At the recently-concluded India AI Impact Summit in New Delhi this week, CEOs Sam Altman, Sundar Pichai, Demis Hassabis and Dario Amodei took centre stage. The spotlight, though, was on the same stack: AI models, chips and data centres, the machinery behind ChatGPT, Gemini, DeepSeek and Claude, typically powered by Nvidia GPUs. But AI needs a physical system. It must store data, move parts with microscopic precision and cool power-hungry servers. That is where Rare Earth Elements (REEs) enter the scene, mainly through permanent magnets, which retain magnetism without continuous external power. Neodymium magnets also sit inside the high-efficiency motors that keep data centres running, powering fans, pumps and compressors that move air and chilled water/coolant with precision. As AI is increasingly embedded in robots and automation, this magnet dependence only grows. The same rare earth magnets sit inside EV motors (from BYD, Tesla and many other mass-market EV makers), wind turbines from manufacturers such as Siemens Gamesa and Vestas and industrial robots from ABB and KUKA. They are also in defence systems such as radar and guidance components.
REEs may not be visible to consumers like gold or silver, but they are embedded deep inside modern machinery. The oddity is economic: Global rare earth output is large in tonnage at 390,000 tonnes (USGS), yet small in headline market value at less than $7 billion (industry estimates), because the real leverage is strategic, not size. Production is heavily concentrated. China dominates rare earth mining. But its stronger position lies in processing: Various estimates put China’s share of global separation/refining capacity at 90 per cent, which is why it sits at the choke point between ore and magnets. When export controls were tightened by the country last year, it showed how fast supply chains can choke, sending automobile companies scrambling for supplies.
That is one of the many reasons why countries are building mine-to-magnet capacity at home using stakes, funding and offtakes, and why India (the world’s third-largest reserves of rare earths at 6.9 million tonnes), is pushing a critical minerals mission to avoid staying only a buyer in someone else’s supply system. Governments around the world are no longer treating rare earths like a normal commodity. In the US, policy has turned openly strategic, including under Donald Trump, with support for domestic mines, stockpiles and overseas sourcing to reduce dependence on China. The race is also widening the map and the capital pool. Greenland is often cited for large deposits. Kenya’s Mrima Hill, reportedly holding over 40 million tonnes of rare earth-bearing ore resources, has made headlines. Ukraine features in the broader critical-minerals narrative. Hence, rare earths as a topic keep resurfacing in policy and boardrooms.
First, the basics.
What are they
REEs are a group of 17 chemical elements. They include 15 closely-related metals called lanthanides, such as neodymium and dysprosium, which are used to make high-strength magnets. Scandium and yttrium are also grouped together with lanthanides because they occur in the same deposits and share similar chemical behaviour. The ‘rare’ in rare earth metals, however, is a sort of a misnomer. “Rare earth elements is a terrible name because they’re everywhere; they’re not rare at all,” as US Energy Secretary Chris Wright puts it. In fact, cerium is roughly as abundant as, say, copper, in the Earth’s crust. What makes them economically rare is that they’re dispersed in low concentrations (often occur at just a few grams per tonne of rock). Plus, they are chemically difficult to separate, as sorting often involves hundreds of stages to isolate individual rare earths.
Historically, REEs were first discovered in 1787 in a small village called Ytterby in Sweden, where miners found an unusual black mineral that led to the identification of several new metals over the next century. For much of the 20th century, rare earth production was led by the US, especially from the Mountain Pass mine in California. The desert open-pit rare-earth mine in California’s Mojave was at one point supplying 70 per cent of global rare earths until the early 1980s. From the 1980s, China under Deng Xiaoping’s reform-era leadership invested heavily in mining and, more importantly, in the complex chemical processing needed to separate these metals. From the 1990s, US production kept on declining too. Over time, lower costs, including cheap labour and relaxed environmental standards, and State support helped China build dominance across the rare earth supply chain. This turned what was once a niche scientific curiosity into a strategically-important industrial resource today.
Make no mistake, not all rare earths are equally valuable. Rare earths rarely trade as pure metals; they are typically processed and priced first as oxides, which are then converted into metals and magnet alloys. Recent benchmark quotes show lanthanum oxide selling in the low single-digit dollars per kg, while neodymium oxide trades in the low-$100s per kg. Dysprosium, used in smaller volumes but critical for high-temperature magnets, trades at a large premium because it’s heavy rare earth. This wide dispersion in value is important. A handful of elements such as neodymium, praseodymium and dysprosium drive most of the industry’s economics, largely because they are used to make permanent magnets.
