英雄

A |     

    Wind turbines operate at a wind farm in Burqin County, Altay Prefecture, Northwest China's Xinjiang Uygur Autonomous Region, on July 25, 2026. Photo: VCG
In economics, overcapacity is a dynamic and relative concept. Supply-demand balance is temporary, and imbalance is the norm. The Western narrative that equates the sheer scale of China's production capacity directly with "overcapacity" is, frankly, economically illiterate. It conflates size with pathology.
Consider the math. China's manufacturing sector accounts for roughly 30 percent of global manufacturing value-added - a dominance unmatched since the post-war US. Yet this reflects not overcapacity but competitive capacity. The critics ignore that China's trade surplus in 2025 reached a record $1.2 trillion, roughly equivalent to the GDP of a top-20 economy like Saudi Arabia. They ignore that Chinese exports carry high import content - foreign value-added accounts for nearly half of gross export value - meaning China's capacity integrates global supply chains rather than displacing them.
The judgment lacks rigor because it applies a static snapshot to a dynamic system. It assumes that if China produces more than it currently consumes domestically, the surplus is automatically "overcapacity." Nonsense. By that logic, Germany's persistent trade surplus in automobiles and machinery would constitute chronic overcapacity. It does not, because German capacity serves global demand. So does China's. The narrative isn't economics. It's politics dressed in economic jargon.
Western critics attribute the price advantages of Chinese green technology products to "government subsidies" and "market distortion." This is a convenient story. It is also wrong. The global competitiveness of China's green technology industries stems from four interconnected factors: massive domestic market scale, decades of sustained infrastructure investment, a highly educated STEM workforce, and fierce intra-industry competition. China's NEV purchase-tax exemptions, extended through 2027, total approximately 520 billion yuan ($77.05 billion). Substantial? Yes. But causally linked to overcapacity? No.
The causal chain critics imagine - subsidies → artificial competitiveness → overproduction - collapses under scrutiny. The EU's IPCEI battery programs allocated over 6 billion euros, plus EIB financing. The US CHIPS Act appropriated $54.2 billion for semiconductors, while the Inflation Reduction Act funnels hundreds of billions into clean energy. All major economies subsidize strategic industries. The difference is that China converted subsidies into systemic efficiency: its nominal lithium-ion cell capacity reached approximately 2.2 TWh by end-2023, driving global average battery pack prices down 20 percent year-on-year to $115 per kilowatt-hour in 2024. That is not distortion. That's innovation at scale.
Chinese firms filed 20,081 European patent applications in 2024 - 10.1 percent of the total, with battery technology patents surging 79 percent year-on-year. Patents reflect innovation, not subsidy dependency. The causal link between subsidies and overcapacity is weak; the link between scale, innovation, and competitiveness is robust.
Meanwhile, global demand for green and low-carbon transition is experiencing explosive growth. The International Energy Agency projects that renewable capacity additions must triple by 2030 to meet Paris Agreement targets. In this context, labeling China's wind, photovoltaic, and battery capacity as "overcapacity" is not merely wrong - it is dangerously counterproductive.
In fact, China's capacity has played an indispensable role in advancing the global energy transition. Chinese solar module production costs have fallen roughly 90 percent over the past decade. Goldman Sachs Research notes that China manufactured approximately 86 percent of global solar modules, 80 percent of lithium-ion batteries, and 68 percent of electric vehicles in 2024. This is not overcapacity dumped on foreign markets; it is capacity the world desperately needs.
The "overcapacity" label assumes a zero-sum game where Chinese production crowds out others. The reality is synergistic: Chinese capacity lowers global clean energy costs, enabling faster deployment worldwide. The "overcapacity" label holds only if one believes the world should slow decarbonization to protect less competitive producers and to empower those that deliberately employ "burn, baby, burn" tactics. That's industrial protectionism intersecting with brown imperialism, and it's disproportionately impacting the Global South, which is more vulnerable to climate breakdown, precisely when it's finally trying to shake off centuries of systemic Western exploitation and suppression.
