China equity strategy: JPMorgan maps four China stimulus paths, retaining an Overweight allocation while favoring AI and policy-linked infrastructure exposure.
The report’s base case is a measured fiscal catch-up within the approved budget, supporting the “Six Networks” investment chain over the next two to three quarters. A broad property-and-consumption stimulus would favor high-beta consumer and property names, while fiscal disappointment would favor low-P/B, high-yield cash-flow stocks.
Summary
The report’s base case is a measured fiscal catch-up within the approved budget, supporting the “Six Networks” investment chain over the next two to three quarters. A broad property-and-consumption stimulus would favor high-beta consumer and property names, while fiscal disappointment would favor low-P/B, high-yield cash-flow stocks.
- JPMorgan maintains Overweight China equities in EM/Asia allocations.
- Base-case fiscal execution is supportive rather than expansionary, with over Rmb7 trillion of 2026 investment planned under the Six Networks.
- China data-center capex is estimated to rise from US$55–69 billion in 2026 to US$95–151 billion by 2029.
- A monetary-easing-only outcome historically favors micro/small caps, growth, healthcare, IT and materials over three months.
- Fiscal under-delivery would shift preference toward low-P/B, high-yield companies with positive free cash flow.
Report Interpretation
Overview
JPMorgan assesses four policy-stimulus scenarios for China equities and pairs each with sector and stock screens. Its central case is improved fiscal execution within the existing budget, favoring infrastructure and AI-related “Six Networks” beneficiaries, while it retains an Overweight China-equity allocation despite reasons for caution in the coming weeks.
Core views
The report frames China equities around four possible policy outcomes. JPMorgan’s base case is not a major fiscal expansion but a measured catch-up in execution within the approved budget: faster government-bond issuance, drawdown of fiscal deposits and policy-bank funding should support infrastructure, urban renewal, project completion and strategic investment. Constraints from weak land revenue, local-government risk aversion and project readiness mean the impulse remains supportive rather than expansionary. If spending follows the year-to-date “Six Networks” framework, the report expects policy to turn into order backlogs, revenue recognition and operating leverage over the next two to three quarters. The Six Networks provide the base-case investment transmission. JPMorgan estimates China data-center capex could roughly double from US$55–69 billion in 2026 to US$95–151 billion by 2029, benefiting AI servers, optical modules, liquid cooling, IDC operators and the domestic AI-chip supply chain; Xinhua News has reported around Rmb4 trillion of computing-network spending over the 15th Five-Year Plan. The new-type grid adds more than Rmb5 trillion of planned spending, about 75% above the prior plan, directed toward UHV equipment, distribution automation, smart meters, storage inverters and virtual power plants. Telecom investment includes 500,000 new 5G-A base stations and is projected to generate Rmb7 trillion of upstream and downstream output. Water supply, underground pipes and logistics are more balance-sheet-intensive sleeves, with more than Rmb6 trillion for water, around Rmb5 trillion for 770,000 km of urban pipes, and a logistics backbone targeted for 95% completion. In the upside case, policymakers would go beyond the approved budget with genuinely incremental central or quasi-fiscal funding, central risk-sharing, quicker project execution, stronger housing-completion support and consumer measures. JPMorgan compares this with September 2024, when a one-month rally was led by appliances, autos, property developers and brokers, second-tier baijiu and selected building-materials companies; export-oriented consumer names and defensive high yielders lagged. The report says a repeat would likely produce another short, high-beta consumer-and-property rotation. It notes that investors see export growth slowing to mid-single digits as a possible precondition, while the October 5th Plenum is debated as a potential catalyst. In the downside case, financing fails to translate into spending: bond proceeds and policy-bank funding arrive slowly, weak land revenue and scarce matching funds constrain local implementation, fiscal deposits remain elevated and part of the policy impulse is deferred. Combined with poor liquidity, this would produce a flight-to-safety regime emphasizing valuation, balance-sheet strength, cash flow and dividends. JPMorgan’s defensive screen requires MXCN membership, three-month average daily turnover above US$30 million, trailing P/B below 1.5x, positive year-on-year free cash flow in the latest fiscal half-year, dividend yield above 4%, and positive 2026 and 2027 consensus EPS growth. The resulting list is concentrated in insurers, selected banks, energy and consumer staples. A fourth scenario is monetary easing without material fiscal follow-through: policy-rate cuts, lower interbank rates and looser liquidity. Historical back-testing suggests micro and small caps lead over three months and growth modestly outperforms value. Cyclicals react first, but healthcare, IT and materials lead over three months while utilities, staples and financials underperform. At six months after rate cuts, the median Wind Micro Cap return was +9.5% and CSI 1000 +5.0%, versus -2.5% for CSI 300 Value. JPMorgan maintains Overweight China equities in EM/Asia allocations and end-2026 targets of 100 for MXCN and 5,200 for CSI 300, underpinned by consensus EPS-growth projections of 14% and 25%, respectively. It views AI as a tier-one growth theme after 2Q26 outperformance and expects the ecosystem to rebound through the remainder of 2H26, though liquidity turbulence may make the path uneven. Favored domestic AI supply-chain names are Iluvatar CoreX, V-Test, JECT, AMEC and NAURA; non-AI diversification picks are Meituan, Bank of China-H, Bank of Ningbo, CICC-H, Innovent, BYD-H and China Resources Land. Near-term caution reflects potentially tighter policy and liquidity, a worsening Asia QMI cycle indicator, US mid-term-election risk aversion, no fresh earnings catalyst before October 3Q results and unfavorable seasonality.
