🚨 In 66 days the Pentagon's magnet rule stops asking "where was it melted?" and starts asking "where was it mined?" 🧲 📜 Today (to 31 Dec 2026): DFARS 252.225-7052 only bars NdFeB magnets if the Nd-Fe-B alloy was melted, or the magnet made (powder → press → sinter → magnetise), in China, Russia, Iran or North Korea. ⛏️ From 1 Jan 2027: "the entire supply chain from mining of neodymium, iron, and boron through production of finished magnets." Mined, refined, separated, melted or produced in a covered country = cannot be delivered under a DoD contract. The clause flows down to every subcontract tier. 📊 The gap it lands on: US demand ~48,000 t of NdFeB in 2025 vs ~300 t domestic supply (Arthur D. Little via Reuters). Capacity on track for ~5,000 t by end-2026. ♻️ The one recycling carve-out: an NdFeB magnet made from recycled material is exempt only if the recycled material is milled AND the final magnet sintered in the United States. For recycled NdFeB processed anywhere else, the clause is silent on how the mine-origin test is applied. 🔓 Escape valves: COTS items under 50% covered material by weight, electronic devices, buys under the simplified acquisition threshold, and a written nonavailability determination if compliant magnets "cannot be procured as and when needed at a reasonable price." My read: that last one does a lot of work in 2027. 🗓️ The date has already slipped once (2026 → 2027 via FY2024 NDAA s.854), and contractors have reportedly lobbied for another delay. 👀 Bottom Line: 66 days from now, every NdFeB magnet in a US defence contract needs a paper trail back to the mine, in a market that supplied 300 t of its own 48,000 t last year. #RareEarths #NdFeB #Magnets #DFARS #CriticalMinerals #DefenseSupplyChain #Recycling #SupplyChain

🚨 Rare earths aren't rare. The factory matters more than the mine. If you want the ten-minute version of what my seven-part series is unpacking, this explainer covers the same ground, from the same sources. I've checked its numbers against USGS, the IEA and the company filings. Most hold. Three things need updating, and you should know who made it. 🧵 🎬 Why it is worth ten minutes It gets the shape right, in order: the name is wrong twice (not rare, not earths); the hard part is separating 17 chemical twins, not digging them up; one magnet type sits in almost everything that spins; China's lead is smallest at the mine and biggest at the factory; export licences turn processing into leverage; and most of the strategic metals are by-products, so you cannot simply open a gallium mine when prices jump. That is the spine of Parts 1 to 7. ✅ The numbers I checked 🌍⛏️World mine output 2025 about 390,000 t REO, China about 270,000 t, "close to 70%" — USGS says 69%. United States second at about 13% (51,000 t), then Australia (29,000 t) and Myanmar (22,000 t). Holds. 🧲Magnet metals (Nd, Pr, Dy, Tb), IEA 2024: China about 60% of mining, 91% of refining, 94% of magnets. Holds. 📍Note the two scopes: ~60% is the IEA's magnet-metal mining share; 69–70% is USGS all-REO. Both are right. My Part 1 board carries the IEA figure, Part 2 the USGS one. 🇦🇺🇺🇸"Just under 10,000 t of NdPr oxide" from Lynas and MP Materials: Lynas FY26 7,260 t plus MP 2025 2,599 t = 9,859 t. Holds. ✈️About 417 kg of rare-earth materials per F-35: the CRS report says "approximately 920 pounds". 920 lb = 417 kg. Holds. 🟡Ytterby's four elements, cerium as common as copper, thulium and lutetium more than 100 times gold: standard crustal-abundance data. Holds. ⚠️ Three things to update 📊The export-control timeline stops in April 2025 and misses the biggest package and the date everyone is now watching. 📅On 9 October 2025 China added five more rare earths (holmium, erbium, thulium, europium, ytterbium), a rule reaching foreign-made products containing 0.1% or more Chinese rare earths by value, and controls on mining, separation, magnet and recycling technology. 📅After the Trump–Xi meeting in Busan, MOFCOM Announcement No. 70 suspended that package until 10 November 2026. On 28 September 2026 MOFCOM said the Kuala Lumpur arrangement behind the suspension is extended to 10 January 2027, but that statement does not name rare earths and no announcement amending the 10 November date has been published yet. 📅🧲So: 10 November on paper, 10 January if the extension is formalised. The April 2025 licences on seven rare earths (Sm, Gd, Tb, Dy, Lu, Sc, Y) and on magnets never stopped. The December 2024 ban on gallium, germanium and antimony exports to the US is also suspended, until 27 November 2026. A suspension is not a repeal; the legal instruments stand. 