By Executive Director Abigail Hunter, Policy Manager Zubeyde Oysul, and Policy Analyst Will Kinsman of SAFE’s Center for Critical Minerals Strategy
In June, the G7 announced aggressive new targets to reduce reliance on non-G7 sources of rare earth elements. This move came amid a broader wave of national and international initiatives designed to address dependence on Chinese mineral supply chains and the common objective of improving supply-chain resilience, including the growing use of stockpiling and strategic reserve tools.
At the same time, various terminology for minerals programs covering commodities, derivatives, and downstream inputs along the critical minerals value chain have blurred the distinction between the two tools. Recent analysis from SAFE’s Center for Critical Minerals Strategy assessed Project Vault, the U.S. government-backed stockpiling initiative, to understand how it may function and key challenges that policymakers will likely face during its implementation.
This article seeks to:
- Distinguish stockpiles from strategic reserves at an international level, focusing on recent efforts among G7 and partner nations
- Examine the policy problems each mechanism seeks to solve
- Compare how major national models target different parts of the minerals value chain
- Assess historic stockpiling initiatives to identify lessons for future policy design.
Current G7+ Stockpiles and Reserves
As discussed in SAFE’s recent Fuse post, strategic reserves may employ stockpiling alongside financial tools such as contracts for difference, market commitments, forward contracts, offtake agreements, price floors, and market-maker services. These instruments give governments flexibility to support both upstream and midstream projects. Canada’s National Minerals Stockpile pilot illustrates this distinction. Rather than building a long-term emergency inventory, it supports strategic domestic projects through offtake arrangements that help guarantee demand and improve the bankability of upstream development projects. Australia’s Strategic Critical Minerals Reserve also focuses on upstream development, while retaining physical stockpiling authority as a potential contingency where required.
Some countries already combine stockpiling and reserve-like tools without labeling them as strategic reserves. Japan’s Rare Metals Stockpiling System (RMSS) is formally a stockpile but sits within JOGMEC, a quasi-government institution that can support overseas extraction and processing projects backed by financing from Japanese businesses. In the United States, several legislative proposals have been offered to establish a dedicated strategic minerals reserve, but even without a dedicated mechanism, existing tools may be available to U.S. policy makers. For example, the U.S. Department of War has demonstrated its use of purchase commitments under its Defense Production Act Title III authorities to provide offtake agreements to critical minerals and materials projects. It’s of note, however, that the DOW has only used its authorities to offer an offtake agreement to one project.
Figure 1. Comparison of Select Active Stockpiling & Strategic Reserves at the National Level
Where in the Value Chain are Governments Targeting?
Different tools commonly employed by strategic reserve mechanisms can address different stages of the minerals value chain under various contexts. Historic stockpiles such as the NDS, Japan’s Rare Metals Stockpiling System, and South Korea’s stockpiling system have focused on ensuring that domestic downstream manufacturers and end-users can access key inputs during disruptions. By contrast, Australia’s Critical Minerals Strategic Reserve and Canada’s pilot under its government’s Defence Production Act (DPA) authorities focus on supporting domestic extraction and processing. These models reflect the interests of both countries in leveraging their mature mining industries to support domestic economic growth while serving as reliable sources of upstream supply to partners seeking to diversify away from China.
Material selection also follows the value-chain objective. Australia’s reserve will initially target minerals such as antimony, gallium, and rare earth elements, where Australia has an active project pipeline and strong national security or economic interests. The United States’ Project Vault, by comparison, focuses on supply chain sourcing for participating OEMs to ensure uninterrupted access to required materials inputs. Recognizing that most participating OEMs tend to be users of refined and semi-finished critical mineral and material forms, Project Vault will generally target processed and refined mineral products to preserve the flexibility to reallocate stockpiled materials across industry if an individual participant no longer needs it. In each case, the market segment a government wants to support informs what specific materials will be the focus.
Emergency vs. Non-Emergency Use Cases: A look at the U.S., Japan, and South Korea
Beyond the choice of what to stockpile, mechanisms also differ in whether they are designed for emergencies, normal market conditions, or both, a distinction that largely reflects the unique, often historic challenges each country was originally trying to solve. Japan’s Rare Metals Stockpiling System (RMSS) was built to ensure continuity of supply for domestic industry in the event of a hypothetical disruption, while the U.S. National Defense Stockpile (NDS) was designed around a narrower objective: sustaining U.S. industrial capacity during wartime engagement and maintaining military readiness outside of wartime. By contrast, South Korea’s non-ferrous metal and rare metals stockpiling programs grew out of broader government procurement functions, but they possess authorization to respond to emergency and supply disruption scenarios as well. As G7 policymakers have increasingly focused on China’s ability to disrupt critical minerals supply chains, they have tried to reorient these historically distinct mechanisms toward that shared threat, with mixed success.
