What is the risk of a US-China decoupling for markets?

The short answer

Decoupling is widely discussed but partially mismeasured. The headline numbers (declining China share of US imports, rising tariffs, restrictions on Huawei and ZTE) suggest a major rupture. The granular data tell a more nuanced story.

Most of what looks like decoupling is actually rerouting. Components made in China travel to Mexico or Vietnam for final assembly, then enter the US under the new origin label — escaping tariffs on the original Chinese product but maintaining Chinese supply chain dependence at the component level.

What’s complicating the picture is that real decoupling is happening in narrow but high-value sectors: advanced semiconductors (where US export controls cut off Chinese fab access to ASML EUV machines), AI chips, certain dual-use technologies. These sectors matter disproportionately for market-cap-weighted indices.

→ New to deglobalization? Deglobalization framework

What the data shows

The numerical record on US-China economic relations (USTR, BEA, IMF, Census Bureau):

• China share of US goods imports: 21.6% (2017 peak) → 13.4% (2024) = -38% relative decline
• Mexico share of US imports: 13.5% (2017) → 16.9% (2024) — Mexico now the #1 US trade partner
• Vietnam imports to US: $46bn (2017) → $137bn (2024), nearly tripled
• Mexican FDI inflows: from approximately $35bn average annually to $40bn+ (BEA, 2022-2024)
• China’s share of advanced semiconductor wafer capacity: 95% (USTR notes), targeted by 50% wafer tariffs proposed in 2024
• EV tariff: 100% on Chinese imports (Sept 2024) — effectively closing the US market to Chinese EVs

The exception that nuances this picture: studies of Chinese value-added in third-country exports to the US (Alfaro-Chor 2023) suggest that Chinese components remain heavily embedded in Mexican and Vietnamese exports. Genuine onshoring of Chinese supply chains is much smaller than headline import share would suggest.

→ Dataset: DXY dataset

Why it happens — the macro mechanism

Three channels explain the gap between decoupling rhetoric and decoupling reality.

Channel 1: tariff design favors transshipment. US Section 301 tariffs apply only to direct imports of listed products under specific HTSUS codes. They do not apply to downstream products under different codes, even if the downstream product contains heavy Chinese inputs. This creates a strong incentive for Chinese exporters and US importers to add a final assembly step in a third country, achieving regulatory compliance without true supply chain separation.

Channel 2: narrow but real strategic decoupling. US export controls (October 2022, expanded 2023-2024) genuinely cut off Chinese access to advanced ASML EUV lithography, Nvidia H100 AI chips, and certain critical inputs. These restrictions are not about tariffs — they are absolute denials. China’s response (massive investment in domestic fabs, the recent breakthrough at SMIC) shows real friction. Critical minerals are the next frontier where decoupling could become real.

The asymmetry: tariffs are reroutable; export controls are not.

Channel 3: financial decoupling lagging trade. Capital flows show partial decoupling: foreign direct investment to China has weakened (some quarters showing net outflows for the first time since records began), and US institutional investors have reduced allocations to Chinese equities. But Chinese holdings of US Treasuries remain large (around $760bn in 2024), and major US banks still operate in Hong Kong and Shanghai. Financial decoupling is happening at the equity allocation level, not yet at the architecture level.

Synthesis by regime. Pre-2018 was a regime of intensifying integration: WTO accession, FDI flows, supply chain optimization. 2018-2024 saw selective tariff escalation alongside maintained trade volume — much more rerouting than separation. Post-2024 with second-Trump tariff proposals (Section 301 across 60 economies, possible 60% China baseline tariff), the regime is shifting toward broader-based decoupling that would be harder to circumvent through transshipment, but the implementation lag is long and the legal challenges are unresolved.

What the headlines call decoupling is mostly rerouting; what is actually decoupling — narrow strategic technologies — does not show up in the trade headline numbers but does show up in the chip earnings.

→ Framework: Geopolitics and macro regime transmission

What it means for different economic actors

Savers indirectly exposed via global ETFs (MSCI ACWI, S&P Global) face limited direct decoupling impact, since the index is dominated by US large caps with modest direct China revenue.

Investors with concentrated emerging market or China-themed allocations face genuine regime risk: capital controls, disclosure restrictions on US-listed Chinese ADRs, and the ongoing risk of additional sanctions or restrictions.

Multinationals and corporate strategists are operating under “China+1” or “China+N” supply chain redesigns, with measurable but slow shifts in capex destinations. Mexico, Vietnam, India and Malaysia are the main beneficiaries; ASEAN as a bloc has gained relative share.

A common error is to apply a “decoupling” frame uniformly across sectors. The data shows that consumer goods, EVs and batteries face direct tariff walls; semiconductors face export-control walls; services and many industrials face minimal disruption. Each sector deserves its own framework.

Practical observation

Go deeper

Related questions

• How do trade wars affect equity sectors differently?
• What is the rising role of BRICS in financial markets?
• What is the impact of sanctions on global financial architecture?
• How do critical minerals fit into the energy transition?

Frequently asked questions

Is the US-China decoupling reversible?

Tariffs are reversible by executive action and have been raised, paused, and modified repeatedly since 2018. Export controls on advanced technology are politically harder to reverse and have bipartisan US support. Foreign direct investment patterns, once redirected, are sticky — Mexican manufacturing capacity built since 2022 will not relocate back to China even if tariffs are removed. The trade flows are partially reversible; the capital deployment is mostly not.

Why does China dominate the rare earths and critical minerals supply chain?

China accounts for approximately 60% of global rare earth mining and 85% of refining capacity. This dominance reflects 30 years of strategic industrial policy combined with environmental tolerances Western producers will not match. China announced export bans on gallium, germanium and antimony to the US in December 2024 — demonstrating that the relationship is not one-sided. Critical minerals are the most likely flashpoint for genuine decoupling because alternatives are not deployable on a 5-year timeframe.

How exposed are major US tech stocks to China?

Direct revenue exposure varies widely. Apple has historically derived around 18-20% of revenue from Greater China. Nvidia derived 15-25% from China before AI chip export controls began biting. Tesla manufactures heavily in Shanghai, with the plant accounting for around 50% of global production. Many semiconductor equipment makers (Applied Materials, Lam Research) have 30%+ China revenue exposure. The US market-cap-weighted indices have approximately 5-7% direct revenue dependency on Chinese end markets.

Last updated — 29 May 2026