Merchant Wholesalers, Durable Goods (U.S.) — NAICS 423

A Histometrics rollup primer for public-market and private investors. This is a subsector (3-digit) in the North American Industry Classification System (NAICS — the U.S. government's standard scheme for grouping businesses). It sits one level below the two-digit sector (42, Wholesale Trade) and one level above the four-digit industry groups. Code 423 gathers nine industry groups — 4231 through 4239 — that all do the same basic job (buy long-lived goods in bulk from producers, take title, and resell to businesses) across nine unrelated product worlds. This page synthesizes the nine child primers plus our ground-truth federal statistics for NAICS 423. Its distinctive value is the contrast across the nine — who is big, who is growing, who owns them, and where an investor can actually buy in. 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. All figures that have to add up are on the 2022 Economic Census / 2023 County Business Patterns basis — the only vintage published consistently for all nine children and the group. Several children now lead with newer or narrower federal figures; those are reported alongside and are never summed.

1. Overview

NAICS 423 is the wholesale-distribution layer for durable goods — the middleman tier that stands between the factories, mills, mines, and importers that make long-lived products and the dealers, contractors, hospitals, factories, utilities, and retailers that use them. "Durable" means goods built to last years, not consumed on use: cars, machinery, steel, lumber, medical equipment, computers, appliances, furniture, jewelry, and scrap. Every firm here is a merchant wholesaler — it buys in bulk, takes title (actually owns the inventory, carrying the financing and price risk), warehouses it, breaks it into the quantities customers need, extends credit, and delivers — as opposed to an agent or broker who arranges a sale for a commission but never owns the goods. [1]

For an investor, 423 is the closest thing to a single index of the physical, professional, and industrial economy's plumbing: a ~$4.9 trillion, thin-margin, high-throughput distribution complex whose demand is derived from how much America is building, driving, manufacturing, treating, computing, furnishing, and recycling. It is a picks-and-shovels family of businesses — it earns a small spread on an enormous flow of goods and lives or dies on inventory turnover, vendor rebates, working-capital discipline, and logistics, not on owning a brand.

The reason to read this subsector level rather than jump straight to a leaf primer is contrast. The nine children share a business model and a statistical roof but almost nothing else — different end-markets, different growth clocks, different ownership structures, and wildly different investability. Some are riding historic secular booms; some are in structural decline; some are in cyclical troughs. Sections 2 and 4 draw those distinctions out; the rest covers the subsector as a whole.

Two things have changed materially since the last pass, and both cut against the headline. First, a large and now-measured share of this $4.9 trillion is not contestable independent wholesaling at all — it is manufacturers' own sales branches, counted inside these codes (Section 2). Second, the listed opportunity set is shrinking, not growing: across the nine children, more distributors left public markets in 2025–26 than joined them (Section 4). Neither fact reduces the subsector's economic importance; both narrow what an investor can actually buy.

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

NAICS narrows step by step: the subsector 423 splits into nine four-digit industry groups, each of which splits again into five-digit industries (covered in the child primers). The nine are not competitors — a lumber distributor and a jewelry wholesaler never bid against each other. They are nine separate product markets filed under one federal heading because they share a distribution engine. The table below is the heart of this page; shares are each child's receipts (sales) as a percentage of the subsector's ~$4.9 trillion.

(OEM = original equipment manufacturer, the company that builds the goods; PE = private equity; HVAC = heating, ventilation, and air-conditioning; ETF = exchange-traded fund. Tickers and multiples are held for Sections 4 and 10.)

The federal numbers behind the contrast (each child's own ground-truth figures, from the child primers; 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"):

Child figures from the nine child primers [3][4][5][6][7][8][9][10][11]; group federal statistics [2].

A large share of this revenue is not independent wholesaling — and it is now measured. This is the single biggest correction the revised children force, and it changes how the $4.9 trillion should be read. Federal merchant-wholesale statistics exclude agents and brokers who never take title (NAICS 425), but they include manufacturers' own sales branches and offices, and the Census Bureau's newer Annual Integrated Economic Survey now discloses that split code by code. The spread across the subsector is enormous. In whole-vehicle wholesaling, $464.8 billion of $838.5 billion of 2023 sales — more than half — came from manufacturers' sales branches rather than merchant wholesalers proper; in new auto parts the split was $69.3 billion of $223.6 billion. [3] In metal service centers, producer-owned branches were $82.5 billion of ~$296.4 billion, about 28%. [7] In machinery, branches ran 22.6% of construction and mining, roughly 22% of industrial machinery, and about 16% of industrial supplies. [10] In electrical goods they were $48.8 billion of $226.8 billion, against about 16% in appliances and under 7% in electronic parts. [8] In plumbing they were $14.5 billion of $100.3 billion; in brick and stone, independent merchants account for only about $22.7 billion of a $37.5 billion code. [9][5] These figures are not additive across children and must not be summed — they come from different survey scopes and vintages, and the children say so explicitly. [3][8][10] But the directional point holds at this level and is important: the pool any independent distributor can actually contest is materially smaller than the headline, and how much smaller varies from under a tenth to more than half depending on which child you are in. Market-sizing done off the $4.9 trillion overstates the opportunity everywhere, and overstates it most in the largest child.

Four contrasts worth internalizing:

1. The subsector is top-heavy but not dominated by one child. The four biggest groups — motor vehicles (21.8%), electrical/electronic goods (18.2%), professional equipment (15.6%), and machinery (14.8%) — together hold ~70% of the $4.9 trillion; the smallest five split the rest, and furniture (3.0%) is a rounding fraction of autos. So "durable-goods wholesaling" is mostly a story about vehicles, electrical/electronic goods, professional equipment, and machinery — the four heavyweights.

2. Concentration is a mirage at every level, not just this one. At the group level 423 looks like one of the most competitive industries in America — the top four firms hold just 9% of receipts and the HHI is 36.5. That is an artifact of pooling nine non-competing markets: a company that dominates auto salvage counts for nothing in medical supply or scrap, so mixing them mechanically dilutes every firm's share. The subsector HHI (36.5) is lower than all nine children's own HHIs (which run from 38.6 up to 454.1). What the revised children add is that the same arithmetic repeats one rung down in eight of the nine: professional equipment's group HHI of 133.7 sits below every one of its six publishing children (151.3 to 581.9); metal's 108.7 sits below both its children; miscellaneous durable goods' 38.6 sits below all four of its publishing children. The lone partial exception is 4236, whose two most concentrated children simply have different leaders. And the real arenas inside the children are tighter still — computer distribution is a national duopoly (its two largest firms supplied roughly 46% of one reseller's 2025 product purchases), hospital med-surg an oligopoly, dental and broad-line lab supply duopolies, auto salvage a duopoly (~50% / ~35%), and roofing distribution a CR4 of 51.3%. Never read 423's HHI as the level of real competition; read the children, then the niches inside them. [2][3][5][6][7][11]

3. Revenue per worker splits the pass-throughs from the handlers. Metal service centers ($2.23M/employee) and whole-vehicle wholesalers (inside the $2.51M motor-vehicle average, itself ~$5.6M in the whole-vehicle child) run enormous dollar volumes of costly, pass-through product across lean headcounts; machinery ($0.91M) and furniture ($0.94M) are more labor- and handling-intensive, and dismantling cars for used parts runs at ~$0.39M. The subsector average is $1.35M of sales per employee — the hallmark of distribution, where a small headcount moves a large dollar value of goods on a thin margin. The same split shows up in pay, which tracks how technical the sale is rather than how expensive the box: roughly $164,000 per employee in electronic-parts distribution and ~$125,000 in medical, against ~$63,000 in scrap and ~$77–79,000 in ophthalmic, office machines, and hardware. [3][8][6][11]

4. Growth direction diverges enormously — nine different clocks. Electrical goods and metal are levered to the AI / data-center / electrification super-cycle; computers, recyclables, and used auto parts have structural tailwinds; medical, roofing, refrigeration, and aerospace parts are defensive growers; furniture, farm equipment, appliances, tires, office machines, and thermal coal are cyclical laggards or in secular decline; photographic has already collapsed and is now stabilizing. A buyer of "durable-goods wholesaling" as a theme is really buying a barbell of secular winners and secular losers — which is exactly why the contrast, not the aggregate, is the point.

3. Size (this level's rollup figures)

These are our ingested ground-truth federal statistics for NAICS 423 specifically [2]:

(CBP = County Business Patterns, the Census Bureau's annual employer-establishment count.) The rollup is not an estimate: the nine children's receipts sum to exactly $4,917.5B, and their establishments, employment, and annual payroll each sum to the group total to the unit — a clean nine-way partition and a strong data-integrity check. The one figure that does not add up is the firm count: the children sum to ~145,768 versus the group's 144,016, a gap of ~1,752, because a company that wholesales in more than one of the nine groups is counted once at the subsector level but appears in each child it operates in. [2]

That reconciliation now holds only on one vintage — read the table accordingly. Four children (4231, 4235, 4236, 4238) and parts of two more (4233, 4237) now lead with the Census Bureau's Annual Integrated Economic Survey for 2023, and several report a second, narrower merchant-wholesaler-only sales basis alongside the Economic Census receipts line. The two bases can differ by tens of billions inside a single code, and the newer vintage does not even move consistently in one direction — industrial machinery, construction equipment, and service-establishment equipment read higher on the 2023 basis while industrial supplies reads lower. The 2022 Economic Census remains the only vintage published consistently for all nine children and the group, and the only basis on which they reconcile. We keep it for anything that has to add up, and we do not sum the newer figures into a fresher group total: different surveys with different collection frames, and adding them would manufacture a statistic. [3][5][7][8][9][10]

A few group-level signatures: roughly $34 million of sales per firm, $1.35 million of sales per employee, about 16 employees per location, and average pay near $95,000 per worker — high for warehouse-and-logistics work because the payroll includes skilled commissioned sales forces and technical/clinical/engineering staff (IT, medical, lab, aerospace) alongside warehouse labor, not a low-wage picking operation. [2]

Undercount caveat — read the $4.9 trillion as a well-measured but mixed floor. Federal statistics classify each establishment by its primary activity and count only firms that take title, so 423 misstates real durable-goods distribution economics in several directions the children detail:

