Goldman: China Has Entered “Go Global 3.0,” Mapping 11 AI Export Battlegrounds Worth $212B
Goldman Sachs published a sweeping 48-page equity research report on August 11, 2026, outlining what it calls China's "Go Global 3.0" era — a structural shift in the country's export ambitions that moves decisively beyond cheap manufactured goods and electric vehicles into AI-enabled industrial technology. The report, authored by a cross-regional team led by Jacqueline Du at Goldman Sachs (Asia) L.L.C., identifies 11 discrete product categories where Chinese companies could emerge as globally relevant export winners, with addressable markets ranging from US$12 billion to US$212 billion by 2030. The timing is not coincidental: as Washington expands regulatory scrutiny of Chinese technology — most recently with the US FCC's July 28, 2026 decision to restrict new Chinese humanoid and quadruped robots — Goldman is telling investors that market access has become as important a variable as product competitiveness.
From "New Three" to AI Industrialization
The report's framing is deliberately evolutionary. Go Global 1.0 was about low-cost manufactured goods. Go Global 2.0 was defined by what Beijing called the "New Three" — electric vehicles, lithium-ion batteries, and solar photovoltaics. Go Global 3.0, Goldman argues, is categorically different: it is characterized by the export of AI-enabled, technology-led industrial capabilities, spanning AI data center power infrastructure all the way to physical AI in the form of robotics and automation.
Goldman's analysts estimate that the 11 Chinese companies they track as sector leaders will derive an average 35% of revenue from overseas markets in 2026, rising to 40% by 2030 — compared to roughly 16% for the broader China-listed universe. But the bank is quick to flag that this aggregate figure obscures wildly divergent trajectories. "Overseas growth actually requires a more demanding playbook," the report states, "with category-specific and player-specific decisions."
Four Archetypes, Four Very Different Outlooks
The report's analytical backbone is a four-archetype framework designed to separate near-term opportunity from durable competitive advantage.
Bottleneck Solvers — covering gas turbines, transformers, switchgear, and uninterruptible power supplies (UPS) — represent Goldman's highest-conviction near-term trade. These are sectors where global ex-China supply shortages of 8%-34% over 2026-2030 are creating immediate demand pull for Chinese suppliers. Goldman is Buy-rated on Sieyuan Electric, Shenzhen Kstar Science & Tech, and Yingliu in this bucket. For Sieyuan, the bank forecasts US transformer market share gains from 2% to approximately 9% by 2030, with gross profit margins in the US running roughly 17 percentage points above its China business — driven by supply tightness, faster lead times of 6-9 months versus 2-3 years for Western peers, and AI data center demand. The caveat is explicit: "Sustainability beyond 2030E will depend on the duration of the supply/demand gap."
Technology Upgraders — Shenzhen Envicool Technology in data center cooling, Megmeet in server power supply units, and Hongfa Technology in relays — are benefiting from technology transitions such as the shift to 800VDC architectures that can partially reset competitive dynamics. Goldman is Buy-rated on Envicool and Hongfa, and Neutral on Megmeet. Envicool's US gross margin is projected to run 17 percentage points above its China business over 2026-2030, supported by liquid cooling products validated by NVIDIA and Google. Hongfa is described as the standout exception in the framework — a Chinese company that has already demonstrated durable competitive strength, with European market share projected to rise from 21% to 32% by 2030.
Established Global players — Sungrow Power Supply in energy storage systems and private-company Unitree Robotics in humanoids — face the starkest tension in the report. Goldman stays Neutral on Sungrow, citing the One Big Beautiful Bill Act's (OBBBA) restrictions on tax credits for projects reliant on Chinese supply chains as a direct headwind. Unitree, meanwhile, is held up as the purest expression of what Goldman calls the "Born Global" phenomenon — a company that has treated the global market as its primary market since inception, shipping over 5,500 humanoid robots in 2025 for a 37% global share per Omdia, at price points (US$4,900-US$100,000) that dramatically undercut Boston Dynamics (US$30,000-US$1,000,000). The July 2026 FCC ruling clouds the outlook, though Unitree states its current major models have already obtained FCC certification.
Idiosyncratic Opportunities — Shenzhen Inovance Technology in industrial automation and Estun Automation in industrial robots — lack a clear structural demand catalyst and must win share through sustained execution. Goldman is Buy-rated on Inovance but has a Sell on Estun's A-shares, noting fierce price competition and an inconsistent earnings track record despite the company's domestic leadership.
The Long-Term Competitive Reality
Goldman's global analyst network is candid about where Chinese companies still fall short. Across 12 product-level competitiveness assessments covering six dimensions — time-to-market, price, R&D iteration speed, product quality, certification/ecosystem, and global service — only three categories (relay, humanoid robots, and ESS) scored at or above parity with Western peers on more than three of six metrics. The structural gap in global service coverage, lifecycle support, and certification remains the defining constraint on long-term durability. As Goldman puts it: "Establishing robust, localized aftermarket support and lifecycle services remain a key differentiator for Western incumbents."
Geopolitical risk is not an afterthought. The report dedicates substantial analysis to trade hurdles — from FEOC compliance risks for Sungrow to national security concerns around Chinese grid equipment in US bulk power systems — and localization challenges. Using Sanhua Intelligent Controls as a case study, Goldman notes that most Chinese industrial tech leaders still retain 80-100% of production capacity domestically, making current global expansion "still predominantly China-centric from a production perspective."
The report's bottom line is nuanced in a way that distinguishes it from simple China bull or bear narratives: the Go Global 3.0 era is real, the product competitiveness is increasingly credible, but the path from mid-term share gains to durable global leadership runs directly through the geopolitical and regulatory gauntlet that Western governments are actively constructing.
Related Coverage: