The Rare-Earth Pause Button: China Sets the Timer for November 2026, Leaving Taiwan a One-Year Window to Reassess

On October 9, 2025, China's Ministry of Commerce issued Notice No. 61, pushing rare-earth export controls to an entirely new level. It didn't just add five more elements — holmium, erbium, thulium, europium, and ytterbium — on top of the seven medium and heavy rare earths already controlled since April ^[12]; it also introduced, for the first time, an "extraterritorial rule" — any product manufactured outside China would need a license from Beijing if it contained China-origin controlled rare earths whose value made up 0.1% or more of the finished product's value; products made outside China using Chinese separation, smelting, or magnet-manufacturing technology were likewise brought into scope ^[1]. This design amounts to a Chinese version of a "de minimis" threshold plus a "Foreign Direct Product Rule (FDPR)" — turning the very tool Washington once used to choke off China's semiconductors back around, and applying it to rare earths instead ^[1][2].

Markets held their breath for a moment. This expanded set of controls was originally scheduled to take effect for its extraterritorial portion on December 1, 2025, and was widely read as Beijing raising the stakes as a bargaining chip ahead of the Trump-Xi meeting ^[2]. In theory, it could reach far beyond the U.S.: any factory in any country whose product contained more than the threshold amount of Chinese rare earths would have to answer to Beijing before it could re-export.

Then this wave of expanded controls was paused.

In late October 2025, U.S. and Chinese trade teams negotiated in Kuala Lumpur, followed by a meeting between Trump and Xi in South Korea on October 30 that produced a "partial truce." China's Ministry of Commerce and General Administration of Customs then issued Notice No. 70, declaring the October 9 expanded controls paused for one year, effective until November 10, 2026^[4][5]. The White House simultaneously stated that China had committed to issuing "general licenses" for the rare earths on the April 2025 list, letting shipments of medium and heavy rare earths, superhard materials, and graphite anode material resume flowing under a general-license mechanism over the next twelve months ^[3][5].

This is a pause, not a repeal. The timer is already running, and no one is promising it won't restart the moment it hits zero.

To make the right call, you first need to separate the two waves of controls. This is the point most easily misread — and the most costly one to get wrong.

• The first wave (April 2025): an export-licensing regime covering seven medium and heavy rare-earth elements — samarium, gadolinium, terbium, dysprosium, lutetium, scandium, and yttrium — and their derivatives ^[12]. This wave has never been paused; it has only shifted to expedited clearance for designated exporters via "general licenses" ^[5][6]. It is still operating.
• The second wave (October 2025): adds five more elements — holmium, erbium, thulium, europium, and ytterbium — on top of the original seven, layering on the 0.1% de minimis clause and the extraterritorial rule ^[1][12]. This is the wave that is actually the subject of the current one-year pause ^[4].

What Beijing withdrew was its newest, sharpest extraterritorial blade — the one that could reach "Taiwanese firms making products with Chinese rare earths and reselling them to a third country." But the underlying licensing-review machine hasn't stopped for a single day. In its one-year review of the controls, CSIS states bluntly that the system's real destructive power lies not in any single embargo, but in China's ability to speed up or slow down the pace of approvals at will, turning the licensing regime itself into a lever that can be stretched or shortened at will ^[6]. Pausing the second wave amounts to retracting the newest section of that lever for now — but the handle never leaves Beijing's grip.

What deserves even more vigilance: what's paused is the "controls," not the "price rally." China Northern Rare Earth Group raised its Q2 2026 rare-earth concentrate trading price by about 44.6% in one move, to roughly RMB 38,804 per tonne ^[14]; neodymium metal, as of early July 2026, was up about 82% from a year earlier ^[16]. The supply deficit is now entering its second year, and with global EV sales projected to reach 22.9 million units in 2026 — up about 28% year-on-year — magnet demand is being squeezed even tighter ^[14]. Even if Taiwanese firms can still get material during the pause, their costs have already taken the hit.