• Agents and brokers are excluded (NAICS 425). The biggest whole-vehicle names — Manheim, Copart, IAA — mostly sell on consignment for a fee and are classified as agents (425120), not here. Broker activity is also material in used aircraft material and in electronic parts. [3][10][8]
• Retailers with large wholesale arms are booked in retail — and now also buy into the channel. The mega auto-parts chains run enormous professional operations counted under retail (roughly half of O'Reilly's sales are to professional customers; domestic commercial was 31.7% of AutoZone's domestic revenue in FY2025). 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. [3][5]
• Manufacturer-owned distribution straddles the boundary — it is not simply "elsewhere." The prior read of this page was too simple. Sales branches located away from the plant are counted inside these wholesale codes and are quantified in Section 2; plant-attached and captive distribution filed under manufacturing is genuinely outside, and that is where Boeing Distribution and Airbus's Satair (aircraft parts), the vertically integrated eyewear giants, camera makers' U.S. arms, most office-machine volume, and some OEM HVAC branch networks sit. [3][6][7][8][9][10]
• Commodity pass-through inflates the dollar figure while the private/nonemployer tail deflates the counts. In metal, scrap, and bullion, receipts book the full value of every ton or ounce, so the $4.9 trillion overstates economic value added — jewelry makes the point starkly, where this 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. Meanwhile thousands of small, family-owned, and nonemployer distributors, recyclers, coin dealers, and independents never fully surface in employer statistics. [7][11]
• Some of it is simply not published, and some of it disagrees with itself. Census suppresses photographic wholesaling's sales, cost of goods sold, gross margin, and concentration ratios entirely; reports computer-distribution receipts as $319.4 billion in one 2022 table and $331.3 billion in another; and suppressed the 2023 all-operating-types tire total. The tire code carries $76.0 billion of receipts against $56.2 billion of own-account sales, unreconciled. Employment differs by program too — office equipment at 83,163 (CBP) versus 66,100 (BLS), and brick/stone at 37,399 (Census) versus 64,200 (BLS). Do not mix federal programs, and do not assume every corner of this subsector can be sized at all. [3][5][6]

So $4.9 trillion is the well-measured title-taking channel — trade flow, not value added, including captive manufacturer branches, and a floor on the products' true distribution footprint.

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

The subsector's $4.9 trillion of revenue does not translate into a broad menu of public stocks, and — crucially — where you can invest does not track where the revenue is. The largest child (motor vehicles) skews to agents and marketplaces rather than classic buy/resell wholesalers, and the smallest (furniture) has essentially no pure-play. Listed value clusters instead in the "picks-and-shovels" distributors levered to technology, healthcare, industry, and infrastructure. There is no exchange-traded fund that targets NAICS 423; the listed names sit inside broad industrial, technology, healthcare, and infrastructure funds. Where the public exposure actually is:

• Electrical & electronic goods (4236) — the deepest bench, and effectively four pure-plays. Electrical distributors WESCO (~$23.5B of 2025 sales) and Rexel (~€19.4B, North America now 46%); electronic-component leaders Arrow (~$27.9B) and Avnet (~$22.2B, roughly 82% semiconductors). The largest operators are not listed: Sonepar (~$35B globally, $17.1B in the Americas), ESOP-owned Graybar (~$12.9B), Border States, CED, family-owned Digi-Key, and Berkshire-owned Mouser/TTI. [8]
• Professional & commercial equipment (4234) — clean pure-plays in the two giants. Computers/IT via TD SYNNEX (~$62.5B FY2025 revenue on $89.4B of gross billings) and Ingram Micro (~$52.6B); medical/dental via Medline (~$28.4B of 2025 net sales; December 2025's $6.26 billion IPO, the year's largest), Henry Schein (~$11.1B distribution segment), and the diversified Cardinal Health / McKesson / Cencora; lab via Avantor ($6.552B, explicitly a turnaround). Office, foodservice, ophthalmic, and photographic are indirect only. [6]
• Machinery, equipment & supplies (4238) — the open-market industrial distributors. Applied Industrial Technologies (~$4.6B), W.W. Grainger (~$17.9B), Fastenal (~$8.2B), MSC Industrial (~$3.77B), DXP, DNOW (~$2.5B pro forma after absorbing MRC Global), plus the coming Motion separation from Genuine Parts (targeted ~Q1 2027); construction/farm dealers are thin (Alta the cleanest construction pure-play, Titan the cleanest ag one) and aerospace runs through AAR, VSE, and HEICO. [10]
• Hardware, plumbing & heating (4237) — quality-compounder distribution. Ferguson (~$30.8B FY2025), Watsco (~$7.2B, the only listed HVAC pure-play), Core & Main (~$7.4B), Fastenal; refrigeration has no pure-play. Hardware's biggest operators — the co-ops Ace ($9.2B of wholesale revenue) and Do it Best (~$6B), plus private Orgill — are not open to outside equity. [9]
• Metal & mineral (4235) — the metal service centers, now fewer and larger. Reliance ($14.29B of 2025 net sales), Ryerson (merged with Olympic Steel in February 2026), Worthington Steel (~$9.5B combined after taking control of Klöckner in June 2026), Canada's Russel Metals; coal/ore has no listed pure-play at all. [7]
• Lumber & construction materials (4233) — clustered in roofing. QXO is the nearest pure-play and is now far larger (~$18B of combined 2025 revenue after Beacon, Kodiak, and TopBuild), plus BlueLinx, Boise Cascade, UFP Industries, Builders FirstSource; the largest operator of all, ABC Supply (~$20.2B, 1,000-plus branches), is private, and brick/stone and "other" have no pure-play. [5]
• Motor vehicles & parts (4231) — the auction/marketplace agents. The salvage duopoly Copart (~$4.6B FY2025) and RB Global (IAA, 2.5M+ vehicles a year), digital platforms OPENLANE (~$1.93B on ~$28.8B of gross merchandise value) and ACV, and parts wholesaler Genuine Parts (~$24B total, ~$9.5B North American Automotive); there is no pure-play buy/resell vehicle wholesaler (Manheim sits inside privately held Cox). [3]
• Miscellaneous durable goods (4239) — wider than this page previously said, but still single-vertical. The prior read that only recyclables and bullion were investable was wrong: four of the five children now have a listed merchant wholesaler — Sims in recyclables (~US$4.9B), Pool Corporation in sporting/recreational goods ($5.29B — but a pool and backyard business), Gold.com (formerly A-Mark) in jewelry and metals ($10.979B of revenue on $210.9M of gross profit), and Alliance Entertainment in "other" (~$1.06B, physical media). Toys has none. Scrap exposure also runs through the electric-arc-furnace steelmakers that own yards (Nucor, Steel Dynamics, Commercial Metals, Cleveland-Cliffs). [11]
• Furniture & home furnishings (4232) — impure proxies. Vertically integrated makers (Hooker, Bassett, La-Z-Boy, Ethan Allen, MillerKnoll, HNI) carry the wholesale function inside; on the furnishings side the proxies are flooring manufacturers who also distribute (Mohawk, Interface, Dixie) plus housewares marketer Lifetime Brands. There is no clean pure-play. [4]

