Commodities
Semiconductors
Advanced and mature-node fabrication, what the concentration figures actually measure, and where overseas capacity has reached.
AssessmentModerate confidence
Not investment advice. This surface explains exposure and transmission channels; it never recommends, ranks or names securities. Figures are estimates and carry their own as-of dates and confidence grades.
Current state
Moderate confidence · Economic data as of 2026-08
Taiwan is not only a military flashpoint. It is a global economic chokepoint. The most advanced semiconductors in existence are fabricated on an island that imports almost all of its energy, sits astride some of the world's busiest shipping, and is claimed by its largest trading partner. That combination means disruption in the strait would not stay in the strait: it would arrive as an industrial supply shock in economies with no involvement in the dispute. This section explains the structures and transmission channels; it is not investment advice and contains no forecasts.
Majority of the world's most advanced logic chip fabrication
Semiconductor position
Share estimates vary by definition of "advanced"; the concentration is not disputed
Largest export destination including Hong Kong, though the share has fallen since roughly 2018
Role in Taiwan's trade
Advanced semiconductor and equipment restrictions on China since 2022, revised repeatedly
Export controls
Key supplier of semiconductor materials and equipment; major consumer of Taiwanese output
Supply-chain role
Commodities
Advanced and mature-node fabrication, what the concentration figures actually measure, and where overseas capacity has reached.
Energy
An overwhelmingly imported primary energy supply, its storage margins, and what generation failure would reach first.
Trade & transport
Container traffic through and around the strait, the rerouting options, and the undersea connectivity beside them.
Finance
What sanctions and financial measures would reach in a crisis, and what is established about their scale rather than assumed.
Sanctions
The restriction architecture aimed at the mainland, and how it regulates Taiwanese firms as a side effect.
Sanctions
Cross-strait interdependence as an instrument — which measures are actually applied, and what the completed cases show about their effect.
Market access
The fastest-moving and least-watched channel: premia and underwriting appetite reprice long before anything is shipped differently.
Economic impacts
Two questions that are not the same question — how exposed the world is, and how a disruption would actually propagate.
Everything below is authored on this module’s economy record. Figures carry the vintage of the release that produced them, not the date of this page: the section is graded Moderate confidence and its figures are stated as of 2026-08.
| Commodity | Why it matters here | Exposure |
|---|---|---|
| Advanced-node semiconductors | Taiwan fabricates the majority of the world's most advanced logic chips, and TSMC's leading-edge capacity is concentrated on the island even as overseas fabs come online. There is no rapid substitute: replacement capacity is measured in years and hundreds of billions of dollars, which is precisely what makes this a strategic dependency rather than a commercial one. | High exposure |
| Mature-node semiconductors | Less discussed and arguably more exposed. Older-generation chips go into cars, appliances and industrial equipment in enormous volume, and Taiwan is a major supplier. The 2021–22 shortages showed that losing mature-node supply halts assembly lines faster than losing cutting-edge supply. | High exposure |
As of 2026-08
Taiwan Semiconductor Manufacturing Company is the reason this module has an economy section this size. It fabricates chips for most of the world's leading designers, and its most advanced processes remain concentrated in Taiwan — principally around Hsinchu, Taichung and Tainan — even as it invests heavily abroad. The supplier, materials and packaging ecosystem around the fabs is harder to relocate than the fabs themselves.
Sources: TSMC investor filings and official statements↗ · Counterpoint Research — global semiconductor foundry market share↗ · Financial Times — semiconductor and supply-chain coverage↗
As of 2026-08
Four different quantities circulate in this argument and they are routinely swapped for one another, which produces claims that sound precise and are not. Stated separately, with what each covers. COMPANY SHARE OF A MARKET SEGMENT: TSMC held about 73 per cent of the pure-play foundry market by revenue in Q2 2026, against roughly 7 per cent for Samsung Foundry, 5 for SMIC, 4 for UMC and 3 for GlobalFoundries. "Pure-play" excludes the in-house foundry operations of integrated manufacturers, so this is a share of the contract-manufacturing segment rather than of all chipmaking. SHARE OF ONE COMPANY'S OWN REVENUE: processes at 7nm and below accounted for roughly 74 per cent of TSMC wafer revenue in Q1 2026. That says how advanced TSMC's mix is; it says nothing about anyone else's capacity. GEOGRAPHIC SHARE: what proportion of the world's sub-7nm capacity physically sits on Taiwan is the figure this module would most want, and Vigil has not found a defensible public number for it. It is recorded as an information gap rather than estimated from the company figures, which do not support the conversion — TSMC now has capacity outside Taiwan, and other firms have capacity inside other countries. ANNOUNCED VERSUS OPERATING CAPACITY: an announced fab, a fab under construction and a fab in volume production are three different things, separated by years. The distinction is made in the overseas-capacity entry below. Where this module says "the majority of the world's most advanced logic chips are fabricated in Taiwan", it is making the geographic claim qualitatively, on the strength of TSMC's dominance and the location of its leading-edge lines, and it does not attach a percentage to it.