Furthermore, the disparity in how the West treats its own industrial policy versus China's is glaring. When the US funnels $52.7 billion through the CHIPS Act for semiconductors, or when the EU subsidizes Airbus, or when American agricultural subsidies distort global food markets, the silence from Western media is deafening. When China gains an edge in manufacturing, it faces media offensives and policy restrictions.
This is not inconsistency. This is strategy. The underlying motivation is the "China threat" rubric - a framework that treats China's development itself as a danger to the Western-led order. Countries wagging their fingers at China do so not because China violates economic rules they themselves follow, but because China succeeds under rules they once wrote to their own advantage.
These accusers possess their own national industrial policies and subsidies. They have long exploited historical and contemporary advantages against China and developing countries. They have not competed well, have not provided themselves with good governance, and have not invested adequately in their own infrastructure, education, and R&D. They have allowed their own leading industries to seek monopolistic positions detrimental not only to others but to their own consumers. Rather than reform themselves, they blame China.
In these contexts, the cry of "China Shock 2.0" is the latest iteration of an old playbook: frame China's development as a threat to global stability rather than a contribution to global progress. The claim that China's advances in renewable energy and AI will send more serious shock waves than traditional manufacturing displacement is analytically flawed and politically motivated.
First, the premise is false. China is not reducing the market share of developed countries through unfair means; it is outcompeting them through innovation, scale, and efficiency. According to Goldman Sachs Research, China's "new three" green technology sectors alone contributed around one percentage point to the country's 5.4 percent nominal export growth in 2025. This is competitive capacity serving global demand, not overcapacity dumped on foreign markets.
Second, the "shock" framing ignores history. China maintained production and trade flows in 2020 while major Western economies crashed, providing essential goods to global markets, likely saving millions of lives while mitigating what would have been greater global economic harm. The "shock" is not to the global economy; it is to Western complacency.
Third, the argument is self-contradictory. The critics want China to be simultaneously weak enough to dismiss and strong enough to fear. If China's high-tech development is genuinely formidable - so innovative and efficient that it threatens to displace developed economies - then it is delivering precisely the technological progress the world needs to meet climate goals. Restricting it through tariffs and barriers would harm global innovation, raise energy costs for developing countries, and slow decarbonization. The critics would be sacrificing the planet to protect their own complacency.
"China Shock 2.0" is a rhetorical device to justify protectionism. It replaces engagement with containment, competition with confrontation. The world does not need another shock doctrine. It needs cooperation, investment, and the recognition that China's rise is not a threat but an opportunity - for green transition, for technological progress, for shared prosperity, and for a shared future. 
The author is a professor of politics and international relations and director of the Center for Ecological Civilization at East China Normal University in Shanghai. [email protected] 
。    全球AI市场规模高速增长。随着GPT、Deepseek等生成式AI问世驱使全球AI事业进入技术奇点时刻，全球AI市场规模进入高速增长期。根据IEA预测，2024年全球AI产业市场空间约234亿美元，到2032年全球市场规模有望达到2745亿美元，CAGR高达36%，成为全球不可小觑的增量市场。                             AI前景虽然看起来很美好，但是目前美国的电力系统却有点跟不上节奏。当前美国数据中心的电力承载能力已接近极限，美国超400吉瓦的数据中心供电申请，已达到全美峰值负荷的57%，但实际落地率仅约20%。                   微软CEO纳德拉在一档热门播客节目中说到：微软手中有大量AI芯片（GPU）处于闲置库存状态，其根源竟然是缺乏具备充足电力和冷却能力的机架空间，导致高性能芯片沦为“废铁”。                   美国AI缺电为中国电网企业带来的出海发展机遇。电力系统由发电、输电、配电、用电四大环节构成，其中配电环节处于发电与用电之间的核心位置。

