Zhihu Review: A-Share Market on June 4, 2026 — Redefining the 'Going Global' Thesis

Fact Verification

✅ Confirmed

1. EU-China trade tensions are escalating.

• NYT (5/29): “Europe Is Edging Closer to a Trade War With China”
• Reuters (5/29): European Commission says China trade relationship is “not sustainable”
• Economist (5/12): EU and China “stumbling into a trade war”
• Guardian (5/28): EU commissioners meeting to discuss new import restrictions on China

The author’s claim that the June 3 afternoon A-share selloff was triggered by EU trade war fears is highly plausible. Europe’s signals were concentrated in late May, and markets reacted in early June — completely normal timing.

2. Zhongji Innolight (中际旭创) Thailand factory confirmed.

• Company statement July 2024: Phase 1 & 2 Thailand capacity reached, successfully mass-producing high-end optical modules
• March 2026 investor relations: 2025 annualized capacity 28M units, 2026 capacity further increasing
• Caveat: Xueqiu analysis notes the Thailand factory is only 67% owned, meaning 33% of profits go to minority shareholders

3. Eoptolink (新易盛) Alpine Optoelectronics acquisition confirmed.

• Agreement signed 2021, completed April 2022, deal value ~$22.2M
• Provided North American R&D resources, silicon photonics capability, and client network

4. Yoke Technology (雅克科技) UP Chemical acquisition confirmed.

• 2016: Established Jiangsu Sunko to acquire Korea’s UP Chemical (96.28% stake)
• UP Chemical has been SK Hynix’s precursor supplier since 2004
• Core team, factory, and supply chain all in Korea — genuinely insulated from China tariff risk

5. TFME (通富微电) AMD Suzhou & Penang factories acquisition confirmed.

• April 2016: $371M for 85% of AMD Suzhou and AMD Penang
• Forms a “joint venture + partnership” model with AMD
• Handles ~80% of AMD’s packaging and testing orders

6. TFME entering Foxconn Industrial Internet (工业富联) as CPO packaging supplier.

• Multiple Chinese financial social media sources (early June 2026): TFME announced at Taipei Computex that it has officially entered the Foxconn Industrial Internet system as a core CPO packaging supplier
• This opens a pathway from AMD to NVIDIA’s supply chain — “building another TFME”

Key Analytical Points

1. “True overseas expansion = overseas factories + overseas clients + overseas operating system”

This is the most valuable insight in the entire article. Traditional A-share narratives conflate “export ratio” with “going global” — but in a trade war environment, high export ratio is actually risk exposure. True globalization means moving production capacity abroad. The distinction is critical: export = “sell from China to the world,” globalize = “produce globally, sell globally.”

2. Clear three-tier company classification

This isn’t speculation — Fuyao Glass’s US factory success is a well-documented precedent. The author’s historical parallel with 1980s Japanese automakers relocating to the US (Toyota→Kentucky, Honda→Ohio, Nissan→Tennessee) is also apt.

3. AI chain three-layer framework

“China produce → global sell” → “China tech → overseas assembly” → “China capital + overseas manufacturing” — a useful progression for understanding valuation divergence across AI export names. However, A-shares that truly qualify for “Layer 3” remain very limited.

Points of Caution

1. “Valuation restructuring” is hard to verify in the short term.

The directional thesis (market preference shifting from “export ratio” to “overseas capacity ratio”) is likely correct, but valuation restructuring takes multiple earnings cycles to materialize. In the short term, overseas factory construction is a heavy-capex “asset-heavy” strategy — depreciation and early operational costs will weigh on near-term profits. Whether the market rewards this long-term logic with a premium depends on the macro environment and capital flows.

2. TFME’s NVIDIA story is still in the expectation stage.

Getting CPO supplier qualification from Foxconn Industrial Internet is a positive step, but moving from qualification to mass production is a long road. TFME needs time and capacity ramp-up to convert this opportunity into actual NVIDIA supply chain deliveries. At this stage, it’s still expectation-driven speculation.

