China equity strategy under four policy-stimulus scenarios: JPMorgan maps four China stimulus paths, retaining an Overweight equity stance while favoring policy-linked infrastructure and domestic AI exposure.

Overview

JPMorgan examines four potential Chinese policy paths and their equity-market consequences. Its central case is a budget-constrained fiscal catch-up that supports infrastructure and AI-related supply chains, while the firm remains Overweight China equities in its EM/Asia allocation.

Core views

The report frames China equity strategy around four policy outcomes. Its base case is not a new large stimulus package, but a measured catch-up in fiscal execution within the approved budget. Faster government-bond issuance, fiscal-deposit drawdown and policy-bank funding should support infrastructure, urban renewal, project completion and strategic investment. However, weak land revenue, local-government risk aversion and limited project readiness constrain the impulse, so it is supportive rather than expansionary. If spending follows the “Six Networks” framework, JPMorgan expects policy to move through order backlogs and revenue recognition into earnings over the next two to three quarters. The base-case investment logic centers on computing, grid, telecom, water, underground-pipeline and logistics networks. China data-center capex is estimated to roughly double from US$55–69bn in 2026 to US$95–151bn by 2029, with about Rmb4trn of computing-network spending over the 15th Five-Year Plan. The new-type grid adds more than Rmb5trn, around 75% above the prior plan; water investment exceeds Rmb6trn; urban-pipe spending is about Rmb5trn across 770,000 km; and telecom has Rmb3.8trn of direct investment, expected to pull through Rmb7trn of upstream and downstream output. These programs support AI servers, optical modules, liquid cooling, IDC operators, domestic AI chips, UHV equipment, distribution automation, smart meters, storage inverters, engineering contractors, pipe makers and logistics operators. Investor feedback cited in the report was most positive on AIDC supply-chain fundamentals. An upside scenario requires genuinely incremental central or quasi-fiscal funding, central risk-sharing, faster project conversion, stronger housing-completion support and consumer measures. JPMorgan compares this case with September 2024, when appliances, autos, property developers and brokers, second-tier baijiu and selected building-material companies led within one month, while export-oriented consumer firms and defensive high-yield stocks lagged. The report attributes the leadership partly to higher upside beta among outperformers. It characterizes such a repeat as a low-probability upside risk and notes that investors see export-growth deceleration to mid-single digits as a possible precondition. A monetary-only easing path would arise if fiscal policy remains restrained while policy rates and interbank rates fall and liquidity becomes looser. Historical back-testing indicates micro and small caps lead over three months and growth modestly beats value. Cyclicals may respond first, but healthcare, IT and materials lead over three months, while utilities, staples and financials underperform. At six months following rate cuts, the median return was +9.5% for Wind Micro Cap and +5.0% for CSI1000, versus -2.5% for CSI300 Value. The downside case is persistent fiscal under-delivery: financing reaches projects slowly, weak land revenue and insufficient matching funds impede local implementation, fiscal deposits remain high and part of the intended impulse is delayed. Combined with weak liquidity, this produces a flight-to-safety regime focused on P/B, balance-sheet strength, cash flow and dividends. JPMorgan’s defensive screen requires MXCN membership, three-month ADT above US$30m, trailing P/B below 1.5x, positive free-cash-flow growth in the latest fiscal half-year, dividend yield above 4%, and positive 2026 and 2027 consensus EPS growth; it is concentrated in insurers, selected banks, energy and consumer staples. JPMorgan maintains an Overweight China-equity allocation and end-2026 targets of 100 for MXCN and 5,200 for CSI-300, supported by consensus EPS growth forecasts of 14% and 25%, respectively. It sees AI as a tier-1 growth theme after 2Q26 outperformance and expects the domestic AI ecosystem to rebound through the rest of 2H26, albeit with a choppy path if liquidity is turbulent. Near-term caution reflects potentially tighter monetary policy and liquidity, a worsening Asia QMI indicator, US mid-term-election risk aversion, a lack of fresh earnings catalysts before October 3Q results, and poor seasonality. The next six months’ stimulus debate is expected to focus on the October 5th Plenum, the early-to-mid-December Central Economic Work Conference and the Two Sessions in early March 2027.

Analysis framework

JPMorgan first defines fiscal, monetary and downside policy assumptions, then maps each assumption to historical market behavior, sector leadership and stock screens. It combines policy-capex estimates, historical post-stimulus and post-rate-cut returns, valuation and cash-flow filters, and company-level exposure to identify scenario-sensitive equities.

