The Rerouting Illusion: How De-Risking Lengthened Supply Chains Without Cutting China Out

Beneath the deglobalisation headline, the trade data show resilience and rerouting rather than retreat: most corporate de-risking has relabelled and lengthened China-dependence rather than removing it, a 2026 to 2029 exposure-mismodelling risk for manufacturers, procurement teams, investors and trade-policy advisers.

The consensus on deglobalisation has hardened into a familiar story: globalisation is retreating, the world is splitting into rival blocs, and companies are decoupling their supply chains from China. The headline data tell a more awkward story. Trade is not shrinking; the WTO, IMF, UNCTAD and UN all record resilient 2025 volumes. And the "de-risking" that boards have been buying is, on inspection, mostly rerouting: production has shifted one country sideways while the China-dependence underneath has been relabelled, lengthened and obscured rather than removed. The weak signal is not deglobalisation. It is that the diversification firms think they have bought is often China-plus-China, and the new chokepoints sit in a handful of connector economies.

Signal Identification

This is a measurement-and-mismodelling signal: the gap between what "de-risking" is assumed to deliver and what the trade data show it delivering. The signal is not that supply chains are unchanged; they have moved. It is that the movement is geographic dispersion of final assembly without a corresponding fall in upstream China-dependence, leaving exposure intact while balance sheets and strategy decks record it as reduced.

Time horizon: 3 to 6 years (rerouting visible in 2025-2026 trade data; transshipment enforcement and rules-of-origin tightening 2026-2028; genuine reallocation or re-concentration legible 2028-2031) Plausibility band: Medium-High Geographic / Jurisdictional Scope: Primary: the United States and the China-plus-one connector economies (Vietnam, Mexico, Cambodia, Indonesia, Thailand, Malaysia, India). Spillover: the EU, pursuing its own de-risking agenda, and developing economies broadly, where rerouting and tariff exposure now concentrate. Sectors exposed: Multinational manufacturers and their boards across electronics, autos and auto parts, apparel, furniture and pharmaceuticals; supply-chain, procurement and sourcing functions; trade-finance and customs-compliance teams; investors pricing "China exposure"; trade-policy and economic-security advisers; logistics operators and connector-economy host governments.

What's Changing

Trade is resilient, not retreating. The WTO's Global Trade Outlook and Statistics (19/03/2026) records world merchandise trade growing 4.6% in 2025 and forecasts a slower but still-positive 1.9% in 2026: a normalisation, not a collapse. The IMF's World Economic Outlook (14/04/2026) describes trade being rerouted through new partners and regional agreements that do not necessarily align with old geopolitical boundaries. The bloc story and the volume data do not match.

The rerouting is concentrated. UNCTAD's Global Trade Update (07/04/2026) reports goods trade grew about 7% in 2025, but with a sharp feature underneath: US-China trade fell roughly a quarter, about $170 billion, while connector economies such as Cambodia, Egypt, Vietnam and Indonesia stepped in as intermediaries. The bilateral line moved; the dependence did not disappear, it relocated.

The dependence is relabelled, not removed. The Information Technology and Innovation Foundation (23/02/2026) finds that China-plus-one strategies have increased geographic diversification of production without meaningfully decreasing supply-chain dependency on China, and that China has explicitly allowed geographic dispersion of production so long as control over key links in value chains remains anchored inside China. The factory moves; the chokepoint stays.

The tariff wall is real but uneven. ITIF's analysis of the Global South (06/04/2026) records China's import share into developing economies rising over 21 percentage points since 2000 while the US share fell 10. And the UN's Financing for Sustainable Development Report (09/04/2026) finds average tariffs on least-developed-country exports surged from 9% to 28% in 2025: the barriers are rising, but they are landing on the connector economies, not closing the reroute.

Disruption Pathway

The pathway runs in three stages. Through 2026, the rerouting shows up in the data as resilience: trade volumes hold, US-China bilateral trade falls, and connector economies absorb the redirected flows while staying upstream-dependent on China. From 2026 to 2028, enforcement catches up: transshipment tariffs, rules-of-origin tightening and customs scrutiny force a choice between genuine value-add relocation and exposure to penalty rates. By 2028 to 2031, the supply map either genuinely reallocates, with upstream capacity built outside China, or re-concentrates as firms conclude the reroute is not worth the compliance cost and return to direct China sourcing.

