China reduced its shipments of dysprosium and terbium to Japan to zero in the first half of 2026, according to customs data cited by Caixin Global. Sixteen years after the first supply cut, the Japanese case shows how costly and time-consuming it is to diversify critical supply chains when a single country controls processing.
The precedent is 2010, when Beijing briefly suspended rare earth shipments to Tokyo and Japan responded with stockpiling, public investment, and supply agreements with producers in Australia and Canada. The dependence persists: Japan remains the world's largest importer of rare earth metals and, according to Foreign Policy, experts consider that diversification took years and remains incomplete. Researcher Mireya Solís of the Brookings Institution describes it as a web strategy in which no single country resolves its supply challenge alone. China processes approximately 85% of the world's rare earths and produces 92% of magnets, a concentration that turns supply into an instrument of trade policy. For Mexico and Latin America, where critical mineral policies for resources such as lithium are under debate, the Japanese case defines the real horizon of any strategy: it cannot be resolved within a single six-year presidential term.
The current effort combines public financing and corporate partnerships. In July, Japan's state agency JOGMEC invested up to 47.7 million Canadian dollars in the Lofdal heavy rare earths project in Namibia. In March, Lynas Rare Earths and Japan Australia Rare Earths signed an agreement for 5,000 tonnes per year of neodymium-praseodymium alloy. Magnet manufacturer Shin-Etsu stopped accepting dysprosium-containing orders and is planning a new refinery, and according to consultancy Argus, Japanese manufacturers are securing barely two-thirds of the supply they need. The key figure: JOGMEC-backed projects will take one to two years to scale and new production is not expected before 2027-2028, according to Caixin.
The lesson Japan leaves for critical mineral policy is a matter of timing: a project financed today takes one to two years to scale and does not reach full production before 2027-2028. For Mexico, this reference point makes clear that building domestic processing capacity takes more than a single presidential term.
This article was written with artificial intelligence assistance from verified sources and reviewed by a human editor before publication.