Taiwan Stress Test: One Baseline, Three Escalation Scenarios

The whitepaper opens with a simple but consequential observation: a conflict over Taiwan would hit companies not through one channel but through two interconnected ones. System A covers technology and supply chains linked to Taiwan, above all chips, electronics and server manufacturing. System B covers business and capital in China: plants, customers, suppliers, receivables, payment channels and local decision rights. Many industrial companies are exposed to both dependencies at the same time.

The figures on slide 6 show the degree of concentration: TSMC reached a global revenue share of 70.4% in chip foundry services in the fourth quarter of 2025 (TrendForce). Taiwanese manufacturers produce more than 80% of the world's complete server systems (TrendForce News, Feb 2025). On the trade side, Germany imported goods worth €170.6bn from China in 2025 and exported goods worth €81.3bn to China (Destatis). The German Economic Institute (IW) counts 229 industry-related product groups in which China's share of imports was at least 50% in 2024.

NEXERY explicitly warns, however, against translating these figures too quickly into supply dependency. The TSMC figure is a revenue share, not a unit count and not a share of production locations. The server figure likewise refers to the worldwide output of Taiwanese companies, which increasingly manufacture outside Taiwan. A Taiwanese headquarters does not tell you where a product is actually made. Concentration only becomes a failure risk at the specific site.

This leads to the central management question of the first chapter: which specific manufacturing sites, components and payment channels would be affected? Every critical dependency must be assigned to a product, a process and a site. Only when it is linked to the available bridging time and the financial impact does it become a robust basis for decisions. Corporate concentration, production sites and substitution options must be examined separately. A second supplier only provides protection if it does not depend on the same bottleneck – the same foundry, the same tooling or the same transport route.

Chapter 2 anchors the stress test in a documented situation picture as of 13 August 2026. The timeline of selected events runs from the live-fire exercises and missile launches after the Pelosi visit in August 2022, through the reporting and inspection claims announced in April 2023, the Joint Sword-2024A and -2024B exercises in May and October 2024 with blockade drills and coast guard involvement, renewed large-scale exercises in April 2025 and the "Justice Mission" exercise in December 2025, to the ship patrols east of Taiwan described by Taiwan's Ministry of National Defense in its report of 1 July 2026. The sources document exercises, control claims and patrols; they do not document an actual blockade.

On this basis NEXERY distinguishes five situations. S0 is normal competition with diplomacy, trade and basic military presence and serves as the reference. S1, persistent grey-zone pressure consisting of military, economic and digital pressure below the level of a blockade, is the baseline on the reference date. Three conditional escalation scenarios build on it: S2, an administrative quarantine with selective reporting and inspection obligations enforced by civilian authorities and the coast guard; S3, a military blockade as an actual interruption of essential sea and air links, modelled without widespread destruction; and S4, war or invasion with large-scale combat operations. The distinction between quarantine and blockade follows the CSIS analyses of 2024.

What matters is what the paper does not claim: the sequence S1 to S4 is neither a forecast nor a mandatory escalation chain. No probabilities of occurrence are given. S3 and S4 can overlap in reality; a blockade can be part of a war. NEXERY also assesses the year 2027 soberly: US reports (DoD 2025, ODNI 2026) describe 2027 as a military capability goal and distinguish capabilities from a political decision to attack. The ODNI assessment of March 2026 names no currently planned invasion in 2027 and no fixed unification timetable. No date for war can be derived from this. Management should therefore align its preparation with lead times and observable disruptions, not with calendar years.

Each situation has a focus: S1 prevention (prepare alternatives early), S2 flow of goods (manage delays and inventories), S3 bridging (secure scarce quantities and liquidity), S4 protection (people, emergency operations and assets).

For each situation the whitepaper provides a description, impact, measures and early indicators.

In baseline S1, military presence, economic pressure and cyber operations increase uncertainty; a lasting interruption of sea and air links is not assumed. Review and planning costs may rise, but individual cyber or supply disruptions can already hit a company hard. The measures focus on prevention: review critical parts and payment channels, qualify alternative sources, justify inventory levels and prepare powers of attorney. In the case of serious individual incidents, crisis management is possible even in S1. Early indicators are export controls, longer exercises, cable or cyber incidents and changed risk premiums.

In administrative quarantine S2, authorities and the coast guard enforce selective reporting, control or inspection obligations. Trade remains partly possible but is impeded. Delays can exceed inventory coverage; transport costs and tied-up capital increase. When supply gaps loom, companies activate prepared sourcing and transport options, pull inventory forward only via secured routes, and align priorities between procurement, sales and finance. Early indicators: confirmed inspections, longer port stays, route changes, new insurance coverage terms.

In military blockade S3, the military and coast guard interrupt essential sea and air links. Missing supplies can stop production once inventories are depleted; energy shortages can additionally burden manufacturing in Taiwan. Sanctions and financing consequences are tested as separate assumptions. In the event of critical failures the crisis team is activated, customers and products are prioritised according to pre-agreed criteria, and remaining stock, emergency operations and liquidity are managed jointly. Early indicators are confirmed interruptions of ports or flight routes, forced rerouting and energy shortages.

War or invasion S4 adds large-scale combat operations, destruction or an invasion. Employees, facilities and data on site are directly endangered; sanctions or countermeasures can further restrict China business and payment channels. People are protected according to local threat levels, prepared decision and communication channels are used, and emergency operations, solvency and data access are secured. Early indicators: sustained attacks, evacuation warnings, destruction and restrictions on payments and data access.

The scenarios are analytical simplifications. Scope, duration, sanctions and follow-on reactions must be varied by each company in its own stress test.