China's dominance of the rare-earth supply chain is a funnel that gets more suffocating the further downstream you go. International Energy Agency (IEA) data shows that in 2024, China accounted for about 91% of global rare-earth refining output, with its share of sintered permanent magnets reaching as high as 94%; its control over heavy rare earths, from mining and separation through magnet manufacturing, exceeds 95% ^[6][7]. The real bottleneck was never "digging ore" — the ore itself isn't especially concentrated — it's the midstream steps of separation, smelting, metallization, and magnet manufacturing. Without that infrastructure, even if other countries dig up the ore, it still ends up shipped back to China for processing, and the dependency remains the same ^[7]. Even the USGS's 2026 Mineral Commodity Summaries lists rare-earth refining and magnets as a structural weakness in the U.S. supply chain ^[15].

And Taiwan stands at the most vulnerable point in that funnel.

According to Taiwanese media and official accounts, more than 95% of Taiwan's rare-earth imports come from China; Taiwan imports about 3,000 tonnes of rare earths a year for manufacturing use, produces almost none of the raw material domestically, and has only one or two firms involved in refining and processing ^[8][10]. That means Taiwan doesn't just lack ore — it barely has a midstream lifeline to fall back on either. Once Beijing slows the pace of approvals, Taiwan has no domestic backup to switch to. This is a vulnerability distinct from both the U.S. and Japan: those countries at least have nascent midstream capacity in the works to cushion the blow; Taiwan is starting from almost zero.

Core judgment: this year is not "risk eliminated" — it is "risk repriced for later." What actually needs watching is whether that extraterritorial blade gets drawn again after November 10, 2026 — and right now there is no written commitment of any kind to renew the pause ^[4][6]. The four supply chains below differ in exposure and substitutability, and each has to be handled separately.

Magnets and motors — the deepest exposure, the slowest to substitute. Neodymium-iron-boron (NdFeB) permanent magnets are the heart of nearly every high-efficiency motor, and China controls more than 90% of the world's high-performance magnet capacity ^[6][7]. If Taiwan's magnet and motor makers want to switch to non-Chinese sources, they face a brutal reality: Western, Japanese, and South Korean magnet capacity is still being built out, short-term supply is limited, and certification alone routinely takes a season or a year. Worse, magnets that need to resist high heat and demagnetization typically require heavy rare earths like dysprosium and terbium — and heavy rare earths are exactly the segment where China's control exceeds 95%, the hardest to route around ^[6][7]. This is the gap that most needs a sprint during the pause — and the hardest one to close.

EV components — high volume, and demand is still expanding. About 85% of EV drive motors use rare-earth-containing permanent-magnet synchronous motors; global EV sales are projected to reach 22.9 million units in 2026, up nearly 30% year-on-year ^[14]. Taiwan's motor, motor-control, power-tool, and component supply chains are deeply embedded in this demand curve, mostly as contract manufacturers for international automakers and Tier 1 suppliers, whose demands for source compliance and traceability will only get stricter. Even with "rare-earth-free motor" or ferrite-substitute paths available, energy density, size, and cost remain trade-offs that make full replacement difficult in the short term — this segment needs both "evaluating alternative designs" and "locking in long-term supply contracts" at once, not an either-or choice.

Defense and aerospace — small volumes, but not a single unit can be missing. Precision guidance, radar, sensors, sonar, and jet engines rely on rare-earth permanent magnets and heat-resistant samarium-cobalt (SmCo) magnets; an advanced fighter jet like the F-35 uses rare earths on the order of hundreds of kilograms, and delivery disputes have previously been triggered by magnets found to contain Chinese alloys ^[11]. The defense supply chain's defining trait is that "a single point of shortage stops the line"; certification of alternative sources is stricter and slower, and one non-compliant magnet can jam an entire production line. For Taiwan, calls for defense self-reliance have grown louder in recent years, but if critical magnets still run through the Chinese supply chain, that self-reliance is built with its foundation resting on an adversary's switch — which is exactly why strategic stockpiling and traceability of material sources matter more here than in any other industry.

