When governments restrict trade, how do stocks react?
We collected every US restriction on chip exports to China from 2018 to 2025, and every Chinese restriction on exporting key minerals from 2023 to 2025. For each one, we looked at how much the related stocks rose or fell over the next two trading days compared with the overall market.
1 · US limits chip sales to China
16 announcements, 2018–2025
2 · China limits exports of key minerals
8 announcements, 2023–2025 — rare earths, gallium, germanium and others
How to read these numbers
- “+4.8%” means the stock did 4.8% better than the overall market over those two days — the market’s own ups and downs are taken out.
- There are only 16 and 8 announcements, and some came at the same time as other US–China news, so treat these as signals, not proof.
- This is research, not investment advice.
Every announcement
| When | What happened | US equipment | China equipment |
|---|---|---|---|
| 2018-10 | Memory-chip maker Fujian Jinhua added to the US restricted list | +0.7% | -5.7% |
| 2019-05 | Huawei added to the US restricted list | -0.1% | -0.5% |
| 2019-06 | Chinese supercomputer makers added to the list | -0.9% | -0.4% |
| 2019-08 | Dozens of Huawei affiliates added | +0.7% | -2.6% |
| 2020-05 | New rule blocks factories using US tools from making chips for Huawei | -7.6% | +10.0% |
| 2020-08 | Huawei restrictions widened to cut off chip supply | -1.6% | -2.3% |
| 2020-12 | SMIC, China’s largest chipmaker, added to the list | +0.5% | +3.1% |
| 2021-04 | Chinese supercomputing chip designers added (incl. Phytium) | -1.6% | +2.5% |
| 2022-10 | Broad new limits on advanced chips and chipmaking equipment | -4.2% | -23.9% |
| 2022-12 | Memory maker YMTC and dozens of Chinese firms added | +1.2% | -4.5% |
| 2023-03 | More Chinese tech firms added to the list | -3.1% | +10.7% |
| 2023-10 | Chip and equipment limits tightened further | +0.8% | +0.9% |
| 2024-12 | Limits on AI memory chips; ~140 Chinese firms added | +4.4% | -2.9% |
| 2025-01 | More Chinese chip and AI firms added | +2.9% | +1.6% |
| 2025-03 | More Chinese server and computing firms added | +0.0% | -0.4% |
| 2025-09 | More Chinese chip-related firms added | +2.4% | +1.2% |
| When | What China announced | MP Materials | Chinese producers |
|---|---|---|---|
| 2023-07 | Export licences required for gallium and germanium | +7.7% | +10.8% |
| 2023-10 | Tighter export rules for graphite | -2.4% | -2.1% |
| 2023-12 | Export of rare-earth processing technology banned | +9.1% | +2.1% |
| 2024-08 | Export licences required for antimony | -0.1% | -0.5% |
| 2024-12 | Gallium, germanium and antimony exports to the US banned | +1.6% | +1.9% |
| 2025-02 | Export controls on tungsten, tellurium, bismuth, molybdenum and indium | +4.7% | +6.4% |
| 2025-04 | Export controls on seven rare earths | +5.7% | +3.4% |
| 2025-10 | Rare-earth controls extended to products made abroad | +12.3% | +3.9% |
Change over the two trading days after each announcement, compared with the overall market.
For researchers: method and full statistics⌄
Pre-registered market-model event studies (registered before any price data was pulled). H1/H2: no support at Bonferroni α = 0.025. China-side study: H7 supported, H6 borderline.
Cross-event results · primary window [0,1]
| Basket | Hypothesis | Expected | Mean CAR | Median | t | p (t) | Sign hits | p (placebo) |
|---|---|---|---|---|---|---|---|---|
| US equipment (AMAT, LRCX, KLAC) vs SPY | H1 | − | -0.34% | 0.27% | -0.47 | 0.321 | 7/16 | 0.294 |
| SOXX vs SPY | H3 | − | -0.66% | -0.39% | -1.53 | 0.073 | 11/16 | 0.061 |
| CN equipment (6 names) vs CSI 300 | H2 | + | -0.82% | -0.37% | -0.43 | 0.663 | 7/16 | 0.758 |
| CNI Semiconductor 980017 vs CSI 300 | robustness | + | -0.27% | -0.57% | -0.49 | 0.683 | 7/16 | 0.628 |
Per-event abnormal returns · CAR[0,1]
Event day = first session closing after the Federal Register public-inspection time, per market (so the Oct-2022 rule lands on 10 Oct in China, after Golden Week). Scope: supply chain = technology-wide control rule; entity = named-company Entity List additions.
