What good is a stockpile if the supply chain behind it is hollow? A missile is only the visible part of an industrial system; the hidden part is the mineral ladder beneath it. When that ladder is made of imported, concentrated, and politically exposed materials, the arsenal looks less like a fortress and more like a bridge built over a river in flood season.
The latest warning comes not from a trader’s screen but from war itself. A RUSI analysis of the Iran war estimated coalition forces expended more than 11,000 munitions in the first 16 days. Replacing the gun ammunition alone took an estimated 4,000kg of tungsten. That is a small number in mining terms and a large one in strategic terms. It is also a reminder that modern rearmament is not just about money, but about whether the right minerals can be pulled, refined, and finished fast enough to matter.
Tungsten is the bluntest example of the problem. The United States produced none of it in 2025, while China accounted for 79% of global tungsten mine production. For a country trying to rebuild war reserves, that is a poor place to stand. Dependency becomes more dangerous when a material serves two masters at once: the weapons themselves and the machines that make the weapons. Tungsten is used in armaments, turbine alloys, and metalworking tools. It sits near the start of the chain, where substitutes are thin and time is slow.
Gallium shows the same fragility in a different form. China accounted for 99% of primary low-purity gallium production in 2025. Gallium’s semiconductor uses support radar and military communications, which means the same mineral that helps a missile think also helps the system that guides it. That kind of concentration is not a market efficiency. It is a strategic choke point. In war, concentration behaves less like scale and more like a single bridge over a canyon.
An August 2025 War on the Rocks analysis divided the risk into two broad tiers. First-tier risks were gallium, the battery-chemicals chain, tungsten, and graphite. The second tier included titanium sponge, germanium, antimony, indium, magnesium, and molybdenum, along with nitrocellulose and finish-line capacity in specialty steels and aerospace aluminum. The point is not that every item is equally scarce. The point is that the weakest segment in a chain can determine the speed of the whole system. Mining is only the first gate. Refining, separation, and specialized finishing often decide whether a domestic ore body is really a domestic supply.
That is why substitution and recycling, while useful, do not erase the problem. A July 2026 Government Accountability Office assessment reported that the United States was fully import-dependent for 12 critical minerals and relied on imports for at least half its requirements for another 34. That is not a temporary inconvenience. It is an architectural fact. You can stock a warehouse, but if the upstream inputs are trapped in foreign systems, you have merely moved the vulnerability back a few steps.
Defense spending can rise fast. Mineral supply cannot. Global military expenditure reached a record US$2.887 trillion in 2025, with European spending rising 14% in real terms. That is a lot of demand chasing a supply base that takes years to expand. Markets like to pretend capital can command anything. History says otherwise. Rome learned that grain routes matter more than speeches. Britain learned that coal, shipping, and steel decide whether an empire can fight a long war. The same rule still applies: money can accelerate procurement, but it cannot instantly create mines, smelters, and processing plants.
S&P Global’s August 2026 assessment on antimony shows the mismatch clearly. Even if announced projects proceed, smelting capacity outside China could remain approximately 27% below projected demand outside China in 2030. It also warned that new processing facilities could face shortages of mine feedstock. That is the part investors often miss. They think in terms of projects. The real constraint is the chain between project and product. A mine without processing is a promise, not a supply source.
Tungsten also shows why securing supply takes more than money. The four tonnes used in the weapons fired in the first 16 days of the Iran war is small relative to the broader market; the United States imported an estimated 11,700 tonnes of tungsten in 2025. Yet even a modest wartime draw can matter when available supply is already spoken for. In July 2026, US processor Global Tungsten & Powders extended its agreement for approximately 90% of Sangdong’s first-phase production to 21 years. That tells you two things at once: new supply is welcome, and it is already heavily committed before it arrives.
China has also tightened access to its own production. In February 2025, Beijing introduced export licensing requirements for selected tungsten products. S&P Global said tungsten prices more than quadrupled between January and July in 2026. Price moves do not solve shortages; they reveal them. In a thin market, price is not a comfort. It is a flare shot into the night sky, exposing how little darkness one can actually see through.
The Pentagon is no longer speaking in abstractions. “Securing resilient domestic and allied supply chains is a fundamental warfighting requirement,” said Mike Cadenazzi, the Pentagon’s assistant secretary responsible for industrial base policy, in an August 2026 announcement committing US$750 million to secure rare-earth supply from Brazil’s Serra Verde. The language matters. Supply chains are no longer being treated as a procurement issue. They are being treated as a battlefield requirement, which is exactly what they are.
The policy moves are becoming more direct. The Pentagon announced a US$174 million investment in planned Australian gallium production in August 2026. It also made a US$750 million commitment for Serra Verde rare earths. Earlier, MP Materials’ July 2025 Pentagon partnership combined equity, a ten-year price floor for neodymium-praseodymium products, long-term magnet purchases, and financing. USA Rare Earth’s June 2026 agreement with the Commerce Department provides access to about US$1.6 billion to support the Round Top project in Texas and expand US metal and magnet production. The US International Development Finance Corporation also moved with US$600 million in the US$1.8 billion Orion Critical Mineral Consortium, then approved a further US$900 million in financing by July.
This is not charity. It is hedging against strategic rupture. A 2026 Hoover Institution policy paper, citing Stanford research, described a defence contractor tracing a titanium supply chain through 13 tiers to Chinese mines and transport infrastructure, ending in “Chinese mines, Chinese roads, and Chinese trucks.” That is the anatomy of a fragile system: long, opaque, and dependent on someone else’s political weather. A chain can be efficient and still be brittle. In game theory, the player with the shorter decision cycle often wins; in industrial policy, the country with the shorter supply chain often survives the shock.
That is why titanium, germanium, antimony, and rare earths matter in the same conversation as missiles and aircraft. Titanium gives military aerospace strong, light material. Germanium supports infrared optics and thermal-sensing equipment. Antimony trisulfide is used in ammunition primers. Magnet rare earths, including samarium, support military aircraft, missiles, and control systems. Each material fills a different slot, but all of them share one trait: the bottleneck is usually not the mine alone, but the chain from mine to metal to finished component.
The practical message is unsentimental. If the United States wants to reload missile stockpiles, it must first reload critical-mineral supply chains. From January 2027, defense procurement restrictions on covered tungsten products extend back to the mining and feedstock stages, generally excluding Chinese-origin material, subject to specified exceptions. That is the government admitting that origin matters, not just end product. It is a late but rational move. In a crisis, provenance becomes a security variable.
There is a temptation to believe stockpiles solve everything. They do not. Stockpiles are only time purchased in advance. If the replenishment system is weak, the stockpile becomes a countdown clock. That is the central fragility now facing rearmament: the West can spend more, but it cannot instantly manufacture resilience. The minerals must exist, the refiners must run, the feedstock must be secure, and the contracts must survive political stress.
Before America can reload its missile stockpiles, it has to reload the mineral systems that feed them. That is a slower, less glamorous task than announcing budgets or buying weapons. But wars are won by the side that can keep reloading after the first burst is gone.