Different Origins
The NDS was established as an emergency measure ahead of the United States’ entry into World War II, and its mandate was extended in 1946 to cover long-term wartime planning during peacetime as well; today it is managed against a list of priority materials maintained independently by the Defense Logistics Agency, though it overlaps significantly with the USGS and DOE critical materials lists. The stockpile contracted sharply after the Cold War as threat assessments changed, but the 2026 National Defense Strategy’s shift from deterrence toward prolonged engagement is again raising questions about whether its material focus needs to evolve.
Japan’s RMSS has a different lineage. Established in 1983 after OPEC oil embargo and the broader commodities crisis in the 1970s, and modeled on Japan’s National Oil Stockpile, it is governed by Japan’s Ministry of Economy, Trade and Industry (METI) through a list of 34 target minerals, with a goal of sustaining the economy for 60 days under most disruptions and 180 days for higher-risk materials like certain rare earths.
South Korea’s system emerged even earlier and more incrementally. The Public Procurement Service (PPS) dates to 1949, originally managing foreign aid supplies before its mandate expanded in 1961 to domestic procurement and in 1967 to economic stabilization, including stockpiling base metals like nickel, copper, zinc, and lead. Korea did not launch a dedicated rare metals program until 2007, when the Korea Mine Rehabilitation and Mineral Resources Corporation (KOMIR) was created. It gradually absorbed PPS’s rare metals stockpiling functions, leaving PPS to manage non-ferrous metals.
The Stockpiling Tools Today
The U.S. National Defense Stockpile’s stricter, wartime-focused mandate has partially limited its use in responding to China’s grip on global mineral supply chains. While NDS materials can be released on demand, the Pentagon and Congress carry out decisions regarding the replenishment or expansion of strategic stocks as part of larger, long-term security assessments. This contrasts with Japan and South Korea’s broader stockpile mandates, which focus on economic stabilization and preparing for disruptions, allowing them to more regularly stockpile needed materials.
For example, Japan’s JOGMEC, which manages the country’s RMSS, can already finance and support overseas extraction and processing projects involving Japanese firms, with recent proposals seeking to expand these authorities. Korea Resources Corporation (KORES) historically had similar powers when it managed South Korea’s rare metal stockpile, but they were revoked in 2016 after loss-making investments led to corruption allegations. In February 2026, however, the government announced plans to restore these overseas investment authorities to KORES’s successor KOMIR, which now operates the country’s rare metals stockpile.
Figure 2. Spectrum of Policy Objectives for Stockpiles and Strategic Reserve Mechanisms
Core Stockpile Design Choices
In designing a stockpile or strategic reserve, policymakers must contend with several design choices that often include trade-offs. To some degree, the overall policy focus of a stockpile or strategic reserve will dictate certain design choices. Key considerations for policy include how much to rely on physical stockpiling, who owns stockpiled materials and what rights other players have to them, which materials to target, what form those materials should be stored in, where they should be stored, how acquisitions and releases should occur, who should manage the mechanism, and overall costs associated with stockpiling.
Whose Stockpile Is It?
Policymakers should clearly define and identify ownership of stockpiled materials and access rights, as ambiguity can undermine effective allocation of resources. Government-led stockpiles, for example, should specify whether materials are primarily intended to support the realization of diverse government policy objectives, or the material needs of industry. In defense or emergency scenarios, governments may restrict releases to specific manufacturers or uses, but this can result in government priorities overriding broader industrial needs during major supply disruptions.
Several models attempt to mitigate this risk. Japan’s approach allows material releases based on users demonstrating an inability to secure supplies and depletion of existing inventories, rather than requiring an official emergency declaration. Project Vault offers another solution by limiting the government’s role largely to funding while assigning the private sector primary responsibility for designing, implementing, and operating the stockpile.
With most countries focusing their strategic stockpile initiatives to counter China’s supply chain dominance, policymakers must also consider how future material releases should be managed, both in response to emergencies and when excess materials are disposed of outside emergency scenarios. Another consideration is whether, and to what extent, disposals should be structured to prevent materials from reaching potential adversaries and instead restricting sales to domestic manufacturers with demonstrated material needs. If sales are conducted without strict conditions on recipients, there is a greater risk that materials could be acquired by intermediaries and then sold to industries in adversary nations.
A participant-led model would mitigate the risk of material making its way into the hands of adversaries. Yet, it does not fully eliminate the risk of misaligned priorities. For example, focusing on large OEMs narrows representation of private-sector interests and may underrepresent the needs of smaller firms. By contrast, South Korea’s metal and non-metal stockpiles managed by the Public Procurement Service explicitly prioritize material releases to domestic SMEs, recognizing their lack of experience in handling overseas procurements, and irregularity and small volume of materials purchases compared to larger enterprises.