| Event | Scope | US equip | SOXX | CN equip | CN index |
|---|---|---|---|---|---|
| E012018-10 Jinhua Addition of an Entity to the Entity List | entity | 0.7% | 1.9% | -5.7% | -1.2% |
| E022019-05 Huawei Addition of Entities to the Entity List | entity | -0.1% | -3.9% | -0.5% | 1.2% |
| E032019-06 Sugon Addition of Entities to the Entity List and Revision of an Entry on the Entity List | entity | -0.9% | -0.1% | -0.4% | -1.3% |
| E042019-08 Huawei affiliates Addition of Certain Entities to the Entity List and Revision of Entries on the Entity List | entity | 0.7% | 0.8% | -2.6% | 1.8% |
| E052020-05 FDPR Export Administration Regulations: Amendments to General Prohibition Three (Foreign-Produced Direct Product Rule) and the Entity List | entity | -7.6% | -2.1% | 10.0% | -1.1% |
| E062020-08 Huawei FDPR Addition of Huawei Non-U.S. Affiliates to the Entity List, the Removal of Temporary General License, and Amendments to General Prohibition Three (Foreign-Produced Direct Product Rule) | entity | -1.6% | -0.3% | -2.3% | -2.7% |
| E072020-12 SMIC Addition of Entities to the Entity List, Revision of Entry on the Entity List, and Removal of Entities From the Entity List | entity | 0.5% | -0.3% | 3.1% | -0.4% |
| E082021-04 Phytium Addition of Entities to the Entity List | entity | -1.6% | -0.9% | 2.5% | 1.9% |
| E092022-10 AC rule Implementation of Additional Export Controls: Certain Advanced Computing and Semiconductor Manufacturing Items; Supercomputer and Semiconductor End Use; Entity List Modification | supply chain | -4.2% | -3.6% | -23.9% | -5.1% |
| E102022-12 YMTC Revisions to the Unverified List and the Entity List | entity | 1.2% | 0.7% | -4.5% | -2.6% |
| E112023-03 Inspur Additions and Revisions of Entities to the Entity List | entity | -3.1% | -2.2% | 10.7% | 3.1% |
| E122023-10 AC update Entity List Additions | supply chain | 0.8% | -0.5% | 0.9% | 2.3% |
| E132024-12 HBM/FDPR Additions and Modifications to the Entity List; Removals From the Validated End-User (VEU) Program | supply chain | 4.4% | 1.8% | -2.9% | -0.7% |
| E142025-01 Sophgo Additions to the Entity List | entity | 2.9% | -1.3% | 1.6% | -1.4% |
| E152025-03 Inspur/Sugon Additions and Modifications to the Entity List | entity | 0.0% | -1.3% | -0.4% | 0.0% |
| E162025-09 SMIC-related Additions and Revisions to the Entity List | entity | 2.4% | 0.8% | 1.2% | 1.7% |
Exploratory observation (post hoc)
The CN equipment basket fell sharply on technology-wide rules — the Oct-2022 advanced-computing rule produced −23.9% over two sessions — while entity-level additions showed no consistent sign. The "domestic substitution" hypothesis may pool two opposite mechanisms.
| CN equipment | n | Mean | Median |
|---|---|---|---|
| Entity additions | 13 | 1.0% | -0.4% |
| Supply-chain rules | 3 | -8.6% | -2.9% |
Descriptive only. Registered as hypotheses v2 and tested solely on events after 22 Sep 2026.
Method & limitations
- Events: Federal Register BIS rules, tagged deterministically, human-reviewed, frozen before prices were pulled.
- Market model, estimation window [-250, -30] sessions; equal-weight baskets; cross-event one-sided t, sign test, 1,000-draw placebo.
- CN basket chosen in 2026 (survivors); names need ≥120 pre-event sessions, plus an index robustness check.
- BIS press releases can precede public inspection (one manual override); no factor adjustment; 16 events.
- A policy-risk study, not a trading signal.