3. Zhongji Innolight’s Thailand factory minority stake issue.

The factory is ~67% owned, meaning 33% of profits go to minority shareholders. This structure has implications: ① overseas profits can’t be 100% consolidated; ② if Thailand becomes the primary export hub, minority partners get a third of the profits; ③ but if the factory is primarily for tariff avoidance, retaining 67% is still worthwhile. Investors need to factor this in.

4. A paradox in the “overseas factory” strategy.

If companies are forced abroad by trade war pressure, the cost advantage of overseas production itself is questionable — they’re being pushed, not leading. Truly successful cases (Fuyao, BYD) pursued globalization proactively, not reactively. Motivation matters for execution quality and long-term returns.

20 A-Share Candidates Following the “True Overseas” Framework

Below are 20 stocks screened by the article’s framework — confirmed overseas production capacity, not just export exposure. Organized by sector.

Group 1: New Energy & Auto (Overseas Manufacturing)

1. BYD (002594) Hungary factory (Q2 2026 mass production, 300K/yr planned), Thailand (operational, 150K/yr), Brazil (operational, 150K/yr), Turkey (planned). Overseas revenue exceeded ¥300B for the first time in 2025. Full supply chain going global — vehicles, batteries, and components.

2. CATL (300750) Germany Thuringia factory (operational), Hungary Debrecen factory (under construction, 100GWh planned). Overseas clients include BMW, Mercedes, VW, Tesla. Global EV battery leader — overseas capacity is essential for maintaining market share.

3. Fuyao Glass (600660) US Ohio factory (world’s largest single-site auto glass plant), Illinois facility. Overseas revenue ~45%. Started building overseas factories in the 1990s — A-share’s most mature automotive globalization story. Directly benefits from the “produce overseas, sell overseas” tariff shield.

4. Luxshare Precision (002475) Vietnam (Apple Watch, AirPods assembly), India, Mexico — ~20-25% of total capacity overseas. Continuing to expand Vietnam with ¥1.7B investment. Most aggressive overseas capacity build among Chinese consumer electronics OEMs.

5. Joyson Electronics (600699) Acquired German Preh (auto electronics), US KSS (auto safety), German TS. Operations in 22 countries. Overseas revenue >70%. Essentially a Chinese-headquartered global auto parts company, not an exporter.

6. Tuopu Group (601689) Factories in Brazil, Poland, Mexico serving Tesla and other overseas clients. Auto chassis components going global — following Tesla’s expansion while building its own global footprint.

Group 2: Home Appliances (Overseas Manufacturing)

7. Haier Smart Home (600690) Global factory network spanning US, Mexico, India, Pakistan, Russia. Overseas revenue >50%. Acquired GE Appliances (US), Fisher & Paykel (New Zealand) for true local-brand globalization. When peers scramble over tariffs, Haier’s global capacity-scheduling ability becomes a moat.

8. Midea Group (000333) Manufacturing bases in Vietnam, India, Egypt, Brazil, Mexico, Thailand. Overseas revenue ~40%+. Accelerating global expansion through acquisitions of Toshiba white goods, Italian Clivet. One of two appliance globalization leaders.

Group 3: Machinery & Tires (Overseas Manufacturing)

9. SANY Heavy Industry (600031) R&D and manufacturing bases in India, US, Germany, Brazil, Indonesia. Overseas revenue exceeded 60% in first 3 quarters of 2025, with overseas gross margins ~9pp higher than domestic. The undisputed Chinese engineering globalization leader.

10. Zoomlion (000157) Production bases in Italy, Germany, India, Brazil, Belarus. Tower cranes and concrete machinery globally competitive. Overseas revenue ~35% and rising.

11. Weichai Power (000338) Acquired German KION (world’s 2nd largest forklift group), Linde Hydraulics, French Baudouin. Heavy-duty powertrain + hydraulics going global — a unique globalization path.