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. Across all nine children, most of the ~144,000 firms and most of the economics sit in private equity roll-ups, family firms, ESOPs, and cooperatives that never trade publicly. More striking is the direction of travel. In 2025–26 alone, GMS was absorbed by Home Depot's SRS (~$5.5B) and TopBuild by QXO (~$17B); Olympic Steel disappeared into Ryerson; MRC Global into DNOW; Distribution Solutions Group agreed to a take-private; ODP and Patterson went private; Owens & Minor's distribution arm went to Platinum; LKQ sold its self-service segment and Toyota Tsusho took Radius Recycling off U.S. markets; Charles & Colvard filed Chapter 11; and DCC opened a formal sale process for the division that owns the appliance channel's best listed proxy. Against that, the offsets are few — Medline's IPO, Asmodee's listing, and the pending Motion separation. The sector is growing while the ways into it are narrowing. [3][4][5][6][7][8][9][10][11]

5. How the money works

Across all nine children this is distribution, not manufacturing — a spread-and-turns business — so the economics of a regulated utility (rate base), a REIT (funds from operations), or a miner (all-in sustaining cost) do not apply. 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. This is a volume-and-logistics game, so the decisive levers are the same everywhere:

• Inventory turns and cash conversion — the real engine. Well-run distributors reach mid-teens-to-20%+ return on invested capital despite thin net margins by cycling inventory fast and managing the cash-conversion cycle (the wholesaler often pays the producer months before the customer pays it).
• Buying scale and vendor rebates — volume-tiered rebates and co-op marketing dollars are frequently the difference between a profitable and an unprofitable year, and they are now quantified: 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. They are also procyclical — GMS's fiscal-2025 margin decline came partly from lower vendor-incentive income as purchase volumes fell, which amplifies a downturn. [8][9][5]
• The line card and fill rate — the roster of manufacturers a distributor is authorized to carry, and the ability to have the right part in stock today, are the moat commodity e-commerce cannot easily bypass.
• Value-added services and the aftermarket — the margin escalator that separates winners: metal processing (roughly half of Reliance's orders now include it), field application engineering in electronics, recurring service and consumables (medical, lab, office copiers), and the razor-and-blades dealer aftermarket. Titan Machinery's disclosure is the cleanest proof in the subsector: 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. [7][10]

The margin range is far wider than a single "distribution margin" implies — and two different margin measures are now in circulation. This is the most important measurement correction the children add, and at least six of them raise it independently: the Census Bureau's "gross margin" and "gross profit" are survey constructs, not GAAP, and cannot be compared to a company's reported margins or applied to agent marketplaces at all. On the Census basis, gross margins across the subsector run from 20.0% (whole vehicles) and 24.0% (lumber) through 28.7% (roofing) and ~34–36% (furniture, brick/stone, "other" construction materials) up to 48.1% (used auto parts) and 48.7% (office equipment). On a GAAP basis the same businesses report far less and span an even wider range — roughly 1.9% in bullion wholesaling and 6–7% in broad-line computer distribution, through 15.1–17.1% (building-products and metal distribution), 25.9–30.7% (equipment dealers, industrial distributors, Ferguson), to 45.0% at Fastenal. [3][4][5][6][7][8][9][10][11]

Three consequences follow, and they are the real rollup insight:

• Gross margin measures how much service is attached; operating margin converges anyway. The selling and overhead cost required to earn a 21% gross margin in electrical distribution is roughly what the extra points buy — so nearly every child lands in the low-to-mid single digits at the operating line and low single digits at the net line. The business is volume × turns, not markup.
• Margin runs opposite to ticket size and throughput. The biggest, most commodity-like, highest-dollar-per-employee children earn the thinnest spreads; the ones that break bulk finely, fabricate, or specify earn the widest. Both the vehicle and building-materials children now demonstrate this as a clean ladder. [3][5]
• Winners do not charge more; they run tighter. The foodservice-equipment benchmarking data is the sharpest illustration in the subsector: typical dealers and high-profit dealers run essentially the same gross margin (23.1% vs 23.3%) but roughly double the pretax margin (3.9% vs 8.2%) — the gap is payroll at 11.9% vs 10.0% of sales, inventory turns of 4.9 vs 4.3, and receivables collected in 28.1 vs 36.3 days. [6]

Two features recur and are worth remembering. First, cash flow is counter-cyclical: because inventory and receivables are the biggest assets, a downturn releases cash as the business runs down stock — Arrow generated over $1.1 billion of operating cash flow in 2024 while revenue fell 16%, and Titan cut inventory by roughly $419 million through a loss-making fiscal 2025. Second, the recurring failure mode is leverage: a thin-margin, working-capital-heavy model punishes debt in a downturn, and the roster now spans the subsector — tire distributor ATD (~$1.9 billion of debt against ~$30 million of cash), foodservice roll-up TriMark, aerospace distributor Incora, hardware co-op True Value, outdoor distributor Big Rock (Chapter 7, January 2026), comics distributor Diamond, and jewelry distributor Charles & Colvard. One footnote worth carrying: the richest profits in this subsector sit closer to the factory than to the resale — Medline's ~12.2% adjusted EBITDA margin partly reflects self-manufacturing, and Avantor earns ~26% in bioprocessing materials against ~11.6% in lab distribution. [3][6][8][9][10][11]

