How Project Vault is being built: Hotter Commodities

A new US-backed structure is attempting to turn critical minerals inventory from a cost burden into a strategic asset. Project Vault combines pooled demand, private governance and capped-return financing to create a shared supply buffer, and potentially the first steps toward price discovery in fragmented markets.

Key takeaways:

  • Project Vault introduces a market-based solution to inventory underinvestment, shifting the burden away from OEM balance sheets.
  • Demand aggregation, not stockpiling, is the core innovation — creating a pooled, industry-led supply buffer.
  • The structure has the potential to generate early-stage price signals and liquidity in fragmented critical minerals markets.

Project Vault: restructuring critical minerals inventory

Washington has finally put structure around a problem the market has spent years sidestepping: nobody wants to pay to hold critical minerals inventory until they need it. Now it has a workaround.

Enter Project Vault.

Launched in February with up to $10 billion in US Export-Import (Exim) Bank financing and a broader blended capital structure, Vault is neither an upstream subsidy nor a traditional strategic stockpile.

Fastmarkets learned more details of the project during a trip to Washington, DC last week.

Vault is a privately run inventory and financing layer designed to sit in the middle of critical minerals supply chains, acting as a pooled demand system and liquidity buffer. Think of it as inventory without the balance sheet pain.

The architecture is deliberate. Exim provides the financing anchor, but Vault itself is structured as an independent, privately governed vehicle with Exim in a non-controlling, board observer capacity.

Original equipment manufacturers (OEMs) will participate directly, while trading houses such as Hartree Partners, Traxys and Mercuria are set to provide sourcing, logistics and working capital capacity. Mining and trading groups, including Glencore, are expected to connect the structure back to upstream production and recycling flows.

The participant set is not fixed and is expected to broaden as liquidity requirements increase, people familiar with the project told Fastmarkets.

From stockpile to demand aggregation

The most important shift Vault introduces is not financial; it’s demand aggregation.

Manufacturers under-hold inventory not because they misunderstand risk, but because capital markets structurally penalize balance sheet inventory intensity.

Just-in-time efficiency and working capital discipline have steadily removed buffer stock from the system, leaving supply chains exposed precisely when liquidity disappears from commodity markets.

Vault is built to reverse that trade-off without forcing it back onto individual OEMs’ balance sheets.

According to the people, OEMs commit to the structure, pay an entry fee, and gain access to a centrally held pool of critical minerals that functions as contingent supply in disruption scenarios such as export controls, logistics breakdowns or geopolitical shocks.

The defining break from legacy stockpiles is governance: allocation, form (oxide, metal, refined product) and cycling are determined by manufacturers rather than a central authority.

In other words: industry, not government, decides what matters and when.

That design converts withdrawals into a replenishment obligation, with material required to be replaced dollar-for-dollar, structurally embedding a recurring demand loop.

In practice, that replenishment flow is expected to skew toward US and allied supply chains, creating a persistent offtake channel for upstream developers.

That’s not a stockpile. It’s a demand engine.

Trading: introducing liquidity into a fragmented market

People familiar with Vault say participants will be able to trade entitlements among themselves.

That means if one OEM is long a material it no longer requires, it can sell exposure to another participant. Those transactions are expected to be transparent, creating observable price points in markets where forward curves are thin and spot pricing dominates.

One detail that materially changes the market interpretation is anonymity: trading counterparties won’t know which OEM sits behind a given position.

Demand will be pooled through a single structure, reducing bilateral information leakage and limiting the ability to infer industrial positioning from order flow.

Perhaps more importantly, by routing storage, drawdown and internal transfers through a single structure, it will create a consolidated layer of liquidity. That allows for material to be reallocated and generate referenceable pricing signals from previously fragmented demand.

Vault won’t create a full futures curve, but it will introduce repeated, structured transactions that could begin to form reference pricing outside dominant Chinese-linked benchmarks, US government sources say.

Not a market — yet. But closer than anything else in critical minerals.

Want to learn more about what is happening at the cutting-edge of critical minerals and battery raw materials? Listen to our Fast Forward podcast series for insight, debate and news from the major players. 

Trading houses bridge physical and financial flows

The role of trading houses is not incidental.

Firms like Hartree, Traxys and Mercuria operate across production financing, storage and marketing, bridging the gap between physical supply chains and capital markets. Glencore-type integrated players extend that further, linking mining output, trading flows and recycling loops. Vault embeds that infrastructure rather than replicating it.

Unlike the Strategic Petroleum Reserve or the National Defense Stockpile, there is no central authority determining what is stored or when it is released. Drawdowns are initiated by OEMs, not bureaucrats.

That distinction matters: Vault is pull-driven, not policy-triggered.

Financing: infrastructure, not arbitrage

The financial structure reinforces that intent.

Returns across both debt and equity are capped, people familiar with the structure say, preventing Vault from functioning as a traditional commodity investment or arbitrage vehicle. Any surplus generated through operating fees will be recycled back to OEM participants if inventory buffers aren’t used.

The goal is not to extract trading profit from volatility, but to recover the cost of maintaining a shared strategic buffer. In that sense, Vault is priced and governed as infrastructure; closer to a utility for industrial resilience than a market-facing trading entity.

The timeframe reflects the underlying constraint: supply chains are being rebuilt on multi-decade horizons, while exposure risk is immediate. Commitments to the project run roughly a decade, with financing extending further, positioning Vault as a bridging mechanism rather than a permanent fixture.

A shift from stock to system

But the more consequential implication sits elsewhere.

If inventory becomes pooled, tradable and partially priced through internal transactions, it stops being passive stock and begins to function as a coordination layer linking manufacturers, traders and upstream producers.

That’s the real shift.

The departure from earlier policy frameworks isn’t the stockpile — it’s the architecture around it.

Inventory has always shaped markets. This is the first attempt in the West to organize it, price it and use it to direct demand rather than just absorb shocks.

In Hotter Commodities, special correspondent Andrea Hotter covers some of the biggest stories impacting the natural resources sector. Read more coverage on our dedicated Hotter Commodities page here.

What to read next
US light vehicle production averaged 10M units per year in 2021 through 2025 with most years finishing above 10M units.
Procurement and supply chain managers face new challenges as wooden pallet prices surge. Understand the key drivers behind this volatility to improve budget planning and supplier negotiations.
South Korea has stepped up its efforts to support its steel sector, amid escalating tensions in the Middle East and tariff pressures elsewhere, by including the sector in a $54 billion support package for key industries in the country, Fastmarkets understands.
As US automotive OEMs localize supply chains and accelerate EV rollout, margin pressure is intensifying across steel, aluminium and battery inputs.
Despite mixed signals on box demand, significant capacity reductions and rising costs are enabling US containerboard producers to push for a second major price increase this year.
The US company EVelution Energy outlined its plans to produce 3,000 tonnes per year of cobalt metal in the United States from Congolese hydroxide, speaking with Fastmarkets on Wednesday May 13 on the sidelines of the Cobalt Congress that took place in Madrid, Spain (May 12-13).