China's Share of Rare Earth Mining Has Fallen From 97% to 58%. Its Share of Refining Has Barely Moved.

China's share of global light rare earth mining has fallen from roughly 97% in 2010 to around 58% by 2024. Its share of refining has barely moved.

That single gap is the finding buried in a new IMF working paper (July 2026), and it reframes the entire Western rare earth push.

The reshoring story is usually told as a mining story — dig the ore out of the ground somewhere other than China. On that measure, it is working: mining has genuinely diversified, but we know now the mine was never the problem.

The problem sits one stage downstream, in separation and refining, where China still holds 85% to 95% of global capacity and has for decades. The paper puts around 58% of the total cost gap between Chinese and Western producers in that refining stage alone. That is where the dependency actually lives.

The pipeline makes the imbalance sharper. The same paper screens 89 publicly listed rare earth companies outside China and evaluates 21 with published feasibility studies. Yet commercial production at scale still comes down to a short list of names: Lynas Rare Earths Ltd, MP Materials, Serra Verde Group.

Here is the detail that ends the argument: Serra Verde, the newest of them, mines in Brazil and ships its concentrate to China to be separated, under an arrangement running to the end of 2026.

Mined in the West. Refined in the East.

New projects outside China face 10 to 15-year lead times before they produce anything, on the IMF's own estimate. So the gap does not close quickly, and it does not close by drilling.

The shortage you can price is upstream. The dependency you cannot escape is downstream. For anyone underwriting this theme, those are two different trades, and most of the capital is chasing the first one.

Seeing which projects actually sit where, and which are close to changing the picture, is a data problem before it is a capital one.

Less searching. More strategising.™