6. Demand drivers

Demand across 423 is derived — it depends on how much the downstream economy is building, buying, and maintaining — and because the nine children serve unrelated end-markets, the subsector's aggregate demand is diversified and partly self-hedging: a downturn in discretionary IT, furniture, or restaurant capex can be cushioned by defensive healthcare, roofing-replacement, and refrigeration demand. The drivers cluster into a few families:

• The construction and housing cycle (interest-rate-driven) moves lumber, brick/stone, hardware, plumbing, HVAC, furniture, and appliances — the single largest cross-cutting switch, and currently the clearest drag. Housing starts ran at a 1.177 million annualized rate in May 2026, 8.7% below a year earlier; total construction put in place eased about 1.4% to $2.16 trillion in 2025; existing-home sales were 4.1 million in 2025, a thirty-year low, with the average seller having owned for a record 8.6 years. The offset is repair-and-remodel at roughly half a trillion dollars a year — though Harvard's indicator expects growth to decelerate toward 0.5% by mid-2027, and the children cite slightly different points on that series, so read it as "roughly half a trillion, growing slowly." [4][5][8][9]
• Industrial production and capital spending moves 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%), nonfuel mineral production up 6% to $112 billion in 2025, and IIJA highway authorizations of $55.7–56.8 billion converting slowly into lettings. [5][7][8][10]
• The AI / data-center / electrification super-cycle is 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 the 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 now force a correction this page previously missed: AI volume is not automatically a margin upgrade. Ingram Micro's fiscal-2025 AI-enablement server volume added growth while diluting gross margin (down 51 basis points on mix), and the electronic-parts child warns that 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. The attractive economics sit in the configuration, networking, security, financing, and services wrapped around the hardware. [6][8]
• Healthcare utilization (an aging population, rising procedure volumes, single-use disposables) drives the defensive medical and ophthalmic channels: 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, taking it to 20.6% of GDP; 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. [6]
• The vehicle fleet and replace-and-repair base — America's fleet of roughly 289 million light vehicles at a record average age of 12.8 years, with 297.5 million registered vehicles covering 3.294 trillion miles in 2024, plus the aging installed base of buildings and equipment (88% of U.S. households on air conditioning; replacement at 80–90% of residential AC unit sales) — makes parts, tires, recycled components, hardware, and HVAC/refrigeration non-discretionary wear-and-replacement demand. [3][9]
• Commodity prices move the dollar value of sales independent of volume in the pass-through children (metal, scrap, bullion, lumber, copper), so these codes can post large year-to-year revenue swings driven by price, not tons moved. The revised children show this decoupling is now unusually wide. Gold ounces sold at the largest bullion wholesaler fell 17.3% while its revenue rose; HVAC distributor sales grew 2.85% in 2025 on price while adjusted unit demand declined; record outdoor participation coincided with flat equipment sales; new powerboat unit sales fell 9.1% while accessory spending held. Dollars pay the bills, but units drive the warehouse — and right now the two are telling different stories in at least four children. [5][7][9][11]
• The offsetting laggards. Farm equipment is the deepest trough — 2026 net farm income is forecast at $153.4 billion, down 2.6% after inflation, with tractor sales down 9.9% and combines down 35.6% in 2025. Office print volume, thermal coal generation, and consumer camera shipments are structural declines, though the last has stopped falling (global shipments rose 11.2% in 2025 off a collapsed base). [6][7][10]

7. Regulation

Wholesale trade is lightly regulated as a business across all nine children — no rate base, no licensed monopoly, no franchise economics, and generally no wholesale-specific license. Barriers to entry come from scale, inventory, capital, and relationships, not from a regulatory grant. The regulation that matters runs through the products the channel carries and the trade policy on their imports, and it varies sharply by child: state dealer/titling, salvage, and odometer law plus the EV transition (autos); building and electrical codes plus export controls on chips (electrical/electronic); FDA medical-device rules and the FTC's Contact Lens and 2024 Eyeglass Rules (professional equipment); dealer-franchise law, aviation traceability, and hazard-communication rules (machinery); firearms licensing (FFL/ATF), FinCEN anti-money-laundering duties, and G7 Russian-diamond sanctions (miscellaneous); formaldehyde, silica, and textile-labeling standards (lumber, furniture); and DOE 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, which the children flag independently: 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, including a $7.4 million penalty against a distributor in 2025). [3][4][5][6][7][8][9][10][11]

