$237M Bet on Recycling to Fix the US Rare Earth Supply Gap

The United States consumes rare earth elements at industrial scale across defence, automotive, and AI infrastructure, yet processes a negligible share of its own rare earth oxides domestically. That gap between consumption and processing capacity is not new. What is new is the capital now moving to close it.

Cyclic Materials closed a $75 million strategic financing round in August 2026, lifting total cumulative equity to roughly $237 million. The raise is not a corporate milestone for its own sake. It is a structural bet that recycling can fill a supply-chain gap that new mines alone cannot address within any realistic permitting timeline.

The question for investors is whether the model behind that capital, a hub-and-spoke rare earth recycling system built on domestic e-waste feedstock, is commercially durable or strategically aspirational. Here is a clear-eyed assessment of the capital, the infrastructure, the feedstock logic, and the demand case, in that order.

The gap between U.S. rare earth consumption and domestic processing capacity has been documented for years, but the capital formation accelerating through 2025 and 2026 reflects a structural shift in how strategically critical that gap has become for both government planners and private investors.

What $237 million in cumulative capital actually signals about rare earth recycling

The $75 million August 2026 round is large for the recycling sector. But its analytical weight sits in what came before it and who wrote the cheques.

Cyclic Materials has now raised approximately $237 million in cumulative equity, positioning it as one of the most heavily capitalised rare earth recyclers in North America. That figure alone separates it from the pilot-stage ventures that dominate the space, where capitalisation typically sits below $50 million.

Approximately $237 million in cumulative equity raised, positioning Cyclic as one of the most heavily capitalised rare earth recyclers in North America.

The round was led by accounts advised by T. Rowe Price Associates, an institution that had already committed capital in 2025 and chose to participate again in this financing. Institutional repeat participation at this scale is commercial trajectory confidence, not early-stage speculation. When an asset manager of that size comes back, it is underwriting a business model it expects to reach cash-generating operations, not a technology it hopes might work.

The composition of the rest of the cap table tells a second story. This is not a round filled exclusively by financial return-seekers. It includes end-users and supply-chain participants with direct, recurring need for what Cyclic produces:

• T. Rowe Price Associates: Institutional lead; repeat investor from 2025, signalling commercial trajectory conviction
• Canada Growth Fund: Sovereign-adjacent capital aligned with Canadian critical minerals strategy
• Microsoft: Hyperscale data-centre operator hedging future rare-earth-bearing hardware waste streams and sustainability requirements
• Amazon: Same rationale as Microsoft; previously backed Kingston, Ontario facility expansion
• BMW i Ventures: Direct exposure to neodymium-iron-boron magnets in EV and hybrid drivetrains
• Jaguar Land Rover InMotion: Automotive supply-chain motivation; securing future recycled magnet supply
• Hitachi Ventures: Industrial and electronics manufacturing exposure to permanent magnet supply
• Fifth Wall: Climate technology and real-asset infrastructure investor
• Energy Impact Partners: Energy transition-focused strategic capital
• ArcTern Ventures: Cleantech venture capital
• ERI: New investor in the 2026 round and simultaneously Cyclic’s feedstock partner, a dual role that aligns financial and operational incentives

For investors evaluating the rare earth recycling space, the composition of this round matters as much as its size. A cap table that includes OEM procurement-motivated investors and hyperscale infrastructure operators provides a demand signal that pure financial rounds do not. BMW, Amazon, and Microsoft are not betting on commodity price. They are securing supply.

How the hub-and-spoke model turns scattered e-waste into industrial-scale rare earth supply

Rare earth magnets are everywhere: inside electric motors, disk drives, wind turbines, MRI machines, and robotics hardware. But once those products reach end of life, the magnets inside them are geographically scattered across thousands of collection points, repair shops, and recycling yards. At any single location, the volume of recoverable magnet material is too small to justify refining infrastructure.

That dispersion problem is what has historically made rare earth recycling uneconomic at small scale. Cyclic’s two-stage architecture is designed to solve it.

