高盛2026中国展望：观点、结构与趋势 / Goldman Sachs China 2026: Views, Structure, and Trends

Executive Summary

• 高盛的核心判断不是“中国已完成再平衡”，而是“出口与先进制造足以在2026年继续托住增长”。 它预计实际GDP增长4.8%，高于同期共识，并预计经常账户顺差升至GDP的4.2%。

Goldman Sachs is not arguing that China has rebalanced; it argues that exports and advanced manufacturing can keep growth resilient in 2026. It forecasts 4.8% real GDP growth and a current-account surplus of 4.2% of GDP.

• 我认为高盛对供给能力的判断有较强证据，但对这种模式的可持续性应更谨慎。 出口竞争力、产业升级和地产拖累的机械收窄可以支撑当期GDP；但家庭需求、就业、资产负债表和外部贸易容忍度决定中期上限。

Goldman’s supply-side case is well supported, but the durability of the model is less certain. Export competitiveness, industrial upgrading, and a smaller mechanical property drag can sustain near-term GDP; household demand, employment, balance sheets, and foreign tolerance set the medium-term ceiling.

• 深层结构是“资源配置偏向生产者、风险承担偏向家庭、需求调节依赖外部市场”。 这不是单一的消费信心问题，而是财政、金融、地方政府激励、社会保障和住房财富共同形成的制度性组合。

The deeper structure is producer-oriented resource allocation, household-heavy risk bearing, and external-market demand absorption. The imbalance is institutional, not merely psychological.

• 最可能的趋势是增长韧性与失衡扩大并存。 高技术出口占比上升、地产拖累下降、通胀仍低、贸易摩擦增加；真正的转折指标不是GDP本身，而是居民收入份额、服务业就业、社会转移支付和经常账户顺差是否同步改变。

The likely outcome is resilient growth alongside wider imbalances. High-tech exports rise, the property drag fades, inflation stays low, and trade frictions intensify; the decisive indicators are household income share, services employment, social transfers, and the current account—not GDP alone.

Evidence boundary / 证据边界

本报告阅读的是高盛公开英文材料，而非其仅向客户开放的完整研究数据库。核心材料包括2026年1月中国宏观展望、2025年11月预测上调说明、2026年1月公开删节版《Top of Mind》和2026年5月出口更新。事实、推断和情景在文中分开表达。

This report reviews Goldman Sachs’ public English-language research, not its full client-only research archive. The evidence base is therefore sufficient to assess the public macro thesis, but not every model assumption or proprietary dataset. Facts, inference, and scenarios are separated throughout.

高盛的乐观来自出口假设，而不是内需复苏 / Goldman’s optimism comes from exports, not a domestic-demand rebound

高盛的4.8%预测比IMF的4.5%和世界银行7月更新的4.4%更高。三者的差异不是对地产与消费问题是否存在的分歧，而是对出口韧性、政策宽松和外部冲击的权重不同。由于发布日期不同，这是一张“预测版本比较”，不是严格的同日共识调查。

Goldman’s 4.8% forecast exceeds the IMF’s 4.5% and the World Bank’s July estimate of 4.4%. The disagreement is not about whether property and consumption are weak; it is about how much export resilience and policy easing can offset them. Because publication dates differ, this is a comparison of forecast vintages, not a same-day consensus poll.

高盛观点及其逻辑链 / Goldman Sachs’ view and causal chain

1. Export-led resilience / 出口驱动的韧性

高盛预计未来数年中国实际出口年均增长约5%–6%，理由是对新兴市场出口扩张、关键矿产形成的谈判杠杆，以及芯片、汽车、零部件等高技术产品竞争力增强。其最具差异化的判断是2026年经常账户顺差由2025年的3.6%升至GDP的4.2%。 Goldman expects roughly 5%–6% annual real export growth over the next several years, supported by emerging-market diversification, critical-mineral leverage, and stronger high-tech exports. Its most distinctive call is a current-account surplus rising from 3.6% of GDP in 2025 to 4.2% in 2026.

2. A smaller property drag is not a recovery / 地产拖累减小不等于复苏

高盛判断全国房价可能仍需下跌约10%，实际房价或到2027年才见底；但由于新开工与投资已较峰值大幅下降，地产对GDP的年度拖累可从2024–2025年约2个百分点开始逐年缩小约0.5个百分点。 Goldman sees another roughly 10% decline in home prices and a possible nationwide real-price bottom only in 2027. Yet because starts and investment have already collapsed, the annual GDP drag can shrink mechanically from about 2 percentage points in 2024–2025 by roughly 0.5 point per year.

