In May 2026, China witnessed a staggering 111% year-on-year surge in integrated circuit exports. This remarkable figure served as a pivotal indicator for Goldman Sachs, signaling a shift in China's export paradigm.

On August 11, the Goldman Sachs Research Department, under the leadership of Jacqueline Du, unveiled a report with a groundbreaking assessment: China has officially stepped into the 'Global Expansion 3.0' era. The Chinese version of this report, widely disseminated in the market, bears the title 'China's Industrial Technology: Global Expansion 3.0 – China Embraces the AI Industrialization Era.'

From Apparel to AI: A Paradigm Shift

Goldman Sachs categorizes China's global expansion into three distinct phases. Phase 1.0 revolved around low-value consumer goods such as clothing and toys; Phase 2.0 encompassed home appliances, mobile phones, and the 'new three' – electric vehicles, lithium batteries, and photovoltaics. Now, Phase 3.0 heralds the export of high-tech, heavy-asset hardware with substantial technical barriers and unit prices, including data center power equipment, computing infrastructure, and industrial automation systems.

Goldman Sachs posits that the driving force behind this latest wave of global expansion differs from previous cycles. While earlier expansions were fueled by China's overcapacity and the quest for new markets abroad, this time, it's the emergence of supply gaps in overseas markets that's propelling the trend. Globally, there's a scarcity of suppliers capable of manufacturing these AI hardware components. Coupled with the dawn of a new technological adoption cycle, Chinese firms have inadvertently found themselves at the forefront of meeting this burgeoning demand.

A notable development is the rise of 'born global' companies. Goldman Sachs specifically highlights Unitree Robotics, a humanoid robot manufacturer. Unlike traditional manufacturing firms that first saturate the domestic market before venturing overseas, Unitree has treated the global market as its primary arena from the outset. Citing data from research firm Omdia, the report notes that Unitree shipped over 5,500 humanoid robots last year, dwarfing the shipments of its three U.S. counterparts – Tesla, Figure AI, and Agility Robotics – which each shipped around 150 units.

11 Products, Four Company Archetypes, with Varying Potential Ceilings

Goldman Sachs' report delineates 11 specific export product directions, estimating that by 2030, the global potential market size (TAM) for these products will range between $12 billion and $212 billion – a disparity of over seventeenfold among different products. Based on this analysis, Goldman Sachs categorizes these opportunities into four archetypal types:

Bottleneck Solvers: These firms stand to benefit most directly from global supply-demand imbalances. Goldman Sachs anticipates that from 2026 to 2030, companies in this category will witness the largest increases in market share and profit margins. The report assigns buy ratings to targets in this direction, including Sieyuan Electric, Kehua Data, and Yingliu Co., Ltd.

Technology Upgraders: These firms gain a competitive edge through rapid engineering iteration and technological advancements, exemplified by companies like Envicool (data center temperature control) and Hongfa (relay components). Goldman Sachs also assigns buy ratings to these firms, noting that Hongfa has successfully navigated overseas validation. However, the report cautions that their long-term success beyond 2030 hinges on the persistence of global supply-demand gaps and their ability to meet the operational standards of overseas large clients – localized after-sales and full-lifecycle services remain formidable advantages for European and U.S. competitors.

Established Global Players: Represented by energy storage and humanoid robot companies, the primary short-term challenge for these firms is market access – overcoming overseas regulatory, safety, and environmental compliance hurdles. Goldman Sachs assigns a neutral rating to Sungrow Power Supply.

Idiosyncratic Opportunities: Represented by industrial robot and automation companies like Inovance Technology and Estun Robotics, Goldman Sachs judges that their initial global expansion speed may be sluggish. Nevertheless, their product competitiveness, R&D speed, and cost advantages endow them with long-term potential for success.

In Line with the 'Long China AI' Thesis

This isn't the only report by Goldman Sachs this year that expresses optimism about China's AI sector. In July, the Goldman Sachs Asia Portfolio Strategy team, led by Kinger Lau, released a report titled 'Long China AI Value Chain,' centered around a 'triple mismatch': China's AI sector accounts for a mere 10% of the global AI market capitalization but contributes 16% of the revenue, while global mainstream funds allocate only 1.2% of their portfolios to Chinese stocks. Goldman Sachs calculates that if this mismatch is gradually rectified, it could usher in over $100 billion in incremental capital inflows, and the economic value unlocked by the AI industry itself could be 50% to 100% higher than current stock price expectations suggest.

Taken together, the two reports present a coherent narrative: the July report focused on the funding and valuation mismatch, advocating for 'Long China AI'; the August report examines industrial and export-side validation, confirming that 'China's AI is genuinely selling hardware abroad.' In Goldman Sachs' own words, China's AI sector isn't a valuation泡沫 (valuation bubble) devoid of fundamental support but rather an undervalued industrial reality.

Implications for Global Expansion Practitioners

For domestic companies engaged in global expansion and their investors, this report conveys at least two pivotal signals.

Firstly, the criteria for stock or sector selection must evolve. Previously, China's exports were viewed through the prism of exchange rates, freight costs, and price advantages; now, Goldman Sachs offers a framework that emphasizes technological certification progress and the visibility of overseas orders. Only those firms that genuinely secure access to overseas data centers and major overseas companies will enjoy valuation re-ratings, rather than merely shipping goods to Southeast Asia.

Secondly, the distinction between 'pseudo-global expansion' and 'true global expansion' will become increasingly pronounced. Among Goldman Sachs' four archetypes, Bottleneck Solvers reap the benefits of supply-demand gap dividends, while Idiosyncratic Opportunities rely on long-term R&D and cost advantages. The time windows for these two types are vastly different – supply-demand gaps will eventually be bridged, and only firms that truly establish technological barriers and localized service capabilities will endure once the gaps close.