Appendix · reaction by period (exploratory, descriptive)
Mean / median CAR[0,1] for events before and after 2022. Not a test and not used to revise the v1 conclusion: US equipment names reacted negatively on average in 2018–2021 but not in 2022–2025.
| Basket | Period | n | Mean | Median | Negative |
|---|---|---|---|---|---|
| CN equip | 2018-2021 | 8 | 0.5% | -0.4% | 5/8 |
| CN equip | 2022-2025 | 8 | -2.2% | 0.3% | 4/8 |
| CN index | 2018-2021 | 8 | -0.2% | -0.8% | 5/8 |
| CN index | 2022-2025 | 8 | -0.3% | -0.3% | 4/8 |
| SOXX | 2018-2021 | 8 | -0.6% | -0.3% | 6/8 |
| SOXX | 2022-2025 | 8 | -0.7% | -0.9% | 5/8 |
| US equip | 2018-2021 | 8 | -1.2% | -0.5% | 5/8 |
| US equip | 2022-2025 | 8 | 0.5% | 1.0% | 2/8 |
The other direction: China's critical-mineral export controls
pre-registered 2026-09-22Do China's item-based export controls on critical minerals (MOFCOM announcements, 2023-07 → 2025-10) move Chinese strategic-materials producers up and the main non-China US rare-earth producer up?
Result: MP Materials supported; Chinese producers borderline
MP Materials rose about 4.8% relative to the market over the two sessions after each announcement (p = 0.014, placebo p = 0.017). The Chinese producer basket rose 3.2% (placebo p = 0.005, t-test p = 0.029 — just above the 0.025 threshold). No single event drives either result, but after dropping the four announcements that coincided with other US–China escalations the effects stay positive and lose significance (n = 4), so the export controls cannot be separated from escalation in general.
| Basket | Hypothesis | Expected | Mean CAR | Median | p (t) | Sign hits | p (placebo) | Excl. C05/C06/C07/C08 |
|---|---|---|---|---|---|---|---|---|
| CN strategic materials (6 names) vs CSI 300 | H6 | + | 3.23% | 2.73% | 0.029 | 6/8 | 0.005 | 2.58% (n=4) |
| MP Materials vs SPY | H7 | + | 4.82% | 5.22% | 0.014 | 6/8 | 0.017 | 3.58% (n=4) |
| REMX vs SPY | H8 | + | -0.38% | 0.18% | 0.639 | 4/8 | 0.648 | 1.60% (n=4) |
Hypotheses
- H6 · CN strategic materials (6 names) vs CSI 300 ↑ — Export licensing restricts supply outside China, raising scarcity rents and pricing power for domestic producers.
- H7 · MP Materials vs SPY ↑ — Non-China supply becomes more valuable; substitution demand and policy support for US producers.
- H8 (secondary) · REMX vs SPY ↑ — Broad strategic-metals exposure; weaker, mixed prior.
Events (MOFCOM web timestamps)
- C012023-07 gallium/germanium
- C022023-10 graphite
- C032023-12 rare-earth technology catalogue
- C042024-08 antimony
- C052024-12 Ga/Ge/Sb ban to US· confounded
- C062025-02 tungsten/tellurium/bismuth/molybdenum/indium· confounded
- C072025-04 seven medium/heavy rare earths· confounded
- C082025-10 extraterritorial rare-earth controls· confounded
A third direction: tariff actions against agricultural futures
pre-registered 2026-10-10Futures are used here and nowhere else in this work for a reason: the other two studies cover export controls on goods with no liquid contract — there is no tradeable gallium, germanium, antimony or rare earth — so a futures test there would measure instruments the policy does not target. Section 301 retaliation hit US agriculture, where the targeted good is the deliverable of a liquid contract.
Result: no support, and the primary came out with the wrong sign
Soybeans benchmarked on corn — the better-identified test of the retaliation channel — gave +0.14% over the two sessions after each action (p = 0.613, placebo 0.630), the opposite direction to the hypothesis. Soybeans against a commodity benchmark had the predicted sign and missed anyway (-0.31%, p = 0.232). Corn fell more than soybeans (-1.06%), which is the reverse of the ordering the story predicts. A 25-event control arm of Section 301 exclusions moves nothing (p = 0.750), so this is a null rather than a pipeline that cannot see. With n = 10 that is weak evidence of no effect, not evidence of absence.
| Basket | Expected | Mean CAR | p (t) | p (placebo) |
|---|---|---|---|---|
| SPREAD_SOY_CORN | ↓ | 0.14% | 0.613 | 0.630 |
| AG_SOYBEANS | ↓ | -0.31% | 0.232 | 0.185 |
| AG_CORN | ↑ | -1.06% | 0.921 | 0.977 |