Actionable insights — Washington's New War on China's Mineral Grip

1. Read a procurement rule as a demand contract, not as sentiment

1. When a government restricts what a buyer may source, skip the geopolitics and answer three mechanical questions: who is legally obliged to comply, by what date, and what must they do instead. A ban with a named compliance population and a hard date is a purchase order with a lag; a "policy priority" without either is noise.
2. Check whether the restriction already existed and was being evaded. A rule that has been on the books but routinely waived is not yet demand — the investable moment is when the escape hatch closes, because that is when compliance becomes cheaper than exemption.
3. Read the replacement obligation literally. If the buyer must now qualify an alternative supplier — a documented, audited process — then every credible non-restricted producer gains a queue of counterparties it did not have to win commercially.
4. Size the compliance population, not the market. "Every prime and subcontractor, at every tier" is a far larger buyer set than the handful of primes a headline names, and each tier must independently source compliant material.

• Waiver/exemption regimes being tightened rather than new bans; the words "at every tier" or "bill of materials" in a rule; rulemaking clocks (here 180 days for the Pentagon to write the implementing rules) that tell you when the obligation becomes concrete; suspension/termination as the enforcement penalty, which converts compliance from optional to existential.

2. Build the policy-cliff calendar and position into the dates, not the news

1. Extract every hard date from the directive stack on both sides of the dispute and put them on one timeline: bans, truce expiries, waiver sunsets, rulemaking deadlines, quota renewal windows.
2. Look for clustering. Two or more cliffs falling within a quarter of each other compress the adjustment period — buyers cannot stage the transition, so procurement is pulled forward and prices move before the date, not on it.
3. Distinguish a cliff you can hedge from one you cannot. A physical substitution deadline (qualify a new smelter, build a magnet line) has a lead time measured in years; if the lead time exceeds the time to the cliff, the shortfall is already locked in regardless of what happens diplomatically.
4. Note the counterparty's own clock. A truce that pauses escalation for a fixed period is a call option the other side holds; the expiry date is a scheduled volatility event.

• Truce/pause expiry dates; statutory waiver sunsets; the gap between a compliance deadline and the build time for the substitute capacity; whether the restricting party has already demonstrated willingness to widen controls (April 2025 → October escalation) rather than merely threaten.

3. Follow the lending authority and the vehicle — not the headline grant

1. For each policy action, separate appropriated money (a fixed grant) from lending or guarantee authority (a balance-sheet capacity that can be drawn repeatedly). The second is an order of magnitude larger and tells you the real ceiling on the buildout.
2. Identify the vehicle, because the vehicle names the eligible borrower. An agency chartered for overseas development lending redirected to domestic mines, or a defense-adjacent capital office opened to mining projects, tells you exactly which asset class just became financeable.
3. Read the eligibility qualifier as the screen. If financing attaches to permitted projects, permitting status — not resource size — is the variable that decides who gets the money.
4. Track the stage of the chain the money targets. Federal capital flowing to processors, refiners and recyclers rather than to mines means the midstream is where margin and strategic scarcity are being created.

• Government equity stakes, loan guarantees and offtake commitments rather than subsidies; the word "permitted" as an eligibility gate; agencies acquiring new mandates outside their charter; physical state assets (land, bases, ports) being opened to private processors — a non-cash subsidy that shortens siting timelines.

4. Find the material a rule binds without naming it

1. Read past the covered-materials list. A trace-to-source obligation catches everything in the product, so a material can be fully constrained without ever appearing on the schedule of controlled items.
2. Test each unlisted input against the same two questions used for the listed ones: does the buyer now have to document its origin, and is the incumbent source in the restricted jurisdiction? Two yeses mean identical economics with less attention on it.
3. Check the post-production side too. An export block on scrap, spent product or byproduct is a supply-side control that keeps domestic feedstock captive — it advantages whoever can process that feedstock at home and disadvantages exporters of it.
4. Stack the other policy levers already aimed at the same material (tariffs, investment incentives, plant awards). Where trade controls, tariffs and capital subsidies converge on one unlisted input, treat it as a covered material in practice.

• Trace-to-source or bill-of-materials language that is material-agnostic; scrap/spent-product export restrictions; tariff structures that pair a punitive rate with an investment-linked discount (a build-here incentive, not just a wall); domestic plant awards in industries with almost no remaining capacity.

5. Rank dependency gaps by how fast they can physically be closed

1. For each material, write down two numbers: the importing country's import share and the adversary's refining share. High on both means the constraint binds at the midstream, where new capacity takes the longest to build.
2. Separate "we don't mine it" from "we can't process it." A country with domestic ore but no refining can respond faster than one with neither — and a country with zero domestic production of a material has no base to scale from at all.
3. Rank by closure time, not by headline dependence. The materials that cannot be re-sourced before the compliance deadline are the ones where price, not policy, does the rationing.
4. Apply the same test to the substitute chain the policy is trying to build: does an ex-restricted-jurisdiction producer exist today at commercial scale, or is the "alternative" still a permit?

• USGS-style import-reliance percentages paired with refining-share data; materials at 100% import reliance with zero domestic output; the mine-share vs refine-share spread as the tell for where the bottleneck actually sits; whether announced ex-China capacity is producing, commissioning, or permitting.

6. Screen at the intersection the policy names, then pick the point on the chain

1. Take the directive's own stated purpose as the screen. When policy is explicitly aimed at where defense and energy infrastructure intersect, that intersection — not the broad commodity — is the universe to search.
2. Add the demand vector that is independent of the policy. A material pulled by both a mandated buyer (defense procurement) and an unmandated one (AI data-center buildout) has two bids under it; policy risk cuts only one of them.
3. Then choose the position on the chain deliberately — mine, smelter/refiner, recycler, fabricator — because the directives hit each stage differently: sourcing rules reward qualified producers, export blocks on scrap reward domestic recyclers, and lending authority rewards permitted projects.
4. Do not let a price near its record talk you out of a policy-driven re-rate on its own; check whether the marginal supply that would cap it can physically arrive inside the policy window.

• A publisher naming the sector and the timing of its own upcoming pick (a pre-announced research focus is itself a dated catalyst); overlapping mandated + commercial demand on one material; which chain stage a given directive actually pays; permitted-and-financeable projects as the shortlist rather than resource-size leaders.