The Magnet Suspension Runs Out on 10 November: What to Place Before It Does

On 7 November 2025 China announced a temporary suspension of the second wave of its rare-earth export controls, running until 10 November 2026, a step recorded in the European Parliament Research Service briefing of 24 November 2025. The word that carries the weight is suspension. The measures were not withdrawn, repealed or renegotiated out of existence. Their text stands, their scope stands, and their application is paused on a clock that expires inside this calendar year.

What is paused is unusually broad. The expanded controls require foreign companies to obtain approval to export magnets containing even trace amounts of China-sourced rare earths, or produced using Chinese extraction, refining or magnet-making technology (Down To Earth, 2026). That second limb is the one procurement teams underestimate. It reaches a magnet that never physically passed through China, on the basis of the process technology used to make it, which means a supplier in a third country can sit inside the control net while presenting a clean country-of-origin certificate.

For an Indian EV, electronics or auto risk manager this is a rare planning situation. Most supply-chain shocks arrive without notice, which is why they are hard to insure and harder to budget. This one has a published expiry date. A dependency map, a wording review and a placement decision can all be scheduled backwards from 10 November rather than reconstructed afterwards. As at mid-September 2026 that leaves roughly 60 days, which is short for a specialty placement and comfortable for a wording review, so the sequencing decisions matter more than the appetite decisions.

India is in a better position than most importers. China lifted curbs on rare-earth magnet exports to India, confirmed during Foreign Minister Wang Yi's visit to India (Trading Economics, 2026). That is a real change in the flow of goods and it should be reflected in how a buyer prices its own risk internally.

It is not, however, a change in the rule. The relief operates on the approval of shipments to India. The extraterritorial technology limb, the trace-content limb and the licence architecture behind them remain in the suspended text, ready to resume on 10 November unless Beijing extends the pause or converts it into something permanent. A bilateral accommodation granted in one diplomatic climate can be reweighted in another, and it does not travel with a magnet bought from a Vietnamese, Korean or Thai supplier whose process line uses Chinese refining or magnet-making technology.

The market view supports treating this as reduced probability rather than removed exposure. S&P Global expects rare earth supply bottlenecks to persist through 2026 (S&P Global, 27 January 2026). A buyer whose board has quietly written the exposure off after the India relief has changed its likelihood estimate without changing its loss mechanism, its inventory position or its contracts. Underwriters will not make that mistake at renewal, and neither should the risk register.

The insurance problem is structural. The workhorse business-interruption cover in the Indian market is Fire Loss of Profits written against a Standard Fire and Special Perils foundation, and its contingent business interruption extension inherits the same material-damage proviso. Both respond where physical loss or damage by an insured peril interrupts the business, whether at the insured's own premises or at a supplier's.

A licence refusal produces none of that. The Chinese processor's plant is intact, staffed and running. The Indian buyer's plant is intact and idle. The proximate cause is an administrative decision by a foreign ministry, and the standard exclusions for government action, confiscation and cessation of supply sit ready to catch anything the trigger clause does not already exclude. The revenue loss is genuine, the standing charges continue, and the flagship policy is structurally blind to it. We set out the full analysis of this gap, including which specialty products do respond, in our earlier piece on insuring supply disruption from China's export controls.

The short version for the covers that answer a licence-driven cutoff: trade disruption insurance written on a named-peril basis that expressly lists government action, embargo and licensing restriction; non-damage business interruption extensions where the damage proviso has been removed for defined triggers; and parametric structures that pay a pre-agreed sum on an objective index event. Each must name the export-control peril. A wording that is silent on regulatory cutoff is read against the buyer at claim stage, every time.

Specialty placements of this kind normally want 90 to 120 days of runway because they are brokered into offshore markets, priced off a bespoke submission, and often reinsured. From mid-September the buyer does not have that. The realistic menu narrows to three routes, in ascending order of difficulty.

1. Mid-term endorsements to the existing programme. Fastest and cheapest, because the incumbent insurer already holds the risk file. The gain is usually narrow, an extension on supplier failure or public authority, and the policy wording has to be read against a foreign export ministry rather than an Indian local authority.
2. Parametric supply-chain cover. Binds faster than indemnity cover because there is no loss-adjusting architecture to negotiate, only an index, a trigger definition and a payout table. The trade-off is basis risk: the index event may not coincide with the buyer's actual loss.
3. A standalone trade disruption placement. Achievable in 60 days only where the submission is already assembled. Starting the dependency map in October and expecting terms by early November is not a plan.

Two mechanical points decide whether a policy bought in October is worth anything on 11 November. The first is inception date, since cover bound to incept on 1 December is irrelevant to an event on 10 November. The second is the waiting period. These covers commonly carry a waiting period expressed in days, so a 30-day waiting period on a cutoff beginning 10 November pays nothing until 10 December, and a buyer holding 45 days of magnet inventory may find the waiting period and the inventory buffer are doing the same job twice.

Underwriters differentiate on evidence and default to the harshest assumption without it. Four documents move terms on a magnet-dependency submission.

• An element-level bill of materials showing which controlled elements enter which products, in what quantity, through which named and unnamed suppliers, and at which tier.
• Inventory days by element, with the reorder lead time for each, since terbium and dysprosium behave differently from bulk neodymium.
• Supplier process provenance, addressing the technology limb rather than only country of origin, including any qualified non-Chinese sources and how far through qualification they are.
• Contract terms with magnet suppliers, specifically whether a licence refusal counts as force majeure, who bears the cost, and how much sits outstanding as prepayment or advance.