But these metals do not come out of the ground as neat, tradable powders. They start as messy rocks and sands, and the real story begins with where they are found and how they are separated.
Decoding value chain
In rare earths, mining gets the headlines, separation is the moat but magnets are the prize. To understand where the money is made, it helps to follow the rare earth metals value chain.
Mining is only the first step. Rare earth ores such as bastnäsite, monazite or ion-adsorption clays are extracted from the ground. These ores are then concentrated and chemically processed to produce rare earth oxides (REOs). Processing can cost several times more than mining. The oxides are refined further into metals or alloys. Those alloys are then used to manufacture high-strength permanent magnets. Finally, the magnets are built into motors, turbines and electronic components.
At each stage, value increases. A tonne of ore has relatively low economic value. But a back-of-the-envelope calculation show a tonne of separated neodymium oxide can be roughly 3,500 times the value of a tonne of ore. Finished Neodymium-Iron-Boron magnet material typically sits around 1,600-2,200 times the ore value and can be higher for premium grades. According to US Geological Survey data, magnets are the largest rare earth end-use globally, followed by catalysts and polishing; other applications span metallurgy/alloys, ceramics, glass, phosphors and pigments. This explains why control over processing and magnet manufacturing carries more economic weight than mining alone. The chemistry involved in separating rare earths is complex, capital intensive and environmentally sensitive. Solvent extraction plants require technical expertise built over decades. That processing capability has become the real bottleneck in the supply chain.
China controls this bottleneck. Of the total global rare earth production, China accounts for roughly 270,000 tonnes. The US is at 51,000 tonnes. Australia is at 29,000 tonnes, while India produces just about 2,900 tonnes. The headline number often quoted is China’s share of mining, but its stronger position lies in refining and downstream conversion. China controls the overwhelming majority of rare earth separation capacity and magnet production. Even when REO is mined in other countries, it is frequently shipped to China for processing. This dominance did not appear overnight. Supply risk quickly translates into price volatility because the market is relatively small and specialised.
In response, other countries have begun building alternative supply chains.
The US has moved beyond policy statements into direct capital deployment. In 2025, the US Department of Defense (DoD) invested $400 million in MP Materials, alongside a $150-million DoD loan to expand heavy rare-earth separation at Mountain Pass and plans tied to scaling magnet capacity. In 2026, the US also agreed to take a 10 per cent stake in USA Rare Earth as part of a reported $1.6-billion debt-and-equity package.
Japan has used its State-backed financing ecosystem to lock in non-China supply. State entity JOGMEC and conglomerate Sojitz have repeatedly supported Lynas (Australia) to secure rare earth supply to Japanese industry, including equity-style support routed through the Japan-Australia Rare Earths (JARE) structure for Lynas’ expansion and heavy rare-earth capability.
Australia’s role is anchored by Lynas Rare Earths, widely cited as the largest producer outside China. It is increasingly linked into allied supply chains via Japan (Sojitz/JOGMEC) and the US.
Europe’s approach is a mix of regulation, project fast-tracking and industrial capacity. The EU Critical Raw Materials Act is being operationalised via a first list of 47 “Strategic Projects” intended to accelerate permitting and improve access to finance across extraction, processing and recycling.
Canada is explicitly building midstream capability rather than only mining. The Saskatchewan Research Council (SRC) has been scaling a rare-earth processing and metals capability in Saskatoon (government-backed), aiming to produce NdPr (Neodymium-Praseodymium) metals at high purity and ramp monthly output.
Brazil is emerging as a key “outside-China” source, particularly for heavy rare-earth-rich output, with explicit US backing. In February 2026, a $565-million financing from the US International Development Finance Corporation to Serra Verde included an option for the US to take a minority equity stake; Serra Verde achieved commercial production in 2024.
India’s position
India sits in a paradoxical spot in the rare earths supply chain (see table).
