S
SHAWNFU

Lao Li Hai Channel · "Goldman Sachs" 20261113: China Economic Outlook for 2027, Strong Externally Weak Internally, Return to Deflation

LongbridgeAII'm LongbridgeAI, I can summarize articles.

Goldman Sachs 2027 China Outlook: The optimism of a 4.5% GDP growth relies on 'less drag' rather than 'improvement'

Takeaway: Goldman Sachs forecasts a 4.5% real GDP growth rate for 2027 (higher than Wall Street's 4.4% and the IMF's 4.0%), but this 'optimism' is not because the economic fundamentals are turning around—the property sector is still declining, broad fiscal policy remains tight, export volumes continue to slow, CPI will drop from 0.8% to 0.5%, and PPI will turn negative again in Q2. The source of the 4.5% lies in two drags simultaneously easing (property -1.8 to -1.3 percentage points, fiscal tightening 1.4 to 0.5 percentage points), creating an extra 1 percentage point on paper, which just fills the gap left by slowing exports.In other words: next year is a year of 'falling slower', not a year of 'getting better'.

Video: Lao Li Hai 20261113 "[Goldman Sachs] 2027 China Economic Outlook, Strong Externally Weak Internally, Return to Deflation"
Research Report Title: Less Drag, More Uncertainty
Original Video: https://www.youtube.com/watch?v=dLRF-KLL1Tk


I. The Arithmetic of 4.5%: Two Accounts of Decreasing Drag

Goldman Sachs's own model actually cannot calculate 4.5%.

Decomposition of Actual GDP in 2026 Percentage Points
Property Drag -0.1 (marginal change relative to baseline)
Export Drag -0.4
Fiscal Contraction Drag -1.1
Model-implied Growth Rate 3.4
Officially Reported (First Three Quarters) 4.6
Unexplained Residual (unexplained) +1.2

Goldman Sachs directly labels this item as 'unexplained' in its charts. The author's understanding is the inventory accumulation effect under production-based statistics: goods equivalent to about 1.2% of GDP were produced but not sold this year, contributing to GDP on the production side but not realized on the income side. This also explains why the residual item in 2027 is instead -0.9 percentage points—the inventory piled up this year needs to be digested next year, and production lines have to slow down.

So the composition of 4.5% next year is:

Supporting Items Calculation
Narrowing Property Drag -1.8 → -1.3 percentage points, equal to a +0.5 positive contribution
Narrowing Fiscal Tightening Broad deficit ratio 11.4% (2025) → 10% (2026) → 9.5% (2027), collecting 1.4 percentage points this year and only 0.5 percentage points next year, equal to a +0.5 positive contribution
Total of Both Items +1.0 percentage point
Corresponding Gap Next year's export pull on GDP drops from 1.2 → 1.0 percentage points, a gap of -0.2; further compounded by the residual turning negative

Conclusion: The GDP figure can hold up, but micro-level sentiment won't change. The author's original words are very straightforward—the feelings people have regarding employment, consumption, and prices this year will likely remain the same next year.

The Other Side of Structural Transition: The Property Sector's Scale Has Halved

Share of GDP (Including Upstream and Downstream) 2015 2025 2030 (Forecast)
Broad Real Estate (Construction + Services + Home Furnishings + Land Finance) ~30% 15% 8%
High-Tech Industries 14% 23% 28%

This is the physical reason for 'decreasing drag': it's not that property has stopped falling, but that it has become so small that its decline no longer drags down the entire economy. Goldman Sachs's real estate team expects land transfer fees and new starts to fall another 20% next year.


II. Three Divergences Against Market Consensus

Divergence 1: Export Surplus to Set New Highs (Goldman Sachs Most Optimistic)

Indicator 2026 2027 (Goldman Sachs) Market Consensus
USD-denominated Export Growth 22.9% 13.7% —
Real Export Volume Growth 8.5% 6.0% —
Goods Trade Surplus $1.25 trillion ~$1.5 trillion —
Current Account Surplus/GDP 4.4% 4.9% Bloomberg 3.6%, IMF 4.0%
Broad Balance of Payments Surplus 2.9%/GDP $838 billion, 3.5%/GDP —
USD-denominated Import Growth 28.3% 12.3% (real volume only 1.6%, price dropped from 18.9% to 4.5%) —

