The Kengly Letter
Xi's Export-Control Ladder — The NdPr Runway

Xi's Export-Control Ladder — The NdPr Runway

7 May 2026 · 6 min · issue

The thesis in one line. The market is pricing the April 2025 heavy rare earth licensing round as a terminal retaliation event, but the equilibrium is a sequenced escalation toward NdPr controls — which means Western refining capacity is undervalued by the duration mismatch.

Beijing's export-control architecture is not reactive retaliation but a calibrated escalation ladder designed to preserve optionality while eroding Western semiconductor supremacy. On August 1, 2023, China imposed export licensing requirements on gallium and germanium products under national-security framing (PRC Ministry of Commerce — Announcement No. 23 of 2023). The sequence tightened on October 20, 2023, when graphite items joined the control list (PRC Ministry of Commerce + General Administration of Customs — Joint Announcement on Graphite Export Controls). By December 3, 2024, Beijing escalated to outright bans on gallium, germanium, antimony, and super-hard materials exports to the United States (PRC Ministry of Commerce — Statement on Export Ban to the United States). The April 2025 addition of heavy rare earth elements — including samarium, gadolinium, terbium, dysprosium, lutetium, scandium, and yttrium — to the licensing regime (PRC Ministry of Commerce + General Administration of Customs — Joint Announcement on Heavy Rare Earth Export Licensing) completes the penultimate rung. Each step correlates loosely with US semiconductor restrictions but the timing is sovereign — which means Xi controls the cadence, not Washington. The ladder forces Western stockpiling that depletes fiscal reserves without triggering immediate FEOC-compliant reshoring completion under the US Inflation Reduction Act (US Congress — Public Law 117-169) or the EU Critical Raw Materials Act (EUR-Lex — Regulation (EU) 2024/1252). This is Managed Friction with Chinese characteristics, and the heavy rare earth basket represents the final warning before NdPr.

Three pillars

One. Xi's dominant strategy is a sequenced export-control ladder that preserves optionality while maximising leverage. He does not seek a total supply cutoff that would trigger immediate FEOC-compliant reshoring completion. Instead, he sequences gallium/germanium → graphite → antimony → US-specific bans → heavy rare earths to test Western inventory buffers without forcing capital allocation to Western refining. Each rung is calibrated to create supply anxiety sufficient to raise input costs for Western tech and defence sectors but insufficient to trigger the IRA's strict FEOC prohibitions that would accelerate non-China refining construction. The spacing between rungs allows Western bureaucracies to exhaust their rapid-response capital, ensuring that when the next control lands, the policy response is slower and more fragmented.

The market views the April 2025 heavy-REE move as reactive to US tariffs, but this is the mispricing. Xi's ladder is offensive, not defensive. He maintains plausible deniability by keeping each step technically below the threshold of declared economic warfare. Which means the next rung is NdPr — the workhorse permanent-magnet inputs for EV motors and wind turbines — once Western heavy-REE inventories deplete and before Mountain Pass stage-2 achieves commercial separation. So the equilibrium here is a Managed Friction regime that escalates by regulatory precision rather than shock, with NdPr controls as the terminal move that forces Western OEMs to pay scarcity rents or halt production. The market is pricing this wrong by assuming the ladder stops at heavy rare earths.

Two. Wang Yi and Pan run complementary strategies that sustain the equilibrium through narrative discipline and monetary stability. Wang Yi frames each control as "national security" rather than retaliation, which means Beijing maintains WTO-ambiguous legal cover while avoiding the diplomatic isolation that would accompany declared economic warfare. This framing forces Western diplomats to respond through technical working groups rather than Article XXI disputes, buying time for the ladder to advance. The "national security" rhetoric also provides domestic cover for SOEs to prioritise domestic downstream users, ensuring Chinese EV and wind-turbine makers retain supply priority as exports tighten. This narrative control prevents the formation of a unified Western counter-bloc that could accelerate CRMA implementation or IRA funding.

Pan manages FX volatility through the retaliation cycles, ensuring yuan stability that prevents capital flight from disrupting the timeline. The PBOC intervenes to smooth volatility when export-control announcements hit, which means the escalation ladder proceeds at Beijing's pace, not Washington's. The market is pricing the heavy-REE shock as a discrete event when it is actually a structural shift in supply-chain bargaining power. So the equilibrium here is a sustained friction state where narrative and monetary policy extend the timeline long enough to make Western stockpiling economically exhausting while keeping Chinese export revenue intact.

