A procurement manager does not care where a shipment of desiccated coconut was grown. They care that container 14 reads the same fat, moisture and grain as container 6, whichever origin it came from. That is the job of a multi-origin contract: hold one specification constant while the loading port changes underneath it.
Most buyers sourcing desiccated coconut, coconut milk powder or coconut oil from more than one country learn that “multi-origin” is a sourcing strategy, not a contract clause. Without a written tolerance band, a defined COA reconciliation method and a substitution trigger, three origins do not average into one spec. They give a quality manager three different arguments with three different mills.
What does a spec-lock clause actually say?
A spec-lock clause fixes a target value, a tolerance band, and a financial remedy for landing outside it, applied identically regardless of country of origin. It is not one number. It is three: basis, maximum, and allowance.
FOSFA International’s Contract No. 26 for CIF oilseed shipments, the trade-association template copra and edible-oil contracts descend from, writes specification clauses this way. Each parameter carries a basis figure, a hard ceiling, and a pre-agreed price adjustment for shipments landing between the two. Deviate within the allowance and price moves. Breach the maximum and the buyer can reject.
Applied to coconut, a workable spec-lock for desiccated coconut sits on the Codex Alimentarius baseline (CXS 177-1991): moisture at or below 4% m/m, oil content at or above 60% m/m for standard grade, ash at or below 2.5% m/m. A buyer’s contract narrows that Codex ceiling into a commercial tolerance band, for example moisture 2.8 to 3.6%, with the allowance formula starting above 3.2%. Origin never appears in the formula. Only the number does.
How wide should the tolerance band be?
Wide enough to admit real crop-to-crop and mill-to-mill variation, narrow enough that a buyer’s finished-goods line never notices which country the coconut came from. Codex sets the outer boundary. The contract sets the working band inside it.
For coconut milk and cream, Codex Standard 240-2003 fixes composition by product tier, not by origin. A Sri Lankan and a Vietnamese processor both operate inside the same coconut cream band on the same Codex sheet. The trader’s job is picking the narrower commercial slice inside it and holding every origin to that slice.
| Coconut milk tier | Total solids (% m/m) | Fat min (% m/m) | Moisture max (% m/m) |
|---|---|---|---|
| Light coconut milk | 6.6 to 12.6 | 5.0 | 93.4 |
| Coconut milk | 12.7 to 25.3 | 10.0 | 87.3 |
| Coconut cream | 25.4 to 37.3 | 20.0 | 74.6 |
| Coconut cream concentrate | 37.4 min | 29.0 | 62.6 |
Source: Codex Alimentarius CXS 240-2003, Section 3.1.3. Codex sets the product-tier boundaries; a buyer’s contract narrows the working tolerance inside them.
Grain size follows the same logic. CXS 177-1991 defines three sieve grades: extra-fine (90% or more passes a 0.85mm sieve), fine (80% or more passes 1.40mm), medium (90% or more passes 2.80mm). A bakery running an extra-fine grade line does not want a shipment landing at the fine boundary just because it technically clears Codex. The contract states the sieve grade by name and ties it to the tolerance band, not to Codex’s outer limit.
How does a buyer reconcile a COA from three different labs?
By agreeing the sampling and analysis method before the first container ships, not after a disputed result arrives. FOSFA’s structure is instructive here even where coconut is not the named commodity: Clause 19 of Contract No. 26 requires buyer and seller superintendents to jointly draw and seal representative samples, with the analyst issuing a certificate carrying a weighted average. Either party can demand a second or third analysis at its own cost, with disputes settled by a mean-of-two or mean-of-closest-two-of-three formula agreed in advance.
That structure matters more, not less, across three origins. A Sri Lankan mill’s in-house lab, a Vietnamese exporter’s contracted inspector and a Philippines superintendent’s report are not directly comparable unless the contract specifies a common sampling standard. Codex’s own General Guidelines on Sampling, CAC/GL 50-2004, point buyers toward ISO 2859-1 attribute sampling plans for this reason: one statistical method at every origin, so a COA from Colombo and a COA from Ho Chi Minh City read against the same acceptance limit.
Name the sampling standard and the tie-break formula in the contract itself, and require every origin’s COA to disclose the method used, not just the result. A number without a method is not comparable across origins.
Can the desk blend product from different origins in one container?
