Steel Embedded in Traded Goods
Much of the world's steel crosses borders not as coil, plate or bar but inside finished goods: cars, machinery, appliances, ships and containers. This indirect steel trade does not appear in steel trade statistics, yet every imported car or excavator represents steel that was bought from a mill abroad rather than at home. The result is indirect steel demand: demand for steel that is met by foreign mills through imported goods, and for a steel-producing country, demand forgone.
The volumes are large. The World Steel Association (worldsteel) estimates that indirect steel exports from the 74 countries it analyses rose from 325 million tonnes (Mt) in 2014 to 410 Mt in 2024, equal to 93% of direct steel exports in that year (worldsteel, January 2026).
This page gives indirect steel demand coefficients for converting trade in steel-containing goods, recorded by Harmonized System (HS) code, into the tonnage of steel those goods embody. It covers 54 product lines, made up of 66 four-digit HS headings and chapter 82 (tools and cutlery), where the steel content is best understood and the trade most significant. The coefficients are consistent with the assumptions behind our steel end-use analysis 2025, so that indirect trade can be set directly against each sector's share of the 1,718 Mt of world steel demand shown in our steel demand analysis 2025.
How to Use This Page
- Get the trade data. Download imports and exports by HS code in net mass (tonnes or kg), not value, from UN Comtrade, Eurostat Comext or your national customs service. Use 6-digit codes where the tables below split a heading, such as 8704 by vehicle weight.
- Net off exports. For each heading, net imports = imports − exports. A negative figure means the country is a net exporter of steel in that product.
- Multiply by the gross coefficient. Net imports × gross steel (%) = finished steel forgone: the steel the foreign manufacturer had to buy, including the scrap lost in stamping, forging and machining.
- Convert to crude steel, if needed. Multiply by about 1.10–1.15 to allow for losses between crude and finished steel at the mill, an estimate by Metals Consulting International (MCI).
- Add up by sector. Each table below matches one sector of the end-use analysis. Adding indirect trade to apparent steel use sector by sector gives true steel use: the steel a country's economy actually consumes.
Worked example. A country imports 100,000 t (net mass) of petrol and diesel cars under HS 8703 and exports none. At 58% net steel, the cars contain about 58,000 t of steel. At 107% gross steel, the carmakers abroad bought about 107,000 t of finished steel to make them, equal to about 118,000–123,000 t of crude steel.
Indirect Steel Demand Coefficients by HS Code
All figures are percentages of the product's net mass as reported in customs data. Headings are grouped by the end-use sectors used in our end-use analysis.
Attention: Comparing with worldsteel
worldsteel defines its steel coefficient as the amount of finished steel products needed to produce one tonne of a manufactured product (definitions and sources). That measures steel bought, not steel contained, so it is closer to the gross steel figures on this page than to the net figures. When comparing your results with worldsteel's indirect trade data, use the gross column.
- Net steel: steel left in the finished product.
- Yield loss: share of the steel bought by the manufacturer that ends up as process scrap.
- Gross steel: net steel ÷ (1 − yield loss), the steel bought per tonne of product. It can exceed 100%.
Construction
| HS Code | Product | Net Steel (%) | Yield Loss (%) | Gross Steel (%) | Confidence |
|---|---|---|---|---|---|
| 7308 | Structures and parts (bridges, towers incl. wind towers, frames, doors) | 95 | 10 | 106 | High |
| 7322 | Radiators for central heating | 92 | 10 | 102 | High |
Source: MCI research. Figures are MCI estimates. Data as of November 2026.
See the construction assumptions in the end-use analysis.
Mechanical Engineering
| HS Code | Product | Net Steel (%) | Yield Loss (%) | Gross Steel (%) | Confidence |
|---|---|---|---|---|---|
| 7307 | Tube or pipe fittings | 95 | 25 | 127 | High |
| 8402 | Steam boilers | 90 | 15 | 106 | Medium |
| 8403 | Central heating boilers | 73 | 20 | 91 | Medium |
| 8417, 8514 | Industrial furnaces and kilns | 80 | 15 | 94 | Medium |
| 8425–8428 | Lifting and handling equipment (cranes, forklifts, conveyors) | 83 | 18 | 101 | Medium |
| 8429 | Bulldozers, excavators and loaders | 83 | 20 | 103 | Medium |
| 8430 | Other earth-moving and mining machinery | 83 | 20 | 103 | Medium |
| 8431 | Parts of lifting, handling and earth-moving machinery | 90 | 30 | 129 | Medium |
| 8432–8433 | Agricultural machinery | 78 | 20 | 97 | Medium |
| 8454–8455 | Converters, casting machines and rolling mills | 90 | 35 | 138 | Medium |
| 8474 | Crushing, mixing and screening machinery | 88 | 20 | 109 | Medium |
| 8480 | Moulds and mould boxes | 92 | 45 | 166 | Medium |
| 8482 | Ball and roller bearings | 90 | 45 | 164 | High |
| 8483 | Transmission shafts, gears and gearboxes | 80 | 40 | 133 | Medium |
| 8501 | Electric motors and generators | 60 | 35 | 92 | Medium |
| 8502 | Generating sets (incl. wind turbines, 8502.31) | 63 | 25 | 83 | Medium |
| 8504 | Transformers and inductors | 60 | 15 | 71 | Medium |
| 8535 | High-voltage switchgear (over 1,000 V) | 53 | 20 | 66 | Medium |
| 8701 | Tractors (incl. road tractors) | 73 | 30 | 104 | Medium |
Source: MCI research. Figures are MCI estimates. Data as of November 2026.