Stockpiling Vs. Midstream Investments: Policymakers must also determine when minerals stockpiling or direct investment in infrastructure—particularly smelters and refineries—is the more effective means of advancing economic and national security objectives. The tradeoff depends in part on the size of the market and the volume of material required. For minerals with relatively small market volumes, such as gallium or germanium, stockpiling may be more cost-effective than building new processing capacity given the relative low volume required to fulfill demand. For high-volume materials such as aluminum and copper, however, maintaining inventories large enough to provide comparable supply security can become prohibitively expensive. Despite the significant project costs of stockpiling major metals like aluminum and copper, initiatives like Project Vault have yet to restrict what minerals can be stockpiled.
For minerals like aluminum and copper, investments in productive capacity may provide greater and more durable supply security. SAFE’s prior analysis demonstrated how investments in U.S. aluminum smelting capacity could potentially secure a much larger and more durable share of future supply than a stockpile covering less than a year of demand.
However, infrastructure investments do not eliminate supply chain risks. Governments must still contend with China’s ability to influence global markets. Because new refining projects require years to develop, privately financed facilities may become uneconomic before becoming operational. For example, Chinese smelting overcapacity has undermined the economics of copper smelting, making new investments unlikely without additional remedial policy action. Further, where smelting economics do not necessarily prohibit new investments, policy makers must contend with the risk that falling prices would curtail production. In most instances, smelting facilities cannot be idled easily, without incurring significant costs, raising the potential cost of investing in a smelter where policies fail to address underlying market dynamics.
Sourcing, Acquisition, and Storage: Sourcing decisions can require difficult trade-offs. Governments seeking immediately usable inputs may have to rely on supply from China in the near term, while efforts to build stockpiles during periods of limited availability or supply shortages could compete with existing industrial demand and further tighten markets. Governments will also face tradeoffs in terms of being able to source materials where availability of supply outside of domestic and allied sources is limited or unavailable. Designing stockpiling programs also means aligning sourcing rules with existing and emerging trade restrictions that serve other, sometimes separate, policy goals.
A prime example is the Pentagon’s DFARS (Defense Federal Acquisition Regulation Supplement) restrictions, which bar defense materials sourced or produced in China, Russia, North Korea, or Iran. This covers rare earth magnets and certain refractory metal alloys and steel where downstream steps like melting or finishing happen in a covered country. Starting January 1, 2027, the restriction expands to prohibit any production step occurring in a covered country. With China dominating virtually every stage of the value chains for rare earth magnets and other minerals such as tungsten implicated by DFARS restrictions, sourcing from China will likely remain necessary until ex-China capacity comes online.
In response, the Trump Administration’s July Executive Order 14415 tightened enforcement by limiting DFARS waiver exemptions, while also creating an alternative path through Project Vault: materials acquired via Vault and transferred to industry won’t be classified as defense articles, effectively exempting China-sourced DFARS materials from normal restrictions when routed through Vault. But such exemptions only help if supply exists. China already restricts exports of many DFARS-relevant minerals, including those used in NdFeB magnets, samarium-cobalt magnets, and refractory alloys. Further tightening of export restrictions by China could further undercut the effectiveness of new stockpiling efforts.
Stockpiling faces market challenges too. Where China’s export controls or other distortions constrain supply, there may simply not be enough material available to stockpile outside of China, and where it is available, stockpiles may need to be designed to be able to source from jurisdictions that would otherwise be difficult for private sector consumers to access, making tools that encourage new supply potentially more valuable than stockpiling.
Reserve models focused on domestic upstream supply, such as Canada’s and Australia’s, partly avoid this problem by directing support toward preferred sources. Japan’s model also reduces sourcing risk because JOGMEC can support overseas projects that may later supply the stockpile. Project Vault faces a different challenge: because it is driven by private manufacturers’ needs, the required minerals and forms may vary widely, and many refined or intermediate products remain available only in limited quantities outside China.
Storage decisions also matter. Governments must choose whether to rely on centralized stockpiling hubs or decentralized warehouse networks, and decide whether to locate inventories near ports, rail hubs, industrial users, or military installations. Centralized warehouses may be more efficient, but they can also become security risks in conflict or cyber scenarios, particularly materials requiring storage under controlled conditions. Governments must also decide whether target materials for stockpiling should be identified by government, industry, or both as well as whether to stockpile ores, concentrates, refined products, or downstream-ready inputs.
