The same three words keep showing up in the same kind of headline. A government restricts an export, a mining stock doubles, and someone says the world is running out of what batteries and magnets are made from. The running-out part is mostly wrong. The worry underneath it is not.
What "Critical" Actually Means
In American law this is a category, not a compliment. The Energy Act of 2020 defines a critical mineral as one essential to the economic or national security of the United States, with a supply chain vulnerable to disruption. Two tests. It matters, and somebody could cut you off from it.
Scarcity is not one of them. A mineral can sit in huge quantity underground and still be critical, if the only plant that makes it usable is in a country you might quarrel with.
The US Geological Survey keeps the list, and the law requires a review at least every three years. It has grown substantially since the first list of 35 entries in 2018, and the November 2025 revision added copper and silver, the two additions most worth knowing. Entries come off as well. The label describes one country's exposure at one moment.
Three Minerals, Three Different Problems
Lithium, cobalt and rare earths get bundled into one story about batteries and China. That is where most coverage goes wrong, because the choke point sits in a different place for each. Lithium's problem is construction time. Cobalt is a leftover from mining something else. Rare earths are a chemistry problem. Buy them as one theme and you own three unrelated bets.
Lithium: The Bottleneck Is Not the Rock
Lithium is not a scarce element. It reaches the market two ways. Brine operations pump salty water from beneath salt flats into evaporation ponds, which takes many months of sunshine. Hard rock operations mine an ore called spodumene, then crush and roast it.
Neither gives you anything a battery can use. The concentrate has to be refined into high-purity lithium carbonate or hydroxide, and that happens in a chemical plant. Those plants sit in a handful of countries, and a new one means permits, engineers and years of construction. A mine takes longer still.
A carmaker, meanwhile, can revise its demand forecast in a quarter. A lithium shortage is usually a scheduling mismatch, not a geological one.
Cobalt: A Metal That Cannot Answer Its Own Price
Apart from a few small exceptions, USGS says, most cobalt is mined as a by-product of copper or nickel. Almost nobody digs for cobalt on purpose.
That breaks the usual supply response. A high price normally pulls out more production. But nobody opens a copper mine because cobalt got expensive; cobalt is a thin slice of the revenue. Output follows the copper and nickel cycle instead, so a cobalt shortage cannot cure itself the way an oil shortage can.
Most mined cobalt comes from one country, the Democratic Republic of the Congo. Some is artisanal, hand-dug by independent miners without equipment or safety rules, including children, and it enters the same refineries as industrial material. That is a human problem first. It is a business risk too: buyers who cannot document the source lose customers.
The engineering answer has been to design cobalt out. Battery makers cut the cobalt share of the cathode, then moved much of production to a chemistry built on iron and phosphate that uses none.
Rare Earths: Seventeen Elements That Refuse to Separate
They are not rare. USGS calls them a relatively abundant group of 17 elements: scandium, yttrium and the 15 lanthanides. Cerium ranks 25th among elements in the earth's crust, and even the scarcest of the group beat gold for abundance.
The trouble is that they occur together and behave almost identically in chemical reactions. You cannot melt one out. Separating them takes solvent extraction, which mixes the solution with chemicals so slightly more of one element crosses into the oil, then repeats that step hundreds of times. It is slow and dirty, and monazite ore carries thorium, so the waste is radioactive.
That is where the concentration actually is. Digging the ore is the easy half. A country can open a rare earth mine and still ship the concentrate abroad, because the separation plants are somewhere else.
The market is really only a few of the 17. Neodymium and praseodymium make the magnets in motors and generators, with dysprosium or terbium added so the magnet survives heat. The rest come out of the ground anyway, wanted or not.
Why These Markets Swing Harder Than Oil
Three things stack up. These markets are small in dollar terms, so one new plant or one halted mine moves prices further than the same event would in oil. Supply arrives slowly: IEA analysis puts well over a decade between a discovery and first output at a typical new mine, so supply cannot answer a demand spike quickly. And demand rests on one or two end uses, so a shift in battery chemistry hits everything at once.
Then there is politics. When a mineral turns strategic, export licenses and quotas are a tool governments reach for again and again, and investors price that in.
What You Can Actually Buy, and Why It Is Awkward
There is no practical futures market here for an individual. Lithium and cobalt contracts exist on major exchanges but are thin, used mostly by producers and traders. Rare earths barely trade in public, and prices come from private contracts and reporting agencies.
So exposure comes through mining companies and thematic funds, which changes what you own. A miner carries a management team, a debt load, a permit that can be denied, a mining code that can be rewritten. Explorers with no revenue behave more like lottery tickets than like metal.
Funds are narrower than they look. Open one and you often find a short list of names, many of them chemical processors or battery makers rather than miners. You can be right about the mineral and lose money on the companies.
Substitution and Thrifting Set the Ceiling
A high price is an instruction to engineers, and they answer it two ways. Thrifting means using less of the expensive input per unit. Substitution means using something else.
Both have happened here already. After rare earth prices spiked in 2010 and 2011, magnet makers cut their dysprosium content sharply, redesigned motors to need less magnet, and switched to cheaper ferrite where they could. Demand fell, and the price followed. Cobalt ran the same arc a decade later.
None of this makes the shortages fake. They are real while they last, and that can be years, long enough to make and lose serious money. But it puts a clock on any thesis resting on the world needing something it cannot get. The question is not whether demand is growing. It is which step is jammed, who is paying to unjam it, and what the engineers on the other side are designing around.