See the mechanical engineering assumptions in the end-use analysis.
Automotive
| HS Code | Product | Net Steel (%) | Yield Loss (%) | Gross Steel (%) | Confidence |
|---|---|---|---|---|---|
| 8703 | Passenger cars – petrol, diesel and hybrid | 58 | 46 | 107[a] | Medium |
| 8703.80 | Passenger cars – battery electric | 50 | 45 | 91[b] | Medium |
| 8704 (≤5 t) | Vans and pickups, up to 5 t gross vehicle weight | 53 | 46 | 98[a] | Medium |
| 8706–8707 | Vehicle chassis with engines; vehicle bodies | 73 | 44 | 129[b] | Medium |
| 8708 | Motor vehicle parts and accessories | 63 | 46 | 116[b] | Low |
Source: MCI research. Other figures are MCI estimates. Data as of November 2026.
- Net steel and yield loss both taken from the MCI end-use workbook (for example 830 kg of steel in a 1,440 kg car).
- Yield loss from the end-use workbook; net steel an MCI estimate.
See the automotive assumptions in the end-use analysis.
Other Transport
| HS Code | Product | Net Steel (%) | Yield Loss (%) | Gross Steel (%) | Confidence |
|---|---|---|---|---|---|
| 8601–8606 | Locomotives, rail wagons and coaches | 83 | 20 | 103 | Medium |
| 8607 | Railway parts (wheels, axles, bogies) | 95 | 30 | 136 | High |
| 8702 | Buses and coaches | 65 | 20 | 81[a] | Medium |
| 8704 (>5 t) | Trucks and dumpers, over 5 t gross vehicle weight | 51 | 20 | 64[a] | Medium |
| 8705 | Special-purpose vehicles (mobile cranes, concrete mixers) | 78 | 25 | 103 | Medium |
| 8714 | Parts of motorcycles and bicycles | 65 | 30 | 93 | Medium |
| 8716 | Trailers and semi-trailers | 78 | 15 | 91 | Medium |
| 8802–8803 | Aircraft and aircraft parts | 6 | 80 | 30[a] | Low |
| 8901 | Cruise ships, cargo ships and ferries | 90 | 23 | 117[b] | Medium |
Source: MCI research. Other figures are MCI estimates. Data as of November 2026.
- Net steel and yield loss both taken from the MCI end-use workbook.
- Yield loss from the end-use workbook; net steel an MCI estimate.
See the other transport assumptions in the end-use analysis.
Domestic Appliances
| HS Code | Product | Net Steel (%) | Yield Loss (%) | Gross Steel (%) | Confidence |
|---|---|---|---|---|---|
| 7321 | Stoves and cookers (non-electric) | 65 | 22 | 83[a] | Medium |
| 8415 | Air-conditioning machines | 45 | 20 | 56 | Medium |
| 8418 | Refrigerators, freezers and heat pumps | 65 | 22 | 83[a] | Medium |
| 8450 | Washing machines | 65 | 22 | 83[a] | Medium |
Source: MCI research. Other figures are MCI estimates. Data as of November 2026.
- Net steel and yield loss both taken from the MCI end-use workbook.
See the domestic appliances assumptions in the end-use analysis.