3. Policy support remains producer-heavy / 政策支持仍偏生产端

高盛认为十五五规划将继续强化传统产业升级、新能源和技术自立，财政、信贷、物流与地方政府配套共同提高制造业竞争力。消费支持会增加，但并非政策优先级的中心。 Goldman reads the 15th Five-Year Plan as reinforcing industrial upgrading, new energy, and technological self-reliance through fiscal, credit, logistics, and local-government support. Consumption support may increase, but it is not the dominant policy priority.

4. Low inflation and weak labor demand persist / 低通胀与弱就业并存

高盛预计2026年PPI仍下降0.7%，到2027年初才转正；CPI大部分时间低于1%。它同时指出招聘指标接近十年低位，2025年三季度城市名义工资增速约3.8%。 Goldman expects PPI inflation of -0.7% in 2026, turning positive only in early 2027, while headline CPI remains mostly below 1%. It also highlights hiring indicators near decade lows and urban nominal wage growth of about 3.8% in the third quarter of 2025.

高盛判断中最强与最弱的环节 / Strongest and weakest links in Goldman’s thesis

高盛最强的部分是对“产能—竞争力—出口”的解释；最需要保留意见的是从出口韧性推导整体宏观可持续性。下表把观点、点评和可证伪指标并列，避免把分析写成不可检验的叙事。

Goldman’s strongest argument is the production–competitiveness–export chain. Its weakest leap is from export resilience to broader macro sustainability. The table pairs each claim with an assessment and a falsification test.

深层结构：生产端强、家庭端弱、外部吸收 / Deep structure: strong producers, weak households, external absorption

事实 / Fact. 2025年中国GDP增长5.0%，最终消费、资本形成和净出口分别贡献2.6、0.8和1.6个百分点；CPI为0，PPI下降2.6%，70个大中城市二手房价格同比均下降。居民人均可支配收入增长5.0%，但财产净收入仅增长1.6%。这些数据支持“实际产出有韧性、名义需求和资产价格偏弱”的组合，而非全面繁荣。

推断 / Inference. 现有结构可概括为四个相互强化的机制： The structure can be summarized as four mutually reinforcing mechanisms:

1. 资本与财政更容易到达生产端。 地方政府、政策性金融、国有金融机构和产业政策更擅长形成工厂、设备、基础设施和技术能力。

Capital and fiscal channels reach producers more easily. Local governments, policy finance, state-linked credit, and industrial policy are effective at building capacity.

1. 家庭承担更多不可保险风险。 住房、教育、医疗、养老与就业不确定性提高预防性储蓄；地产下跌又削弱财富效应。

Households bear more uninsurable risk. Housing, education, healthcare, pensions, and job uncertainty raise precautionary saving; falling property prices weaken wealth effects.

1. 高技术制造提高生产率，但未必同步创造足够就业。 自动化和资本密集型产业能增加产出与出口，却可能降低每单位投资带来的就业和工资扩散。

High-tech manufacturing raises productivity without necessarily creating enough jobs. Capital intensity can lift output and exports while weakening employment transmission.

1. 国内储蓄超过国内可盈利吸收能力，差额转化为外部顺差。 因而经常账户顺差不是孤立的贸易现象，而是国内收入分配、投资回报和需求结构的镜像。

Excess domestic saving becomes an external surplus. The current account is the mirror image of domestic income distribution, investment returns, and demand weakness.

点评 / Assessment. 高盛正确识别了这一结构的短期稳定性，但对其政治经济反馈讨论不足：顺差越大，贸易伙伴采取关税、本地化要求、补贴和供应链去风险的激励越强，外部市场就越难无限承担中国的再平衡成本。 Goldman correctly identifies the structure’s near-term stability but gives less weight to its political-economy feedback: the larger the surplus, the stronger the incentives for tariffs, localization mandates, subsidies, and de-risking abroad.

结构约束下的六个趋势 / Six trends implied by the structure

1. 出口升级而非简单外迁。 劳动密集型产品对关税更敏感，高技术产品和“新三样”继续提高出口权重。

Export upgrading, not simple relocation. Labor-intensive exports are tariff-sensitive, while high-tech products and green technologies gain share.

1. 市场多元化伴随贸易壁垒扩散。 对美依赖下降不等于外部约束下降；摩擦会从美国扩散至欧洲、亚洲和新兴市场。

Market diversification coexists with wider barriers. Lower US exposure does not remove external constraints; frictions spread across regions.