Three. Western refiners — MP Materials, Lynas Rare Earths, and JOGMEC offtake partners — pursue counter-bloc reshoring under statutory mandates, but the market applies standard mining discounts rather than strategic scarcity premiums. The IRA's FEOC definitions and the EU CRMA's 2030 domestic-processing targets create inelastic demand for non-China separated rare earths, yet Mountain Pass stage-2, Lynas Kalgoorlie, and Iluka Eneabba remain 2026-2027 commissioning stories. Western OEMs have not yet secured offtake for NdPr outside China at scale. The refining bottleneck is not mining but separation chemistry, and Chinese engineers dominate solvent-extraction intellectual property. Western projects face technical and permitting delays that compress the probability of early commissioning.

When NdPr controls land, the scarcity rent will accrue to the only available capacity. The market is pricing this wrong by treating Western refining equity as commodity beta rather than the sole alternative supply source during a structural China embargo. The discount implies commissioning certainty that does not exist in mining jurisprudence — which means the optionality value of early-stage Western refining is being given away for free. So the equilibrium here is scarcity-rent capture by the only players with commissioned capacity before the NdPr threshold is crossed. The mispricing is in the duration mismatch between Western policy urgency and equity-market patience.

Position card

Long: Ex-China rare earth refining basket — MP Materials (NYSE:MP), Lynas Corporation (ASX:LYC), VanEck Rare Earth/Strategic Metals ETF (AMEX:REMX). Size for 12-24 month conviction.

Short: Single-name EV motor manufacturers without disclosed 3-year NdPr reshoring plans.

Nash thesis: The equilibrium is sequenced escalation toward NdPr export controls, which means Western refining capacity captures scarcity rent as the only available supply source during the 2026-2027 commissioning gap.

Falsifier: Wrong if China rolls back the heavy-REE export licensing regime within 6 months OR Western refining capacity (Mountain Pass stage-2 + Kalgoorlie + JOGMEC offtake) hits commissioned production milestones >12 months ahead of schedule.

Time horizon: 12-24 months — matched to the project commissioning cycle for Western non-China refining.

Red-team responses

Response to Attack 1 — "The Ladder Has Already Peaked"

The argument that NdPr controls would force a coordinated G7 response strong enough to deter Beijing assumes a unified Western counter-bloc exists. It does not. The heavy-REE round in April 2025 produced co-ordinated rhetoric but disjoint operational responses — DPA Title III tranches arrive in months, EU CRMA targets are 2030-anchored, JOGMEC offtake is project-specific. Beijing reads this fracture clearly. NdPr controls would be calibrated like every prior rung: technically national-security-framed, narrowly scoped (specific element categories or end-use carve-outs), with domestic SOE allocation preserved. The G7 invocation cost remains below the leverage gain. The "ladder peaked" attack also requires Beijing to forfeit the highest-value rung voluntarily, which contradicts every observed step in the sequence. Position survives.

Response to Attack 2 — "Reshoring Arrives Before Escalation"

This is the strongest attack and the one to take seriously. MP shipped first NdPr metal in 2024, Lynas's Kalgoorlie cracking-and-leaching is operational, and Iluka has DPA Title III funding. The fair pushback: shipping first metal is not commercial separation at scale. Mountain Pass stage-2 commercial production timelines have been adjusted multiple times in MP investor disclosures. Lynas Kalgoorlie is upstream of separation chemistry — the bottleneck is solvent-extraction IP and permitting, not feedstock. Western capacity arriving "before" the next escalation has been the consensus thesis since 2023 and has been wrong each time China has stepped up the ladder. The position is calibrated to a 12-24 month window precisely because shorter horizons over-index on commissioning press releases and longer ones lose to recession-cycle macro beta. If reshoring genuinely arrives in 12 months, the falsifier triggers and the trade exits.

Response to Attack 3 — "The Equity-Beta Trap"

REMX correlation with broader risk-off is real and is the single biggest non-thesis risk. The pure-play physical alternative — NdPr inventory or NdFeB magnet stockpile contracts — is illiquid, custody-constrained, and operationally outside the readership's portfolio practice. Equity proxies are the practical expression. The mitigation is the falsifier itself: a sharp REMX drawdown driven by macro beta (Fed cuts compressing all commodity-equity multiples) does not invalidate the structural thesis; a price-based stop-loss does not apply here. The trade is sized for conviction, not for stop-driven exit. Investors uncomfortable with macro-beta correlation should treat the basket as a smaller satellite position, not a core allocation. Equity-beta exposure is acknowledged, not denied.

Sources

Get next week's issue

One equilibrium read per week. Free during the soft-launch.