Stress concentrates at four points. The first is the connector economy itself: Vietnam, Mexico and others have built export volumes on cheap Chinese inputs, and a hard transshipment rule threatens large chunks of those exports. The second is the multinational's balance sheet: a China-plus-one investment recorded as de-risking may be, in the ITIF framing, another node in a China-dependent value chain. The third is the upstream layer: final assembly is mobile, but intermediate inputs, components and process know-how are not, and that is where the leverage sits. The fourth is the data itself: aggregate trade statistics conflate genuine reallocation with rerouting, so firms and policymakers are navigating with a distorted map.

Adaptation will sit at three levels. Operationally, firms move from counting countries to tracing value: mapping where the intermediate inputs and the irreplaceable process steps actually originate, not just where the final box is assembled. At the policy level, governments shift from tariff walls to rules-of-origin and content thresholds, the instruments that actually test whether a reroute is a relocation. Financially, investors and lenders re-underwrite "China exposure" to include indirect, second-degree dependence through connector economies, rather than reading a changed country-of-origin label as reduced risk.

Why This Matters

For multinational boards, supply-chain and procurement leaders, investors and trade-policy advisers, the decision architecture under pressure is the equation of "moved production" with "reduced risk." That equation underpins most China-plus-one capital allocation, most supplier-diversification reporting, and most investor assessment of China exposure, and the trade data say it is frequently wrong. Boards should ask not how many countries their supply chain now touches but where the non-substitutable inputs and know-how still sit. Investors should treat a changed country-of-origin label as a question, not an answer. Procurement teams should expect transshipment enforcement to convert today's cheap reroute into tomorrow's penalty rate. Policymakers should recognise that a tariff on a connector economy without a rule-of-origin test mostly moves the paperwork. The common thread: diversification on the map is not diversification in the value chain.

Decision-action posture for this signal: Prepare. The rerouting is measurable now and enforcement is tightening on a known timetable, so the task is value-chain mapping and scenario planning against transshipment-rule triggers, not wholesale supply-chain restructuring this cycle.

Counter-Argument

The strongest objection is that the reroute is a transition, not an illusion, and that enforcement will resolve it. The OFW Law analysis of 2026 trade enforcement (04/02/2026) documents a US regime moving hard against exactly this: a 40% transshipment tariff, AI-powered customs supply-chain mapping, False Claims Act actions, and a USMCA review tightening rules of origin. On this reading, the rerouting illusion is self-correcting: firms either build genuine value-add and upstream capacity outside China, which is real de-risking, or they are penalised out of the reroute. The first trade war, after all, did produce durable manufacturing capacity in Vietnam and Mexico, not only transshipment.

That is the optimistic path, and it is possible, but it does not dissolve the signal; it sets the test. Enforcement closing the reroute only produces real de-risking if upstream capacity, components and process know-how actually move, and the ITIF evidence (23/02/2026) is that the upstream layer is the hardest and slowest to relocate, with China actively working to keep it anchored. The more likely near-term outcome of hard enforcement is not clean reallocation but higher costs and, for some firms, a quiet return to direct China sourcing. Either way, the board that recorded its China-plus-one move as completed de-risking has mismodelled its exposure.

Implications

This is a measurement and strategy problem with a real economic core, not a transient data quirk. The inflection window is 2026 to 2029, set by how fast transshipment enforcement and rules-of-origin tighten and whether upstream capacity genuinely moves. The IMF (14/04/2026) is explicit that as flows are rerouted through connector countries, the less effective the policies driving fragmentation may be in achieving their stated objectives, which is the structural point: the fragmentation is real at the bilateral level and much shallower at the value-chain level. Deglobalisation as a headline and re-concentration as a reality can coexist, and the gap between them is where the strategic risk lives.

This signal is not a claim that nothing has changed: trade has visibly reorganised, and connector economies have genuinely gained assembly capacity. It is also not a claim that decoupling is impossible: it is a claim that most current de-risking has not achieved it, and that the upstream layer is the binding constraint. And it is not a prediction that the reroute lasts: enforcement may close it, but closing it does not automatically produce de-risking. Competing interpretations include: that enforcement forces genuine reallocation and the illusion resolves into real diversification, or that fragmentation deepens into capital and technology as well as goods, making the goods-trade resilience the misleading number.

Early Indicators to Monitor

Disconfirming Signals

Strategic Questions

Keywords

Deglobalisation; regionalisation; connector economies; China-plus-one; supply chain de-risking; trade rerouting; transshipment; rules of origin; geoeconomic fragmentation; value chain dependence; reshoring; US-China trade

Bibliography