Separated rare earth oxide contracts use a benchmark formula or negotiated mechanism for each oxide, with examples covering Nd, Pr, Dy and Tb and contracts extending for 15 years and 100% of Phase 1 output [S1].
The commercial terms determine who bears price movements, production shortfalls, contamination and qualification risk when mixed rare earth feedstock is converted into individual oxides [S4][S7].
Separated Oxides Require Oxide-by-Oxide Pricing
Separation converts mixed rare earth feedstock into individual oxides, and refined oxides must then be converted into metals before many magnet alloys can be produced [S4][S6]. A contract written for a mixed concentrate or a multi-element basket therefore does not automatically establish a comparable price for NdPr oxide, dysprosium oxide or terbium oxide.
The units and product definitions are central to price comparability. A 15-year offtake covering 100% of Phase 1 Nd, Pr, Dy and Tb production must identify the four oxides separately, while another arrangement covers separated NdPr oxide above 99% purity and a mixed SEG+ carbonate containing dysprosium and terbium [S1][S5]. Purity, assay, moisture basis, delivery location and permitted impurities belong in the product schedule because a nominal price per kilogram has limited meaning without them.
Contract administration should also distinguish oxide value from later conversion economics. The buyer may take delivery as an oxide, toll the material into metal, or use a share of output that remains in intermediate form [S5]. Those are different transactions: metallization adds processing cost and recovery risk, while toll treatment shifts part of the capital burden to the processor and can add return-material terms, processing charges, yield assumptions and loss allocation.
Price Floors, Ceilings and Hybrid Structures
One documented structure uses guaranteed minimum floor prices for Nd, Pr, Dy and Tb, with the associated offtake running for 15 years and covering 100% of Serra Verde's Phase 1 production of those elements [S1]. A floor protects the producer below a specified market level, but it does not by itself determine the final invoice price.
A workable formula normally separates the reference price from the contractual adjustment. If the market benchmark sits below the floor, the floor becomes the payable amount, subject to the agreement's definitions. If the benchmark is above the floor, the contract may require payment at the benchmark, a stated share of it, or another negotiated rate. Ceilings protect the buyer from unusually high prices. Contract-for-difference mechanisms are another proposed tool because they can transfer a defined price difference between the contract price and a reference price [S3].
The main comparison is therefore not simply floor versus no floor. The buyer should compare protection on the downside, exposure on the upside, benchmark liquidity, the duration of the reference and the credit quality of the counterparty. A floor with strong downside protection but no ceiling may not meet a buyer's affordability requirement. A ceiling with no floor can leave the producer exposed to a price collapse, while a two-sided mechanism allocates risk more evenly but requires a reliable settlement process and clear treatment of quality discounts.
Volume, Ownership and Delivery Terms Shape the Effective Price
An oxide price is not economically meaningful without a defined volume and delivery obligation. The 15-year arrangement cited by USA Rare Earth covers 100% of Phase 1 Nd, Pr, Dy and Tb production, while the Phalaborwa memorandum of understanding gives Neo offtake rights to 40% of annual separated NdPr oxide above 99% purity and 65% of annual SEG+ carbonate [S1][S5]. These percentages establish quantity rights but do not disclose all commercial price terms.
For a buyer, the take-or-pay comparison is useful: capacity reservation may preserve access to a specified volume without requiring full payment, while take-or-pay generally obliges the buyer to pay for contracted quantities even if it does not take them. The same distinction is explained in Capacity Reservation vs Take-or-Pay: Spec-Level Differences for Industrial Off-Take. Rare earth contracts may instead use minimum purchase obligations, annual caps, output shares or tolling volumes, so buyers should compare the actual default remedy rather than rely on the label.
Ownership transfers when the contractual risk and rewards pass, and that point should be distinguished from physical delivery. Incoterms, shipping responsibility, insurance, demurrage, assay exchange, title passage and rejection rights all affect the delivered cost. Contracts should also specify whether the producer bears freight and whether a buyer may nominate an alternate assay laboratory. A price stated as an index plus a premium can produce materially different net costs if logistics and assay administration are left open.
Reference Indexes, Baskets and Quarterly Adjustments
Rare earth prices differ by element and specification, so a benchmark must match the oxide and the quality basis used in the contract. One illustrative Phase 1 economic model used average TREO production after lanthanum removal of about 4,400 metric tons per year and a basket price of about $190 per kilogram, based on Argus projections as of December 2025 [S1]. That $190 per kilogram figure is a modeling assumption for a basket, not a disclosed contractual price for each of Nd, Pr, Dy and Tb.
This distinction is commercially important. A TREO basket, an NdPr oxide index, a dysprosium oxide price and a terbium oxide price are different pricing objects. Using one as a proxy for another exposes the contract to basis risk, particularly when chemical compositions, impurity tolerances and delivery terms differ. A more auditable structure names the benchmark publisher, assessment date, currency, unit, location, quality adjustment and the exact arithmetic used to convert the assessment into the contract price.
Adjustment frequency must match procurement and inventory cycles. Monthly or quarterly reset can reduce the period in which the contract price diverges sharply from the market, while annual or fixed pricing gives more budget certainty but leaves greater exposure to price drift. Long-term contracts should also address stale assessments, unavailable observations, index discontinuation and disputes over laboratory results. Those provisions matter as much as the headline formula once a contract has a 15-year term.