Sources: Counterpoint Research — global semiconductor foundry market share↗ · TSMC investor filings and official statements↗ · Congressional Research Service — Taiwan reports↗
As of 2026-08
TSMC's Arizona investment is the largest single test of whether concentration can be reduced, and the company's own published position is more specific than the reporting around it. Total announced investment is $265bn, described by TSMC as the largest greenfield foreign direct investment in American history, covering six logic wafer fabs, two advanced packaging facilities and an R&D centre. The first fab has run high-volume production on N4 since the fourth quarter of 2024. The second, on N3, targets volume production in the second half of 2027. The third, on N2 and A16, targets volume production by the end of the decade. A fourth entered early construction in 2026. Two cautions. First, secondary reporting has circulated a figure of roughly 30 per cent for the share of TSMC's 2nm-and-beyond capacity that will sit in Arizona once announced fabs are complete; TSMC's own Arizona materials make no such comparison, and Vigil treats it as a projection attributed to reporting rather than a company statement. Second, and more important analytically, Taiwan-based advanced capacity is expanding at the same time — 3nm output in Taiwan has been reported rising toward 180,000 wafers a month through late 2026, and further advanced fabs are reported planned on the island. Diversification and concentration are both increasing, and only the ratio between them settles the question. Overseas fabs reduce concentration at the margin and over years; they do not change the dependency's essential character on any horizon this module covers.
Sources: TSMC investor filings and official statements↗ · Financial Times — semiconductor and supply-chain coverage↗
As of 2026-07
The Hsinchu Science Park and the corridor running south through Taichung and Tainan hold not just fabrication but the research institutes, equipment servicing, specialty chemicals, substrate and advanced packaging capabilities that make leading-edge production possible. This ecosystem density — decades of accumulated tacit capability and skilled labour — is the part that cannot be duplicated by capital alone, and is the real content of the phrase "silicon shield". That shield is double-edged: it raises the cost of disruption to everyone, which deters, while also making Taiwan more valuable to control and giving outside powers an interest in relocating capacity away.
Sources: TSMC investor filings and official statements↗ · Congressional Research Service — Taiwan reports↗
What to watch
| Commodity | Why it matters here | Exposure |
|---|---|---|
| LNG and energy imports | Taiwan imports the overwhelming majority of its primary energy, and LNG — with storage measured in days — is the tightest link. Energy import continuity is the single clearest transmission channel from maritime disruption to economic and civil effect. | High exposure |
As of 2026-07
Taiwan's near-total energy import dependence is an economic vulnerability before it is a civil one: industry, and semiconductor fabrication in particular, is enormously energy-intensive and intolerant of interruption. A fab losing power does not pause; it loses work in progress. Energy continuity is therefore directly coupled to the global chip supply the rest of this section describes — the two dependencies are one system.
Sources: MOEA Energy Administration — energy statistics and security stockpile requirements↗ · IEA electricity and energy-security analysis · Taiwan Power Company — generation, coal stocks and reactor filings↗
What to watch
| Commodity | Why it matters here | Exposure |
|---|---|---|
| Container shipping and the strait transit | A very large share of global container traffic passes through or near the Taiwan Strait. Even short disruption would reroute traffic, lengthen voyages and reprice freight globally — as unrelated chokepoint disruptions have repeatedly demonstrated. | High exposure |
As of 2026-07
The Taiwan Strait and the approaches around the island carry a very large share of world container traffic and a substantial share of LNG movements to Northeast Asia. Traffic can route east of Taiwan through the Philippine Sea, so the geography is not a single point of failure — but rerouting costs time, fuel and capacity, and in a constrained market that translates quickly into global freight rates. The relevant risk is disruption and repricing, not closure.
