Supply chains are the circulatory system of global trade. Goods and data flow like lifeblood to keep the world's economy alive. The problem is that this system has a critical vulnerability — its main arteries are concentrated in the hands of a few. One company in the Netherlands is the only source of the machines that make advanced chips. One country refines 19 of 20 critical minerals. One strait carries a fifth of the world's oil. When the circulatory system has blockages, the whole body suffers. This letter is about what those blockages are, who controls them, and how the world is building immunity.
Not investment advice. Data sourced from ASML filings Q1 2026, US Critical Minerals Strategy, Pax Silica framework documents, IMO nuclear shipping regulatory documents, and Radia/LTA Research company disclosures. All figures current as of June 2026.
The human circulatory system is a model of elegant engineering. The heart pumps blood through a hierarchy of vessels — from the great arteries that carry massive volumes at high pressure, through progressively smaller branches, down to the capillaries that reach every individual cell. The system is redundant at the margins and concentrated at the centre. A blockage in a capillary is local. A blockage in a coronary artery is fatal.
The global supply chain follows exactly this architecture. At the arterial level: the great shipping lanes — the South China Sea, the Strait of Hormuz, the Suez Canal, the Panama Canal — through which the bulk of global trade flows. These are the coronary arteries. At the mid-level: the major industrial clusters — Taiwan's semiconductor fabs, China's battery manufacturing, Germany's chemical plants — that take raw materials and transform them into the components that the rest of the world needs. At the capillary level: the individual suppliers, logistics companies, port operators, and last-mile distributors that connect production to consumption. The system reaches everywhere. It is also dangerously concentrated at its core.
For forty years, the global economy optimised this circulatory system for efficiency. Just-in-time inventory management — the principle that goods should arrive exactly when needed, with minimal stockpiling — reduced costs across every industry. Global sourcing — the principle that components should be manufactured wherever they can be produced most cheaply — reduced costs further. The result was the most efficient supply chain system in human history. It was also the most fragile. COVID-19 was the stress test that revealed the fragility. The world failed it comprehensively. A semiconductor shortage shut down automotive plants. A container ship wedged in the Suez Canal disrupted global trade for a week. PPE shortages killed healthcare workers. The most efficient system ever built turned out to be a system that could not absorb a shock.
The Strait of Hormuz is a 21-mile-wide passage between Iran and Oman through which approximately 20% of the world's oil supply travels daily — roughly 17–18 million barrels. It is also the only maritime exit for the oil production of Saudi Arabia, the UAE, Iraq, Kuwait, and Iran. There is no realistic alternative route for most of this volume.
The NGE Commodity Compass has tracked this chokepoint through the Iran nuclear deal of June 2026 — when Hormuz reopening compressed Brent oil from $126 to $75–82, saving India alone an estimated $67 billion annually on its import bill. That single event demonstrates the scale of economic consequence that flows from control of 21 miles of water. When Hormuz is open, $67 billion stays in India's economy annually. When it is closed — or even threatened — oil prices spike, shipping costs surge, and every economy that runs on energy (which is every economy) feels the pressure within weeks.
Hormuz is not unique. The Suez Canal carries approximately 12% of global trade. The Panama Canal carries 5% of global trade and is now subject to water level restrictions driven by climate change — a physical chokepoint being tightened by a physical force that no government controls. The Strait of Malacca — through which 40% of global trade passes between the Indian Ocean and the South China Sea — is the primary artery of Asian commerce. Each of these passages is a point where a single actor, a single event, or a single climate pattern can sever the flow.
In a small city in the Netherlands called Veldhoven, a company called ASML manufactures the only machines in the world capable of producing advanced semiconductor chips. Extreme ultraviolet lithography — the technology that etches transistors at nanometre scale onto silicon wafers — requires light at a wavelength of 13.5 nanometres. Generating that light requires lasers hitting droplets of molten tin at the right energy level. The engineering is so complex that after Nikon and Canon both attempted it and failed, ASML was left as the sole supplier.