Semiconductors and specialty materials — the underrated link. Wafer polishing, grinding, and certain processes and equipment use rare-earth compounds (such as cerium-based polishing powder, yttrium, scandium, and others); the volumes are far less visible than in magnets, but this is Taiwan's lifeline industry. Reports such as CommonWealth have already flagged Taiwan's semiconductor sector's hidden dependence on Chinese rare earths ^[10]. The homework here is turning "hidden exposure" into "an auditable list of part numbers" — don't wait for a shortage to discover which process step or which piece of equipment is stuck on rare earths. Semiconductors aren't short on capacity; they're short on the discipline to map the supply chain down to fine enough detail.

One question is the most personally relevant to Taiwanese firms, and also the hardest to verify: is Taiwan covered by China's 0.1% extraterritorial rule?

By the design of the text, the extraterritorial rule is "attached to the object, not the country" — it targets "whether the product contains Chinese rare earths," not "which country the product is exported from" ^[1]. On that reasoning, magnets or modules made in Taiwan using Chinese rare earths, if re-exported to a third country, would in principle fall within this rule's reach (inference; medium confidence). But public information so far shows no case of China naming Taiwan specifically, nor any verifiable instance of a Taiwanese firm actually being blocked, so "exactly how far enforcement against Taiwan would go" remains to be verified. The one thing that can be said for certain is this: the pause has put this blade back in its sheath, extraterritorial reach and all, until November 2026 ^[4][5].

And precisely because the withdrawal of that reach is "temporary," Taiwanese firms that use this window to push a product's Chinese rare-earth content below 0.1%, or to build a switchable non-Chinese source, are effectively buying insurance for their post-November-2026 selves — this is exactly the thing most worth doing during the pause, and also the easiest thing to keep putting off. One important reminder: the 0.1% figure is a value ratio, not a weight ratio — in other words, a high-value finished product can trip the threshold even with just a small amount of mixed-in Chinese rare-earth content. Audits should always be calculated by "share of value," not by the intuitive assumption that "a small quantity means it's fine."

One line of argument holds that rare-earth controls hurt the target but also hurt the wielder — China's own mining companies need to make a living too, this card can't be played many more times, and the pause could very well just keep extending indefinitely. This view has a real basis and deserves to be taken seriously:

• Every swing of the blade whittles down its own long-term leverage. Every time China acts, it accelerates diversification investment in Europe, the U.S., Japan, South Korea, and Australia — effectively pushing its own customers toward alternative supply chains with its own hands ^[6].
• Buyers are also growing teeth. Vietnam banned exports of unprocessed rare-earth ore starting January 1, 2026, aiming to transform itself from a raw-material supplier into a midstream power; Vietnam's rare-earth reserves are about 3.5 million tonnes, sixth-largest in the world ^[13]. The midstream is becoming multipolar, and China's monopoly won't last forever.
• Price itself is a double-edged sword. Rising rare-earth prices certainly hurt buyers, but prices that get too high stimulate substitute technologies and new capacity, and also invite "demand destruction" — buyers simply switching to low-rare-earth or rare-earth-free designs instead ^[14].

But treating "it might not come back" as "no need to prepare" mistakes an observation for a conclusion. Three counterarguments make that bet too expensive:

1. The written commitment only runs to November 2026. Whether it gets extended depends on the ebb and flow of U.S.-China relations and negotiating leverage — it's not Taiwan's call, and there's no promise either way ^[4]. The second wave could revive the moment it expires, or even get redeployed as a bargaining chip in the next round of negotiations.
2. The first wave's licensing regime is still running. Even if the second wave isn't extended, the lever Beijing holds for "adjusting the pace of approvals" never went away ^[6]; prices climbing steadily through the pause is itself evidence that this lever is still turning ^[14][16].
3. Alternative supply chains take years to build. Midstream separation and magnet capacity need years and large amounts of capital to catch up ^[7][15]. If action waits until the blade comes back out in November, the gap in between becomes a shortage period — and Taiwan has no domestic midstream capacity to ride out that window.

