1.9 million of sales per employee" is the signature of pass-through distribution — a small, skilled, capital-supported workforce moving an enormous dollar volume at razor-thin markups. Payroll is only about 4.8% of receipts (about 3.3% in 424 alone); the cost of the product (and, in fuel and tobacco, embedded excise tax) dwarfs labor. Average pay near $90,500 — ~$95,000 in 423, ~$85,500 in 424 — is high for warehouse-and-logistics work because the payroll includes skilled commissioned sales forces and technical, clinical, and engineering staff, not a low-wage picking operation. Each firm runs about 1.4 establishments and each location about 16 employees — the multi-branch signature of a consolidating industry. [2][3][4]

Undercount and boundary caveats — read the $11.9 trillion as a channel figure, not the total value of goods distributed in America. The distortions run in several directions at once:

• The $11.9 trillion mixes two non-identical things. Subsectors 423 and 424 report title-taking sales of goods they actually bought and resold; subsector 425 reports the gross transaction value of goods it merely brokered (goods that also flow through other channels). The sector figure blends the two as the Census reports them, but they are not economically equivalent — 425's real revenue is commissions, a small fraction of its $786B line. [5]
• Manufacturer-owned distribution straddles the boundary — it is not simply "elsewhere," and the previous version of this page said otherwise. Sales branches and offices located away from the plant are counted inside 423 and 424 and are quantified in Section 2. Plant-attached and captive distribution filed under manufacturing is genuinely outside — integrated refiners' captive fuel terminals, meatpackers, pharma and apparel makers, aircraft-parts arms, the vertically integrated eyewear giants, most office-machine volume. So the sector overstates the independent-middleman share on one side of the line while excluding real distribution economics on the other. [3][4]
• Retailers with large wholesale arms are booked in retail — and now also buy into the channel. Mega-retailers that self-distribute (Walmart, Kroger, Costco) and chains with big professional operations (roughly half of O'Reilly's sales are to professional customers; domestic commercial was 31.7% of AutoZone's domestic revenue in FY2025) book that flow under retail. The home centers cut both ways: they sell pro volume booked in retail and, since 2024–25, own major wholesale distributors outright through Home Depot's SRS/GMS and Lowe's Foundation Building Materials. So the true economic footprint of "distributing goods" is larger than $11.9 trillion. [3][4]
• Small and one-person operators are missed — worst in the broker layer. County Business Patterns counts only employer establishments, so nonemployer sole proprietors fall below the line. This is severe in 425, where independent manufacturers' reps working alone are the norm: a private research house (IBISWorld) counts roughly 103,000 agent/broker businesses versus the 34,962 employer establishments in the federal file, and a federal regulatory analysis finds ~99% of wholesale-agent firms employ fewer than 125 people. The same gap appears in 424's produce, seafood, livestock, and cut-flower lanes — USDA regulates roughly 4,600 registered livestock dealers against the 463 livestock "firms" the Census sees. Treat the counts as an employer core. [4][5]
• Receipts overstate value added, and some dollars double-count. The figure is gross throughput; in commodity lanes (metal, scrap, bullion, fuel, grain) the full value of every ton, ounce, or barrel is booked, and redistributors that sell to other wholesalers can have the same goods counted more than once — a hazard USDA has warned about explicitly. Jewelry makes the point starkly: that channel's $76.2 billion of wholesale receipts is roughly the size of the entire downstream U.S. retail jewelry market, arithmetically impossible until you know bullion houses book full metal value. [3][4]
• Do not mix federal programs, and do not assume every corner can be sized at all. The children document outright suppression (photographic wholesaling's sales and concentration; several 424 sub-line HHIs; 425's own HHI), unreconciled figures inside a single program (computer receipts of $319.4B and $331.3B in two 2022 tables; tire receipts of $76.0B against $56.2B of own-account sales), and cross-program employment gaps of 20% or more. In 425 the frames differ by design: Census counts 243,788 employees while BLS occupational data put employment at 502,330 on a different basis, and a BLS recoding tied to its 2018 benchmark moved about 336,000 jobs out of this subsector — so long-run charts of the broker layer are not a clean read on growth or decline. [3][4][5]

No values are suppressed in the federal file at the 42 level; the sector figures above are complete. Suppression appears below it.