The one regulatory theme that touches nearly every child at once is trade policy, and it is now quantified rather than assumed. Section 232 duties on steel and aluminum rose to 50% in June 2025, with 2026 proclamations adding full-value, derivative, and copper treatment (an April 2026 framework applying 50% to specified articles, 25% to derivatives, and 15% to certain industrial and grid equipment through 2027). On top of that sit 25% Section 232 duties on autos and parts (effective for parts no later than May 3, 2025); Section 232 duties on wood products and furniture — a 25% rate on imported upholstered wooden furniture from October 14, 2025 with a step-up postponed to January 2027, though the two furniture children describe the measures at different resolutions and should be read together rather than reconciled; combined softwood-lumber duties at about 35% from August 2025; Section 301 tariffs including a 100% rate on syringes and needles and additional 10–12.5% rates across 60 trading partners on toys; antidumping orders on tires, quartz, mattresses, and furniture; and tightening semiconductor export controls, which in 2025 briefly placed a listed distributor's own Chinese subsidiaries on the Entity List. The company-level bill is large and visible: Medline disclosed roughly $290 million of adverse 2025 pretax impact with ~$200 million more expected in 2026; Goodyear forecast ~$300 million of annualized 2026 tariff cost; Deere flagged ~$600 million in 2025 and ~$1.2 billion projected for fiscal 2026 before a June 2026 cut in equipment tariffs to 15%; and consumer-technology importers paid $23.5 billion of tariffs in 2025 against $4.0 billion the prior year, with the average rate rising from 1% to 7%. [3][4][5][6][7][8][9][10][11]

Two newer regulatory patterns are worth carrying at this level. First, hard-dated efficiency and refrigerant deadlines now sit on the calendar — general-service lamps in July 2028, washers and dryers from March 1, 2028, distribution transformers from April 23, 2029, refrigeration in 2029–30, commercial gas water heaters from October 2026 — which turns inventory obsolescence from a speculative risk into a scheduled one. Second, the enforcement calendar is a moving target even where the direction is settled: DOE proposed rolling back or postponing more than a dozen appliance standards in 2025, and the EPA proposed eliminating or extending the December 2025 refrigerant installation deadline, deprioritized enforcement of the January 1, 2026 ban, and relaxed several commercial-refrigeration deadlines in May 2026. The 4237 children disagree on whether that installation bar is binding — the HVAC child treats it as effective, the refrigeration child documents the relaxation — and we carry both. The honest read is that the direction of travel is settled and the timing is not, which makes inventory positioning genuinely risky in both directions. The countervailing case is right-to-repair, which has gone from pending to landed: the FTC and five states settled with Deere in July 2026, requiring ten years of dealer-equivalent repair access for farmers and independent repairers — a direct assault on the aftermarket margin that anchors the dealer model. The federal REPAIR Act for vehicles remains a bill. [8][9][10]

For a group-level investor, regulation here is a cost-and-inventory risk plus, in a few children, a revenue tailwind — not the license-to-operate regulation of a utility.

8. Consolidation

The defining structural fact for the whole subsector is fragmentation being slowly rolled up. With ~144,000 firms and ~225,000 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: in industrial machinery alone, firm counts fell from 22,773 in 2017 to 18,795 in 2022 while sales grew. But the group-level fragmentation (CR4 9%, HHI 36.5) is partly an artifact of aggregation: the leaders in each child are different companies, so pooling them dilutes concentration. Consolidation is a within-niche game with a different engine per child:

• Private-equity buy-and-build of aging-owner independents — the dominant story in professional equipment, machinery, miscellaneous, lumber, and flooring distribution. [5][6][10][11]
• Salvage duopoly in auto auctions — Copart (~50%) and RB Global's IAA (~35%) — over a fragmented tail, with digital entrants reshaping the whole-car side. [3]
• 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 in September 2025; QXO's ~$10.6 billion Beacon takeover in April 2025, then Kodiak (~$2.25 billion) and TopBuild (~$17 billion, July 2026), against an openly stated ~$50 billion roll-up ambition; Lowe's $8.8 billion Foundation Building Materials deal, completed October 2025; and Ferguson, Core & Main, and Watsco compounding in plumbing/HVAC. [5][9]
• Cooperatives and buying groups in hardware and appliances — Ace, Do it Best (which absorbed True Value's wholesale platform for $153 million after its October 2024 Chapter 11), Nationwide, and the electronics buying co-ops. [8][9]
• Backward integration by end-users — steelmakers buying scrap yards to lock in feedstock (Nucor/DJJ, Steel Dynamics/OmniSource, Toyota Tsusho's $907 million purchase of Radius in July 2025). The pace has accelerated: roughly 18 scrap deals in 2021–2025 against 11 across the whole 2005–2020 span. [11]
• Strategic and vertically integrated acquirers buying the channel outright — a force the previous version of this page understated. Sysco's $969 million purchase of foodservice dealer Edward Don (with Restaurant Depot pending); Cencora's February 2026 agreement to combine MWI Animal Health with Covetrus at a $3.5 billion valuation; vision insurer VSP's acquisition of frame maker Marcolin; Xerox's ~$1.5 billion Lexmark combination; Worthington Steel taking control of Klöckner; and Ryerson's merger with Olympic Steel. [5][6][7][8]

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 primer reports building-products deal volume down about 21% in 2025 on tariff uncertainty while the roofing primer 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. [5][10]

The countervailing force — also universal — is disintermediation: manufacturers selling direct to large accounts, e-commerce marketplaces (Amazon Business, WebstaurantStore), hyperscaler cloud marketplaces, OEM captive branches, and direct-to-consumer flows all threaten to route around the middleman's spread. That threat is no longer merely feared: the captive-branch shares in Section 2 are a direct measurement of how much distribution has already bypassed the independent channel. The strategic response, seen in every child, is to get stickier — same-day delivery, technical service, private label, vendor-managed inventory, vending, and digital ordering, which at scale is an advantage rather than a threat (Sonepar booked €12.3 billion of online sales in 2025 alongside, not instead of, its branch network). [3][6][8][10]

9. Risks

The children share a common risk stack, weighted differently:

• Cyclicality. Big-ticket, deferrable durable goods swing hard with their underlying cycle — sharpest in furniture, construction materials, farm equipment, and appliances; cushioned in medical, roofing-replacement, refrigeration, and aerospace aftermarket. The amplitude is measurable: real output in furniture and home-furnishing wholesaling fell about 27% peak-to-trough in the last recession, and Titan's U.S. agriculture same-store sales fell 17.4% in fiscal 2026. [4][10]
• Thin margins plus heavy working capital and leverage. A model with low-single-digit net margins and inventory-and-receivables-heavy balance sheets is rate-sensitive and punishes debt in a downturn — the failure roster (ATD, TriMark, Incora, True Value, Big Rock, Diamond, Charles & Colvard) now spans four different children. [3][6][9][10][11]
• Disintermediation. Manufacturer-direct selling, e-commerce, marketplaces, and captive branches compressing the distributor's cut — the single biggest long-run structural threat, and the one now quantified by the branch-share data in Section 2.
• Tariffs and trade policy. Near-total import dependence in several children makes the whole subsector exposed, and pass-through is usually possible but incomplete — Alta explicitly reported that 2025 tariff costs were not fully recovered, contributing to a 100-basis-point decline in equipment margins. [10]
• Commodity-price swings. In the pass-through children (metal, scrap, bullion, lumber, copper), unhedged or mis-timed inventory turns thin spreads negative and creates markdown risk; the framing-lumber composite's 73% three-month crash in 2021 is the reference case. [5][7][11]
• Concentration and contract risk — and it points in opposite directions by child. Upstream: Watsco's ten largest suppliers were 85% of 2025 purchases (62% Carrier alone); WESCO's top ten were ~32%; Genuine Parts drew ~55% of U.S. automotive inventory purchases from ten suppliers and booked a $151 million credit-loss reserve when one of them failed; Titan took ~75% of new-equipment revenue from a single manufacturer. Downstream: Home Depot and Lowe's together were 43.4% of Hillman's 2025 revenue, Alliance Entertainment's top three customers ~40%, and two customers were 33% of a major roofing manufacturer's revenue. Loss of authorized-dealer status can gut a distributor; loss of a line review can gut a supplier. [3][5][8][9][10][11]
• Secular decline in mature niches. The EV transition (auto parts), disintermediation of tires, falling office print volume, and thermal-coal retirement are structural, not cyclical, headwinds. Photographic belongs on this list historically but not prospectively — its collapse has already happened. [3][6][7]
• Cyber and operational risk — newly visible. Two children raise it independently: Ingram Micro's July 2025 ransomware incident took systems offline, temporarily impaired its ability to process and ship orders, and cost $6.2 million; in lab distribution, customers wire procurement systems straight into distributor platforms, so an outage halts order entry, inventory visibility, invoicing, and regulated-product traceability at once. For a distributor, a cyber event is a fulfillment failure, not just a data one. [6][8]
• Data opacity as an investor risk. Photographic wholesaling's sales and concentration are suppressed outright, computer receipts differ by $12 billion across Census tables, tire receipts carry two unreconciled figures, and office employment differs by 20% between programs. Parts of this subsector cannot be sized or benchmarked from federal statistics alone. [3][6]

10. How to invest & outlook

Public routes are narrower than the subsector's $4.9 trillion implies — narrower still than a year ago — and cluster in the picks-and-shovels distributors levered to technology, healthcare, industry, and infrastructure, not in the largest children by revenue. The cleanest listed exposure sits in electrical/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, MSC, DNOW), the plumbing/HVAC compounders (Ferguson, Watsco, Core & Main), metal service centers (Reliance, Ryerson, Worthington Steel), roofing (QXO), the auto-salvage agents (Copart, RB Global), and — wider than this page previously allowed — four single-vertical distributors inside miscellaneous durable goods (Sims, Pool Corporation, Gold.com, Alliance Entertainment). These are cyclical-value and quality-compounder distribution names — judged on organic growth, return on invested capital, cash conversion, inventory turns, and disciplined M&A, not growth multiples. There is no dedicated NAICS-423 ETF, several children (furniture, tires, coal/ore, refrigeration, brick/stone, office machines, photographic, toys) have no pure-play at all, and the listed set keeps thinning as strategics and sponsors take distributors private. One diligence discipline applies subsector-wide: normalize gross-versus-net accounting and survey-versus-GAAP margins before comparing anything — headline revenue mixes principal and agent activity, gross merchandise value, commodity pass-through, and Census constructs that are none of the above. (Reserve valuation multiples and dividend decisions for company-level diligence.) [3][4][5][6][7][8][9][10][11]

Private routes are where most of the subsector's capital actually participates, because 423 is fundamentally a private-market business: backing or rolling up regional independents, family-business succession, PE buy-and-build platforms, participating in cooperatives or ESOPs, owning the distribution-yard and warehouse real estate, and lending against inventory and receivables. The ~144,000 mostly small firms — a majority of the economics — are reachable only this way, and the largest single operators in several children (ABC Supply, Sonepar, Manheim, Ace, Orgill, Digi-Key, Winsupply, US LBM) are not for sale on any exchange. Diligence is unusually similar across children: supplier authorizations and change-of-control terms, rebate quality and normalization, inventory aging and transition-sensitive stock, receivable quality, territorial exclusivity, branch and route density, demonstrated tariff pass-through, and how much of the franchise walks out with the selling owner. [5][6][9][10][11]