The unit economics of rare earth magnet recycling have shifted materially as hydrometallurgical processing costs have fallen and secondary feedstock volumes have grown, creating the conditions under which a hub-and-spoke aggregation model can generate positive margins rather than requiring subsidies to survive.

Spoke facilities: separating magnets at scale

Spoke facilities receive end-of-life products and use Cyclic’s proprietary MagCycle technology to separate rare earth magnets from the devices that contain them. The spokes process volume but do not refine.

The Arizona spoke facility is the operational benchmark: capable of processing up to 25,000 tonnes per year of end-of-life components, with operations projected to commence in Q3 2026. It serves as a magnet-separation centre for American e-waste and industrial scrap, isolating magnet material that is then shipped to hub facilities for hydrometallurgical processing.

Hub facilities: refining to industrial-grade rare earth oxides

Hub facilities apply Cyclic’s REEPure hydrometallurgical processing platform to convert isolated magnet material into high-purity mixed rare earth oxides (MREO), a concentrated form of the rare earth elements that manufacturers need. The performance of these recycled outputs is designed to rival virgin mined supply. By-product metals including copper, aluminium, nickel, steel, and boron improve unit economics by creating additional revenue streams from the same feedstock.

The architecture’s economic logic is aggregation: collecting dispersed feedstock at multiple spoke locations and channelling it into centralised refining achieves throughput that no single collection point could sustain alone.

The McBee scale-up pathway from 2,000 to 6,000 tonnes per year is the section’s most important data point. It shows whether the economic model depends on hitting the expansion threshold to be viable, and at what timeline. More than $82 million was committed to the South Carolina campus in a January 2026 announcement, funding what will become Cyclic’s largest hub to date.

The ERI partnership and why feedstock is the real constraint in rare earth recycling

Most rare earth recycling ventures have not failed because the chemistry did not work. They have failed because they could not secure reliable, high-volume feedstock to keep their processing assets economically loaded. Separation and refining infrastructure sitting idle is the fastest route to negative unit economics.

This is the bottleneck that the ERI partnership, announced in July 2026, is designed to address.

Up to 25,000 tonnes per year of end-of-life electronics directed from ERI’s national U.S. collection network to Cyclic’s Arizona spoke facility.

ERI operates one of the largest electronics recycling and processing networks in the United States. Under the partnership, ERI directs magnet-bearing components from its national collection infrastructure into Cyclic’s Arizona operations. The feedstock categories flowing through that network include the product types richest in rare earth magnets:

• Electric motors and drivetrain components from end-of-life vehicles
• Disk drives and data storage hardware from enterprise and consumer electronics
• Wind turbine generator components
• MRI machines and other magnet-intensive medical equipment

What makes this arrangement analytically distinct from a standard offtake letter of intent is ERI’s dual role. ERI joined as a new strategic investor in the 2026 financing round. That means the feedstock supplier holds equity in the company it is supplying. Misaligned incentives that typically allow feedstock commitments to erode under commercial pressure are partially neutralised when the supplier’s financial return depends on the processor’s success.

For investors assessing rare earth recycling plays, feedstock security is frequently the variable that separates viable businesses from conceptual ones. The ERI structure provides a more durable feedstock floor than reliance on spot collection or unanchored supply agreements.

Which demand sectors are pulling hardest, and what that tells investors about timing

The capital tells you the supply side is funded. The infrastructure tells you the processing model exists. The feedstock partnership tells you input is secured. The remaining question is output: who is buying recycled rare earths, and on what timeline?

Four demand vectors are pulling at different speeds, and their order of near-term volume credibility shapes the investment thesis:

1. Automotive: The most immediate volume driver. Rising EV penetration combined with the first large cohorts of neodymium-iron-boron magnet-rich vehicles approaching end of life creates simultaneous feedstock growth and output demand. BMW i Ventures and Jaguar Land Rover InMotion are the investor proxies for this thesis; their participation is a procurement signal, not a financial bet. Automotive provides the most credible near-term pull on Cyclic’s MREO output.

Automotive rare earth dependency is more concentrated than most OEM procurement teams publicly acknowledge; neodymium and dysprosium alone account for the majority of magnet content in traction motors, and supply-chain diversification away from Chinese processing has been slower than EV adoption curves would require.