Methodology notes

• Industry AnalysisUpstream-Midstream-Downstream Transmission

  Policy-funded Six Networks investment is traced into order backlogs, revenue and earnings for relevant supply chains.

  The report links planned spending in computing, power grids, telecom, water, pipelines and logistics to suppliers and contractors expected to benefit over subsequent quarters.

• Quantitative, Factor, and Portfolio TheoryStyle factor analysis

  Historical back-testing of market-cap and growth-versus-value performance after policy-rate cuts.

  The analysis compares index and sector returns following rate cuts to identify the styles and sectors that historically led under monetary-only easing.

• Valuation methodsPB valuation

  A defensive equity screen using trailing P/B below 1.5x alongside free cash flow, dividend yield and earnings-growth conditions.

  The screen identifies stocks the report considers suited to a flight-to-safety environment if fiscal execution disappoints.

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

  Exposure to the domestic AI ecosystem.

  Risks

  Liquidity turbulence could make the AI rebound choppy.

• V-Test (688372.SH)

  Favored domestic AI supply-chain player.

  Strengths

  Exposure to domestic AI supply-chain growth.

  Risks

  Liquidity turbulence could make the AI rebound choppy.

• JECT (600584.SS)

  Favored domestic AI supply-chain player.

  Strengths

  Exposure to domestic AI supply-chain growth.

  Risks

  Liquidity turbulence could make the AI rebound choppy.

• AMEC (688012.SS)

  Favored domestic AI supply-chain player.

  Strengths

  Exposure to domestic AI supply-chain growth.

  Risks

  Liquidity turbulence could make the AI rebound choppy.

• NAURA (002371.SZ)

  Favored domestic AI supply-chain player.

  Strengths

  Exposure to domestic AI supply-chain growth.

  Risks

  Liquidity turbulence could make the AI rebound choppy.

• Meituan (3690.HK)

  Non-AI top pick for diversification.

• Bank of China-H (3988.HK)

  Non-AI top pick and a constituent of the defensive screen.

  Strengths

  Meets the report’s defensive-screen characteristics.

• Bank of Ningbo (002142.SZ)

  Non-AI top pick for diversification.

• CICC-H (3908.HK)

  Non-AI top pick for diversification.

• Innovent (1801.HK)

  Non-AI top pick for diversification.

• BYD-H (1211.HK)

  Non-AI top pick for diversification.

• CR Land (1109.HK)

  Non-AI top pick for diversification.

Key data

• China data-center capexUS$55–69bn in 2026; US$95–151bn in 2029JPMorgan estimate; roughly doubles by 2029.
• Grid investment>Rmb5trn15th Five-Year Plan spending, about 75% above the prior plan.
• Water and urban-pipe investment>Rmb6trn; ~Rmb5trn across 770,000 kmPlanned water-network and urban underground-pipeline investment.
• Telecom investmentRmb3.8trn direct investment; Rmb7trn output pull-throughExpected upstream and downstream economic effect.
• Post-rate-cut six-month median returnsWind Micro Cap +9.5%; CSI1000 +5.0%; CSI300 Value -2.5%Historical comparison cited for monetary-only easing.
• End-2026 index targetsMXCN 100; CSI-300 5,200Maintained base-case targets, underpinned by 14% and 25% consensus year-on-year EPS growth, respectively.

Impact & implications

The report argues that the policy mechanism matters more than headline stimulus size. Budget-funded network investment favors infrastructure and domestic AI supply chains over coming quarters; broad incremental fiscal support favors high-beta property and consumption; monetary easing favors smaller growth companies; and weak fiscal delivery favors cash-generative, low-valuation dividend payers.

Risks

• Fiscal financing may fail to convert promptly into project spending because of weak land revenue, insufficient matching funds, cautious local implementation and limited project readiness.
• The upside stimulus scenario requires genuine incremental funding, central risk-sharing and rapid spending conversion rather than a larger headline package alone.
• Near-term China-equity conditions may be pressured by tighter liquidity, a worsening Asia QMI indicator, US mid-term-election risk aversion, weak seasonality and limited earnings catalysts before October 3Q results.
• Liquidity turbulence could make an AI-led rebound uneven.

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.
• Government-bond issuance, fiscal-deposit drawdown, policy-bank funding and actual project execution.
• Whether export growth decelerates to mid-single digits, a condition investors flagged for broader stimulus.
• Progress in Six Networks investment and evidence that backlogs are converting into revenue and earnings.
• Liquidity conditions and October 3Q earnings results, particularly for the AI ecosystem.