Loading workbench page
Fetching primary parquet sources and computing exhibits.
Loading chart...
Fetching primary parquet sources and computing exhibits.
Batteries, wind turbines, solar panels and grid infrastructure each demand a narrow set of non-substitutable inputs: cobalt, lithium, nickel, copper, rare-earth elements, graphite, silicon, tungsten, manganese, aluminum. The IEA (2024) Global Critical Minerals Outlook projects demand for lithium to rise roughly 8x by 2040 in the net-zero scenario, while nickel and cobalt roughly double; Bazilian et al. (2020) in Nature warn that the choke-points are on the supply side, because refining and processing are concentrated in a handful of countries that mine-to-market price shocks cannot dislodge. This page tracks the twenty HS6 lines that constitute that basket across BACI bilateral trade, 1996-2024.
We sum BACI export value across every HS6 in the basket, each year, 1996-2024. Current-USD values reflect both price and quantity movements; the 2008, 2011 and 2022 peaks align with commodity-price cycles (Deaton & Miller 1995, Jacks, O'Rourke & Williamson 2011). The IEA (2023) Critical Minerals Market Review reports that the market value of energy-transition minerals doubled over 2017-2022 alone, reaching US$320B in 2022, our basket is broader than the IEA's six-mineral core but the trajectory tracks.
SELECT year, SUM(export_value)*1000 AS usd FROM 'country_year_product/**/*.parquet' WHERE product_code IN (basket_20_hs6_codes) AND export_value > 0 GROUP BY year ORDER BY year;
Bazilian, Bradshaw, Gabriel, Goldthau & Westphal (2020) in Nature argued that supply concentration, not geological scarcity, is the binding constraint on the energy transition. Filtering the basket to HS6 lines where a single exporter commands more than 50% of world exports in 2022 yields the figure below. The list is not the same as the USGS mining-production ranking: the BACI series captures cross-border trade, so producers that consume domestically (China's own rare-earth refining, Australia's domestic lithium chemicals) show up differently in trade than in mined-production statistics.
Supply concentration is not static. For the six family HS6 lines below, cobalt ores, lithium carbonate, nickel ores, copper ores, rare-earth compounds, graphite, the top-1 share has moved in opposite directions since 2000. Lithium carbonate has seen its top-exporter share rise since 2018 as Chile and China consolidated; copper ores and rare-earth compounds, by contrast, have diversified, their leading exporters' shares falling since 2000 as new producers scaled. USGS Mineral Commodity Summaries (annual) tracks the mining side of this story; the BACI series below tracks the export side.
Across the six family lines, the top-6 producers in each year (by combined 2015+2024 share) show a clear story of reshuffling. Indonesia's 2014 raw-ore export ban did the reverse for ore: its nickel-ore exports (HS 260400) collapsed to near zero, while the subsequent Tsingshan-led smelting build-out pushed Indonesian nickel trade into nickel mattes (HS 750110, not shown in this HS6-specific family), where Indonesia is now a leading exporter. The DRC's position in cobalt ores strengthened further; Chile's in lithium carbonate consolidated; the copper picture is more diffuse.
The flip side of exporter concentration is importer exposure. For each country we compute the basket's share of its total merchandise imports in 2024. Chowdhry & Felbermayr (2023) define strategic dependence as a product-country pair where imports are dominated by one supplier and substitutes are scarce; the share shown below is the aggregate exposure. The European Commission (2023) Critical Raw Materials Act and the US Department of Energy (2023)Critical Materials Assessment each identify countries with elevated exposure and mandate diversification targets.
An HS6 line is maximally fragile when it is both concentrated (few exporters) and distant (the exporters are far from the world's consumers). Distance raises shipping costs and shipping-lane vulnerability; concentration eliminates redundancy. We plot the Herfindahl index across exporters (HHI) against each mineral's share-weighted mean export distance, computed from the population-weighted distance in Legacy gravity_bilateral table (exact release provenance unverified) between each exporter and its bilateral trading partners (2020 snapshot). Minerals in the upper-right of the chart are both concentrated and long-haul.