4. Investable universe (where value concentrates across the children)

The sector's $11.9 trillion of throughput does not translate into a broad menu of public stocks, and — the key point for an allocator — where you can invest does not track where the revenue is. There is no exchange-traded fund that targets NAICS 42; listed names sit inside broad consumer-staples, health-care, technology, industrials, and infrastructure funds. Public exposure clusters in three pockets:

• Nondurable merchant wholesalers (424) — the cleanest large-cap access. Public value concentrates in the three biggest groups, which are also ~73% of that subsector's dollars: the drug-distribution oligopoly (McKesson, Cencora, Cardinal Health) that handles 90%+ of U.S. prescription-drug volume and roughly $900 billion of combined fiscal-2025 revenue; the grocery broadliners (Sysco, US Foods, Performance Food Group, United Natural Foods, plus specialty names); and the petroleum terminal/midstream operators (ONEOK, Kinder Morgan, MPLX, Plains, Phillips 66, plus fuel marketers and propane). The other six groups — chemicals, farm raw materials, alcohol, apparel, paper, and a miscellaneous residual — are private or reachable only through indirect proxies, and even the largest physical assets are often private (Buckeye's 130-plus terminals and ~125 million barrels exceed any listed network). [4]
• Durable merchant wholesalers (423) — the deepest "picks-and-shovels" bench, but only in select lanes. Listed compounders cluster in electrical and electronic distribution (WESCO, Rexel, Arrow, Avnet), computers (TD SYNNEX, Ingram Micro) and medical (Medline, Henry Schein), lab (Avantor), the open-market industrial distributors (Grainger, Fastenal, Applied Industrial, MSC, DNOW), plumbing/HVAC (Ferguson, Watsco, Core & Main), metal service centers (Reliance, Ryerson, Worthington Steel), roofing (QXO), and the auto-salvage agents. Wider than this page previously allowed, four single-vertical distributors are now listed inside miscellaneous durable goods (Sims, Pool Corporation, Gold.com, Alliance Entertainment). Furniture, tires, coal, refrigeration, brick/stone, office machines, photographic, and toys have no pure-play. Most of 423's ~144,000 firms and most of its economics sit in private-equity roll-ups, family firms, ESOPs, and cooperatives — including the largest operators of all (ABC Supply, Sonepar, Graybar, Digi-Key, Ace, Orgill, Winsupply). (HVAC = heating, ventilation, and air-conditioning.) [3]
• Agents and brokers (425) — a narrow, idiosyncratic bench, narrower than before. Public exposure concentrates in vehicle remarketing (OPENLANE, Copart, RB Global, ACV Auctions), plus one-off franchises in surplus-asset marketplaces (Liquidity Services), B2B manufacturing (Xometry), and consumer-goods brokerage (Advantage Solutions). Premier, the listed healthcare group-purchasing organization this page previously named, was taken private by Patient Square Capital in a $2.6 billion transaction completed in November 2025 — healthcare purchasing is now private-market exposure alongside member-owned Vizient and HealthTrust. The single largest asset in the subsector — Cox Automotive's Manheim auction network, roughly 8 million vehicles and ~$57 billion of value a year — remains inside privately held Cox, and the vast manufacturers'-rep economy is essentially not publicly investable. [5]

The through-line, and it has sharpened: this is a private-markets universe first, a stock-market one second — and the listed slice is contracting in all three children at once. Across the sector, the great majority of the 265,000-plus firms — and much of the economics — sit in private equity, family ownership, ESOPs, and cooperatives that never trade publicly. What is new is the direction of travel. In 423, GMS was absorbed by Home Depot's SRS and TopBuild by QXO, Olympic Steel disappeared into Ryerson and MRC Global into DNOW, and ODP, Patterson, and Distribution Solutions Group left public markets, against few offsets beyond Medline's December 2025 IPO — the year's largest — and a pending Motion separation. In 424, paper lost its last U.S.-listed participant with ODP's take-private, grocery lost SpartanNash and Calavo to acquisition, a paint proxy divested its owned store network, and the miscellaneous residual's clearest proxy is moving its distribution arm into a joint venture in which it retains 20%. In 425, Premier went private. The sector is growing while the ways into it are narrowing, and a second-order consequence is that private owners face fewer exit routes and investors have fewer clean listed marks to price against. Company-by-company scale, tickers, and revenue splits are in each child primer. [3][4][5]

5. How the money works

Two engines run under one roof, and telling them apart is the core skill for this sector.

The merchant wholesalers (423 + 424) run a spread-and-turns engine. A wholesaler buys at a landed cost (producer price plus freight, duties, and tariffs), marks it up, resells to the trade, and repeats as inventory cycles. The economics of a regulated utility (rate base), a real-estate trust (funds from operations), or a miner (all-in sustaining cost) do not apply. The decisive levers are the same in durable and nondurable alike: inventory turns and cash conversion (the real engine — a well-run distributor reaches mid-teens-to-20%-plus return on invested capital despite thin margins by cycling inventory fast); buying scale and vendor rebates, now quantified rather than asserted — supplier volume rebates equalled 1.4% of WESCO's 2025 sales against a ~5–6% operating margin, and Ferguson carried $471 million of supplier rebates receivable at fiscal year-end, and they are procyclical, amplifying a downturn; the line card and fill rate (the roster of authorized brands and the ability to have the right item in stock today); and value-added services that separate winners. [3][4]

