The Critical-Minerals Midstream Trap: Why New Mines Do Not Equal Supply Security
By Robert Tanaka, Head of Operations & Supply Chain Research
For a decade, mineral supply security has been measured in mines. Governments have subsidised extraction in allied jurisdictions and counted each new pit as diversification. Yet the International Energy Agency (IEA) reports, in its Global Critical Minerals Outlook 2026, that the average market share of the leading refining country — rare earths excluded — increased to 72% in 2025.
That figure reframes the question. Supply security is a property of the complete chain — extraction, refining, component manufacture, processing equipment and skilled operation — and concentration is highest at the stages hardest to replicate. The defining decision for boards and ministries is no longer where to dig, but which midstream chokepoints justify paying a premium to route around.
The Numbers Point the Wrong Way
Four IEA findings, read together, describe a midstream trap: diversification of the cheapest stage while the real chokepoints stay intact.
- Refining concentration is rising, not falling. The 72% share is an increase, recorded after a decade of friendshoring. In our reading, mine-count was the wrong security metric all along.
- Growth compounds the problem. The IEA finds dominant suppliers accounted for more than three-quarters of recent refined-supply growth; absent intervention, 72% trends higher.
- The pipeline is mismatched. The IEA finds planned refining and component capacity outside dominant suppliers is not keeping pace with announced mining projects — a pipeline forecast, not built capacity, but it makes the trap explicit: diversified ore reverts to incumbent processors.
- Capital is retreating at the wrong moment. The IEA records a 9% decline in critical-mineral investment in 2025 — first-cut data, subject to revision. Price weakness in lithium, nickel and cobalt, driven partly by incumbent oversupply, is rationing the higher-cost diversified projects security policy needs.
Why the Midstream Resists Diversification
The binding constraint is economics, not geology. A World Bank study of smelting and refining documents why processing is structurally harder to diversify than extraction: high capital intensity and scale thresholds, heavy energy and environmental cost burdens, thin, volatile treatment margins, and dependence on specialised equipment, licensed process intellectual property and scarce metallurgical skills.
The same analysis is the honest counterweight to reflexive onshoring. A smelter without competitive energy, credible offtake and a trained workforce is a stranded asset, not a security asset; the World Bank's economics imply some producing countries capture more value through beneficiation partnerships or equity in offshore processing than through subscale national smelters.
The Chain Is Longer Than It Looks
Even with nominally diversified refined output, the dependency chain runs further: component manufacture (cathodes, anodes, magnets), the few equipment houses and process licensors behind each flowsheet, and the metallurgists who keep plants running. Ownership matters too — in our reading, a third-country facility owned by a dominant supplier, or entangled in its sanctions exposure, hollows out apparent geographic diversity. Our supply-chain rewiring analysis argued resilience requires mapping full network topology, not first-tier geography; the mineral midstream is where that lens bites hardest.
| Chain stage | Concentration signal | Principal barrier | Diversification difficulty |
|---|---|---|---|
| Extraction (mining) | Most geographically dispersed stage | Permitting timelines, price cycles | Moderate |
| Refining / chemical processing | Leading country averaged 72% share in 2025, ex-rare-earths (IEA) | Scale economics, energy cost, incumbent price power | High |
| Component manufacture | Ex-dominant-supplier pipeline lagging announced mines (IEA) | OEM qualification cycles, integrated ecosystems | High |
| Processing equipment and IP | Few licensors and equipment houses per flowsheet | Licensed IP, proprietary process control | Very high |
| Skilled operation and metallurgy | Concentrated where plants have run for decades | Tacit knowledge; no capex shortcut | Very high |
Difficulty ratings are the World Research Institute's interpretation; concentration descriptions beyond the IEA's 72% figure are qualitative.
Policy Is Buying the Easy Stage
Restriction risk is accumulating where the capacity sits. The OECD's inventory of export restrictions on critical raw materials documents a fivefold increase in restrictions since 2009 — an observed count of measures, though each measure's impact is an estimate. Set that rise against a 72%-concentrated refining stage and the exposure is worse than either number alone suggests.
Yet public money still flows disproportionately to mine announcements. The pattern echoes what our semiconductor industrial-policy scorecard found: governments can buy capacity announcements far more easily than competitive, complete chains. The instruments that address the IEA's pipeline mismatch are different — bankable long-term offtake with floor prices, stockpiles of refined chokepoint products rather than ores, processing-skills programmes and shared infrastructure — reserved for projects that pass the World Bank's commercial tests.
What Could Rewrite the Map
- Restrictions moving up the chain. Export controls extending from raw materials to processing technology, equipment and know-how would shut the hardest-to-diversify stages. Watch the next OECD inventory update and any licensing regime touching refining flowsheets.
- Price-suppression attrition. With investment already down 9%, sustained low refined-product pricing could cull the non-dominant pipeline before it commissions. Watch final investment decisions and cancellations over the next 12–18 months.
- Substitution, thrifting and recycling. Chemistry shifts and scaled recycling could re-rank which minerals deserve a security premium, stranding midstream bets made on today's exposure map. We frame this as uncertainty, not prediction.
What This Means for Decision-Makers
- Audit the chain, not the map. Score exposure at every stage, including processor ownership and sanctions entanglement of nominally third-country facilities. A mine in a friendly jurisdiction feeding a dominant-supplier-owned refinery is undiversified supply wearing a diversified flag.
- Price the security premium explicitly. Where a non-dominant processing route costs more, the delta is an insurable-risk price — rational for chokepoint-exposed inputs, irrational as blanket policy. Boards and ministries should decide which minerals warrant it, and cap it.
- Shift support from mine announcements to midstream bankability. Offtake guarantees, refined-product stockpiles and skills pipelines address the mismatch; per-mine subsidies do not.
- For producer countries, sequence rather than reflex. Domestic refining is a case-by-case investment decision, not an entitlement of geology.
Evidence and data cut-off: 1 September 2026. Third-party figures are as published by the cited sources; Logistics Index readings are the World Research Institute's own. See our Research Standards.
The World Research Institute provides supply chain risk assessment, industry analysis and commissioned market intelligence. Contact our team to commission tailored research.