Concentration and distance only matter if prices respond to them. The World Bank Pink Sheet (Commodity Markets Outlook data release) is the longest-running nominal price series for the metals complex. We index ten base- and precious-metal commodities that overlap the critical-mineral basket to 2015 = 100 and read off 2024. Lithium and cobalt are not in the Pink Sheet because the World Bank covers them in a separate series; battery-grade lithium prices collapsed more than 80% from their Nov-2022 peak (Benchmark Mineral Intelligence, Fastmarkets) and are not reproduced here. The visible story is (i) the Ukraine-war spike of 2022 in nickel and aluminum (World Bank, 2022 CMO), (ii) the 2023-2024 softening driven by slower Chinese construction demand, and (iii) the precious-metal rally in gold and silver as central-bank buying and inflation-hedge flows lifted both above their 2015 base. The ten-year cumulative change on the right is the transmission channel from Figures 2-6 concentration into battery-cell, grid-infrastructure and wind-turbine bills-of-materials.
The EU Critical Raw Materials Act (Regulation (EU) 2024/1252, adopted 11 April 2024) Annex II lists 34 critical raw materials; Annex I separately lists the 17 strategic raw materials (a subset). Article 5 sets a binding concentration target: no more than 65 per cent of any strategic raw material's EU consumption from any single third country by 2030. The heatmap below cross-tabulates the strategic-list HS6 codes in our basket against the ten largest EU-member importers in 2024, with each cell shaded by the world-level top-exporter share (a lower-bound proxy for third-country single-origin concentration since member-resolved bilateral HS6 data is not in this build). Red cells are HS6 lines where world concentration already exceeds the CRMA 65 per cent ceiling, so EU single-origin concentration cannot be below it.
Lithium-ion cell manufacturing concentrates in seven economies: China (CATL, BYD, CALB), Korea (LG Energy Solution, Samsung SDI, SK On), Japan (Panasonic, GS Yuasa), the US (Tesla, LGES, Panasonic JVs under IRA Section 45X), and the three largest European cell-assembly hubs, Germany, Hungary, and Poland (BloombergNEF 2024 Battery Manufacturing Assessment). Their active-material basket is the same six-element set (Li, Co, Mn, Ni, graphite) drawn from eight HS6 lines: lithium hydroxide (HS 282520) and carbonate (283691), cobalt ores (260500) and oxides (282200), manganese ores (260200), nickel ores (260400) and mattes (750110), and natural graphite (250410). The bar chart below reports each producer's total 2024 import bill across the eight HS6 lines; the composition stacks below show which elements dominate for each producer. Import dependence on these lines is the numerator of the IEA (2024) Critical Minerals Market Review supply-risk index for battery chemistries.
Figure 5 mapped dependence (basket imports as a share of each country's total imports). Figure 10 shows the absolute scale: the ten largest gross-USD buyers of the same 20-HS6 critical-minerals basket in 2024. Concentration on the demand side determines who sets the marginal price and writes the rules. The IEA (2024) Critical Minerals Market Review stresses that mid-stream refining capacity follows demand-side concentration with a 5-7 year lag (Fig. 3.2), so today's top buyers are tomorrow's mid-stream processors absent a deliberate diversification policy. This is the empirical anchor for the EU CRMA Art. 5 strategic partnerships and the US IRA Section 30D free-trade-partner sourcing rules in Figure 8.