1. AI infrastructure and hyperscale data centres: The medium-term strategic vector. Microsoft and Amazon are constructing data-centre fleets whose motors, actuators, drives, and cooling systems will eventually return as high-grade rare-earth-bearing waste. Their investment in Cyclic functions as both a sustainability commitment and a hedge on future feedstock access. The demand timeline here is longer, but the hardware refresh cycle means the base grows over a decade rather than plateauing quickly.
2. Defence and advanced manufacturing: The national-security demand layer. Permanent-magnet motors and guidance systems in military hardware benefit from domestic, non-Chinese processing streams, adding a policy-tailwind dimension to commercial demand.

The national security premium on domestic processing

Chinese dominance of rare earth separation and refining is a recognised U.S. strategic vulnerability. Critical mineral recovery, including the reclamation of end-of-life rare earth magnets and manufacturing scrap, has been formally designated a national security priority by the U.S. government. Recycled rare earths sourced from domestic waste streams create a parallel supply chain that bypasses import exposure entirely, because the end-of-life products and manufacturing scrap originate inside the United States and Canada. Cyclic’s feedstock is inherently domestic.

The White House’s Section 232 actions on processed critical minerals formalised what the market had already priced in: foreign dependency on rare earth separation and refining constitutes an active national security exposure, not merely a supply-chain inefficiency.

1. Medical and robotics: MRI machines, surgical robotics, and semiconductor-adjacent hardware round out the target market. Volume here is smaller in near-term scale but contributes to feedstock diversity and output demand breadth.

The staggered nature of demand timing matters. Automotive provides credible near-term volume, but AI infrastructure’s long-tailed hardware refresh cycle means the demand base for recycled rare earths grows over a decade. Reading the cap table as a forward-looking demand signal, each strategic investor represents a sector that either generates feedstock, consumes output, or both.

What comes next, and which milestones will tell investors whether the model is working

The thesis is funded, the infrastructure is specified, the feedstock is contracted, and the demand vectors are mapped. What remains is execution, and the milestone sequence over the next 24 months will determine whether Cyclic’s hub-and-spoke model delivers at industrial scale or stalls at demonstration scale.

Five milestones, in chronological order, define the confirmation pathway:

1. Arizona ramp-up (Q3 2026 onward): Stable throughput, recovery rates, and unit economics at a 25,000-tonne facility are the first data the market will see from this system operating at commercial scale. This is the earliest available proof point.
2. McBee groundbreaking (Q4 2026): Moving from announcement to site preparation demonstrates that permitting and construction timelines are on track. Delays here would signal execution risk in the most capital-intensive facility.
3. Kingston scale-up to approximately 500 metric tons: Integration of feedstock from ERI and industrial partners will reveal how effectively the company aggregates dispersed magnet waste into centralised refining capacity.
4. First formal offtake agreements (timeline unspecified): No publicly disclosed formal supply agreements with automotive OEMs, defence contractors, or hyperscale operators have been announced. Securing these would move Cyclic from infrastructure-builder to embedded supply-chain participant, and their absence remains the most significant open variable.
5. McBee expansion from 2,000 to 6,000 tonnes per year: The expansion threshold is where the unit economics either justify scale or reveal dependency on subsidy and policy support. If South Carolina reaches initial operations on schedule in 2028, Cyclic will have demonstrated hub-and-spoke rare earth recycling at a scale no North American company has achieved before.

Each milestone is a decision point for investors following this space. Arizona throughput data in late 2026 and early 2027, McBee groundbreaking signals in Q4 2026, and operational confirmation in 2028 create a concrete sequence for evaluating whether the model is working, not just whether it is funded.

For investors wanting to pressure-test the hub facility economics against alternative processing routes, our deep-dive into rare earth midstream processing economics covers the cost structures, throughput thresholds, and margin profiles that determine whether a hydrometallurgical refinery reaches commercial viability without sustained policy support.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Forward-looking statements regarding facility timelines, production targets, and expansion plans are subject to change based on market developments and company performance.