Outlook (judgment, not fact). As an aggregate, 423 is a self-hedging barbell whose weighted trajectory is modestly constructive but with a widening internal divergence. The clearest tailwind remains the AI / data-center / electrification super-cycle, which lifts the two largest technology-and-power-facing children plus metal — roughly a third of the subsector — while defensive children (medical, roofing, refrigeration, plumbing) and the aftermarket floor across all nine provide ballast. The laggards (furniture, farm equipment, appliances, tires, office machines, thermal coal) drag the average down. Three cautions this pass adds. First, volume growth is not margin growth — AI hardware grew computer distribution while diluting its gross margin, and the richest returns in this subsector sit closer to manufacturing than to resale. Second, the contestable pool is smaller than the headline, because captive manufacturer distribution is inside these codes and, in the largest child, is more than half of it. Third, the listed opportunity set is contracting even as the sector grows. The clearest shared wildcards are tariff-driven cost volatility, the housing and construction cycle, disintermediation, and a regulatory calendar whose direction is settled but whose enforcement timing is not. The most durable structural theme, common to all nine children, 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. For the full company detail and section-by-section depth, follow the nine child links (4231–4239). These are projections, not certainties.

---

Sources

This is a nine-child rollup. The size, count, and concentration figures in Sections 2–3 are Histometrics' ingested ground-truth federal statistics for NAICS 423, on the 2022 Economic Census / 2023 County Business Patterns basis; the remaining substance is synthesized from the nine child primers, which carry the fuller source lists and the newer federal vintages cited above.

1. U.S. Census Bureau — 2022 NAICS Definitions, subsector 423 Merchant Wholesalers, Durable Goods and its nine industry groups (merchant-wholesaler scope; exclusions vs. agents/brokers 425 and retail 44–45). 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 423. 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 4231, Motor Vehicle and Motor Vehicle Parts and Supplies Merchant Wholesalers (receipts $1,072.3B; CR4 39.2% / HHI 454.1; 2023 AIES showing $464.8B of $838.5B whole-vehicle sales from manufacturers' sales branches; Census gross-margin ladder 20.0%/24.3%/35.0%/48.1%; unreconciled tire figures $76.0B vs $56.2B; ~289M light vehicles at 12.8 years; salvage duopoly; Section 232 auto-parts tariffs).
4. Histometrics rollup primer — NAICS 4232, Furniture and Home Furnishing Merchant Wholesalers (receipts $149.7B; CR4 12.9% / HHI 65.9; Census gross margins 34% and 38.4% on a narrower merchant-only base; sources disagree on which half is softer; real sectoral-output amplitude; Section 232 furniture tariffs at two resolutions).
5. Histometrics rollup primer — NAICS 4233, Lumber and Other Construction Materials Merchant Wholesalers (receipts $324.0B; CR4 16.7% / HHI 124.9; two federal sales bases per child; Census gross margins 24.0–36.2% inversely ranked against throughput; roofing CR4 51.3%; GMS and TopBuild removed from public markets; Home Depot–SRS price disagreement; QXO ~$18B combined).
6. Histometrics rollup primer — NAICS 4234, Professional and Commercial Equipment and Supplies Merchant Wholesalers (receipts $766.5B; CR4 15.2% / HHI 133.7 — below all six publishing children; computers + medical ≈82–84%; photographic sales and concentration suppressed; computer receipts a $319–331B band; office employment contested; AI dilutes distributor gross margin; foodservice high-profit-dealer benchmarking; Medline IPO; Ingram Micro ransomware).
7. Histometrics rollup primer — NAICS 4235, Metal and Mineral (except Petroleum) Merchant Wholesalers (receipts $329.3B; CR4 15.6% / HHI 108.7; 2023 AIES ~$296.4B with $82.5B producer-owned branches; group Census gross margin 24.2% of own-account sales; Reliance 28.7% vs Ryerson 17.1%; Ryerson–Olympic Steel and Worthington–Klöckner closings; Section 232 steel/aluminum 50%; no coal/ore pure-play).
8. Histometrics rollup primer — NAICS 4236, Household Appliances and Electrical and Electronic Goods Merchant Wholesalers (receipts $895.2B; CR4 31.0% / HHI 331.1; 2023 AIES branch shares from under 7% to ~a fifth; gross-margin ladder with converging operating margins; electricity demand 1.7%/yr; ECIA authorized Americas +7.1% to $30.9B; WESCO supplier concentration and rebates; DCC Technology sale process; 2028–2030 efficiency deadlines).
9. Histometrics rollup primer — NAICS 4237, Hardware, and Plumbing and Heating Equipment and Supplies Merchant Wholesalers (receipts $298.6B; CR4 19.8% / HHI 177.1 — below both publishing children; 2023 AIES plumbing $100.3B with $14.5B branches; 2025 growth was price not volume; Ferguson/Watsco/Core & Main; hardware co-ops and True Value; DSG take-private; refrigerant-rule disagreement between children).
10. Histometrics rollup primer — NAICS 4238, Machinery, Equipment, and Supplies Merchant Wholesalers (receipts $725.6B; CR4 10.1% / HHI 40.9; 2023 AIES branch shares 16–23%; Titan's 73.1%-of-revenue / 33.8%-of-gross-profit equipment split; firm counts down from 22,773 to 18,795 between 2017 and 2022; DNOW/MRC and DSG removing listed exposure; Motion separation ~Q1 2027; FTC–Deere right-to-repair settlement; farm trough).
11. Histometrics rollup primer — NAICS 4239, Miscellaneous Durable Goods Merchant Wholesalers (receipts $356.3B; CR4 8.5% / HHI 38.6 — below all four publishing children; four of five children now have a listed distributor — Sims, Pool Corporation, Gold.com, Alliance Entertainment — none a code pure-play; scrap captive to steelmakers; jewelry receipts inflated by bullion gross value; scrap M&A acceleration; Big Rock, Diamond, and Charles & Colvard failures).