ASML has 100% market share in EUV lithography — the only technology capable of producing the advanced logic and high-bandwidth memory chips that power AI. Every advanced chip in every AI data centre in the world, every smartphone using a sub-7nm process, every cutting-edge military system — all depend on machines that come from one company in one city in one country. ASML's full-year 2025 revenue was €32.7 billion at 52.8% gross margin, with a backlog of €38.8 billion. It guides to €44–60 billion in 2030 revenue as High-NA EUV ramps.
The geopolitical stakes are explicit. The Netherlands joined the Pax Silica AI supply chain group in 2026, making ASML's export policy a formal instrument of allied technology strategy. China — which accounted for 36% of ASML system sales in Q4 2025 — has been progressively cut off from EUV access, with its share falling to 19% by Q1 2026. China is now attempting to develop its own EUV capability at Huawei's Dongguan facility, using laser-induced discharge plasma technology — targeting mass manufacturing in 2026. The race to replicate ASML is the semiconductor race that defines the decade.
The green economy and the digital economy both require specific minerals in large quantities. Lithium for batteries. Cobalt for cathodes. Rare earth elements for magnets in wind turbines and electric motors. Gallium and germanium for semiconductors. Graphite for battery anodes. China refines approximately 19 of the 20 minerals identified as critical and strategic by the US and EU — giving Beijing leverage over the energy transition and the digital economy simultaneously.
This leverage is not theoretical. In July 2023, China imposed export restrictions on gallium and germanium — two materials critical to semiconductor manufacturing — in direct response to US export controls on semiconductors. The message was precise: cut us off from advanced chips, and we cut you off from the materials needed to make them. In December 2024, China expanded restrictions to cover antimony, critical for defence applications, and rare earth processing technologies. Each restriction was a demonstration that the material chokepoint is as powerful as the technological one.
The response has been building for years but is accelerating. The US Inflation Reduction Act provides incentives for domestic rare earth production and processing. Australia — the world's largest lithium producer — is building refining capacity. The Clarion-Clipperton Zone of the Pacific floor, as documented in Letter 81, holds more cobalt, nickel, and manganese than all known terrestrial reserves. Indonesia's nickel gambit (Letter 73) is the template — refusing to export raw material and forcing processing to happen domestically. The mineral chokepoint is being dismantled, but it takes 10–20 years to build a refining industry, and China has had 30.
The structural shift from efficiency to security is the most important change in global trade policy since the end of the Cold War. For decades, the doctrine of comparative advantage — the economic principle that every country should specialise in what it produces most efficiently and trade freely for everything else — governed supply chain decisions. The result was extraordinary: global poverty fell by more in the 30 years after 1990 than in the preceding 300. The efficiency gains from global supply chains funded development on a scale that no aid programme ever matched.
But efficiency-maximised supply chains have a fatal flaw: they optimise for normal conditions. They have no redundancy, no buffers, no alternative routes. When conditions become abnormal — and geopolitical conditions are increasingly abnormal — they break. "In the old paradigm, the cheapest chip won. In the Pax Silica framework, the trusted technology wins." This is not a subtle change. It is a restructuring of the decision criteria by which trillions of dollars of supply chain investment is made. The question is no longer "who can produce this most cheaply?" It is "who can produce this reliably, under what geopolitical conditions, with what supply chain exposure, with what regulatory risk?"
The US-led framework creating a trusted supply chain for the full AI stack · Minerals to models
Pax Silica is the emerging architecture of friend-shoring applied to the AI supply chain. The framework — led by the United States and now including the Netherlands, Japan, South Korea, Taiwan, Australia, the UK, and the EU — creates a "trusted zone" for the full AI supply chain from mineral extraction through chip manufacturing through model development. Allied nations specialise and share. No single external actor can exert leverage over the whole chain.
The logic is explicitly geopolitical. If China controls rare earth refining, allied nations must build alternative refining capacity — even at higher cost. If Taiwan is the primary location for advanced chip manufacturing, allied nations must develop alternative fabs — even at significantly higher cost. If ASML is the only EUV lithography supplier and it is Dutch, then Dutch export policy becomes allied technology policy — which is why the Netherlands joined Pax Silica.