📈The prices quoted (neodymium about $186/kg, dysprosium about $1,000, terbium about $4,400, gallium over $3,000, germanium almost $12,000) are the publisher's own indicative retail reference prices, which its pricing page says include a display markup and are not order prices. 🇦🇺Lynas no longer simply "ships its ore to Malaysia". Mt Weld concentrate now also goes through the Kalgoorlie plant in WA, which produces mixed rare-earth carbonate, before separation in Malaysia. And the 2030 forecast (China's magnet-metal mining share below 55%, refining at 80% or more) is the publisher's call. 🔍 Who made it EarthRarest is not a research institute. It is a brokerage that helps private investors buy physical strategic metals and rare earths, operated by Finofin Limited, a company registered in Israel; its site names Tradium GmbH in Germany among its suppliers and offers storage in Switzerland. I am sharing the explainer because it is good and the sources are the right ones. 👀 Bottom Line The explainer gets the shape right: not rare, hard to ⚗️separate, and the grip tightens downstream. That is exactly the shape of this series. 📍Part 1 drew the map. 📍Part 2 showed the mine: 🇨🇳China at 69% of ore, two state groups holding 100% of the quota, and a heavy-rare-earth quota frozen at 19,150 t since 2018 while Myanmar digs up more than that next door. 📍Part 3 goes where the 91% lives: separation and refining, and who can actually do it outside China. #RareEarths #CriticalMinerals #SupplyChain #NdFeB #Dysprosium #Terbium #Gallium #Germanium #Antimony #ExportControls #China #USGS #IEA #Lynas #MPMaterials

🚨 China mined 69% of the world's rare earths in 2025, yet mining is the layer where its grip is weakest. It is also the layer where Beijing has held its own heavy-rare-earth quota at 19,150 t a year since 2018 while Myanmar, next door, dug up more than that. Part 2 of 7: who digs what, who controls the tap, and who is actually building outside China. 🧵 The board below is Part 2 of the map from Part 1: xitter.cf/roblun1/status/2113476… same six stages, same loops, with MINING lit up. Four numbers at the top, four columns underneath: who mined what in 2025, reserves against production, how the quota works inside China, and the producing-versus-developing list outside it. ⛏️ Who mined what in 2025 (USGS, tonnes REO) 🇨🇳China 270,000 🇺🇸United States 51,000 🇦🇺Australia 29,000 🇲🇲Myanmar 22,000 🇹🇭Thailand 4,800 🇮🇳India 2,900 🇲🇬Madagascar 2,700 🇷🇺Russia 2,600 🇧🇷Brazil 2,000 🇳🇬Nigeria 1,500 🌍World 390,000 → China 69% 📊Straight from the USGS table. Myanmar, Thailand and Madagascar are estimates derived from China's import data, so they measure what crossed into China rather than what any named mine reported. And US output is essentially one mine, Mountain Pass. 🪨 Reserves are not the constraint 🇨🇳China 44M t (~52% of the world's 85M+) → 270,000t mined 🇧🇷Brazil 21M t → 2,000 t 🇦🇺Australia 6.3M t (3.3M t JORC-compliant) → 29,000t 🇷🇺Russia 3.8M t → 2,600 t 🇻🇳Vietnam 3.5M t → 150 t 🇺🇸United States 1.9M t → 51,000 t 🇬🇱Greenland 1.5M t, 🇹🇿Tanzania 0.89M t, 🇿🇦South Africa 0.86M t, 🇨🇦Canada 0.83M t → 0 🇧🇷Brazil holds about a quarter of the world's reserves and produced about half a percent of 2025 output. Vietnam has 3.5M t in the ground and 150 t out. Rock in the ground is not the constraint. What is built on top of it is. 🏛️ Inside China: two groups, one tap 🇨🇳China's mining comes from three places, in effect: Bayan Obo in Inner Mongolia (the world's largest deposit, light rare earths recovered alongside iron ore, feeding China Northern Rare Earth at Baotou); Maoniuping in Sichuan (light, the world's number two); and the ion-adsorption clays of Jiangxi, Guangdong, Guangxi, Fujian, Hunan and Yunnan (dysprosium, terbium, yttrium, feeding China Rare Earth Group at Ganzhou). ⛏️The 2024 quota is the last one published: 270,000 t of mining and 254,000 t of smelting and separation, issued to just two groups. China Northern Rare Earth received 188,650 t, all light. China Rare Earth Group received 62,200 t of light plus 19,150 t of ion-clay. That is roughly 70/30, and it is 100% of the quota. In 2023 there were still four quota-holders; from 2024 there were two. 📊Since then the numbers have gone dark. The 2025 quotas were issued quietly in July 2025 without a public figure, and no 2026 quota has been published either. Tonnages you see quoted for 2025 and 2026 in the trade press are not official and I have left them off the board. 