UK metals trading: the billions flow through tiny desks, the work happens in warehouses

About 150 UK metals trading and stockholding companies publish a full profit-and-loss, booking £19.1bn of combined turnover — and almost none of that number means what it appears to mean. The most dramatic line on the map is a departure: Kaz Minerals, the Kazakh copper group long run through a London holding company, sold its entire operating business — the mines and roughly 14,000 employees — to Dubai-based owners in May 2024, paid out roughly £1.45bn of dividends along the way, and left behind a residual London company with no operations. The trade that remains splits into two businesses that share a label and nothing else: trading desks, where Derek Raphael & Company books £365M of metal through three people, and stockholders, where Barrett Steel needs 1,107 people, and a national network of warehouses and saws, to book £422M. Figures are approximate — verify against a company’s own accounts before relying on any single number.

Two trades behind one label

Before comparing any two lines below, separate the models — because in this market, turnover per employee runs from £122M a head to under £300k a head, and that ratio tells you which business you’re looking at:

• Trading desks buy and sell cargoes of metal — often concentrate or ingot that never touches Britain — and book the full contract value as turnover. Derek Raphael (£365M, 3 staff), Minmetals (U.K) (£775M, 10 staff — the London arm of China’s state metals trader) and Stratton Metal Resources (£334M, 9 staff) are desks. Their sub-2% “margins” aren’t thin — they’re the normal arithmetic of pass-through commodity turnover, and their profits swing with positions, not payroll.
• Stockholders and service centres hold metal in racks, cut it to order and deliver it — Aalco (703 staff), Barrett Steel (1,107), National Tube Stockholders (214). Turnover per head sits at £300–700k, and the 2–7% margins are genuine operating economics: warehouse productivity, stock turn and the buying cycle.
• Producer-owned arms — Voestalpine High Performance Metals UK, Press Metal UK, VSMPO Tirus (the Redditch distribution arm of Russia’s titanium producer VSMPO-AVISMA, still trading at £69.9M) — distribute their parent’s output, so their margins carry whatever transfer price the group sets.

Never compare a desk’s margin with a warehouse’s. And treat the map’s combined turnover as a measure of flow, not of a British industry’s size: one desk’s cargo can be another’s, and the same tonne of copper can appear in two companies’ revenue lines.

The giants — one group counted twice, a departed miner’s sales desk, and one 1929 conglomerate

Two of the top three lines are one organisation: Ocean Partners Holdings is the parent of Ocean Partners UK — the same Maidenhead concentrate trader appearing at group and subsidiary level (both report in dollars; we’ve converted the pair on the same basis). And Kaz Minerals Sales sells what a mining group that is no longer run from Britain digs up. Count distinct groups and the top of this market is a concentrate trader and a scatter of desks.

The name conspicuously absent from the top of that table is the bigger story. Until 2024, Kaz Minerals — the London holding company of the Kazakh copper group — would have bent every average on the map, supplying most of its combined profit at the kind of pre-tax margin only an ore body earns — mining economics, not warehouse economics. Its latest accounts record the exit instead: in May 2024 it sold the entire operating group — the mines and roughly 14,000 employees — to Dubai-based KAZ Minerals International, booked about $6.7bn of dividend income against $7.1bn of impairments for a $453M loss, and paid $1.9bn (≈£1.45bn) of dividends. The group is now headquartered in the UAE; what remains in London is a residual holding company with no operating businesses, which we’ve excluded from the map’s combined figures. Its sales arm, still on the map, books £1.96bn here — down 34% on the year.

The quieter structural story is that several mid-table names that look independent are stablemates. Aalco and Righton & Blackburns (£136.4M, £9.8M PBT) share a Wednesbury base and a parent — Amari Metals — while Smiths Metal Centres (£121.9M, £11.5M PBT) sits under Cobham Aluminium & Stainless Holdings. Aalco alone paid £60M of dividends against £23.1M of pre-tax profit in its latest year — a balance-sheet distribution up the Amari group, not a trading result. Amalgamated Metal Corporation, the London trading group incorporated in 1929, runs a stable of its own — including William Rowland, the Victorian-rooted Sheffield alloys name further down this report, which is 100% AMC-owned. These groups’ entities also file to different dates, so their figures are not all from the same year.

The shape of the market

There is no small tier here — just 7 companies below £5M of turnover, against 138 above it. Metal is heavy, stock is capital, and the businesses big enough to publish a full profit-and-loss start at warehouse scale. Every band is 76–84% profitable: this is a mature, mostly-profitable trade whose losers are the exception.

The year the metal got cheaper

99 of the 135 companies with a prior-year comparison saw turnover fall. But look at the pattern among the stockholders: Aalco −9% turnover with staff up 3%; Thyssenkrupp Materials (UK) −16% with staff flat; Rainham Steel −12%, staff +1%; Sebden Steel −21%, staff −2%; National Tube Stockholders −12%, staff +3%. A demand collapse empties warehouses and cuts headcount. Revenue falling while the workforce holds is what a price fall looks like: steel and aluminium came off their post-2022 peaks, and since a stockholder’s turnover is tonnes × price, the whole map deflated together. Read every growth and decline figure in this report with that in mind — much of it is the commodity cycle passing through the revenue line, in both directions.

The one large-cap genuinely cutting both is Barrett Steel (−8% turnover, −8% staff, and a £2.0M profit on £421.8M — half a penny in the pound), a reminder that commodity steel stockholding at national scale is a punishingly thin business even for its biggest family-owned name.

Where the margins are: the periodic table, not the warehouse

The best-run mid-market operators cluster around one pattern: the further from commodity steel, the fatter the margin. Nickel superalloys, titanium, aerospace-certified stock and cobalt all pay; mild steel doesn’t.

