Investors have been buying a story. The story goes like this: tariffs, geopolitical pressure and boardroom risk committees have combined to pull manufacturing out of China and into friendlier jurisdictions. Vietnam is booming. Cambodia is assembling. Mexico is reshoring. The supply-chain risk that the first instalment of this series described, economies caught between China as supplier and America (and Europe perhaps) as customer, is being resolved. The data, on this reading, is clear: US-facing export share is rising across the new production hubs, and money has followed.
Half of that story is true. The export data does show what it is supposed to show. The problem is that the investment thesis ultimately depends less on where finished goods are going than on where production inputs are coming from. And on that question, the story falls apart for most of the economies at its centre.
The test that most people are not running
The claim that supply chains are diversifying away from China is, at root, a claim about production dependence. China+1 is usually described as a diversification of manufacturing and sourcing footprints. If the thesis is structural, China should matter less not only as the place where final goods are assembled, but also as the source of the components, machinery and intermediate goods that make those exports possible. On the other hand, export share can only show the first half of that story – it only tells us where finished goods are sold.
That is why the import side matters. Export share here means a country’s exports to the US as a share of its total exports; import share means imports from China as a share of its total imports. The import figure is the second half of the diversification story: the half export data alone can’t show.
A country that has genuinely restructured its supply chain away from China should show falling or flat China import share alongside its rising US export share. A country where China import share is rising alongside US export share is displaying the pattern one would expect from a different outcome: final assembly or routing has shifted while the upstream input base has not. The “made in China” label on the product changed, but the supply chain behind it may not.
The question therefore is whether the aggregate trade profile moves in the direction the diversification thesis requires: if a country is gaining US export share while becoming more dependent on Chinese imports, the pattern looks less like supply-chain diversification and more like rerouting.
A buffer: the test is deliberately conservative. Rising US export share and rising China import share can occur for unrelated reasons: domestic demand, capital-equipment imports, consumer goods, or sectoral shifts that do not belong to the same production chain. To safeguard the analysis, the rerouting signal is counted only when the China import-share increase is both directionally aligned and large relative to the US export gain. Below that buffer, the result is classified as inconclusive. The cutoff is set at 30% of the US export gain; re-running the classification at 20% and 40% could potentially move countries between buckets, including Thailand, Malaysia, and Nepal. Hence, the deep-rerouting group is stable across this range, while the partial/inconclusive boundary remains blurry.
The cohort
The rerouting logic is straightforward: genuine restructuring should weaken the link to China upstream, alongside strengthening the link to the United States downstream. When a country gains US export share while also becoming more dependent on Chinese imports, the pattern looks less like a supply chain moving away from China than a China-linked production system finding a new route to market.
Between 2018 and 2023, 23 countries in the dataset gained more than two percentage points of US-facing export share. This cutoff was chosen because it sits above the sample’s natural break. They are the economies who gain US export. Their trade profiles most visibly match the first half of the China+1 story, with a larger share of their exports going to the American market after the tariff shock.
Where the diversification winners actually stand: US export gains vs China import gains, 2018–2023
Change in merchandise export dependence on the US vs. change in import dependence on China (percentage points, 143 countries)
The two-percentage-point threshold tests the diversification narrative where it is strongest, among countries with export gains large enough to anchor an investment case. If the import-side test fails here, it fails where it matters most.
What the import side shows
Of the 23 meaningful US export gainers, 15 show a rerouting signal on the import side: China import share rising by more than 30% of the US export gain, the threshold at which the movement is large enough to be analytically meaningful rather than noise. Canada and Iceland show the opposite pattern, with China import share falling as US export share rose. Six are inconclusive: the China import increase exists but is too small relative to the US export gain to read as a directional signal.
The fifteen are not evenly weighted. Cambodia sits furthest from every reference line on the chart: the single clearest case of rerouting in the dataset. Ghana is a distant second. Of course, this should be read with the caveat that both are small-volume traders, where a given dollar shift produces a larger percentage-point swing than it would for Mexico or Vietnam. The remaining thirteen cluster closer to the partial-rerouting boundary, including the countries most central to the China+1 narrative: Vietnam, Thailand, Korea, Taiwan and Mexico.