The key lies in 'volume resilience, price retreat': exports surged this year, mainly due to sharp price increases—price hikes in AI-related semiconductors, integrated circuits, and memory, combined with rising oil prices driven by the Strait of Hormuz (referred to as the Iran war in the report), pushed export prices up by over 20% year-on-year; while the share of AI hardware in real export volumes is actually negligible. What truly supported the 9.5% volume increase was the 'New Three' (electric vehicles, lithium batteries, solar panels) and power equipment, machinery. Next year, chip prices will peak in Q4 and oil prices will fall, eliminating the price contribution, but volume can still rely on cost advantages and a complete industrial chain to capture emerging market shares, with export volumes likely stabilizing at 5%-6% from 2027 to 2030.

Divergence 2: Inflation Much More Pessimistic than Market (Completely Opposite Direction)

Indicator 2026 2027 (Goldman Sachs) Market Consensus / IMF
CPI 0.8% 0.5% 1.4% / 1.5%
Core CPI 1.0% 0.6% —
PPI 2.9% 0.3% (turns negative in Q2 single month) —
Brent Crude Oil — $88/barrel —

This year, PPI turned positive in March after 41 consecutive months of declines, reaching 2.9% for the full year, and the GDP deflator ended 12 quarters of negative growth with +1.5% in Q2—but this is imported inflation, not demand-driven. Upstream price hikes do not transmit to the terminal: both CPI and core CPI remained below 1% throughout the year due to weak domestic demand. Next year, with falling oil prices and high base effects for chips, PPI will return directly to 0.3%, turning negative again in Q2. This judgment aligns with Bank of America, Citigroup, and Barclays.

Divergence 3: RMB Appreciation to 6.3

Timepoint USD/RMB
End of 2026 6.7
+3 Months 6.6
+6 Months 6.5
End of 2027 6.3

This implies an annual appreciation of 3%-5%, with a path significantly faster than the forward market implied path. Three drivers: trade surplus increasing from $1.3 to $1.5 trillion, policymakers' inclination towards strong RMB internationalization, and strong willingness of exporters to settle foreign exchange. Counter-forces are also present: the US-China interest rate spread remains large and widening; the central bank has already raised the foreign exchange deposit reserve requirement ratio from 2% to 4%, with an attitude of controlling the pace, not changing the direction.


III. Domestic Demand: Weakness Across Housing, Jobs, and Consumption

Real Estate: Sixth Year of Decline, Core Tier-1 Cities Are the Only Exception

New starts, sales area, and development investment have fallen 60%-90% compared to the peaks in 2020-2021; the pre-sale system reform (shifting from pre-sales to completed-home sales) in late August this year, along with tightened project fund supervision, will continue to suppress these indicators in the coming years. The language on housing prices is restrained: only tier-1 cities like Shanghai and Shenzhen, where inventories are already very low, 'may recover'; others, especially lower-tier cities, will continue to fall, with national second-hand home prices not bottoming out until 2030. The Goldman Sachs macro team did not cite its own real estate team's view of 'Shanghai and Shenzhen cumulative gains of 20% over two years (15% six months ago)' throughout the report.

Employment: Stable Official Unemployment Rate, Very Pessimistic Roadshow Feedback

Three suppressions: ① AI replacing jobs; ② mandatory new social security policies increasing labor costs and suppressing hiring willingness; ③ structural issues—the labor intensity of the high-tech industry is far lower than that of real estate and services, heavy investment in technology can drive industrial production and GDP, but cannot drive employment and resident income, only translating into corporate-side income. Resident income-increasing policies are expected to only include 'gradual small moves' (small subsidies for labor-intensive services and offline entities, moderate increases in minimum wage, lowering the threshold for social security contributions for riders/ride-hailing drivers), with no substantial employment or income-increasing policies to be introduced.

Consumption: Next Year Relies on 'Low Base', Not Policy

Item 2026 2027
Real Final Consumption Growth 4.0% 4.3%
Of Which Resident Consumption 4% 4.2%
Of Which Government Consumption 4% 4.5%
Nominal Retail Sales Growth 1% 2%
Final Consumption Pull on GDP 2.3 percentage points 2.4 percentage points (3.6 in 2019)
Capital Formation (Investment) Growth 2% 2.5% (increment mainly from 'Six Networks')

Trade-in subsidies phase out annually: 300 billion in 2025 → 250 billion in 2026 → likely to decrease further in 2027 (author estimates around 200 billion). New energy vehicle purchase tax remains halved (5%) in 2027, but returns to 10% in early 2028, bringing forward purchases at the end of 2027, boosting year-end retail sales.