Chapter 3 links external events to a company's own operations. Six impact channels structure the analysis: semiconductors and electronics, raw materials and intermediate products, China business and assets, transport and insurance, financing and liquidity, and legal and regulatory obligations. All of them feed into the corporate impact on revenue, profit and liquidity. The channels count together: a missing component can stop deliveries and delay payments at the same time; customer prioritisation, contractual consequences and additional capital needs arise from the same bottleneck. The stress test captures these interdependencies and avoids counting the same effect twice.

The industry matrix on slide 14 maps twelve industries against the four situations – 48 qualitative fields in total. Examples: in semiconductors and electronics, S1 means more effort for export reviews, S2 inspections that delay chip deliveries, S3 an interrupted chip supply from Taiwan and S4 endangered fabs and know-how on site. In the automotive industry, missing parts complicate plant planning in S2, chip shortages can stop assembly lines in S3, and China plants and sales can fail in S4. In mechanical and plant engineering, control units arrive late in S2, no deliveries are possible without electronics in S3, and service and China receivables are at risk in S4. Pharma and medical technology face missing spare parts for medical devices in S3 and regional supply gaps in S4. In aerospace and defence, specialised electronics arrive late in S2, missing parts delay construction and maintenance in S3, and allocation of critical electronics is possible in S4. Further rows cover chemicals and materials, energy and electrical engineering, telecoms, cloud and data centres, logistics, shipping and aviation, retail and consumer goods, banks and insurers, and food and packaging.

NEXERY stresses that the fields describe risks, not certain consequences. They are conditional assessments given corresponding Taiwan or China exposure; affected sites, inventory coverage and substitution options determine the strength and timing of the impact. The consequence for leadership: priority goes to processes with a gap between bridging capacity and recovery. Existing hardware can keep running despite missing supplies – the bottleneck arises where new demand, spare parts or expansion meet interrupted supply routes.

At the heart of the whitepaper is an illustrative calculation that shows how a supply stop translates into revenue, profit and liquidity. All values are NEXERY assumptions, not a geopolitical forecast of duration.

The assumptions: on day 0 a complete supply stop begins for one product. Inventory carries operations for 30 days; no other approved source of supply exists. From day 90 supply is available again; partial quantities are not possible before then. This results in an unsecured period of 60 days (90 minus 30). Lost production is not made up later.

The impact: at 100 units per calendar day, 6,000 units are lost. At €500 revenue per unit, revenue falls by €3.0m. Against this stand €2.1m of avoidable variable costs (€350 per unit). Contribution margin, and thus profit before countermeasures, falls by €0.9m with fixed costs unchanged. Assuming all differences are cash-effective within the period and adding €0.2m of additional cash outflows for measures, the liquidity gap is €1.1m versus plan. The €0.9m profit impact is already included in the liquidity figure and must not be added again.

The calculation is deliberately simple: 6,000 × €500 = €3.0m; 6,000 × €350 = €2.1m; 3.0 − 2.1 = 0.9; 0.9 + 0.2 = 1.1. In a real model, payment terms, tied-up inventory, fixed costs and tax and financing effects must be mapped over time.

What matters is the decision derived from it: management resolves to finance additional inventory coverage or an earlier alternative source. Procurement and finance jointly present costs, approval date and remaining gap. The supply gap thus becomes a financing question – with budget, owners and a latest decision date. The methodological framework follows ISO 31000, ISO 22301 and 22313 and BSI Standard 200-4.

In practice this means: the first process without timely replacement determines the bottleneck. The stress test translates its failure into lost revenue, profit impact and liquidity needs – and thereby shows which precautions are worth the money.

Chapter 4 translates the analysis into leadership action. The external situation is an input signal; risks, available bridging time and financial consequences determine which measure a company prepares, finances or activates. Three tasks secure implementation.

Risk identification creates transparency. The dependency chain runs from the critical product (revenue, margin, customer) via the business process (plant, line, IT, staff) and the sub-supplier site (material, tooling, IP rights) to the failure impact in time, euros and contracts. Four review fields structure the analysis: concentration (how many real process sites, tools and approved capacities exist), substitutability (how long technical, regulatory and commercial approval of an alternative takes), bridging (which inventory, transport, liquidity or contractual solution carries for how long) and data quality (what is documented, what is assumed, and where information on deeper supply tiers is missing). Outputs for leadership are a map of critical products and real production sites, priorities by bridging capacity and financial impact, open assumptions with owners and clarification dates, and early indicators with source, interval and threshold.

Risk management finances prevention. Four risk treatment options are available: avoid (reduce the dependency or end the activity), reduce (prepare an alternative source, inventory or product change), share or transfer (contractual risk sharing, suitable insurance) and consciously accept (justify and monitor the residual risk). Clear decision rules make this operational: a trigger is, for example, a recovery time exceeding inventory coverage. Procurement presents the alternative option, finance checks the funding requirement, and management decides on budget and residual exposure. The latest start date equals the required deployment date minus implementation time.

Crisis management activates prepared decisions. The decision cycle has five steps: activate, assess the situation, decide, implement and communicate, review impact. Management appoints a crisis lead and deputy and defines powers of attorney, reporting lines and decision cadence. Prepared alternatives exist for endangered processes, including secure communication, data access and payment channels. Activation follows operational thresholds rather than a rigid scenario assignment: if a delivery is delayed, procurement reconciles the confirmed date with inventory coverage; if bridging is insufficient, management approves substitution or allocation; if liquidity falls below the minimum buffer, finance presents immediate measures; if people or key systems are acutely endangered, the crisis lead acts immediately. NEXERY recommends practising the process annually and after significant changes – because a second source of supply or secure data access often cannot be established at short notice once an event occurs.