Quality, Contamination and Recovery Risk Enter the Invoice
Separated oxide contracts must define ownership, contamination limits and pricing remedies because the technical and commercial terms are inseparable [S7]. The NdPr product in the Phalaborwa arrangement is specified at more than 99% purity, and the SEG+ material is a carbonate intermediate containing Dy and Tb together with Sm, Eu, Gd, Ho, Er, Tm, Yb, Lu and Y [S5]. A price for separated NdPr oxide cannot be compared directly with a price for mixed SEG+ carbonate without a stated conversion, recovery assumption and treatment of the remaining 35%.
Quality adjustments should cover the assay, elemental profile, permissible impurities, particle size, moisture, physical condition, sampling method and approved laboratories. Off-spec material needs an enforceable path: rejection before unloading, rework, re-separation, price discount or replacement. The contract should also assign responsibility when a carrier, sampling error or processing defect causes the failure.
Recovery assumptions affect both price and volume. A nominal tonne of contained element at the mine is not a saleable tonne of separated oxide. Solvent extraction yields, metallization recovery, reagent consumption, waste treatment and process losses determine the quantity that reaches the contract delivery point. Overruns or shortfalls should be handled through transparent reconciliation rather than an open-ended obligation to supply the same physical tonnes regardless of recovery.
Who Uses Each Pricing Structure
Long-term oxide offtake is suited to magnet manufacturers, metal producers and strategic buyers that need predictable access to qualified material. The arrangement between Neo and Rainbow illustrates the model: technical support for a separation circuit is linked to offtake rights for 40% of annual NdPr oxide and 65% of annual SEG+ carbonate, with toll treatment of the remaining 35% SEG+ under consideration [S5].
It is less suitable for small buyers unable to meet minimum-volume commitments, buyers requiring immediate delivery before a project reaches commercial operation, or counterparties that need identical oxide grades on a spot basis. A 15-year, 100%-of-production structure also differs from a partial annual share because the buyer bears a much larger exposure to the producer's operating performance and the seller's concentration risk.
Buyers also differ from sellers in their preferred risk allocation. The producer normally prioritizes a floor, a credible buyer, a long tenor and a defined quality discount. The buyer normally prioritizes a ceiling or sharing mechanism, credible volume, an enforceable delivery obligation and access to replacement material if the project underperforms. A bankable contract should align those positions without pretending the floor eliminates all project risk.
Total Cost Extends Beyond the Oxide Price
Purchase price is only one component of the delivered cost of a separated oxide. Processing, assay, freight, insurance, inventory, financing, metallization, yield loss and qualification testing all add cost. The 35% SEG+ share considered for toll treatment illustrates this distinction: further separation adds a service charge, recovery exposure and a return-material obligation, while the separated Dy, Tb and Y would be returned to Rainbow for marketing [S5].
Total cost of ownership also includes the cost of failed qualification and interrupted magnet production. A cheaper oxide with a wider impurity profile can be more expensive if it requires additional separation, causes rejection or delays a qualified magnet line. Conversely, a premium oxide may carry lower processing and inventory costs if its assay is stable and the delivered schedule is dependable. Buyers should therefore model net landed cost and working capital, not compare invoice prices alone.
Certification and volume tier can also move commercial terms. New processing projects must qualify products with downstream metal and magnet customers before they can deliver dependable commercial volumes. Larger contracted volumes can support dedicated separation, metallization and logistics arrangements, but they also concentrate the buyer's exposure if the project misses schedule or quality targets. A contract with bankable demand may lower financing risk for the producer and permit more competitive delivery terms, although the research does not establish a standardized discount for any particular volume tier.
Contract Clauses That Survive Price Volatility
A defensible separated-oxide offtake should identify each element, the commercial product form, purity, delivery basis, benchmark, floor, ceiling, adjustment frequency, volume, title transfer, assay method, rejection remedies, ownership of material in processing and treatment of force majeure. The 15-year Serra Verde structure establishes tenor, production coverage and oxide-specific floors, while the Phalaborwa arrangement establishes annual share, product form and purity [S1][S5].
Governance matters as much as the initial formula. Independent assay laboratories, periodic price and volume reconciliation, reporting for lost production, and a process for amending benchmarks prevent a long-term contract from becoming ambiguous after a plant, counterparty or market structure changes. Buyer protections should include replacement rights, force-majeure treatment and transparent access to operating performance data. Seller protections should include payment security, defined rejection deadlines and a dispute process that does not suspend undisputed amounts.
Demand-side mechanisms can reinforce private offtake by creating a market for qualified non-China supply. A proposed plurilateral economic security agreement would use a capacity-linked tariff-rate quota for suppliers holding 60% or more market share across the rare earth permanent magnet supply chain, while a residual quota would cover demand that qualified supply cannot yet meet [S3]. That policy concept does not set oxide contract prices, but it could affect buyer alternatives, bargaining power and the commercial value of long-term supply security.
The next trackable signals for each contract are the named benchmark or formula, the floor and ceiling for every oxide, the annual volume or production share, and the assay and rejection rules. Those four fields reveal the actual risk transfer better than the headline label, whether a buyer is dealing with a fixed price, an indexed price, a floor-protected offtake or a capped hybrid.