That last point creates a separate exposure that is often missed. Where an Indian OEM has paid deposits against a magnet supply contract and a licence action frustrates delivery, the loss is financial rather than operational, and it belongs with contract frustration or structured credit cover rather than with business interruption. It also carries a disclosure duty: known supplier concentration that was not disclosed at inception is the fastest route to a declined claim.

A structured resilience assessment turns this material into something an underwriter can rate rather than a set of assertions, and our framework for supply-chain resilience scoring sets out how to build one that maps to underwriting questions rather than to procurement dashboards.

India is not only buying its way out of this. The country is close to approving a Rs 7,300 crore incentive scheme for rare-earth magnet manufacturing, alongside Rs 1,500 crore for critical-mineral recycling under the National Critical Minerals Mission 2026-2031 (Down To Earth, 2026). Money at that scale produces a pipeline of first-of-a-kind plants, and each of them is an insurance placement twice: once during construction, once in operation.

The construction phase is an erection all risks placement with three components that must be sized together. The EAR policy covers the physical works. A marine cargo and inland transit section covers the imported process equipment, which for a magnet line means strip-casting units, jet mills, vacuum sintering furnaces, inert-atmosphere handling systems and coating lines, much of it long-lead and single-sourced. A delay in start-up or advanced loss of profits section covers the revenue consequence of a covered physical loss pushing commissioning back.

Delay in start-up is where these projects are most commonly underinsured. The sum insured for a DSU section is gross profit over the anticipated delay period, and on a first-of-a-kind plant with imported long-lead equipment a realistic replacement timeline for a damaged sintering furnace can exceed the indemnity period a promoter instinctively picks. There is a second exposure that DSU does not answer at all: incentive schemes are milestone-linked, and a missed capacity or commissioning milestone can cost scheme benefits that no standard wording indemnifies. That belongs in the risk register as a retained exposure, stated plainly to the board.

The testing and commissioning clause deserves close reading. Cover during hot testing is frequently limited, sub-limited or conditional, and on a plant where the first magnet pressing run is also the first time the powder handling system operates at rate, that is exactly the window in which loss is most likely.

Once the plant runs, the placement becomes property, machinery breakdown, business interruption and liability, and the rating is driven by process hazards that most Indian property underwriters have not surveyed before.

Magnet manufacturing

Sintered neodymium-iron-boron production passes through hydrogen decrepitation and jet milling, which convert alloy into a very fine powder. Fine rare-earth alloy powder is reactive and is handled under inert atmosphere for that reason, so the underwriting questions are about dust hazard analysis, inerting integrity, explosion venting or suppression, hydrogen supply and zoning, and physical separation between powder handling and the rest of the plant. Vacuum sintering furnaces are high-value, long-lead items, which pushes them into the machinery breakdown conversation and makes spare-part lead time the real determinant of the business-interruption indemnity period.

Recycling

A critical-mineral recycling line has a different shape. Incoming feedstock is variable and poorly characterised, which is a hazard in itself. Processing typically involves shredding, thermal demagnetisation and chemical recovery, so the exposures move towards chemical handling, effluent, storage of recovered material and pollution liability alongside the fire and explosion questions.

Neither process has an Indian loss history to rate against. Underwriters will lean on international benchmarks and, more importantly, on survey-driven conditions and warranties, which means the operator who engages early with a risk engineer buys better terms than the one who submits a plan drawing at renewal. The pattern is the same one seen on India's first cell plants, set out in our EV battery gigafactory risk profile, where absence of domestic loss data pushed the whole placement onto engineering evidence.

The work divides cleanly between buyers exposed to a cutoff and promoters building capacity. For a magnet-dependent EV, electronics or auto buyer, the sequence for the remaining weeks is:

1. Build the element-level dependency map and days-of-cover inventory position. Two weeks, procurement-led, no insurer involvement needed.
2. Read the existing property, BI and CBI wordings against a licence-refusal scenario specifically, and write down where each fails. This is a wording exercise, not a market exercise, and it can run in parallel.
3. Decide the retention. Some of this exposure will be retained whatever is bought, and deciding that deliberately is better than discovering it at claim stage.
4. Take the map to market for the narrow, fast options first, mid-term endorsement and parametric, before committing to a standalone trade disruption submission that may not bind in time.
5. Check inception date and waiting period against 10 November on anything quoted, and confirm the export-control peril is named rather than implied.
6. Set the review for the first week of November so the board sees a position before the date, not after it.

For a promoter building a magnet or recycling plant under the National Critical Minerals Mission, the sequence is different and longer. Engage a risk engineer during design rather than at commissioning, size the DSU indemnity period against genuine equipment replacement lead times rather than a default twelve months, place marine cargo and EAR as one conversation so the long-lead imported equipment is not covered twice at one end and not at all at the other, and state the milestone-linked incentive exposure explicitly as retained.

Both exercises rest on the same thing, which is knowing exactly what the wordings say. The difference between a programme that answers a licence freeze on 11 November and one that declines on the material-damage proviso sits in a handful of clauses, and those clauses vary between insurers and between placements.