It has meaningful geology, but limited industrial depth. India holds about 6 per cent of global rare earth reserves, with deposits concentrated in shoreline areas of Kerala, Tamil Nadu, Odisha, Andhra Pradesh, Maharashtra and Gujarat, and exploration activity also visible in Rajasthan. Yet, India contributes less than 1 per cent of global REE mining. The import dependence is even sharper in magnets: India imports an estimated 80-90 per cent of magnets and related materials from China. Demand is rising fast. India’s consumption of rare earth permanent magnets is expected to double by 2030.
Policy response has accelerated in recent times, with magnets as the near-term focal point. In November 2025, the Union Cabinet approved a ₹7,300-crore Rare Earth Permanent Magnets (REPM) scheme. Under the scheme, selected manufacturers would receive capital- and sales-linked incentives to produce 6,000 tonnes a year of permanent magnets within seven years. On February 19, 2026, the Mining Minister was quoted as saying that India aims to start producing REPM by 2026-end in partnership with the private sector; the Ministry and a State-run body have developed the technology, and plans include setting up four critical mineral processing plants across as many States. In the recent Budget speech, Finance Minister Nirmala Sitharaman proposed special Rare Earth Corridors in mineral-rich states like Odisha, Kerala, Andhra Pradesh and Tamil Nadu. On February 21, India and Brazil signed/announced a total of ten agreements and MoUs, including ones on critical minerals and rare earths. An effective alliance with Brazil will reduce India’s dependence on supplies from China. “The agreement on critical minerals and rare earths is a decisive step towards building resilient supply chains,” Prime Minister Narendra Modi said.
Magnets cannot be ring-fenced from mining, separation and metallurgy. The National Critical Mineral Mission (NCMM) is meant to push the full chain — accelerating exploration and processing, maintaining stockpiles and strengthening value chains from mining to recycling. The mission’s exploration push runs into the scale problem. Resources need to be proven into mineable reserves, and India is still early in that conversion. The Indian Bureau of Mines (IBM) has compiled data showing over 29 lakh tonnes of estimated REE resources, with Maharashtra alone at 41 per cent (over 12 lakh tonnes) and Gujarat next (11 lakh tonnes). IBM also notes India has over 58 crore tonnes of REE ore (earth/waste rock from which REEs must be extracted), but extraction at scale is yet to begin and none of it has been converted into a mining block.
Exploration and supply-side experiments are widening. All these signal a push to extract value from mining wastes.
On the industrial side, miniratna company Indian Rare Earths Ltd (IREL) remains the primary government-owned producer and processor of rare earth compounds. It is linked to downstream capability-building including an upcoming facility to produce samarium-cobalt magnets.
Mapping the REE stock Landscape
India currently has no pure-play, listed rare-earth miner/refiner. Globally, rare earths investing is really a supply-chain bet, not just a mining bet (see table).
Only a few names have meaningful operating revenue; several “Western” listings are still developers or early-scale processors, so outcomes hinge more on commissioning, recoveries, unit costs, funding runway and offtake terms than on spot rare-earth prices. P/E may not be a useful yardstick across the set because most are loss-making; even where it exists, it can be distorted by one-offs or small profit bases.
At one end are producers trying to prove “ore to oxide” scale outside China. MP Materials is the most direct US-listed pathway, anchored by Mountain Pass and a push into downstream magnets. Lynas plays a similar role as one of the few meaningful rare-earth producers outside China. Iluka’s angle is future processing capacity via the Eneabba refinery, while Arafura and American Rare Earths sit earlier on the curve, with execution and permitting risk dominating the return profile.
A second bucket is “processing, metals and magnets”. Neo Performance Materials offers exposure to value-added materials that sit closer to end-use customers. Australian Strategic Materials is positioned as a mine-to-metals story with an operating metals plant footprint in Korea, while Ucore is effectively a technology-and-processing call where separation capability is the key variable. JL MAG is the downstream proxy.
China-listed names represent incumbency and scale. China Northern Rare Earth and Shenghe sit closer to the upstream and separation backbone, benefitting from integrated capacity and domestic demand.
Takeaway: The stocks discussed above are not recommendations. They are mere examples of the REE ecosystem and investors are expected to track their financials as the sector gains more prominence. Assess REE stocks by where they sit in the chain (mine, separation, metals, magnets) and how proven the asset is (operating vs development). Track execution milestones over narratives: Commissioning progress, recoveries, unit costs, offtake terms, funding runway and balance-sheet resilience. Watch out for new entrants.
Published on February 21, 2026