The single most useful correction the revised children supply is that gross margins diverge enormously while operating margins converge — and that two different margin measures are in circulation. On the Census Bureau's survey basis (a construct, not GAAP, and not comparable to a company's reported margins or applicable to agent marketplaces at all), 423's gross margins run from 20.0% in whole vehicles to 48.7% in office equipment. On a GAAP basis the same businesses span an even wider range — from roughly 1.9% in bullion wholesaling and 6–7% in broad-line computer distribution up to 45.0% at Fastenal. In 424 the driver is customer mix and service intensity rather than product category, and the cleanest proof is inside a single company: Sysco's US Foodservice unit earned a 19.1% gross margin on $57.0 billion while its chain-focused SYGMA unit earned 7.9% on $8.4 billion — same firm, same trucks. Yet at the operating line nearly everything collapses into a narrow band: each of the Big Three drug distributors earned 1.09%–1.23% in its core segment, US Foods turned 17.4% gross into 3.0% operating and 1.7% net, convenience and tobacco distribution runs ~1.3–1.7%, petroleum bulk stations 3–4%, and grain merchandising as low as 0.7% pretax. Gross margin measures how much service is attached; the cost to deliver that service eats the difference. The exceptions are earned by technical depth or an entrenched franchise, not scale alone (Bunzl 7.7%, Univar 7.3%). [3][4]

Two corollaries recur in both merchant children and are worth carrying at sector level. First, the profit hides in the tail, not the headline line — Titan Machinery's equipment produced 73.1% of revenue but only 33.8% of gross profit while parts and service produced 24.9% of revenue and 63.2% of gross profit, and in drug distribution generics are ~15% of revenue but ~77% of gross profit. Second, cash flow is counter-cyclical: because inventory and receivables are the biggest assets, a downturn releases cash as stock runs down — Arrow generated over $1.1 billion of operating cash flow in 2024 while revenue fell 16%. The recurring failure mode is the mirror image, leverage: a thin-margin, working-capital-heavy model punishes debt, and the failure roster now spans both merchant subsectors (ATD, TriMark, Incora, True Value, Big Rock, Diamond, and Charles & Colvard in durables; the long-time No. 2 wine-and-spirits distributor's Chapter 11 filing in July 2026 in nondurables). [3][4]

The agents and brokers (425) run a take-rate engine. They earn the spread between fee income and a light, mostly people-and-technology cost base — no inventory to finance or write down. The master gauge is the take rate (fee as a percentage of transaction value): mid-to-high single digits of gross merchandise value for online vehicle marketplaces, 1.5%–3% of member purchasing for group-purchasing organizations, and 5%–15% of net invoiced sales for manufacturers' reps (the underlying rep surveys date to 1999–2005 and should be read as directional, not a current tariff). Revenue = volume × take rate; digital marketplaces enjoy strong operating leverage, while traditional rep agencies behave like a billable-people model. Headline revenue is not comparable across this group, because several of these companies act as agent on some flows and principal on others: Liquidity Services' consignment transactions were 81.3% of gross merchandise value but only 29.0% of revenue, while purchase-model transactions were 18.7% of GMV and 65.1% of revenue. [5]

One diligence discipline therefore applies sector-wide: normalize before comparing anything. Headline revenue across these three children mixes principal and agent accounting, gross merchandise value, commodity pass-through, embedded excise tax, and Census survey constructs that are none of the above. Two further mechanics cut across all three. First, revenue tracks commodity and product prices, so nominal sales swell when fuel, drugs, metal, grain, or resin get expensive even when physical volume is flat — a perpetual source of confusion between "growth" and inflation, and one that cuts both ways, since deflation shrinks gross-profit dollars even when the percentage spread holds. Second, the dominant private-market value mechanic is roll-up arbitrage: buy fragmented regional distributors or rep agencies at modest earnings multiples, fold in purchasing scale and route density, and re-rate the combined platform. [3][4][5]

6. Demand drivers

Demand across all of wholesale trade is derived — it depends on how much the downstream economy is building, buying, and maintaining — and because the three children serve unrelated end-markets, the sector's aggregate demand is diversified and partly self-hedging. The drivers cluster by child:

• Defensive, non-discretionary consumption (mostly 424). People eat, medicate, drive, and clean in every economy. Total U.S. food expenditure reached $2.58 trillion in 2024, a record 58.9% of it eaten away from home — but only 0.4% real growth, so record nominal spending is not case growth. Prescription volume, an aging population, and away-from-home hygiene provide the rest of the floor, and the structural growth pockets sit here too: the GLP-1 (glucagon-like peptide-1) drug boom and the specialty/biosimilar wave. [4]
• The construction, housing, and capital-spending cycle (mostly 423) — currently the clearest drag. Housing starts ran at a 1.177 million annualized rate in May 2026, 8.7% below a year earlier; existing-home sales were 4.1 million in 2025, a thirty-year low; total construction put in place eased ~1.4% to $2.16 trillion in 2025. Industrial production and capex move machinery, industrial supplies, and metal service centers, with reshoring, factory automation, and infrastructure as structural tailwinds (U.S. factory construction near $234 billion in 2024, up 21%). [3]
• The AI / data-center / electrification super-cycle (concentrated in 423). The standout secular pull, lifting both electrical distribution (the power to run the racks) and electronic-parts distribution (the chips inside them) — roughly a third of that subsector — plus grid, EV-charging, and reshoring-driven metal demand. U.S. electricity demand grew about 1.7% a year from 2020 through 2025, against 0.1% a year from 2005 through 2019. But the children force a correction this page previously missed: AI volume is not automatically a margin upgrade. One large computer distributor's AI-enablement server volume added growth while diluting gross margin by 51 basis points on mix, and leading accelerators and memory are often sold direct or allocated through a few chosen channels — so a distributor's AI leverage has to be read off its line card, not inferred from the theme. [3]
• Healthcare utilization (both merchant children). U.S. health spending reached ~$4.9 trillion (17.6% of GDP) in 2023 and is projected to grow 5.4% a year through 2034; hospital med-surg supply costs rose from about $40 billion to $57 billion between 2020 and 2025. The mix is shifting faster than the total — 2025 distributor sales grew 5.1% to hospitals but 9.3% to home care and 12.6% to treatment centers — which changes which distributors win. [3]
• The replace-and-repair base. America's fleet of roughly 289 million light vehicles at a record average age of 12.8 years (297.5 million registered vehicles covering 3.294 trillion miles in 2024), plus the aging installed base of buildings and equipment, makes parts, tires, recycled components, hardware, and refrigeration non-discretionary wear-and-replacement demand — a floor under both merchant subsectors. [3]
• Transaction volume and the outsourcing calculus (425). Brokers' and reps' fees ride on their principals' sales, so industrial output and B2B goods volume set the ceiling; a commission-only, variable-cost sales force is cheaper than a fixed in-house team; used-vehicle supply drives the auto-auction platforms; healthcare spending drives GPO fee income; and buyer consolidation cuts both ways, favoring scaled agencies while concentrating their revenue. [5]
• Commodity prices move the dollar value of sales independent of volume in the pass-through lanes (fuel, metal, scrap, bullion, grain, lumber, resin, cotton), so those codes can post large year-to-year revenue swings driven by price, not tons moved. That decoupling is currently unusually wide: gold ounces sold at the largest bullion wholesaler fell 17.3% while its revenue rose, and HVAC distributor sales grew 2.85% in 2025 on price while adjusted unit demand declined. Dollars pay the bills; units drive the warehouse. [3][4]
• The offsetting laggards. Farm equipment is the deepest trough (2026 net farm income forecast at $153.4 billion, down 2.6% after inflation; tractor sales down 9.9% in 2025); U.S. gasoline consumption fell ~1% in 2025 and ~4% against 2019, with a further ~1% decline projected for 2026; alcohol volumes are falling, with only 54% of U.S. adults reporting they drink, the lowest reading in a series beginning in 1939; the U.S. cattle herd stood at 86.2 million head in January 2026, the smallest since 1951; and office print, cigarette units, and thermal coal are in structural decline. [3][4]

7. Regulation

Wholesale trade is lightly regulated as a business — there is no rate base, no franchise, no tariff schedule, and generally no wholesale-specific operating license. Barriers to entry come from scale, inventory, capital, and relationships, not from a regulatory grant. Regulation instead falls on the wholesaler as the licensed and taxed point in the chain for whatever product it carries, and it varies sharply by child:

• Nondurable products (424) carry the heaviest product-specific regimes: Drug Enforcement Administration registration and the Drug Supply Chain Security Act track-and-trace rules for drug distributors (the wholesaler deadline fell on August 27, 2025); the post-Prohibition three-tier system and state Alcohol Beverage Control licensing for alcohol (where regulation is the moat); food-safety law (FDA Food Safety Modernization Act, USDA inspection, the Perishable Agricultural Commodities Act trust, state grain-dealer bonding) for grocery and farm products; and environmental/hazmat rules (Toxic Substances Control Act, OSHA hazard communication, spill-prevention standards) for chemicals and petroleum. One correction worth carrying up: the Packers and Stockyards "packer" definition reaches wholesale brokers, dealers, and distributors, so some meat wholesalers are regulated parties, not bystanders. [4]
• Durable products (423) run through building and electrical codes, FDA medical-device rules and the FTC's Contact Lens and Eyeglass Rules, dealer-franchise and aviation-traceability law, firearms and anti-money-laundering rules, and Department of Energy efficiency and EPA refrigerant phase-downs that actually lift selling prices in plumbing, HVAC, and refrigeration. Two exposures now land squarely on the distributor rather than the manufacturer: the CPSC's 24-hour reporting duty for qualifying defective products, and EPA enforcement against aftermarket defeat devices (172 civil cases and $55.5 million in penalties FY2020–23). [3]
• Agents and brokers (425) face light, pocketed oversight: the healthcare group-purchasing safe harbor under the Anti-Kickback Statute (fees generally 3% or less, disclosed, with a written customer agreement), state sales-representative prompt-payment statutes, auto-auction licensing and odometer rules, USDA PACA and Packers and Stockyards licensing for produce and livestock intermediaries, and IRS worker-classification tests — which matter unusually much here, since the independent-contractor model dominates the rep population. [5]