…and 8 more profitable operators between £5M and £100M clear a 2% margin. The specialists — Neonickel at 13.8%, United Performance Metals and Voestalpine High Performance Metals at 10.1%, Darton at 10.4% on cobalt with eight staff — earn three to twenty times the margin of the commodity-steel names (Premier Steel Stockholding 0.2%, A.J.N. Steelstock at a small loss, Barrett at 0.5%). The moat is metallurgy and certification: aerospace-grade traceability, exotic-alloy stock nobody else holds, and customers who can’t switch on price.

One name the raw ranking would put at the top doesn’t belong in it: Senior Architectural Systems (£50.2M, 26.8% margin) designs and makes aluminium window and door systems — a manufacturer’s margin, not a stockholder’s, so we’ve left it out of the comparison. And Maritime House (£73.1M, 6.5%) appears in the growth section below for reasons that matter more than its margin.

Growth, read with care

In a year when three-quarters of the market shrank, treat triple-digit growth as a structural event until proven otherwise. Maritime House’s +235% is a group company whose revenue tripled with staff up only 13% — a jump of that shape usually reflects an acquisition or a first-time consolidation rather than like-for-like trading; read its accounts before reading its growth rate. Voestalpine Specialty Metals UK was incorporated in 2023, so its +126% compares a full year against what appears to be a short first period. The genuine signals are smaller: United Performance Metals (+41%, staff +13%, 10.1% margin — aerospace alloys, hiring behind real demand) and CCB Sheet Steel (+26% with headcount doubled). Konexus is the opposite trade: +41% turnover, +49% staff, and £725k of profit on £188.7M — buying growth at four-tenths of a penny in the pound. Tangent Trading’s 21.4% margin on £9.7M is desk economics on the year’s positions, not a repeatable operating margin.

Market structure: concentration that isn’t

On paper the top five hold nearly half the market. In practice that head contains the Ocean Partners group twice, and its second-biggest line — Kaz Minerals Sales — is the sales arm of a copper-mining group now run from Dubai, not a competitor to anyone else on the map. Net those out and the distinct trading-and-stockholding market is roughly £15.4bn, earning about £300M before tax between all of them — around 2p in the pound, which matches the trade’s median margin almost exactly. Below the desks, physical stockholding is genuinely fragmented: dozens of regional service centres in the £20–150M band, many family-owned, competing on stock range and delivery radius.

An old trade in old hands

This is one of the oldest maps we’ve drawn: 60 of the 150 companies predate 1990, and the register reaches back through Amalgamated Metal Corporation (1929) to Sheffield and Black Country names whose roots are Victorian — like William Rowland, which despite the independent-sounding name is itself a 100% AMC subsidiary. Ownership matches: 88 of the 150 are individually or family owned against 56 corporate-owned, and only 14 carry a Holdings/Bidco/Topco-style name — the buyout industry has largely left metals stockholding alone. The 2021+ cohort is just 8 companies; nobody starts a stockholder from scratch when the incumbents own the freeholds and the stock.

What the map shows

1. The map’s would-have-been biggest profit-maker left the country. Kaz Minerals — the Kazakh copper group long run through a London holding company — sold its mines and roughly 14,000 employees to Dubai-based owners in May 2024 and paid ≈£1.45bn of dividends along the way; what’s left in London is a residual holdco. The actual UK metals trade earns about £300M on £15.4bn of distinct-group turnover: 2p in the pound.
2. Turnover per head is the honest ratio. £122M a head (Derek Raphael, 3 staff) is a trading desk; £400k a head (Barrett Steel, 1,107 staff) is a warehouse business. Their turnover lines measure different things and their margins are never comparable.
3. Last year’s decline was the metal price, not demand. 99 of 135 companies shrank, but the stockholders held or grew headcount while revenue fell — tonnes × a cheaper price, not empty warehouses.
4. Margin lives in the periodic table. Nickel superalloys, titanium, cobalt and aerospace-certified stock earn 8–14%; commodity steel earns 0–1% even at £400M scale.
5. Concentration is a bookkeeping artefact. One group counted twice plus a departed miner’s sales desk make the top five look like nearly half the market; the physical trade underneath is fragmented, family-owned and old — 60 of 150 companies predate 1990, and only 8 have appeared since 2021.

Methodology and caveats

This covers only the 150 UK metals trading and stockholding companies that publish a full profit-and-loss; smaller merchants file abridged accounts with no figures and don’t appear, and metals distribution that sits inside builders’ merchants, engineering groups or producers’ own manufacturing companies is mapped elsewhere. Kaz Minerals, whose latest accounts record the May 2024 sale of its entire operating group to Dubai-based owners, is treated as a residual holding company with no trading turnover and is excluded from the combined figures, though its sales arm remains on the map. Group structures report at several levels — Ocean Partners appears at both group and subsidiary level, and the Amari Metals and Amalgamated Metal Corporation stables span several separately-reporting companies — so the £19.1bn combined turnover overstates the distinct-group total, and entities within one group can file to year-ends a year apart. Several of the global trading houses report in US dollars; their figures are converted to sterling at a single recent rate of roughly $1.34 to the pound rather than each company’s own year-end rate, which understates December-2024 year-ends (when the pound bought about $1.25) by around 6% — growth rates and margins are unaffected. Trading-desk turnover is pass-through contract value and is never comparable with a stockholder’s sales, nor their margins with each other; commodity-price moves pass straight through revenue lines in both directions, so year-on-year growth conflates price and volume. Extreme proportional outliers are excluded from the charts, and business descriptions are directional. Figures are approximate — verify any specific figure against the company’s own accounts. This is analysis, not financial advice.