Where the diversification winners actually stand: US export gains vs China import gains, 2018–2023
Change in merchandise export dependence on the US vs. change in import dependence on China (percentage points)
Beyond the ratio
The rerouting ratio is a country-level fingerprint. It cannot show that Cambodian or Vietnamese exports actually contain Chinese components, only that the two trends move together. Closing that gap needs evidence this dataset doesn’t have: enforcement records, investment data, value-added decomposition.
Rerouting Ratio: Ranked Table of Meaningful US Export Gainers, 2018–2023
Ratio = change in China import share ÷ change in US export share. Calculated only within the filtered cohort (Δ exp → USA > 2 pp). Flags indicate where country structure limits the inference. Click any column header to sort within category.
| Country | Δ Exp→USA (pp) |
Δ Imp←CHN (pp) |
Rerouting ratio | USA exp. share 2023 (%) |
CHN imp. share 2023 (%) |
Category | Flag |
|---|
UNCTAD merchandise trade dependence ratios · 23-country cohort = all economies with Δ export share to USA > 2 pp between 2018 and 2023
The regulatory record backs the top of the table. In August 2023, the US Commerce Department determined that solar cells completed in Cambodia, Malaysia, Thailand and Vietnam “using parts and components produced in the People’s Republic of China” were circumventing existing China duties. Cambodia’s penalty reached a 125% across-the-board dumping rate; cooperating Vietnamese and Cambodian producers faced individual rates as high as 542%. Vietnam shows up again in a separate case: the Commerce Department confirmed that cold-rolled steel produced in Vietnam from Chinese-origin substrate remained subject to China’s antidumping duties. Mexico has its own version: in July 2024 the US and Mexico jointly moved to close a steel and aluminum loophole that, in a White House official’s words, “countries like China use to avoid US tariffs by shipping their products through Mexico”. Mexican steel now qualifies for duty-free treatment only if it is “melted and poured” in North America. Three of the table’s flagged or highest-ratio countries, Cambodia, Vietnam, Mexico, have each had the mechanism this instalment is testing for confirmed by regulators, in at least one sector.
The capital flows are broadly consistent with the trade flows.McKinsey puts Chinese manufacturing FDI into Southeast Asia at $24 billion in 2023. Rhodium Group tracked the mechanism directly: after the solar ruling, Chinese firms “redirected investment and production away from jurisdictions facing steeper or more immediate duties, such as Thailand, toward those with more favorable treatment, notably Vietnam”. By Q2 2023, 80% of US solar imports came from the four named countries. That is capacity being built specifically to serve the rerouting corridor.
The strongest counter-evidence says the ratio is wrong for Vietnam. IMF research via CEPR tested six Asian “connector” economies using value added instead of gross trade share, asking the same question this instalment asks: genuine production (reallocation) or pass-through (rerouting)? In Vietnam, domestic value added in its “electrical and machinery” exports to the US ran 10-12 percentage points above its synthetic counterfactual by 2022, while Chinese value added in those same exports fell. The study calls this “trade reallocation rather than trade rerouting”, a direct, sector-matched contradiction of what the ratio implies for Vietnam. Federal Reserve research finds a softer version of the same signal: US investment shifting toward Vietnam since 2018 but calls its own evidence early and partial, so it corroborates without confirming.
Both things can be true in the same country. Vietnam’s solar sector was found, component by component, to be passing through Chinese product with minor processing. Its machinery and electronics sectors may simultaneously show genuine, rising domestic value added. The rerouting ratio cannot tell these apart, and that is a real limitation. Cambodia has no such tension: a high ratio, a small commodity-adjacent export base, and a confirmed circumvention finding all point the same way. Vietnam’s evidence splits by sector. That split is the clearest argument for treating this instalment’s table as a screen for where to look closer.
The investment implication
The rerouting ratio was never going to identify a single trade. What it does is convert one country-level pattern into three separate questions an investor can act on: where should investor look, what makes the signal actionable, and what should investors own or hedge.
Where should investor look.The ratio already did this work. Deep rerouting is a policy-catalyst watchlist; partial rerouting needs sector selection; inconclusive stays a research queue; cleaner restructuring is the strongest China+1 long.