Four themes now touch nearly every child at once. First, trade policy — and its current level is genuinely unsettled. Section 232 duties on steel and aluminum rose to 50% in June 2025, with 2026 proclamations extending derivative and copper treatment; 25% Section 232 duties apply to autos and parts; wood products and furniture carry Section 232 measures (25% on imported upholstered wooden furniture from October 14, 2025); combined softwood-lumber duties run about 35%; Section 301 duties include a 100% rate on syringes and needles; and semiconductor export controls keep tightening. The company-level bill is large and visible — roughly $290 million of adverse 2025 pretax impact at one medical distributor, ~$600 million flagged at one equipment maker in 2025, and consumer-technology importers paying $23.5 billion of tariffs in 2025 against $4.0 billion the prior year. But the children describe different corners of the landscape rather than one settled state: 423's children treat Section 232 and 301 duties as live and quantified, while one of 424's children reports that the U.S. Supreme Court struck down the IEEPA tariffs 6–3 in February 2026, with the refund mechanism unresolved. These are different legal authorities and different moments — a ruling on IEEPA does not disturb Section 232, Section 301, or antidumping orders — so they are not in conflict; the practical upshot is that the level of border cost across the sector is unsettled and every tariff figure here should be read as a snapshot. The end of the sub-$800 de-minimis exemption on August 29, 2025 is common ground. [3][4]

Second, antitrust: because consolidation is the sector's defining trend, the Federal Trade Commission and Department of Justice are effectively the live regulators of the investment thesis itself, reviewing the mergers that public and private buyers underwrite and imposing conditions. Newly prominent is Robinson-Patman price-discrimination enforcement — the FTC sued the leading wine-and-spirits distributor in December 2024, alleging it charged independent retailers 12%–67% more than favored chains for identical products (allegations, not findings) — a regime that, if tightened, would cut pricing flexibility across the whole tier. [3][4][5]

Third, product-stewardship rules are the fastest-growing shared burden: PFAS restrictions and packaging Extended Producer Responsibility laws now reach paper, apparel, and chemicals from three different directions, and in chemicals "manufacture" includes import, so a distributor can acquire reporting obligations without formulating anything. The common effect is assortment churn and stranded-inventory risk. [4]

Fourth, hard-dated efficiency and refrigerant deadlines now sit on the calendar (general-service lamps in July 2028, washers and dryers from March 2028, distribution transformers from April 2029, refrigeration in 2029–30), which turns inventory obsolescence from a speculative risk into a scheduled one — but the enforcement calendar is a moving target even where direction is settled, with DOE proposing rollbacks and EPA relaxing several refrigerant deadlines in 2025–26, and 423's own children disagreeing on whether one installation bar is binding. Running the other way, right-to-repair has landed: the FTC and five states settled with Deere in July 2026, requiring ten years of dealer-equivalent repair access — a direct assault on the aftermarket margin that anchors the dealer model. For a sector-level investor, regulation here is a cost-and-inventory risk plus an antitrust overhang, with a revenue tailwind in a few lanes — not the license-to-operate regulation of a utility. [3]

8. Consolidation

The defining structural fact for the whole sector is a very long tail of small, regional, family firms being slowly rolled up under a few national or global giants — a barbell shape present in every child. With ~265,000 firms and ~382,000 establishments, there are far more locations than firms — the signature of multi-branch consolidators — and thin margins plus better vendor terms at scale make roll-ups the dominant corporate strategy everywhere. The clearest evidence sits inside the federal data itself: in industrial machinery alone, firm counts fell from 22,773 in 2017 to 18,795 in 2022 while sales grew. The low sector-level concentration (CR4 6.7%, HHI 26.8) says nothing about this dynamic; it is an aggregation artifact, and the real story is dozens of firms per lane being folded into a handful of scaled platforms. The consolidation engine differs by child:

• Private-equity buy-and-build of aging-owner independents — the dominant story across durable-goods distribution (professional equipment, machinery, lumber, flooring) and much of nondurables (chemicals bolt-ons, paper). [3][4]
• Public serial acquirers and megadeals — Home Depot's SRS purchase (the children cite both $18.0 billion and $18.25 billion; we carry both) followed by SRS's ~$5.5 billion acquisition of GMS; QXO's ~$10.6 billion Beacon takeover in April 2025, then Kodiak (~$2.25 billion) and TopBuild (~$17 billion, July 2026); Lowe's $8.8 billion Foundation Building Materials deal; Sysco's ~$29.1 billion agreement for Jetro Restaurant Depot (announced March 2026) and C&S's completed ~$1.77 billion purchase of SpartanNash; ONEOK's ~$18.8 billion acquisition of Magellan and Sunoco's ~$7.3 billion purchase of NuStar, then Parkland; Bunge's ~$8.2 billion merger with Viterra, closed July 2025; and Apollo's ~$8.1 billion take-private of Univar. [3][4]
• Cooperatives and buying groups in hardware and appliances (Ace, Do it Best — which absorbed True Value's wholesale platform for $153 million after its Chapter 11) and in grocery and farm supply. [3][4]
• Backward integration by end-users — steelmakers buying scrap yards to lock in feedstock, at an accelerating pace: roughly 18 scrap deals in 2021–2025 against 11 across the whole 2005–2020 span. [3]
• Strategic and vertically integrated acquirers buying the channel outright — a force this page previously understated: Sysco's $969 million purchase of a foodservice dealer, Cencora's agreement to combine MWI Animal Health with Covetrus, a vision insurer buying a frame maker, Worthington Steel taking control of Klöckner, and Ryerson's merger with Olympic Steel. In drugs the movement is vertical rather than direct-to-customer — the Big Three have spent more than $16 billion buying physician-practice management-services organizations to steer purchasing to themselves. [3][4]
• Platform consolidation and digitization in the broker layer — Ritchie Bros. + IAA forming RB Global in 2023, Acosta absorbing CROSSMARK and Product Connections in 2024, OPENLANE selling its U.S. physical auctions to go digital-first, and Patient Square Capital taking Premier private in 2025. [5]

Two refinements matter at this level. First, consolidation increasingly runs through the public market rather than into it — most of the deals above removed a listed way in (Section 4). Second, the pace has turned, and the children read it slightly differently: the lumber child reports building-products deal volume down about 21% in 2025 on tariff uncertainty, the roofing child describes M&A as having "cooled somewhat but stayed near its long-run average," and the machinery child reports industrial-distribution activity down considerably. All point the same way — fewer deals, undiminished strategic intent — but a roll-up thesis has to underwrite lumpy, cycle-and-financing-dependent deal flow rather than a metronome. [3]

The countervailing force — also universal — is disintermediation: manufacturers selling direct to large accounts, mega-retailers self-distributing, captive branches, and B2B e-commerce marketplaces (Amazon Business) compressing the middleman's spread. That threat is no longer merely feared: the branch-share measurements in Section 2 quantify how much distribution has already bypassed the independent channel. But the children also make clear it is two-directional, not a one-way exit — a major athletic brand has decided to reinvest in wholesale after overemphasizing direct channels, a coatings maker sold its owned U.S./Canada store network into private hands, and one of the largest electrical distributors booked €12.3 billion of online sales alongside, not instead of, its branch network. The strategic response across every child is to get stickier — same-day delivery, technical service, private label, vendor-managed inventory, and digital ordering — and, in the broker layer, to own data, liquidity, or financing rather than just pass orders. [3][4][5]

9. Risks

The three children share a common risk stack, weighted differently:

• Thin margins over heavy working capital and leverage (merchant wholesalers). Low-single-digit net margins and inventory-and-receivables-heavy balance sheets are rate-sensitive and punish debt in a downturn; the failure roster now spans both merchant subsectors and, in nondurables, reached the No. 2 wine-and-spirits distributor's Chapter 11 filing in July 2026. [3][4]
• Disintermediation (all three, sharpest for brokers). Manufacturer-direct selling, retailer self-distribution, captive branches, and online marketplaces erode the intermediary's reason to exist — the single biggest long-run structural threat, existential for pure order-passing agents, and now quantified by the branch-share data in Section 2. [3][4][5]
• Cyclicality (sharpest in 423). Big-ticket, deferrable durable goods swing hard with their underlying cycle — real output in furniture and home-furnishing wholesaling fell about 27% peak-to-trough in the last recession — while nondurable staples and non-discretionary aftermarket demand provide ballast. [3]
• Commodity-price and inventory swings. In the pass-through lanes (fuel, metal, scrap, bullion, grain, lumber, resin), unhedged or mistimed inventory turns thin spreads negative and creates markdown risk; the framing-lumber composite's 73% three-month crash in 2021 is the reference case. Deflation is the underrated half — it shrinks gross-profit dollars even when the percentage spread holds. [3][4]
• Tariffs and trade shocks. Near-total import dependence in several lanes makes the sector exposed, pass-through is usually possible but incomplete (one equipment dealer explicitly reported 2025 tariff costs were not fully recovered, costing 100 basis points of equipment margin), and the legal level of border cost is currently unsettled (Section 7). [3][4]
• Customer, supplier, and contract concentration — pointing in opposite directions by child. Upstream: one HVAC distributor's ten largest suppliers were 85% of 2025 purchases; one auto-parts distributor booked a $151 million credit-loss reserve when a supplier failed. Downstream: CVS Health is ~24% of one drug distributor's and 30% of another's revenue, and two home centers together were 43.4% of one hardware supplier's 2025 revenue. Loss of authorized-dealer status can gut a distributor; loss of a line review can gut a supplier. In 425 the same risk lands on a small agency losing a single principal's line. [3][4][5]
• Secular decline in mature niches. The EV transition (auto parts), tire disintermediation, the gasoline-volume plateau, alcohol and cigarette-unit decline, falling office print volume, and thermal-coal retirement are structural, not cyclical, headwinds. [3][4]
• Automation, environmental, and cyber tail risks. BLS projects wholesale and manufacturing sales-rep employment growing only ~1% over 2024–2034, below average, citing e-commerce and AI. Spills and fires haunt the fuel and chemical lanes. And cyber is now a fulfillment risk, not just a data one, raised independently by both merchant children: a July 2025 ransomware incident took a major technology distributor's systems offline and cost $6.2 million, and a 2025 breach cost a grocery distributor approximately $400 million in sales and forced manual order processing. [3][4][5]
• Data opacity as an investor risk — new at this pass. Whole corners of the sector cannot be sized or benchmarked from federal statistics alone: photographic wholesaling's sales and concentration are suppressed outright, computer receipts differ by $12 billion across two Census tables, tire receipts carry two unreconciled figures, 425's HHI is suppressed, and employment differs by 20% or more between programs. [3][5]
• Fewer public exits and fewer listed comparables. As listed participants leave all three children (Section 4), private owners face narrower exit routes and investors have fewer clean marks to price against — stated outright in 424's paper child and moving the same way in grocery, miscellaneous, alcohol, durable goods, and the broker layer. [3][4][5]

10. How to invest & outlook

Public routes are narrower than the sector's $11.9 trillion implies — narrower than a year ago — and cluster in defensive, scalable distribution. The cleanest large-cap access is in nondurable staples distribution: the drug-distribution oligopoly (defensive, volume-driven compounders returning capital through buybacks), the grocery broadliners, and the petroleum terminal/midstream names (toll-like cash flow, though the master limited partnerships among them issue a Schedule K-1 with tax wrinkles inside retirement accounts). In durable goods, the bench is the "picks-and-shovels" distributors levered to technology, healthcare, industry, and infrastructure (electrical/electronic, computers, medical, lab, industrial supply, plumbing/HVAC, metal, roofing, auto salvage), plus — wider than this page previously allowed — four single-vertical distributors inside miscellaneous durable goods. In the broker layer, exposure is idiosyncratic and concentrates in vehicle remarketing, and several of those names trade more like marketplace or software growth stocks than "wholesale" stocks. These are cyclical-value and quality-compounder names, judged on organic growth, return on invested capital, cash conversion, inventory turns (or take rate and gross merchandise value), and disciplined M&A — not on growth multiples. There is no dedicated NAICS-42 ETF; large parts of the sector (furniture, tires, coal, refrigeration, brick/stone, office machines, photographic, toys, alcohol distribution, paper, and most manufacturers' reps) have no pure-play at all; and the listed set keeps thinning as strategics and sponsors take distributors private. One discipline applies sector-wide before any comparison: normalize gross-versus-net accounting and survey-versus-GAAP margins (Section 5). (Reserve valuation multiples and dividend judgments for company-level diligence.) [3][4][5]

Private routes are where most of the sector's capital actually participates — the frame that matters most for a private-credit or BDC/CEF audience. With 265,000-plus mostly small firms, fragmented lanes, thin-but-durable margins, aging owners, and a small-business bar most firms clear, wholesale trade is structurally a private-capital asset class: private-equity and search-fund roll-ups of regional distributors and rep agencies; private-credit and business-development-company (BDC) lending against cash-generative, working-capital-heavy distributors; asset-based and factoring finance against inventory and receivables; adjacent real assets (cold storage, fuel terminals, distribution-yard real estate); and cooperative or ESOP membership. The largest single operators in several lanes are not for sale on any exchange. Notably, all three children converge on the same underwriting instruction: key off the unit of value the intermediary actually keeps — gross profit per case, per stop, or per gallon in nondurables (stripping out excise tax in fuel and tobacco), rebate quality and inventory aging in durables, and actual commissions and fees, never facilitated merchandise value, in the broker layer — then test the contract that creates it for termination, assignability, territory exclusivity, change-of-control, and how much of the franchise walks out with the selling owner. The fragmentation that makes public exposure hard is precisely what makes private buy-and-build attractive, and the wine-and-spirits bankruptcy is actively creating distressed sellers under court supervision. [3][4][5]