Sources: Joint War Committee — Listed Areas (marine war-risk)↗ · Public market and insurance data · CSIS China Power Project and ChinaPower analysis↗
What to watch
As of 2026-07
A serious cross-strait crisis would raise sanctions questions of a different order from any precedent: China is a far larger economy and more deeply embedded in global trade and finance than any previously sanctioned state, and Taiwan's own economy would be inside the blast radius of measures intended to protect it. Research institutions have published estimates of global cost running into trillions of dollars, with wide ranges and heavy assumption-dependence. Vigil cites the existence and rough scale of such estimates and explicitly does not endorse any specific figure.
Sources: Congressional Research Service — Taiwan reports↗ · US / EU / UK sanctions designations · Public market and insurance data
As of 2026-07
Since 2022 the United States has progressively restricted advanced semiconductor and equipment exports to China, with Japanese and Dutch alignment on tooling, and has adjusted the rules repeatedly — a January 2026 revision moved certain advanced AI accelerators from a presumption of denial to case-by-case review. Taiwanese firms sit inside this framework rather than beside it: TSMC serves Chinese customers for a modest share of revenue, operates older-node capacity in China under licences that have been revoked, renewed and renegotiated, and complies with controls set in Washington. Export control policy is therefore a live variable in Taiwan's economy and one of the fastest-moving files in this module.
Sources: Congressional Research Service — Taiwan reports↗ · Financial Times — semiconductor and supply-chain coverage↗ · MERICS — China analysis↗
What to watch
As of 2026-07
Beijing has repeatedly used targeted trade measures in response to political events — agricultural and food import suspensions, tariff investigations into the ECFA framework, and restrictions on group tourism — generally calibrated to be politically legible and economically survivable. Since late 2025 comparable pressure has been applied to Japan following statements on a Taiwan contingency. This is the most frequently used instrument in the whole pressure system and among the least covered; the module tracks it as a standing indicator.
Sources: PRC government and Taiwan Affairs Office statements↗ · MERICS — China analysis↗ · Reuters — Taiwan and China coverage↗
As of 2026-07
China, including Hong Kong, remains Taiwan's largest export destination, and Taiwanese firms have invested in the mainland for three decades — though the direction of travel since roughly 2018 has been diversification toward Southeast Asia, the US and Japan under the "New Southbound" framing and supply-chain pressure. The relationship is asymmetric in both directions: Taiwan's export exposure gives Beijing leverage, while Chinese industry's dependence on Taiwanese chips limits how far that leverage can be pushed without self-harm. Neither side has resolved that tension, and both are working to reduce their side of it.
Sources: WTO and IMF trade and economic data↗ · MERICS — China analysis↗
What to watch
| Commodity | Why it matters here | Exposure |
|---|---|---|
| Marine insurance and war-risk premia | The fastest-moving and least-watched channel. War-risk premia and underwriting withdrawal can suppress shipping well before any physical interdiction — meaning an economic blockade effect can precede, or substitute for, a military one. | Moderate exposure |
As of 2026-08
Coverage of Taiwanese semiconductors habitually runs two distinct questions together, and they have different answers and different evidence. The first is TAIWAN'S OWN RESILIENCE: can fabrication on the island keep running under pressure? That is largely an energy and continuity question rather than a market one — fabs are among the least interruptible industrial loads anywhere, a fabrication process cannot be stopped and restarted cleanly, and the binding constraint is the eleven-day gas stockpile behind roughly half of Taiwan's generation rather than anything about chips. The second is GLOBAL EXPOSURE: what happens to everyone else if Taiwanese output falls? That is a substitution question, answered in years and hundreds of billions of dollars of replacement capacity, and it applies whatever the cause of the interruption. The two are connected — Taiwan's energy dependence is the physical route by which its domestic vulnerability becomes a global one — but they are not interchangeable, and a measure that improves one may do nothing for the other. Overseas fabs reduce global exposure without improving Taiwanese resilience at all; grid hardening and reserve policy do the reverse.
Sources: TSMC investor filings and official statements↗ · Congressional Research Service — Taiwan reports↗
As of 2026-07
Exposure to a Taiwan disruption is far wider than the set of actors with a security role. European automotive and industrial manufacturing, Japanese and Korean electronics, and essentially every economy running on modern computing depend on Taiwanese output at one or two removes. This is why the European Union, with no security role in the strait, has become an active participant through chip subsidies, economic-security instruments and export-control alignment — economic dependency creating policy engagement in the absence of any military one.
Sources: WTO and IMF trade and economic data↗ · Financial Times — semiconductor and supply-chain coverage↗ · MERICS — China analysis↗
An advanced, trade-dependent economy that occupies a position in global technology supply chains out of all proportion to its size — and imports almost all of its energy to sustain it.