The cost of Pax Silica is deliberately accepted. The CHIPS Act ($52 billion in US semiconductor manufacturing subsidies) and the EU Chips Act (€43 billion) are not economically efficient investments. They are national security investments that happen to take the form of industrial policy. The calculation is: the cost of building resilient trusted supply chains is less than the cost of dependency on an adversary who can sever them. Whether that calculation is correct depends entirely on whether the adversary relationship intensifies or moderates over the next decade — which is the central geopolitical uncertainty of our time.
Deep-tier data transparency — knowing not just who your supplier is but who your supplier's supplier's supplier is — is now technically possible through AI. Previously, companies knew their Tier 1 suppliers well, their Tier 2 suppliers partially, and their Tier 3 suppliers not at all. A disruption three tiers deep was invisible until it became a crisis.
AI systems can now map supply chains to Tier 5 and beyond, identify concentration risks (two different products depending on the same obscure mineral processor in one Chinese city), run scenario planning for tariff changes, conflict events, or natural disasters, and suggest alternative sourcing routes in real time. The intelligence that used to arrive as a crisis is now arriving as a forecast.
Friend-shoring — sourcing from politically aligned nations — and onshoring — bringing production home — are the two primary resilience strategies being deployed by governments and corporations simultaneously. The US is offering price guarantees on domestic rare earth production. The EU is mandating minimum domestic capability in critical technologies including semiconductors and pharmaceuticals.
This is explicitly not efficient. A semiconductor fab in Arizona costs 3–4× more to build and operate than one in Taiwan. A rare earth processing facility in the US produces at higher cost than one in China. The premium is the insurance premium on supply chain resilience. Governments have decided that the premium is worth paying. The supply chain is being deliberately made less efficient to make it more survivable.
At the corporate level, the response to supply chain concentration is widening the supplier base — sometimes to countries and suppliers that are less efficient, more expensive, or less proven, specifically because they are not the dominant supplier.
Apple sourcing from Vietnam and India alongside China is the canonical example. Vietnam cannot replace China's manufacturing capacity overnight. But having 15–20% of production in Vietnam means that a complete China shock can be absorbed partially rather than catastrophically. The redundancy has a cost — coordination, quality management, logistics complexity. The cost is accepted. Vulnerability costs more.
The data flows in supply chains — orders, inventory levels, logistics tracking, payment flows — are as important as the physical flows of goods. Blockchain-based provenance tracking, real-time logistics visibility platforms, and AI-powered demand forecasting are all making the supply chain's nervous system more sophisticated.
The goal is to make the supply chain's "data blood" flow as reliably as its physical blood. A supply chain that cannot see itself cannot adapt. The digital infrastructure being built now — across shipping, customs, inventory, and payments — is the immune system's diagnostic capability. You cannot fight what you cannot see. The investment in supply chain visibility is the investment in knowing where the blockages are before they become crises.
Both South Korea and China are developing designs for nuclear-powered commercial cargo ships using Small Modular Reactors (SMRs) and Thorium Molten Salt Reactors (TMSRs). The strategic logic is direct: a nuclear-powered cargo ship operates for years without refuelling, runs on a sealed reactor module swapped out every 10 years, and is entirely immune to oil price shocks and maritime fuel supply disruptions. It physically bypasses the fuel dependency that makes maritime trade vulnerable to Hormuz and other energy chokepoints.
One design uses a supercritical carbon dioxide propulsion system achieving approximately 5% better thermal efficiency than conventional nuclear marine propulsion. The reactor module concept — pre-fuelled, sealed, swapped like a battery — addresses the operational complexity that has historically made nuclear marine propulsion impractical for commercial vessels outside military submarines and icebreakers.
The barriers are real: IMO and IAEA regulatory frameworks for commercial nuclear vessels do not yet exist at the required scale. Port reception infrastructure for nuclear vessels is limited. Upfront capital costs are astronomical compared to conventional vessels. And public acceptance of nuclear-powered cargo ships in commercial ports remains uncertain. But the 2030s timeline for early commercial deployment is credible if regulatory frameworks develop alongside the technology — which is precisely what the IAEA is working on.