Outlook (judgment, not fact). Treat sector 42 as three overlapping stories, not one. The nondurable half (424) is a mature, defensive base whose steadiness comes from diversification across nine partly independent demand curves — one structurally growing lane (drugs) offsetting managed decline elsewhere. The durable half (423) is a self-hedging barbell whose weighted trajectory is modestly constructive, with the clearest tailwind being the AI / data-center / electrification super-cycle lifting technology- and power-facing distribution, and defensive medical/roofing/refrigeration demand as ballast against cyclical laggards. The broker layer (425) is a share-shift, not a rising-tide story — aggregate goods value roughly flat, with digital marketplaces taking share from human reps. Across all three, returns come from scale, working-capital discipline, cost-out, favorable mix, and disciplined acquisition — not from end-market growth; where mix stays commodity, operating margins converge on low single digits regardless of size. Three cautions this pass adds. First, volume growth is not margin growth — AI hardware grew computer distribution while diluting its gross margin, GLP-1 volume lifted one distributor's revenue without meaningfully contributing to segment profit, and the richest returns sit closer to manufacturing than to resale. Second, the contestable pool is smaller than the headline, because captive manufacturer branches sit inside the merchant codes and 425's dollars are transaction value rather than fees. Third, the listed opportunity set is contracting in all three children even as the sector grows. The most durable structural theme, common to every child, is consolidation: expect the fragmented middle to keep shrinking and the scaled, service-embedded operators to keep taking share, though at a lumpy rather than steady pace, while tariff-driven cost volatility, the construction cycle, and disintermediation remain the shared wildcards. For the full company detail and section-by-section depth, follow the three child links: 423 · 424 · 425. These are projections, not certainties.

---

Sources

This is a three-child rollup. The size, count, and concentration figures in Sections 2–3 are Histometrics' ingested ground-truth federal statistics for NAICS 42, on the 2022 Economic Census / 2023 County Business Patterns basis; the remaining substance is synthesized from the three child primers, which carry the fuller source lists and the newer federal vintages cited above.

1. U.S. Census Bureau — 2022 NAICS Definitions, sector 42 Wholesale Trade and its three subsectors (the merchant-wholesaler title-taking definition and the boundary between merchant wholesalers, 423/424, and agents and brokers, 425). https://www.census.gov/naics/
2. U.S. Census Bureau — 2022 Economic Census — Concentration by Largest Firms (receipts, firms, CR4/CR8/CR20/CR50, HHI) and County Business Patterns 2023 (establishments, employment, annual and Q1 payroll), NAICS 42. Histometrics ingested ground-truth federal statistics. https://www.census.gov/programs-surveys/economic-census.html; https://www.census.gov/programs-surveys/cbp.html
3. Histometrics rollup primer — NAICS 423, Merchant Wholesalers, Durable Goods (receipts ~$4,917.5B; 144,016 firms; CR4 9.0% / HHI 36.5 — below all nine children's; nine industry groups; manufacturers' sales branches measured inside the codes via 2023 AIES, from under 7% to more than half by child; Census gross margin a survey construct, 20.0%–48.7%, against GAAP 1.9%–45.0%; listed set contracting in 2025–26; four listed distributors now inside miscellaneous durable goods; private-markets-first ownership; AI/electrification super-cycle barbell with the volume-is-not-margin caveat). primer-423-DRAFT.md
4. Histometrics rollup primer — NAICS 424, Merchant Wholesalers, Nondurable Goods (receipts ~$6,166.6B; 90,578 firms; CR4 12.5% / HHI 73.6 — below all nine children's; nine industry groups; petroleum/grocery/drugs ~73%; manufacturers' branch shares of 46–56% in the three lines that publish the split; gross margins diverging while operating margins converge on 1–3%; sales per employee spanning ~$1.0M to ~$19M; public surface narrowing in paper, grocery, and the miscellaneous residual; drug oligopoly, grocery broadliners, petroleum midstream; IEEPA tariff ruling and PFAS/EPR burden). primer-424-DRAFT.md
5. Histometrics rollup primer — NAICS 425, Wholesale Trade Agents and Brokers (~$786B transaction value on the 2022 Economic Census basis and $856.4B of 2023 AIES sales; 32,160 firms; 34,962 establishments; 243,788 employees; $17.5B annual payroll; CR4 25.4% / CR50 44.8%, HHI suppressed; single-child branch, coextensive with 4251/42512/425120; commission/take-rate model at 1.5%–15% depending on channel; agent-versus-principal revenue recognition; vehicle-remarketing public bench; Manheim private and Premier taken private in November 2025; ~103,000-business nonemployer undercount and the BLS reclassification of ~336,000 jobs). primer-425-DRAFT.md