What should a country hoard?

Critical minerals occupy an unusually prominent place in the global economy. Gallium is used in semiconductors and power electronics, graphite in batteries, and rare earths in the permanent magnets found in electric vehicles and wind turbines. Yet many of their supply chains are concentrated in a handful of countries. China alone accounts for more than 90% of global refining supply for gallium, graphite, manganese and rare earths, according to the International Energy Agency (IEA).

A disruption can reverberate far beyond the minerals market. The IEA’s Global Critical Minerals Outlook 2026 estimates that full implementation of rare-earth export restrictions could put $6.5 trillion in annual production outside China at risk. To reduce that exposure, governments are building reserves to keep factories running when supplies are interrupted.

While strategic oil reserves provide a well-established model for guarding against supply shocks, critical minerals are a more difficult proposition. Some deteriorate, others require specialized storage, and government purchases can influence prices in small markets. Choosing what to stockpile therefore means judging which shortages pose the greatest economic risk, how much material would be needed and what that protection is worth.

What makes a mineral worth storing?

The case for stockpiling a mineral can change surprisingly quickly. The U.S. Department of Energy’s Critical Materials Assessment 2023 examined 23 materials against expected demand for energy technologies and the risks surrounding their supply. Seven were classified as critical in the short term, including gallium, natural graphite and the rare earths neodymium, dysprosium and terbium. 

Notably, natural graphite had not been considered critical in the department’s previous assessment, while rhodium and palladium were no longer included as catalytic converters, the emissions-control devices used in conventional vehicles, declined in importance. The materials most vulnerable to shortage today are therefore not necessarily those most vulnerable several years from now.

The value of a stockpile also depends on how quickly industry could adapt if supplies disappeared. Even a mineral concentrated among a few producers may pose a manageable risk if manufacturers can readily substitute another material. The DOE, for example, classified praseodymium and neodymium, two rare-earth metals used in high-strength permanent magnets, differently in the short term, partly because praseodymium can be more readily substituted. Where substitutes are limited and new supply takes years to develop, a reserve can bridge the gap until production adjusts.

From mineral to usable material

Choosing which minerals to stockpile is only the first decision; governments must also determine what form would be useful during a shortage.

The IEA’s January 2026 analysis of strategic stockpiling argues that countries should generally hold materials in a form they can use without further foreign processing. A stock of rare-earth ore, for example, is of limited value during a disruption if domestic manufacturers require permanent magnets and the country lacks the capacity to turn the 

ore into them.

Some materials are also considerably harder to store than others. Lithium hydroxide is highly sensitive to humidity and has a shelf life of around six months, according to the IEA, while lithium carbonate can be stored much longer. Gallium, meanwhile, melts at about 30 degrees Celsius, requiring storage that accounts for relatively modest changes in temperature. These added precautions increase the cost and complexity of maintaining a reserve.

Even materials that store well may not be interchangeable from one manufacturer to another. Graphite anode material, used in lithium-ion batteries, can be tailored to the specifications of individual producers, limiting the usefulness of a single standardized government inventory. One alternative is to keep government-backed reserves with companies themselves, allowing material to pass through normal production rather than remain in a central warehouse until it deteriorates.

The price of buying time

Buying the material requires substantial upfront spending, but the purchase price overstates the long-term cost of a reserve. The minerals remain valuable assets that can eventually be used or sold, although the money committed to them cannot be spent elsewhere in the meantime.

The sums vary sharply by mineral. In a January 2026 analysis, the IEA modeled reserves sufficient to replace six months of imports from China for several high-risk materials. It estimated annual net costs of about $800,000 for gallium metal, nearly $90 million for rare-earth permanent magnets and just under $300 million for lithium hydroxide. 

Much of the difference comes from scale, since covering six months of lithium hydroxide imports requires far more material than doing the same for gallium. Across 11 high-risk materials, the IEA estimates that maintaining six-month reserves in countries dependent on a dominant foreign supplier would carry net annual costs of less than $900 million in total

Yet cost is not the only constraint. Critical-mineral markets can be relatively small, so large government purchases could themselves push up prices. Releases can distort the market too. If governments routinely sell reserves when prices rise, producers may have less incentive to invest in new supply. The IEA therefore argues that stockpiles should primarily address serious supply disruptions rather than ordinary price movements.

There is no single reserve target that works across critical minerals. The amount depends on how quickly alternative supplies can emerge, whether manufacturers can substitute another material and how long production can continue without new imports. Reserves cannot replace investment in mining, processing and recycling. Their narrower purpose is to buy time during a disruption, at a cost weighed against the economic losses that shortage could cause.