📜Two rule changes sit behind this. The Rare Earth Management Regulations, in force since 1 October 2024, state that rare-earth resources belong to the State and that no enterprise may mine or separate outside the designated groups. And MIIT Order 71, in July 2025, brought imported ore and monazite inside the quota system, so the clay coming over the border from Myanmar is now counted alongside domestic ore. 📊 The number almost nobody shows: 19,150 t China's own heavy-rare-earth mining quota, the ion-clay line, has been 19,150 t a year since 2018. On the official numbers, any growth in China's heavy-rare-earth feed since then has come from imports, above all Myanmar. 🇲🇲Myanmar's ion-clay output was about 22,000 t in 2025 on the USGS estimate, more than China's entire domestic ion-clay quota, and almost all of it is trucked across the border for separation in China. Laos is now rising as a second source. In September 2026, Myanmar and Laos supplied 97% of China's imports of unlisted rare-earth oxides (3,509 t and 1,369 t of 5,021 t). So the heavy-rare-earth story of the next two parts, separation and pricing, starts here: the dysprosium and terbium feed is mined outside China and processed inside it. 🌏 Outside China: producing now 🇺🇸MP Materials, Mountain Pass, California: 50,692 t of REO in concentrate and 2,599 t of NdPr oxide in 2025. 🇦🇺Lynas, Mt Weld, WA, with processing at Kalgoorlie and in 🇲🇾Malaysia: 13,089 t of REO ready for sale in FY26, including 7,260 t of NdPr. 🇧🇷🇺🇸Serra Verde, Pela Ema, Goiás, Brazil: an ionic-clay operation producing since 2024, ramping to about 6,400 t TREO a year by the end of 2027. Owned by USA Rare Earth since 3 September 2026. 🇲🇲Myanmar and 🇱🇦Laos, as above: producing, but into 🇨🇳China. That is the whole producing list of any scale outside China and its neighbours. Three companies. 🚧 Outside China: still building 🇦🇺Iluka, about 1M t of monazite stockpiled at Eneabba, WA; the Eneabba refinery was about 60% built at the half-year, with commissioning in 2027. 🇦🇺🇲🇼Lindian, Kangankunde, Malawi: first monazite concentrate targeted for December 2026, with offtake to Iluka. 🇦🇺Arafura, Nolans, NT: final investment decision 21 May 2026, ground broken 8 October 2026, about three years of construction ahead. 🇦🇺Hastings, Yangibana, WA: fully permitted, 40% Hastings and 60% Wyloo, no final investment decision; Wyloo is running a sale process for its 60%. 🇦🇺Northern Minerals, Browns Range, WA: heavy rare earths, concentrate committed to Eneabba, first production targeted for late 2027 subject to funding. 🇧🇷🇦🇺Brazil: Meteoric (Caldeira), Viridis (Colossus) and Brazilian Rare Earths (Bahia). Lynas 🇦🇺🇺🇬Ionic Rare Earths, Makuutu, Uganda: an ionic-clay project, 60% held; mining licence granted 28 December 2023; a strategic review of the project has been running since 31 July 2026. 🇬🇧🇦🇴Pensana, Longonjo, Angola: construction more than 30% complete; on 25 September 2026 the company extended its schedule, with first output now guided for late 2028. 🇨🇳Ngualla, Tanzania, no longer a Western developer: Shenghe Resources completed its takeover of Peak Rare Earths on 30 September 2025. 🇺🇸USA Rare Earth, Round Top, Texas; American Rare Earths, Halleck Creek, Wyoming. 🇬🇱Tanbreez, Greenland approved the mining and closure plans (announced 5 October 2026); construction permits and financing still to come, first ore targeted for late 2028 or early 2029. 🇬🇱Kvanefjeld is not on the board: its exploitation licence was refused and its exploration licence was not renewed, and the dispute is in the courts. Every one of these is a 2026–2029 story, and none of them produces a tonne of REO today. On the public record the first new ore from this list arrives from ⚗️Lindian (Dec 2026 target) and the first new refined oxide from Iluka (2027). 👀 Bottom Line ⛏️Mining is the least concentrated layer of the chain: China at 69% of ore, against 91% of refining and 94% of sintered magnets. It is also the layer where the West already has real producers, with the US, Australia and Myanmar together mining 102,000 t last year. But look where the heavy feed comes from and where it goes. Beijing has not raised its own ion-clay quota in eight years, and my read is that it has not needed to, the dysprosium and terbium come in by truck from Myanmar and Laos and are processed in Ganzhou. The grip tightens downstream, and that is where the next part goes. Part 3, separation and refining, next. #RareEarths #CriticalMinerals #Mining #SupplyChain #NdFeB #Dysprosium #Terbium #China #Myanmar #USGS #Lynas #MPMaterials #IXR #ASX