Three 'low base' reasons support 4.2%: ① The negative wealth effect of falling housing prices peaked in 2024-2025, with diminishing marginal drag; ② The overdraft effect of trade-ins was fully digested this year; ③ Purchase tax policy timing.

The real concern is the household balance sheet: Seasonally adjusted savings rate rose from 32.7% in Q2 to 33.4% in Q3 (pre-pandemic trend was about 30%); household deposits have reached nearly 180 trillion yuan, and despite repeated rate cuts, they haven't been forced out. Goldman Sachs highlights the risk that sustained weakness in domestic demand could lead to a 'structural downward shift in the consumption growth center'—and the author believes this has already happened: the trend growth rate of retail sales dropped from an average of 9% annually from 2015-2019 to 3.5% post-pandemic, with the 15th Five-Year Plan's implicit 60 trillion yuan target for 2030 corresponding to an annualized rate of only 3.7%.

Service consumption is the only relative bright spot: service retail sales increased by +5% in the first three quarters, 4 percentage points faster than goods retail sales; China's service consumption accounts for 46% of resident consumption, compared to about 60% in the US and Japan. The consensus among government and institutional experts during roadshows in Beijing was: consumption-promoting policies in 2027 will be very limited—because no one dares to guarantee that any policy can sustainably boost consumption (if it doesn't work, they'll bear responsibility), and fiscal and tax system reforms involve central-local relations and interest games, with vested interests unwilling to voluntarily cut the cake.


IV. Policy: Fiscal Continues to Tighten, Monetary Eases Once in Q4

Policy Variable 2026 2027
Broad Fiscal Deficit Ratio ~10% 9.5% (first time below 10% since 2019)
Policy Rate (Overnight Reverse Repo) 1.25% 1.15% (cut 10bp in Q4, after the 21st Congress)
Social Financing Stock Growth — 6.2% (year-end, continuing to step down)
Debt Resolution Quota 2.8 trillion 0.8 trillion (annually in 2027-2028, 2 trillion less than this year)

The batch of incremental policies at year-end—adding 550 billion yuan to local government debt settlement limits and deploying 800 billion yuan in new policy financial tools—Goldman Sachs judges that the purpose is merely to achieve this year's GDP target, belonging to short-term policy padding, and does not change the trend of broad fiscal contraction.

Two structural constraints are worth remembering:
1. LGFVs must complete delisting and market-oriented transformation by mid-next year, coupled with further deterioration in land transfer fees after the pre-sale system reform, leaving local governments 'out of money';
2. The 21st Congress in autumn next year, under high-pressure anti-corruption and local leadership transitions, officials adopt a 'do nothing, make no mistakes' approach, and evaluation priorities are now clearing enterprise arrears and paying back owed civil servant salaries, not launching projects—local governments are 'unwilling to spend money'.

Interest rate benchmark change: In November, the central bank shifted the policy rate from the 7-day reverse repo to the overnight reverse repo at 1.25%, with a corridor upper/lower bound of 1.5%/1%. Goldman Sachs insists that this implies a 5-10bp rate cut—because the difference between 7-day and overnight is only 6 days, it shouldn't create a 15bp spread on the curve (compared to the 7-day vs. 10-year spread being over 9 years, the spread is also less than 20bp).


V. 'More Uncertainty': Four Variables

Two External
1. Whether global AI capital expenditure can maintain intensity—this is an exogenous variable for China's export prices and manufacturing prosperity.
2. US-China relations—four headmeetings this year, tariff truces, and rare earth export control agreements have all been extended, but expire on January 10 next year; with Democrats sweeping both houses in the midterm elections, next year's variables are far greater than this year's.

Two Internal
3. LGFV delisting + debt resolution quota cut from 2.8 trillion to 0.8 trillion (LGFV funding directly decreases by 2 trillion next year).
4. The political cycle before the 21st Congress—stability first, high-pressure anti-corruption, official risk aversion, determining that fiscal policy remains relatively tight.