Metal Products
| HS Code | Product | Net Steel (%) | Yield Loss (%) | Gross Steel (%) | Confidence |
|---|---|---|---|---|---|
| 7309 | Tanks and vats over 300 litres | 95 | 12 | 108 | High |
| 7310 | Tanks, drums and cans up to 300 litres | 95 | 10 | 106 | High |
| 7311 | Containers for compressed or liquefied gas | 95 | 15 | 112 | High |
| 7312, 7314, 7317 | Wire rope, wire mesh and fencing, nails | 98 | 9 | 107[a] | High |
| 7315 | Chain | 98 | 10 | 108 | High |
| 7318 | Screws, bolts, nuts and washers | 98 | 9 | 107[a] | High |
| 7320 | Springs | 98 | 8 | 106 | High |
| 7323 | Kitchen and household articles | 88 | 30 | 125 | Medium |
| 7324 | Sanitary ware (sinks, baths) | 83 | 25 | 110 | Medium |
| 82 | Tools and cutlery (chapter 82) | 80 | 30 | 114 | Medium |
| 8609 | Freight containers | 90 | 17 | 108[b] | High |
| 9403 | Metal furniture (9403.10, 9403.20) | 75 | 15 | 88 | Medium |
Source: MCI research. Other figures are MCI estimates. Data as of November 2026.
- Yield loss from the end-use workbook; net steel an MCI estimate.
- Net steel and yield loss both taken from the MCI end-use workbook.
See the metal products assumptions in the end-use analysis.
Oil and Gas
| HS Code | Product | Net Steel (%) | Yield Loss (%) | Gross Steel (%) | Confidence |
|---|---|---|---|---|---|
| 8905 | Drilling platforms and floating production structures | 90 | 17 | 108[a] | Medium |
Source: MCI research. Other figures are MCI estimates. Data as of November 2026.
- Yield loss from the end-use workbook; net steel an MCI estimate.
See the oil and gas assumptions in the end-use analysis.
Defence
Source: MCI research. Other figures are MCI estimates. Data as of November 2026.
- Yield loss from the end-use workbook; net steel an MCI estimate.
- Net steel and yield loss both taken from the MCI end-use workbook.
See the defence assumptions in the end-use analysis.
Swipe the tables sideways for more columns; the first column stays in place.
How the Coefficients Are Built
Net or gross? Net steel measures the steel a product contains, which is the right basis for steel stocks in use and future scrap. Gross steel measures the steel bought to make it, which is the right basis for the demand a domestic steelmaker has lost. A car body stamping uses only about 56% of the sheet bought, and a forged crankshaft about 45% of the bar, so for a car the gross figure is nearly twice the net. The same yield assumptions underlie the automotive figure in our end-use analysis.
Consistency with the end-use analysis. Where the end-use analysis gives a steel content or yield for a product, the coefficient here uses the same figure: cars, vans, trucks, buses, ships, containers, appliances, wire, fasteners, defence equipment and aircraft. The 15% allowance for automotive spare parts is left out of 8703, because spare parts are traded separately under 8708.
Which headings are shown. The tables show headings rated high or medium confidence, plus three large flows (vehicle parts, aircraft and munitions) whose yield losses come from the end-use workbook. Headings that are mostly cast iron or non-ferrous metals (engines, pumps, valves), very mixed (8479, 7326) or carry little steel (electronics, cable) are left out, as are vessels for breaking up (8908), which are a source of scrap rather than demand forgone.
What the Coefficients Do Not Capture
- Substitution. Steel forgone assumes the goods could have been made at home with domestic steel. Where a country has no car or shipbuilding industry, imports displace no local steel demand.
- Quality of trade data. Net mass is missing or estimated for some lines, and ships are often reported by number or gross tonnage. Where only values are available, convert them with average unit values first.
- Re-exports and processing trade. Goods that pass through a country, or are assembled from imported parts, can be counted twice unless the flows are netted.
- Product mix. A 4-digit heading can hold very different goods. The coefficients are averages; 6-digit codes and national product mixes will give better results.
- Steel quality. The figures cover all steel, including stainless and engineering steels; they do not split it by product.
Health Warning and Terms of Use
These coefficients are a best-efforts assessment by Metals Consulting International Limited, based on public information, the MCI end-use workbook and MCI estimates. They are indicative averages and may be revised.
MCI and SteelOnTheNet make no warranty as to the accuracy, completeness or fitness for purpose of these figures, and accept no responsibility or liability for any loss arising from their use. They are not investment, financial, legal or engineering advice. Users should make their own checks before relying on any figure.
Use of this page is subject to our terms of use.
How to Cite This Page
Metals Consulting International (2026) 'Indirect steel demand coefficients: steel content of goods by HS code', SteelOnTheNet. Available at: https://www.steelonthenet.com/resources/market-data/indirect-steel-demand.html (Accessed: 20 November 2026).
Indirect Steel Demand Analysis for Your Market
We can estimate indirect steel trade for any country or region from customs data, by HS code and end-use sector, and set it against apparent steel consumption to show true steel use and the steel demand lost to imports.
Need This Analysis for Your Own Plant or Market?
152 projects in 48 countries since 2001
Free, confidential first conversation No obligation Reply within one working day