Yes, but blending changes what “country of origin” means on the paperwork, and the rule that decides it is a customs rule. Under the EU Customs Code (Regulation 952/2013, Article 60), the country of origin for a processed good is wherever the “last, substantial, economically-justified processing” happened, not wherever the raw material grew. A container milled and packed at one facility from nuts grown across two or three countries can, for customs purposes, carry the origin of the processing country.
Consumer-facing labelling is stricter. Under EU Implementing Regulation 2018/775, if a food’s stated origin differs from the origin of its primary ingredient, that mismatch must be disclosed. Coconut qualifies as a primary ingredient under Regulation 1169/2011 whenever it exceeds 50% of the finished food or is the ingredient the product is named after.
The desk’s practice: blending happens at the processing facility under one quality system, documented lot by lot with each input origin traceable through the mill’s intake records, even after the nuts are mixed. Traceability does not stop where blending starts. It moves from the field to the mill’s log.
What triggers an origin substitution mid-contract?
A defined event, stated in the contract, not a phone call when a shipment runs late. FOSFA’s Force Majeure clause (Clause 23) draws a line worth copying into any coconut contract: where goods of a specific origin are sold with the option of shipment from alternative ports, a seller can only invoke force majeure for the specific port actually blocked, so long as another loading port in that origin is still shipping. A Prohibition clause (Clause 24) covers a government export ban at the country level, extending the shipment window by a fixed period, commonly 30 days, before either party can cancel.
Applied to a three-origin coconut contract, a port closure in one region does not excuse the whole contract, only the shipments tied to that port. A substitution to Vietnam or Indonesia supply keeps the spec unchanged while the origin line on the shipping documents changes. Commodity lawyers writing on force majeure practice note that a party invoking the clause carries a duty to mitigate: it must show alternate-origin supply was genuinely sourced, not simply assert the clause.
The paperwork trail a substitution needs: trigger date, cause, clause reference, and confirmation the replacement shipment meets the original basis/max/allowance figures. Skip any one and a substitution looks like a unilateral change, not a contractual right.
Why does the same spec cost different amounts by origin?
Because origin, not spec, is where the price risk sits. As of late August 2026, Mundus Agri reported desiccated coconut FCA Netherlands pricing at roughly €2,500 per tonne for Indonesian high-fat fine grade against roughly €2,750 per tonne for Philippines origin of the same grade, with Vietnamese origin offered lower still. Same Codex grade. Same fat content. A 10% price spread on origin alone.
That spread is why the spec-lock clause exists. A spec that travels with the number, not the country, lets a buyer move volume toward whichever origin prices favourably that quarter without touching a line of the quality terms. India’s coconut product exports rose 62% to USD 824.6 million in the 2025/26 season on roughly 31% of global coconut production, a reminder of how fast the cheapest origin can shift. A contract anchored to spec, not source, is built for that movement.
What this means for buyers
- Write the tolerance band as basis/max/allowance, not a single target number. A single number gives a buyer no formula when a shipment lands close to the edge.
- Name the sampling standard in the contract, not just the acceptance limit, so a Sri Lankan COA and a Vietnamese COA are comparable.
- Separate the customs-origin question from the labelling-origin question before blending. The two rule sets can produce different answers for the same container.
- Document substitution triggers as they happen, with a date, a named cause and a clause reference.
- Track price by origin, separately from spec. The spread between origins on an identical grade is where the savings live.
FAQ
Does a multi-origin contract mean lower and more consistent quality? No. It means one contractual quality band applied to every origin, with a defined method for reconciling COAs and a defined remedy for shipments outside the band. Consistency comes from the tolerance mechanism, not from averaging origins together.
Can a buyer reject a shipment that meets Codex but misses the contract’s tighter tolerance? Yes, if the contract’s basis/max/allowance figures are narrower than the Codex ceiling. Codex sets the regulatory floor; the commercial contract sets the working target.
Who pays for a disputed COA analysis? Under the FOSFA-style structure, the party demanding a second or third analysis pays for it, with the dispute settled by a pre-agreed averaging formula rather than a fresh negotiation.
Sourcing three origins against one spec is a contract-drafting exercise before it is a shipping exercise. For how the underlying contract structures compare, see spot, forward and indexed pricing for coconut contracts, and for the sampling this piece assumes, see reading a desiccated coconut COA. Need a spec locked across origins? Send the desk an RFQ.