Off-Take Structures for Critical Minerals: What EU and US Buyers Should Actually Ask For

Last month I sat across from a procurement lead from a German battery-materials group. He'd flown into Islamabad with a two-page term sheet and an assumption that we'd sign it by Friday. We didn't. And the reason we didn't is exactly why I'm writing this.

The term sheet was built for a Chilean copper concentrate deal from 2019. It didn't reflect what a critical mineral off-take out of Gilgit-Baltistan actually needs to look like in 2025 — not for us, and honestly not for him either. He just hadn't been told that yet.

So here's the operator's view. What buyers in Brussels, Berlin, Washington and Pittsburgh should actually be asking for when they sit down with a Pakistani producer on antimony, tungsten, molybdenum or copper concentrate.

The four structures that actually get signed

There are really only four off-take structures I see closing on critical minerals right now. Everything else is a variation.

Fixed-volume, fixed-price. Rare. Buyers love the certainty, producers hate it. On a five-year antimony contract with metal at $38,500/t last week and $15,000/t two years ago, nobody sane signs a fixed price. If someone offers you one, ask why.

Fixed-volume, index-linked. This is the workhorse. You agree tonnage — say 1,200 tonnes of tungsten concentrate (65% WO3 min) annually — and price against Fastmarkets APT Rotterdam or the Argus antimony assessment, with a treatment charge or refining charge deducted. Simple. Bankable. Most of what I sign looks like this.

Volume flex with floor and ceiling. Buyer commits to a minimum offtake (say 70% of nameplate), with an option up to 120%, priced against index but with a floor price that lets the mine service debt. Lenders love this. If you're a buyer trying to help a project reach FID, this is the structure that unlocks — sorry, that gets the project financed.

Streaming or prepay. Buyer advances capital against future deliveries. Typical in Pakistan right now because local project finance is expensive (KIBOR plus 3-4 is not cheap when you're building a flotation circuit). A prepay of $8-15M against 24 months of concentrate deliveries is a structure I've had three serious conversations about this year alone.

What EU buyers need that US buyers often don't ask about

The Critical Raw Materials Act changed the conversation. If you're an EU buyer and you want the tonnage to count toward the 2030 benchmarks, the mineral supply contract needs to show a few things clearly.

Origin traceability from mine to port. Not a certificate someone printed. Actual chain-of-custody documentation — mine gate weighbridge tickets, SGS or Alfred H Knight sampling at loadout, sealed containers to Karachi, bill of lading matched to assay certificates. We do this on every shipment now because German and Dutch buyers started asking for it around Q3 2023 and it hasn't stopped.

ESG documentation aligned with the EU Battery Regulation and, increasingly, the German Supply Chain Act (LkSG). This means a proper ESIA on file, community agreements documented, and — this is the one people forget — evidence that the concession holder actually holds the licence. Gilgit-Baltistan mineral titles are issued by the GB Minerals Department under the 2003 rules as amended. Ask for the lease number. Verify it. Any serious producer will hand it over on day one.

US buyers under the IRA and Defense Production Act care about origin too, but the questions come at it differently. FEOC (Foreign Entity of Concern) status matters more than carbon accounting. If you're a Pentagon-adjacent buyer looking at antimony for ammunition primers or tungsten for penetrators, the question you'll be asked back home isn't "what's the CO2 per tonne" — it's "who owns the mine and where does the concentrate physically go between the pit and your smelter."

Both questions are answerable. But the contract needs to be structured to answer them, with audit rights, site visit clauses, and — this matters — a clean shipping route that doesn't transit through jurisdictions that will get the deal flagged later.

The clauses people get wrong

A few things I've seen buyers push for that don't actually work in the Pakistani context, and a few they should push for harder.

Force majeure needs to include specific regional language. Karakoram Highway closures during winter are not force majeure in the traditional sense — they're expected. A well-drafted contract has a shipping window clause (typically April to October for bulk moves from Skardu and Chilas down to Karachi) and a separate emergency force majeure for landslides, security events, or port disruption at Karachi/Port Qasim.

Assay disputes. Umpire lab should be named up front. We usually agree ALS Loughrea or SGS Rotterdam. Splitting samples three ways at loadout (seller, buyer, umpire) needs to be written in, not assumed.

Payment terms. Irrevocable LC at sight against shipping documents is standard. But for a critical mineral off-take of any size, buyers should expect to post the LC 30 days before the shipping window opens, not on the day of loading. Banking in Pakistan is workable but not fast, and dollar liquidity has been tight since 2022. Plan for it.

Duration. Honestly, I think three to five years is the sweet spot right now. Ten-year deals sound good in a boardroom but neither side really knows what antimony or tungsten pricing looks like in 2033. Shorter tenor with renewal options gives both sides room.

One thing I got wrong

Early on I used to think the price mechanism was the hardest thing to negotiate. It isn't. Index-linked pricing is a solved problem — pick your benchmark, agree the discount, move on.

The hard part is the operational clauses. Delivery points, quality specs, penalty schedules for out-of-spec material, what happens when a shipment sits at Karachi for eleven days because of a customs backlog. That's where deals actually live or die, and that's where I now spend most of the negotiation.

If you're a buyer looking at Pakistan and you want a term sheet that reflects the reality on the ground rather than a template from a Chilean copper deal, that conversation is open. What are you actually trying to secure — tonnes, or supply certainty? Because the answer changes the contract.