Laying both sides out, the neutral judgment is: the pause lowers the "short-term probability of a supply cutoff," but it does not lower the "medium-to-long-term structural risk" — and it has even let cost risk materialize early, through the continued rise in prices. The rational move isn't to bet on one side, but to treat this year as a window to build resilience once, while material sources are relatively smooth. Bet it won't come back, and the stake is the entire supply chain; prepare for it and it genuinely doesn't come back, and the cost is just some extra inventory and certification — the downside risk on the two sides isn't symmetric.

Taiwan's government has already started moving. Taiwan's Minister of Economic Affairs announced in February 2026 a plan to build a pilot rare-earth production line within three years, targeting about 50% of domestic demand, and released roughly NT$405 million in subsidies to encourage companies to invest in critical-mineral R&D ^[8]. Taiwan and the U.S. have also designated critical minerals as a pillar of cooperation under the Economic Prosperity Partnership Dialogue (EPPD) framework, with the direction focused on rare-earth mining and preliminary separation, and a working group has been established ^[9].

These are the right first steps, but their limits need to be seen honestly: both the pilot line and international cooperation take shape on a timescale of years, while the pause-button countdown only runs to November 2026. The clock of policy and the countdown of the pause are two separate races. More critically, if official cooperation stops at "mining and preliminary separation," it still sidesteps the "refining and magnets" segment where China's dominance runs deepest ^[7] — without pushing into the midstream, Taiwan will remain trapped at the narrowest point of the funnel. The pilot line's target of "50% of domestic demand" also depends on which stage's 50% is meant: if it's only 50% of preliminary separation, there's still a large gap before it reaches "able to supply magnets" — and neither officials nor industry should be overly optimistic about this in public.

The national level — treat the window as a war-readiness period, not a diplomatic grace period.

• Keep pushing the pilot rare-earth production line and U.S.-Taiwan critical-mineral cooperation, but pragmatically put the focus on the real bottleneck — "midstream separation and magnets" — rather than just striking a good pose on the mining end ^[7][9].
• Set clear national strategic-stockpile thresholds for high-risk rare earths (samarium, dysprosium, terbium and other heavy rare earths, and neodymium-praseodymium magnets), with a delivery timeline and audit mechanism to be in place before November 2026; make critical defense magnets the top priority.
• Build mutual-aid arrangements and information-sharing on material sources with Japan, South Korea, Australia, Vietnam, and other players in the multipolarizing midstream, incorporating shifts such as "Vietnam banning raw-material exports to build its own midstream" into Taiwan's diversification map ^[13], to avoid simply swapping one dependency for another by tying itself to a single alternative source.

The industry-intermediary level — tier the response, don't spread it evenly.

• Trade associations should lead a "part-number-level audit of Chinese rare-earth exposure," quantifying each product's 0.1% extraterritorial exposure by share of value, and build a traceable inventory list.
• Jointly submit samples to non-Chinese magnet and rare-earth suppliers, sharing the cost and time of certification to flatten a barrier that no single firm could bear alone; collective bargaining can also help hedge against rising prices ^[14].
• Establish a tiered stockpiling principle: build strategic inventory for irreplaceable heavy rare earths and NdFeB magnets, while pushing hard on dual-source certification and alternative designs for substitutable items; resources are limited, so don't stockpile everything evenly.

The SME level — ask three questions first, then decide what to do.

• Does your product contain Chinese rare earths? By value, could the share hit 0.1%? If supply is cut off, is there an alternative material? If you can't answer these, spend this year finding the answers first.
• Get dual-source certification finished while the pause makes material sourcing relatively smooth — the time certification takes is often more fatal than the shortage itself.
• Don't just wait for a large firm or the government to bail you out: locking in long-term supply contracts and minimum safety stock for critical magnets and motors is the one moat an SME can control on its own before the next shortage hits; if cash and warehouse space are limited, prioritize protecting the handful of parts where a shortage would stop your line.

The pause button gave Taiwan one year. This year is for sprinting, not for taking a vacation. And while the countdown runs, prices are still climbing. The number on the timer is November 10, 2026.