What makes the signal actionable. A bucket turns tradable only when a signal fires. Legal/enforcement triggers, AD/CVD rulings, transshipment penalties, rule-of-origin changes, court rulings on tariff authority, are dated events, suited to tactical positioning. Flow, value-added, and policy-alignment signals are slower trends, suited to scaling in. The same headline can be either: a factory announcement is a flow signal if it expands assembly, a production signal if it builds upstream capacity instead.
What should investors own or hedge. Each matrix cell pairs a bucket with an economic function: production (who makes the goods, hit first), corridor (who moves them, profitable even when exporters are fragile), compliance (who proves origin, paid by scrutiny itself, not its outcome), aggregate (FX, ETFs, sovereign risk, for whole-bucket repricing). The rule: deep rerouting is a split trade, short production, long corridor and compliance; partial rerouting is dispersion, not a country call; inconclusive is a watchlist; cleaner restructuring is the one row where the production long is the thesis.
Now, we matrix the ratio bucket and the channel. Each cell in the matrix combines a ratio bucket with an economic function. The row tells us what the country-level pattern suggests. The column tells us who might capture or lose value. The cell is a posture conditional on the right signal appearing. Only deep rerouting and cleaner restructuring resolve to a determinate posture across all four channels here; partial rerouting and inconclusive are deliberately left open: they resolve at the sector level, which this instalment’s country-level ratio cannot do.
Trade expression matrix: from ratio bucket to investable exposure
The ratio sets the starting posture; the signal activates the trade; the channel determines what to own, avoid or hedge.
|
Ratio bucket ↓ Channel → |
Production Who makes or supplies the goods? | Corridor Who moves, stores, powers or finances the flow? | Compliance Who proves origin and legality? | Aggregate How to express the whole-bucket view? |
|---|---|---|---|---|
|
Deep rerouting
|
Short / Avoid Shallow assemblers, tariff-sensitive exporters, China-input-dependent firms Trigger: AD/CVD petition, circumvention ruling, rules-of-origin tightening | Selective Long Ports, logistics, industrial parks, trade finance Trigger: Chinese FDI continues, export-zone expansion, port throughput rises | Long Testing, inspection, certification, customs tech, traceability Trigger: Customs audits, duty-evasion agreement, origin-documentation requirements | Hedge FX, country ETF, sector ETF, importer/exporter baskets Trigger: Country-level enforcement, broad tariff shock, legal regime hardening |
|
Partial rerouting
|
Selective Long genuine local-content sectors; avoid pass-through sectors Trigger: Sector value-added data, HS-level matching, FDI composition | Selective Long Logistics, industrial parks, power serving confirmed growth sectors Trigger: Capacity growth in sectors with durable demand | Selective Long Origin-verification and compliance providers Trigger: Rules-of-origin tightening or product-specific audits | Selective Sector ETFs, paired exporter baskets; long clean / short fragile Trigger: Policy separates winners from pass-through sectors |
|
Inconclusive
|
Watch No position yet; monitor exporters Trigger: Ratio crosses threshold or sector evidence emerges | Watch Flows Logistics, ports, warehouses only if flows accelerate Trigger: Sustained port, export-zone or factory expansion | Watch Policy Compliance providers only if enforcement appears Trigger: First audit, investigation or new origin rule | No Trade None unless macro repricing begins Trigger: New data update or official enforcement signal |
|
Cleaner restructuring
|
Long Domestic suppliers, manufacturers, automation, industrial banks, utilities Trigger: China input dependence flat/falls while exports rise | Long Enablers Logistics, power, industrial infrastructure tied to local production Trigger: Capacity expansion tied to local value-added growth | Neutral Compliance less central; useful for premium market access Trigger: Stricter documentation standards emerge | Long China+1 Country ETF, FX, sector ETF, local supplier basket Trigger: Persistent export gains with low China input dependence |
Legal durability modifies position size and holding period. Final AD/CVD rulings, Section 301 actions and durable bilateral agreements support larger and longer positions; contested or temporary tariff authorities require smaller and more tactical trades.
The matrix can be viewed as a pre-requisite of the due diligence instead of replacing the due diligence. It tells investors whether to begin with the exporter, the corridor, the compliance layer or the aggregate instrument. The ratio narrows the search field. The signal determines timing. The channel determines where the money should sit.
Investors have been buying a story about where exports go. The rerouting ratio and the framework ask a different question: where should capital sit while regulators decide where those exports really came from?
Cover photo by Fabio Jock on Unsplash