Three concentration statistics recur across these figures. The Herfindahl-Hirschman Index (HHI) is Hirschman (1945, National Power and the Structure of Foreign Trade), with the 0-to-10,000 scaling popularised by Herfindahl (1950). HHI = Σi si2 where si is exporter i's share of world HS6 exports. The US DOJ/FTC (2023) Horizontal Merger Guidelinesflag HHI > 1,800 as 'highly concentrated' and HHI between 1,000 and 1,800 as 'moderately concentrated'. CR4 and CR8 are the cumulative share of the top 4 and top 8 exporters, respectively; the IEA (2024) Critical Minerals Outlook chapter 3 reports CR3 for mined vs refined capacity. For our twenty-HS6 basket in 2022, the median HHI sits at 2331, in the DOJ moderately-to-highly concentrated range. Because the monopoly countries differ across minerals (DRC for cobalt, Chile for lithium-carbonate, China for rare-earth compounds) a co-ordinated disruption is not plausible, but any single closure is near-terminal.
The figures above are the empirical base behind the latest wave of critical-mineral statecraft. The US Inflation Reduction Act Section 30D vehicle credit makes eligibility conditional on a minimum share of critical-mineral value being extracted or processed in a US free-trade partner; the accompanying Section 45X advanced-manufacturing credit pays producers per kg of battery-grade lithium, cobalt, nickel, graphite and rare-earth oxides produced domestically. The US Department of Energy's 2023 Critical Materials Assessment lists 15 materials as 'critical' for energy, including lithium, cobalt, nickel, graphite, REEs, Mn and Si. The EU Critical Raw Materials Act (Regulation (EU) 2024/1252, adopted 11 April 2024) sets explicit 2030 benchmarks: 10% of annual consumption from domestic extraction, 40% from domestic processing, 25% from recycling, and the binding concentration cap of no more than 65% from any third country for each strategic raw material. China's 14th Five-Year Plan (2021) and the dual-circulationstrategy mirror the move from the other direction: accelerate domestic consumption of refined outputs and use export controls on gallium, germanium (MOFCOM Announcement No. 23/2023, effective 1 August 2023) and graphite (MOFCOM Announcement No. 39/2023, effective 1 December 2023) as leverage. Figures 2 - 5 say whether each of those instruments is working at the HS6 level.
Caveat on interpretation. BACI captures cross-border trade, not domestic production. DRC's 74% mining share of cobalt (USGS 2024) does not equal its BACI export share because (i) cobalt hydroxide leaves DRC for China for refining and is re-exported from China under a different HS line (282200); (ii) China's domestic rare-earth refining is invisible in rare-earth HS6 trade but shows up in downstream magnets and batteries; (iii) spodumene, lithium carbonate and lithium hydroxide are three different HS lines tracing the same upstream material through refining. Any policy use of these numbers, diversification targets, stockpile rules, tariff design, must pair BACI with the USGS Mineral Commodity Summaries production series and country-level refining capacity data.
WITH tot AS (SELECT product_code, SUM(export_value) AS t FROM cyp_2022 WHERE product_code IN (basket) GROUP BY product_code) SELECT product_code, country_code, export_value/t.t AS share FROM cyp_2022 JOIN tot USING (product_code) QUALIFY ROW_NUMBER() OVER (PARTITION BY product_code ORDER BY share DESC) = 1 AND share > 0.5;
Scroll horizontally to view the full chart.
Scroll horizontally to view the full chart.
Scroll horizontally to view the full chart.
Scroll horizontally to view the full chart.