1. 企业出海从销售转向本地生产。 为绕开关税和获得市场准入，更多中国企业会在海外建厂、合资、采购和本币融资。

Outbound expansion shifts from selling abroad to producing abroad. Firms localize production, partnerships, sourcing, and financing.

1. 地产从“增量冲击”变为“存量拖累”。 GDP拖累缩小，但库存、地方财政、抵押品价值和家庭信心修复更慢。

Property shifts from flow shock to balance-sheet drag. The GDP drag fades faster than inventories, local finances, collateral values, and confidence heal.

1. 低通胀持续压缩名义增长和利润分布。 “反内卷”可能改善部分行业价格，但若家庭需求不变，限制供给只能重新分配利润，不能自动创造总需求。

Low inflation keeps nominal growth and profits uneven. Anti-involution policies may improve sector pricing, but supply restraint alone does not create aggregate demand.

1. 政策会增加消费支持，但真正再平衡仍慢。 以旧换新和服务补贴能改善短期消费；更持久的变化需要社会保障、户籍公共服务、财政转移和家庭收入份额上升。

Policy will support consumption, but true rebalancing remains slow. Durable change requires stronger social protection, portable public services, fiscal transfers, and a higher household income share.

监测框架与情景 / Monitoring framework and scenarios

基准情景 / Base case. 出口和高技术投资支撑约4.5%–4.8%的增长区间，地产拖累继续收窄，CPI与PPI温和，贸易摩擦增加但未形成同步封锁。 Exports and high-tech investment support roughly 4.5%–4.8% growth; property drag narrows; inflation remains subdued; trade frictions rise without becoming coordinated containment.

上行情景 / Upside. 服务业就业、工资、社会转移和居民消费同时改善，顺差占GDP比重稳定或下降；这意味着供给能力开始转化为更平衡的国内收入循环。 Services employment, wages, transfers, and household consumption improve together while the current-account share stabilizes or falls.

下行情景 / Downside. 房价继续快速下跌、民营就业走弱、贸易伙伴同步加码壁垒，导致出口与内需同时承压；此时生产端刺激的边际回报会明显下降。 Home prices fall faster, private employment weakens, and trading partners coordinate barriers, pressuring exports and domestic demand simultaneously.

Recommended next steps / 建议关注

• 每季度同时跟踪实际GDP、GDP平减指数、居民可支配收入、服务业就业、房价和经常账户，不单独依赖工业增加值或出口。
• 区分“地产拖累减少”与“家庭资产负债表修复”；前者是流量变化，后者决定消费。
• 对行业与资产判断，应优先寻找具备海外本地化能力、真实服务需求和现金流定价权的企业，而非只看产量增长。

• Track real GDP, the GDP deflator, household income, services employment, home prices, and the current account together.
• Separate a smaller property-flow drag from household balance-sheet repair.
• For sector analysis, prioritize overseas localization, genuine service demand, and cash-flow pricing power over production growth alone.

Further questions / 后续问题

• 十五五规划的财政支出中，直接流向家庭和公共服务的比例是否持续提高？
• 海外本地化生产会扩大中国企业全球份额，还是降低国内出口与就业乘数？
• “反内卷”能否提高工资和居民收入，还是主要改善头部企业利润？
• 若经常账户顺差继续上升，主要贸易伙伴的政策反应将在何处先形成硬约束？

• Does the 15th Five-Year Plan durably raise the share of fiscal resources going to households and public services?
• Will overseas localization expand Chinese firms’ global share while weakening domestic export and employment multipliers?
• Will anti-involution policies raise wages and household income, or mainly improve leading firms’ margins?
• If the current-account surplus keeps rising, where will foreign policy constraints bind first?

Caveats and assumptions / 局限与假设

• 预测来自不同发布日期，不能视为严格的同日可比；世界银行7月预测已包含更多2026年实际信息。
• 高盛公开网页是客户研究的摘要，部分模型、情景概率和专有指标不可见。
• “深层结构”部分是基于公开证据的分析性推断，不是高盛原文，也不是因果识别结果。
• 本报告不构成投资、法律、税务或政策建议。

• Forecast vintages differ; the World Bank’s July forecast incorporates more 2026 observations.
• Public Goldman pages summarize client research and omit some proprietary models and scenario probabilities.
• The structural interpretation is an analytical inference, not a Goldman Sachs statement or a causal estimate.
• This report is not investment, legal, tax, or policy advice.