Taiwan's export exposure to China gives Beijing economic leverage, and Beijing has used it — suspending agricultural imports and tariff concessions in response to political events. But Chinese industry's dependence on Taiwanese chips limits how hard that lever can be pulled without self-harm. Both sides are working to reduce their own side of the asymmetry, and neither has succeeded.
Near-total energy import dependence is Taiwan's sharpest economic vulnerability, and it is coupled directly to the semiconductor industry: advanced fabrication is enormously power-intensive and intolerant of interruption. Energy continuity and chip supply are one system, not two, which is why storage capacity and grid resilience are economic-security questions.
The world's second-largest economy, Taiwan's largest export market, and simultaneously dependent on Taiwanese semiconductors it cannot yet replace — leverage and vulnerability in the same relationship.
Beijing has repeatedly used targeted trade measures — agricultural and food import suspensions, ECFA tariff investigations, group-tourism restrictions — calibrated to be politically legible and economically survivable. This is the most frequently used instrument in the whole system and among the least covered, which is why the module tracks it as a standing indicator rather than an occasional story.
Chinese industry depends on Taiwanese chip supply, and any disruption to the strait would damage China's own economy alongside everyone else's — while inviting sanctions of a scale without precedent against an economy far more embedded in global trade than any previously targeted. Indigenous substitution efforts are substantial and constrained by export controls. This is a genuine restraint on Beijing, and one of the strongest structural arguments against the imminence readings common in popular coverage.
Washington's most consequential instrument in this system is not military but regulatory: export controls and industrial policy that reshape where the world's most advanced chips are made.
US policy simultaneously treats Taiwanese chip concentration as a vulnerability to be reduced through onshoring and as a source of deterrent value. Those aims are in tension: capacity moved off the island lowers global exposure to disruption, which is desirable economically and arguably reduces the incentive to protect Taiwan. Whether the "silicon shield" is being deliberately dismantled is actively debated; the module notes the tension rather than resolving it.
A major exposure to Taiwanese semiconductors, a supplier of critical materials and tooling to the same industry, and since late 2025 a target of Chinese economic coercion over Taiwan-related statements.
Japan sits on both sides of the semiconductor dependency: it supplies materials, chemicals and equipment the Taiwanese industry cannot readily replace, and it consumes Taiwanese output across its electronics and automotive sectors. That dual position gives Tokyo unusual weight in export-control alignment, and unusual exposure to any disruption.
Modest direct economic exposure to the strait, but significant human exposure through a large Filipino workforce in Taiwan.
The Philippines' most immediate exposure to a Taiwan crisis is not commercial but human: a substantial Filipino workforce in Taiwan whose safety and repatriation would become an urgent government responsibility. Contingency planning for that has been publicly discussed at official level.
The economic centre of gravity of the entire module — and the reason disruption in the strait would arrive as an industrial shock in economies with no stake in the dispute.
A disruption would not stay regional. It would propagate as component shortage into automotive, electronics, industrial and defence production across Europe, North America and Asia, on timescales set by inventory rather than by the crisis. Research institutions have published global cost estimates running into the trillions; the ranges are wide and heavily assumption-dependent, and Vigil cites their existence and rough scale without endorsing any specific figure.
Advanced fabrication is enormously power- and water-intensive and intolerant of interruption — a fab losing power does not pause, it loses work in progress. Taiwan's near-total energy import dependence is therefore not a separate vulnerability from the chip supply; it is the same one.
A resource exporter whose experience of Chinese trade coercion is the module's best available precedent for how that instrument performs over time.
Australian exporters found alternative markets more successfully than expected, and the measures were eventually unwound without the concessions that prompted them. The cautious inference — one case, in specific commodities, with a particular political context — is that economic coercion imposes real costs while frequently failing at its stated political objective. That is directly relevant to reading current pressure on Japan and on Taiwan.
Exposure without presence — and one genuine point of structural leverage in the semiconductor supply chain.
A Taiwan disruption would reach European industry as component shortage within months, on timelines set by inventory. That is the entire basis of European involvement — and it is why the module treats semiconductor dependence as a global file rather than a regional one.
Resilience spending is a growing and politically contested share of Taiwan's security budget, sitting alongside conventional defence rather than inside it.
Including resilience investment in headline security-spending figures is analytically defensible — grid hardening and cable redundancy contribute to deterrence by denial of effect — and politically convenient, since it raises the headline percentage without buying weapons. The module notes both readings and tracks the components separately where reporting permits.