Reverse Risk (Worth Noting Separately): Goldman Sachs writes in its risk warnings that if the economy next year is better than expected, besides exports or US/EU AI capital expenditures exceeding expectations, there is another scenario—the economy is too poor. Severe weakness in domestic demand, exports significantly missing expectations, systemic financial risks, or large-scale unemployment leading to social instability would force policymakers to introduce stimulus. Translation: Stability is paramount at the 21st Congress next year; as long as ordinary people can endure, policy won't move.


VI. Market Implications and Actionable Conclusions

Goldman Sachs is net long across equities, bonds, and FX, but equities offer no beta, only alpha.

A-Shares / HK Stocks
- Overweight rating for Chinese equities maintained, with expected returns of several dozen percentage points (about 10%) for 2027, but the driving logic shifts: tech hardware profit growth driven by AI computing power this year will slow next year, and institutional positions in AI hardware are 'very crowded', posing liquidity and sentiment volatility risks with continued downside in early next year.
- Alpha direction is clear: AI downstream applications + Internet (the latter purely because valuations are too low). The essence of this recommendation is a rotation from upstream hard tech to downstream applications—the author points out that reading by geography means rotation from A-shares to HK stocks (Goldman Sachs doesn't explicitly state this).
- Mapping to tickers: HK internet is the direct carrier of this logic—Tencent (700.HK), Alibaba (9988.HK), Meituan (3690.HK), Baidu (9888.HK); the flip side of AI hardware crowding is amplified volatility for 'hardware + terminal' tickers like Xiaomi (1810.HK) in early next year.
- Export chain: Volume remains (+6%), but price contribution disappears, benefiting share, suffering from lost price elasticity—new energy vehicles and batteries (BYD 1211.HK) need to guard against tightening EU trade policies, with Goldman Sachs judging that China-EU relations won't spiral out of control: mutual dependence (China has rare earths, EU wants to protect EV market share in China), and restrictions are more about 'changing export forms' (forcing EV makers to build factories in Europe, which 反而 boosts intermediate goods like lithium battery exports) rather than cutting total volume.
- RMB appreciation of 3%-5%: Direct beneficiaries are industries with heavy USD liabilities and high import costs—aviation (Air China 753.HK), papermaking, some refining; losers are export manufacturers primarily earning USD with weak bargaining power (home appliances, solar modules), but the moderate appreciation keeps hedging costs manageable.
- PPI returning to 0.3% and turning negative in Q2: Upstream cyclicals (coal, steel, chemicals, oil services) face earnings pressure, while mid/downstream cost ends benefit (utilities, consumer manufacturing); oil falling to $88 is negative for oil extraction but positive for aviation/logistics/midstream chemicals.

Bonds
- Long-end treasury yields continue to hit new lows, four reasons: weak growth momentum, return to deflation, credit growth dropping to 6.2%, broad fiscal tightening reducing net treasury supply. Long-duration rate bonds remain the 顺势 direction.
- Changes in bond tax exemption policy (levying 25% income tax on interest from treasuries and local bonds held by commercial banks): Goldman Sachs believes the announcement will push yields jump a notch, but won't change the long-term downward trend, with the bulk of the tax burden ultimately borne by investors.

US Stocks / Chinese Concepts
- Global AI capital expenditure is the largest exogenous variable for China's export prices and manufacturing prosperity next year, $NVIDIA(NVDA.US) and $Broadcom(AVGO.US)'s capex guidance are effectively leading indicators for A-share AI hardware chains—if guidance is revised down, the 'price contribution' of Chinese exports will recede again.
- Valuation repair for Chinese concept internet (e.g., $PDD(PDD.US)) relies more on policy and domestic demand turnaround, with Goldman Sachs citing 'valuations are too low' rather than earnings inflection points, making it a odds play rather than a trend play.

One-sentence operational meaning: Next year's money isn't in 'economic recovery', but in the intersection of 'decreasing drag + loose liquidity + low valuation'—HK internet and AI applications, long-duration rate bonds, and USD liability assets benefiting from RMB appreciation; avoid crowded AI hardware and upstream cyclicals with disappearing price elasticity.


Personal study notes, not investment advice. Data cited from Goldman Sachs's 2027 China Economic Outlook research report relayed in the video;口径 and assumptions subject to the original report; the 'author's understanding' section reflects video viewpoints, not original report text.

The copyright of this article belongs to the original author/organization.

The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.