Analysis framework
JPMorgan constructs policy scenarios, links each macro assumption to likely market leadership and then screens stocks using exposure, liquidity, valuation, cash-flow, dividend and earnings-growth criteria. It also uses historical market reactions to the September 2024 stimulus and policy-rate-cut back-testing to compare likely style and sector outcomes.
Methodology notes
Policy-scenario analysis using the September 2024 stimulus episode as a reference point.
The report compares possible new fiscal and consumption support with the prior stimulus episode to identify sectors likely to lead or lag in a short-term market rotation.
Up-beta and post-rate-cut historical return analysis.
JPMorgan uses stocks’ sensitivity in rising markets and historical median returns after policy-rate cuts to assess which market-cap, style and sector groups could lead.
Defensive low-P/B, cash-flow and dividend screen.
For the fiscal-under-delivery scenario, the report identifies liquid MXCN stocks with P/B below 1.5x alongside positive free cash flow, dividend yield above 4% and positive consensus EPS growth.
Asset mapping & comparison
Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).
- Iluvatar CoreX (9903.HK)Favored domestic AI supply-chain player.
- Strengths
- Included among JPMorgan’s preferred domestic AI names.
- Risks
- Liquidity turbulence could make AI-theme performance choppy.
- V-Test (688372.SH)Favored domestic AI supply-chain player.
- Strengths
- Included among JPMorgan’s preferred domestic AI names.
- Risks
- Liquidity turbulence could make AI-theme performance choppy.
- AMEC (688012.SS)Favored domestic AI supply-chain player.
- Strengths
- Included among JPMorgan’s preferred domestic AI names and the AIDC screen.
- Risks
- Liquidity turbulence could make AI-theme performance choppy.
- NAURA (002371.SZ)Favored domestic AI supply-chain player.
- Strengths
- Included among JPMorgan’s preferred domestic AI names and the AIDC screen.
- Risks
- Liquidity turbulence could make AI-theme performance choppy.
- Meituan (3690.HK)Top non-AI diversification pick.
- Strengths
- Named as a preferred non-AI diversification holding.
- Bank of China-H (3988.HK)Top non-AI diversification pick and defensive-screen constituent.
- Strengths
- Named as a diversification pick; represented in the low-P/B, high-yield defensive screen.
- Risks
- Would be affected by the report’s broader liquidity and macro risks.
- BYD-H (1211.HK)Top non-AI diversification pick.
- Strengths
- Named as a preferred non-AI diversification holding.
- China Resources Land (1109.HK)Top non-AI diversification pick.
- Strengths
- Named as a preferred non-AI diversification holding.
- Risks
- Property-linked performance depends on the strength of policy support.
Key data
- Six Networks investment planned for 2026More than Rmb7 trillionBase-case planned investment for 2026.
- 15th Five-Year Plan Six Networks investmentRmb23–26 trillionEstimated aggregate investment over the plan period.
- China data-center capexUS$55–69 billion in 2026; US$95–151 billion by 2029JPMorgan estimate; roughly doubles over the period.
- New-type grid spendingMore than Rmb5 trillionAbout 75% above the prior plan.
- Water-network spendingMore than Rmb6 trillion15th Five-Year Plan-related expenditure.
- Urban-pipe programAround Rmb5 trillion and 770,000 kmNewly built or rebuilt urban pipes.
- Post-rate-cut six-month median returnsWind Micro Cap +9.5%; CSI 1000 +5.0%; CSI 300 Value -2.5%Historical comparison cited for the monetary-only easing scenario.
- End-2026 index targetsMXCN 100; CSI 300 5,200Supported by consensus 2026 year-on-year EPS growth projections of 14% and 25%, respectively.
Impact & implications
The report’s central implication is that improved fiscal execution should favor AI computing, telecom, grid and other Six-Networks beneficiaries over the next two to three quarters. A larger property-and-consumption package would shift leadership toward high-beta consumer and property chains, whereas weak fiscal conversion would favor defensive value, dividends and cash flow; monetary-only easing would tilt market leadership toward smaller, growth-oriented shares.
Risks
- Fiscal financing may not convert into spending because of weak land revenue, limited matching funds, local-government caution and project-readiness constraints.
- Near-term market conditions may be challenged by tighter liquidity, a worsening Asia QMI cycle indicator, US mid-term-election risk aversion, weak seasonality and limited earnings catalysts before October 3Q results.
- AI-theme performance could be volatile if liquidity turbulence persists.
What to watch
- The 5th Plenum in October, the Central Economic Work Conference in early to mid-December, and the Two Sessions in early March 2027.
- Evidence of faster bond issuance, fiscal-deposit drawdown and policy-bank funding reaching projects.
- Whether policymakers provide incremental central or quasi-fiscal funding, housing-completion support and consumer measures.
- Export-growth deceleration toward mid-single digits, identified by investors as a possible condition for stronger broad stimulus.
- October 3Q earnings results and liquidity conditions.