Transparency: Transparency can improve predictability but may expose vulnerabilities. Since 1979, the U.S. Defense Logistics Agency has published annual materials plans for the NDS, partly to reassure industry that disposals will not destabilize markets. Yet public tenders, auctions, and sales can potentially reveal supply chain weaknesses to adversaries. JOGMEC generally does not disclose RMSS acquisitions or releases, reducing visibility into Japan’s vulnerabilities but increasing the importance of coordinating internationally to ensure that stockpile releases do not harm the market. Designers must also implement safeguards alongside strategic reserve and stockpile models that mitigate panic buying, insider-trading risks, or market instability.
Comparative Governance Models
Stockpiling and strategic reserve models can best be contrasted based on ownership, management, and financing structures. Government-led systems can provide clearer oversight and alignment with public priorities, but require resources, expertise, and political accountability for market decisions. Industry-led models can draw on private-sector expertise and better reflect market needs, but provide less public control and may allow commercial incentives to diverge from policy goals. Policymakers can potentially hedge risks associated with specific models through several means, including pursuing initiatives that are based on public-private partnership models, or managed by independent/quasi-independent agencies. Policymakers may further hedge risks by pursuing dual-track stockpiling systems as can be seen in the case of the United States, Japan, and Korea.
Figure 3. Comparative Dimensions of Stockpile and Strategic Reserve Models
Key: NDS (US): National Defense Stockpile; Vault (US): Project Vault, DPA Pilot (CA) Canadian Defence Production Act national mineral stockpile pilot; CMSR (AU): Australia Critical Minerals Strategic Reserve; PPS NFMS (SK): South Korea’s Public Procurement Service-led non-ferrous metals stockpile; KOMIR RMS (SK): KOMIR-led Rare Metals Stockpile; JOGMEC RMSS (JP): JOGMEC-managed Japan National Rare Metals Stockpiling System; JRMA VRMS : Japan Rare Metals Association-coordinated voluntary rare metals stockpiling system
Implications for Multilateral Stockpiling
As G7 and partner countries continue implementing new stockpiling and reserve mechanisms, multilateral coordination remains underdeveloped. Novel mechanisms including Project Vault and Australia’s Critical Minerals Strategic Reserve target different parts of the value chain—downstream demand and upstream supply, respectively—underscoring the need to align sourcing, customers, and release rules between partners. The G7’s new rare earth targets emphasize the fact that upstream projects in Australia and Canada, midstream refining capacity in Japan and South Korea, and downstream manufacturing in the United States and Europe will all be needed to meet shared diversification goals.
Economic stockpiles already require close engagement with domestic industry to ensure governments acquire materials that companies can actually use, avoid competing unnecessarily with private-sector demand, and release inventories without distorting markets. Coordinating these decisions across multiple countries, with different industries, supply vulnerabilities, and strategic priorities, would add another layer of complexity. A more immediate priority is therefore ensuring that G7 and partner governments have visibility into one another’s stockpiling activities and market conditions so that national actions do not inadvertently work at cross-purposes.
The G7 has begun to respond. At the Evian Summit in June, leaders announced steps to strengthen stockpiling coordination, including an information- and data-sharing platform administered by the International Energy Agency’s (IEA) Critical Minerals Security Program. The G7’s communique highlights soliciting JOGMEC’s expertise, in particular.
These efforts are important first steps, but they also raise design questions. Without coordination, national stockpiles and reserves could compete for scarce supply, and poorly timed releases could undercut efforts to build upstream extraction or midstream processing capacity. The IEA’s experience coordinating emergency oil releases shows that joint release mechanisms are possible, but minerals will be harder to manage across potentially overlapping voluntary G7, EU, and NATO initiatives.
The effectiveness of information sharing will depend in part on which countries participate in information sharing. Many countries with important mineral resources or midstream capacity—including South Korea and Australia—sit outside the G7, NATO, and the EU, while smaller developed and emerging economies may pursue joint stockpiling arrangements to pool demand and increase purchasing power. Without visibility into these parallel efforts, governments could unknowingly compete for scarce supply or pursue acquisition and release strategies that work at cross-purposes, making multinational joint stockpiling a complicated proposition.
Matching the Tool to the Objective
Stockpiles and strategic reserves are related but not interchangeable tools. Stockpiles work most effectively when governments need to ensure access to immediately usable materials during supply disruptions, especially for downstream users. Strategic reserves are better suited to address persistent market distortions and support upstream or midstream project development through financial and market-shaping tools. Policymakers should match the tool to the objective by first identifying the problem they want to solve, then selecting the appropriate value-chain target, mineral scope, sourcing and storage model, governance structure, and level of allied coordination. Past stockpiling models offer useful lessons for policymakers in the design and implementation of strategic reserve and stockpiling mechanisms both today and tomorrow. At the same time, the scope and depth of Chinese control of mineral supply chains have raised the prospect that the long-term effectiveness of stockpiling and strategic reserve efforts will be determined by the interoperability of such mechanisms at the multinational level.
Twitter