SELECT commodity, year, price_nominal
FROM 'pink_sheet_annual.parquet'
WHERE commodity IN ('Aluminum','Copper','Nickel','Tin','Zinc','Lead','Iron ore, cfr spot','Platinum','Gold','Silver') AND year BETWEEN 2015 AND 2024
AND price_nominal IS NOT NULL ORDER BY commodity, year;| 29% |
| 29% |
| 29% |
| 29% |
| 29% |
| 29% |
| 282520 Lithium hydroxide/oxide | 76% | 76% | 76% | 76% | 76% | 76% | 76% | 76% | 76% | 76% |
| 283691 Lithium carbonate | 75% | 75% | 75% | 75% | 75% | 75% | 75% | 75% | 75% | 75% |
| 280530 Sc/Y (rare-earth metals) | 38% | 38% | 38% | 38% | 38% | 38% | 38% | 38% | 38% | 38% |
| 284610 Cerium compounds | 37% | 37% | 37% | 37% | 37% | 37% | 37% | 37% | 37% | 37% |
| 284690 Rare-earth compounds | 34% | 34% | 34% | 34% | 34% | 34% | 34% | 34% | 34% | 34% |
| 260400 Nickel ores | 40% | 40% | 40% | 40% | 40% | 40% | 40% | n/a | 40% | 40% |
| 750110 Nickel mattes | 44% | 44% | 44% | 44% | 44% | 44% | 44% | 44% | 44% | 44% |
| 260300 Copper ores | 29% | 29% | 29% | 29% | 29% | 29% | 29% | 29% | 29% | 29% |
| 740311 Copper cathodes | 22% | 22% | 22% | 22% | 22% | 22% | 22% | 22% | 22% | 22% |
| 260200 Manganese ores | 51% | 51% | 51% | 51% | 51% | 51% | 51% | 51% | 51% | 51% |
| 250410 Graphite, natural | 44% | 44% | 44% | 44% | 44% | 44% | 44% | 44% | 44% | 44% |
| 280461 Silicon, >99.99% | 33% | 33% | 33% | 33% | 33% | 33% | 33% | 33% | 33% | 33% |
| 261100 Tungsten ores | 15% | 15% | 15% | n/a | 15% | 15% | 15% | 15% | 15% | n/a |
| 810110 Tungsten powders | 27% | 27% | 27% | 27% | 27% | 27% | 27% | 27% | 27% | 27% |
| 260600 Aluminum ores | 71% | 71% | 71% | 71% | 71% | 71% | 71% | 71% | 71% | 71% |
SELECT iso3, product_code, world_top1 * 100 AS dep_pct, imp_usd
FROM eu_member_imports x JOIN world_top1_share w USING (product_code)
WHERE product_code IN ('260500','282520','283691','284690',...,'260600');Scroll horizontally to view the full chart.
Scroll horizontally to view the full chart.
Scroll horizontally to view the full chart.
Scroll horizontally to view the full chart.
Scroll horizontally to view the full chart.
Scroll horizontally to view the full chart.
Scroll horizontally to view the full chart.
WITH ctry AS (SELECT iso3, MIN(code) AS code FROM 'countries.parquet' GROUP BY iso3),
cyp AS (
SELECT country_code, product_code, import_value
FROM 'country_year_product/year=2024/*.parquet'
WHERE import_value > 0
AND product_code IN ('282520','283691','260500','282200','260200','260400','750110','250410')
)
SELECT ct.iso3, cyp.product_code, SUM(cyp.import_value)*1000 AS imp_usd
FROM cyp JOIN ctry ct ON ct.code = cyp.country_code
WHERE ct.iso3 IN ('CHN','KOR','JPN','USA','DEU','HUN','POL')
GROUP BY ct.iso3, cyp.product_code;WITH ctry AS (SELECT iso3, MIN(code) AS code FROM 'countries.parquet' GROUP BY iso3),
totimp AS (SELECT country_code, SUM(import_value) AS total
FROM 'country_year_product/year=2024/*.parquet'
WHERE import_value > 0 GROUP BY country_code),
cmimp AS (SELECT country_code, SUM(import_value) AS cm
FROM 'country_year_product/year=2024/*.parquet'
WHERE import_value > 0 AND product_code IN (basket_20_hs6)
GROUP BY country_code)
SELECT ct.iso3, cm.cm * 1000 AS cm_imp, t.total * 1000 AS total_imp
FROM cmimp cm JOIN totimp t ON t.country_code = cm.country_code
JOIN ctry ct ON ct.code = cm.country_code
WHERE t.total > 1000000
ORDER BY cm_imp DESC LIMIT 10;