Forget the Hindenburg. The new generation of cargo airships bears as much resemblance to the Zeppelin era as a modern smartphone bears to a 1970s telephone. Radia's WindRunner is a 108-metre behemoth with 12 times the volume capacity of a Boeing 747 — designed specifically to carry wind turbine blades directly to remote installation sites with no runways, no port infrastructure, and no conventional logistics chain. Google co-founder Sergey Brin backs a similar venture through LTA Research.
The French company Voliris, working with Michelin, is developing the NATAC — a hybrid aircraft/airship that uses hydrogen for both lift and fuel, can carry 30 tonnes, and ships in 10 standard containers for deployment anywhere. Speed: approximately 80 mph — slow by air freight standards but faster than sea freight for medium distances. Range: intercontinental. Infrastructure required: a mooring point and sufficient open space.
The geopolitical win is explicit: airships can deliver mission-ready assets — radar systems, mobile hospitals, construction equipment, humanitarian supplies — directly where needed, bypassing port dependency entirely. In a world where ports are chokepoints, an aircraft that doesn't need a port is a bypass technology of the highest strategic value. Helium scarcity and airship size limitations are the primary constraints; hydrogen-lift designs address the helium problem while introducing their own handling complexities. The commercial timeline is the 2030s, but government and military interest is already creating deployment pressure.
The metaphor that best captures what is happening to global supply chains is not revolution — it is the transition from a single great river to a delta. A great river is powerful and efficient, but it has one channel. If that channel is blocked, everything stops. A river delta has dozens of channels. It is less efficient — the water travels farther and slower through more complex paths. But if one channel is blocked, the water finds another. The global supply chain is being redesigned from a river to a delta. The efficiency loss is deliberate. The resilience gain is the point.
The investment implications are significant and specific. Companies building the infrastructure of the delta — the alternative semiconductor fabs, the domestic rare earth processors, the diversified logistics networks, the AI supply chain intelligence platforms, the friend-shored manufacturing clusters — are building assets that the efficiency-maximised world did not need and the resilience-maximised world cannot do without. The CHIPS Act fabs in Arizona and Ohio. TSMC's facilities in Japan and Arizona. The rare earth processing investments in Australia, Canada, and Greenland. The nuclear-powered shipping R&D in South Korea. The airship cargo ventures backed by serious capital.
These are not speculative bets on unproven technology. They are the inevitable infrastructure of a world that has decided that resilience is worth the premium. The question for investors is not whether this infrastructure gets built — geopolitical necessity guarantees it will. The question is who builds it, on what terms, and who captures the value of the premium that resilience commands over efficiency.
The transition from efficiency to resilience is expensive, slow, and contradictory. It is expensive because redundancy by definition means paying for capacity that is not used in normal times. It is slow because building new supply chain infrastructure — mines, refineries, fabs, ports — takes 10–20 years. And it is contradictory because the same governments advocating for supply chain resilience are also under pressure to reduce the cost of living — and resilience costs more than efficiency.
The Pax Silica framework and friend-shoring represent a geopolitical bet. The bet is that the adversarial relationship with China will intensify or persist long enough to justify the cost of building alternative supply chains. If the geopolitical environment moderates — if US-China relations stabilise, if Taiwan remains peaceful, if mineral access agreements are reached — then much of the resilience investment will turn out to have been an expensive insurance policy on a risk that didn't materialise. Insurance policies are not wasted money. But they are money spent on something that doesn't produce returns if the insured event doesn't happen.
The honest answer to "is the supply chain transition good?" is: for resilience, yes. For efficiency, no. For global development in the poorest countries — which benefited enormously from efficiency-maximised supply chains that created manufacturing jobs — the answer is complicated. The delta is more resilient than the river. It is also more expensive to navigate, more complex to coordinate, and less accessible to the countries at its margins. The transition that makes the wealthy world more secure may make the developing world less prosperous. That tension is real and largely absent from the mainstream conversation about supply chain resilience.
Long-horizon thinking on capital, technology, and the forces shaping the next decade of wealth creation. Written from first principles. Not consensus. Not noise.