12. Sailun Tire (601058) Nine production bases across Vietnam, Cambodia, Indonesia, Mexico. Most overseas production bases among Chinese tire makers. Serves the US market primarily through Vietnam and Cambodia — tariff-free channel. Tires were among the first Chinese industries to bypass trade barriers via overseas capacity.

Group 4: AI Supply Chain (Tech + Overseas Capacity)

13. Zhongji Innolight (300308) Thailand factory Phase 1 & 2 operational, capacity expanding in 2026. Extremely high North American client concentration. Can deliver from Thailand to bypass tariffs. Note: 67% ownership stake means 33% of profits go to minority shareholders.

14. Eoptolink (300502) Acquired US-based Alpine Optoelectronics for North American R&D and technology resources. Pushing Southeast Asia capacity build. Dual logic: silicon photonics + overseas production.

15. Yoke Technology (002409) Wholly owns Korea’s UP Chemical (acquired 2016), a leading semiconductor precursor supplier. Tech team in Korea, factory in Korea, clients (SK Hynix, Samsung) in Korea. The cleanest “Chinese capital + overseas manufacturing” case on this list — the operating entity is Korean, completely insulated from China-origin tariffs.

16. TFME (002156) Acquired 85% of AMD Suzhou and AMD Penang for $371M in 2016. Deeply tied to AMD supply chain (~80% of AMD’s packaging orders). Recently entered Foxconn Industrial Internet as a CPO packaging supplier, opening a pathway to NVIDIA. Penang, Malaysia is a global packaging hub — capacity was always overseas.

17. Will Semiconductor / OmniVision Group (603501) Acquired US-based OmniVision in 2019, the world’s 3rd largest CMOS image sensor maker. HQ in US, globally distributed R&D and operations. Transformed from distributor to global semiconductor design house through acquisition.

Group 5: Consumer Electronics & Light Manufacturing

18. Transsion (688036) Factories in Ethiopia, India, Bangladesh, Pakistan. >40% Africa smartphone market share, #1 in multiple South Asian markets. Unique: Transsion was built from day one as a “produce overseas, sell overseas” model — almost zero export tariff exposure.

19. Huaqin Technology (603296) World’s largest ODM (original design manufacturer), with production base in Vietnam. Serves Xiaomi, Samsung, Lenovo. Overseas capacity build ongoing.

20. Sunwoda (300207) Battery module factories in Hungary and India. Dual engine: consumer electronics + EV battery. Overseas clients include Renault-Nissan — overseas capacity is a prerequisite for winning these accounts.

Important Caveat

All 20 names above were screened by the article’s “true overseas = overseas factory + overseas client + overseas operating system” standard, verified against publicly available information. They span 5 different sectors and do not move in unison — AI chain follows compute cycles, appliances and machinery depend on overseas demand, NEVs face geopolitical risks.

This is not a buy list. Overseas factory construction is a high-capex long-term strategy. Near-term financials may be pressured by depreciation and ramp-up costs. Each stock requires independent assessment of valuation, industry cycle position, and management execution. The value of this list is in providing a candidate pool for research as the market potentially shifts its preference from “export ratio” to “overseas capacity ratio.”

Overall Assessment

Score: 8.0/10

One of the highest quality A-share analysis articles I’ve seen recently on Zhihu.

• Solid data: All four specific stock claims (Zhongji Innolight, Eoptolink, Yoke Technology, TFME) verify independently. No fabricated facts.
• Valuable framework: The pure export vs. overseas manufacturing distinction, AI chain three-layer progression, and historical Japan analogy — all useful analytical tools.
• Logical coherence: From “EU trade friction” catalyst → “overseas logic restructuring” core thesis → specific stock picks, the chain is complete.

Weaknesses are mainly in execution: valuation restructuring needs time to prove out, TFME’s NVIDIA story is still expectations, and Zhongji Innolight’s minority stake is a hidden risk.

For investors following A-share export/globalization themes, the article’s core value isn’t in stock recommendations — it’s in redefining what “true overseas expansion” means. Going forward, “overseas production capacity ratio” may indeed be a more important metric than “overseas revenue ratio.”