B | 数据中心供配电系统目的在于实现市电对 IT 负载的供电并在市电断电时提供紧急备电功能。

C |                              配电系统是数据中心Capex中价值量占比最高的部分。参考施耐德 2023 年数据中心          投资成本计算器，假设 5MW 数据中心，单机柜功率 8kW，2N 配电、2N 冗余 UPS 架构下，配电系统价值量在建设成本的占比为 55%，制冷系统价值量占比为 19%。

D | 在配电系统中，UPS、发电机组、开关柜、关键配电的占比分别为 42%、23%、14%、 21%。

E |                              随着AI算力需求的增长和高性能GPU的部署，IT负荷功率大幅提升，同时数据中心物理空间有限，供配电功率密度大幅提升，供电系统向 800V/±400V 高压直流系统发展。远期AIDC向GW级发展，SST（固态变压器）等集成供电架构有望成为主流。                             1、思源电气                    思源电气于 1993年成立，深耕电气设备三十载，发展成为国内民营电力设备龙头。公司产品覆盖面广，具备一、二次设备生产和 EPC 能力。公司业务包括开关类、线圈类、 无功补偿类、智能设备类、工程总承包等。                             （基于功能分类，电力设备可划分成一次设备与二次设备，其中：（1）一次设备是指直接用于生产、输送和分配电能的高压电气设备，包括发电机、变压器、断路器、隔离开关、自动开关、接触器、刀开关、 母线、输电线路、电力电缆、电抗器、电动机等；（2）二次设备指对一次设备的工作进行监测、控制、调节、保护的电力设备，包括如测控仪表、控制开关、继电保护装置等。相比于一次设备更加侧重材料与工艺可靠性）                    近几年，思源电气的海外业务继续保持高速增长，2023 年实现海外收入 21.58 亿元，同比增长 15.71%，新增海外订单 40.1 亿元，同比增长 34%，海外订单占比达到 24%。值得注意的是，2023 年变压器海外订单增速超过 50%，显示出公司在核心产品领域的竞争优势正在显现。                   2024 年成为思源电气海外业务发展的关键年份，全年海外营收达到31.22 亿元，同比增长 44.67%，占总营收比例提升至 20.2%。公司在英国、意大利、沙特、科威特等多个高端市场陆续突破，欧洲市场（英国、意大利）、中东市场（沙特、科威特）新增订单增速均超过 50%。                   2025 年上半年，海外业务更是呈现爆发式增长，营收达到28.62 亿元，同比增长 88.95%，占总营收比例跃升至 33.68%，创历史新高。这一增长轨迹表明，公司的海外战略布局已进入收获期，其全球化战略的成功实施为未来的持续增长奠定了坚实基础。                   截至2025三季度末，公司合同负债（即预收款项）达28.96亿元，同比增长29.52%，表明客户预付的订单款项大幅增加，业务增长确定性较强。                             2、金盘科技                    变压器在输配电系统各环节中起着重要作用。

F | 变压器是通过电磁感应的原理来改变交流电压的装置，是输配电的重要基础设备，包括运行在主干电网的升压变压器和运行在终端的配电变压器两大部分。                   1 - 9月中国变压器出口额62.2亿美元，同比增长39%，高压开关出口额增长31.2%，9月单月增速55.7%，与海外电网升级需求相关，中国企业凭借供应链优势抢占海外市场。                             变压器按绝缘及冷却方式可分为干式变压器和液浸式变压器。干式变压器是指铁芯和线圈不浸在绝缘液体中的变压器，依靠空气对流进行自然冷却或增加风机进行冷却，具备环保无污染、安全性高、防火性好、免维护等优势。

G | 液浸式变压器是将铁芯和绕组浸在绝缘油中的变压器，电压等级可涵盖 10kV 至 1000kV 全系列，冷却性能更优。                             干式变压器是金盘科技切入AI数据中心市场的拳头产品。其干式变压器凭借低损耗、高过载能力和高可靠性，能很好地满足数据中心对电力供应稳定性和能效的苛刻要求。另外，公司不仅提供变压器，还配套提供低压柜、电力模块等产品，旨在为数据中心客户提供完整的强电侧供配电解决方案。面向未来AI算力基础设施的高功率密度趋势，金盘科技正积极研发固态变压器（SST） 和高压直流（HVDC）等新一代供电技术，以适应更高效率的供电架构需求。                   最近几年，金盘科技在AI市场取得了不错的业绩。2025年前三季度，公司在AIDC及IDC等数据中心行业收入达9.74亿元，同比飙升337%。这一增长速度，远超公司其他业务板块。更值得关注的是，AIDC业务在公司总收入中的占比已提升至约20%，成为不可或缺的重要组成部分。                             2025年前三季度，公司营收增长8.25%，而净利润增幅却高达20.27%，公司的盈利质量正在提升。2025年前三季度公司外销收入15.9亿元，同比增长约16%，而内销收入35.65亿元，同比增速仅为5%。海外市场的高增长不仅带来了规模扩张，更改善了公司的盈利水平。与此同时，公司的现金流状况大幅改善。前三季度经营活动产生的现金流量净额为1.78亿元，相比去年同期的-8715.77万元，实现由负转正。                              注意：文中均为长期逻辑。