JPMorgan sees China ecommerce shifting from subsidy-led GMV growth to margin-led earnings

Overview

JPMorgan examines how China’s new platform-price rules, fading shopping-festival benefits and a smaller trade-in subsidy program are changing ecommerce competition. Its central view is that slower GMV growth can be a profit positive if regulation lowers industry spending; Alibaba and JD are viewed as the clearest beneficiaries, while PDD faces domestic rule constraints and uncertain Temu outcomes.

Core views

For 15 years, China ecommerce growth was driven by offline share gains, shopping festivals and subsidies. JPMorgan argues that all three engines weakened between April 2025 and June 2026: traditional-platform share loss slowed to roughly 1ppt annually from 3-5ppts in 2021-23, comparable 618 GMV growth fell to about 1% despite a 37-day festival period, and state trade-in support was reduced and redesigned. The report interprets this not simply as a growth problem, but as the end of uneconomic demand creation in a mature market where subsidies mostly move demand among platforms, pull purchases forward or encourage coupon arbitrage. The report’s core thesis is that the Internet Platform Price Conduct Rules can provide external enforcement for lower competitive spending. Issued on 20 December 2025 and effective from 10 April 2026, the rules prohibit merchant coercion into markdowns or participation, require subsidy transparency, and prohibit forcing below-cost selling outside legitimate clearance. A 25 May 2026 regulatory meeting involving nine agencies and 17 platforms reinforced the policy before the 618 event. JPMorgan expects lower acquisition costs, merchant concessions and fulfillment spending to outweigh the near-term revenue headwind, shifting competition toward service, merchant tools and supply-chain efficiency where scaled incumbents have advantages. The supporting condition is slowing disruption from short-video commerce. JPMorgan estimates that Taobao/Tmall, JD, PDD and Vipshop will represent roughly 71% of China online physical-goods GMV in 2026, down only about 1ppt year on year. Douyin remains a share gainer, but its annual gains have slowed to roughly 1ppt from 4-5ppts in 2021-23. The report notes that more than 60% of Douyin GMV is below RMB120 average selling price; moving into higher-value categories increases the importance of logistics, returns, trust and after-sales service. Taobao/Tmall is still the main source of historic share loss, falling from 67% in 2017 to an estimated 33% in 2026, while JD has held near 17% since 2023 and PDD near 20%. Shopping festivals are no longer viewed as a major incremental-demand engine. Syntun data show 618 GMV rising from RMB578.5bn in 2021 to RMB855.6bn in 2025 while the event expanded from 19 days in 2023 to 37 days in 2025; the report calculates 95% more days for only 7.1% more GMV. In 2026, Syntun measured RMB863.6bn of platform GMV from 13 May to 18 June, or roughly 1% comparable growth, with quick commerce separately disclosed at RMB62.8bn. Longer promotion windows, everyday low prices, weaker merchant ROI and daily content-commerce discovery have reduced urgency. This saves promotion costs but also weakens Alibaba’s historically high-margin festival advertising density; JPMorgan’s base case is that year-round fees, AI advertising tools, memberships and traffic monetization more than offset the loss. The report also separates policy-driven demand from underlying operating trends. China’s 2025 trade-in program used RMB300bn of ultra-long special treasury bonds, generated over 360mn subsidy claims and more than RMB2.6tn in related goods sales, and contributed about 0.6ppt to total retail-sales growth according to NDRC. The 2026 program falls to RMB250bn, narrows appliance coverage and lowers the appliance cap to RMB1,500. At the 2025 realized multiplier, the RMB50bn funding cut alone implies roughly RMB430bn less sales. JPMorgan’s base case is a RMB725-850bn year-on-year decline in subsidy-driven retail sales, including RMB400-550bn online, producing a roughly 1.5-2ppt swing in online-retail growth. JD faces the hardest comparisons in 2Q26 and 4Q26 because it captured the most trade-in demand; the report argues this can mask better underlying margins and share. Quick commerce is the major investment overhang and the report estimates Alibaba, Meituan and JD consumed roughly RMB180-200bn of profit from 2Q25 through 1Q26. JPMorgan ranks Alibaba first, Meituan second and JD third in observable returns. Alibaba’s China ecommerce adjusted EBITA fell from RMB193.2bn in FY25 to RMB107.5bn in FY26, with management attributing most of the RMB85.7bn decline to quick-commerce investment; Taobao DAU rose from roughly 375mn in April 2025 to 437mn in September 2025, and Double 11 generated more than 100mn orders from new Shangou users. Its return now depends on conversion into CMR, software fees and 88VIP economics. Meituan’s group adjusted earnings swung from a RMB35.8bn profit in 2024 to a RMB23.4bn loss in 2025, but it retained more than two-thirds of orders above RMB15 and over 70% above RMB30. JD’s new-business operating loss narrowed from RMB14.8bn in 2Q25 to RMB9.9bn in 2Q26, though JPMorgan estimates full-cycle losses above RMB40bn; its network spans 350 cities and could extend its 1P service proposition into higher-value instant delivery, but the earnings payoff remains unproven. JPMorgan assesses quick-commerce returns through cost per share point, cost per incremental engaged user and monetization conversion. Its retention scenarios are 25% bear, 45% base and 60% bull after 12 months, with two to three clean quarters expected to reveal whether traffic persists. For Alibaba, the report expects its 0.6% software service fee, Quanzhantui AI advertising tool and more than 50mn 88VIP members to help offset weaker festival bidding. Monthly active consumers were up 25% year on year after the quick-commerce campaign; the March 2025 quarter delivered 12% CMR growth and FY26 like-for-like CMR grew 8%. The FY27E base case assumes 45% Quanzhantui penetration, 45% cohort retention and a 10% decline in festival ad-bidding intensity, supporting high-single-digit like-for-like CMR growth. For JD, the report argues that restrictions on coordinated discounting and automatic price matching should allow delivery, installation, after-sales service and trade-in capabilities to command greater value. Its more than 1,500 warehouses and direct procurement support high-ticket categories, while quick commerce provides optionality in appliances, electronics and supermarkets. For PDD, everyday-low-price positioning reduces festival dependence, but its domestic model is more directly exposed to restrictions on automatic price tracking, mandatory merchant participation and “ten-billion subsidy” practices. Core advertising grew roughly 5% year on year in 4Q25 and low-single digits in 1H26, while total revenue rose 8% in 2Q26. Temu adds material uncertainty: the loss of the US de minimis exemption, potential European and Latin American restrictions, and New Pinmu supply-chain investment—initially RMB15bn and potentially rising toward RMB100bn over three years—could constrain near-term margins. AI is viewed as a near-term efficiency tool and a longer-term interface risk. Alibaba’s Quanzhantui, JD’s LLM-supported ad creation, and automated warehousing can improve merchant advertising ROI and logistics costs. Yet Alibaba’s connection of Qwen to more than 4bn Taobao and Tmall items, roughly 300mn monthly active users across its consumer surfaces, and about 130mn first-time AI-shopping experiences raises a longer-term question: agent-led search could erode paid-ranking economics if users demand unbiased recommendations and agent-native advertising does not emerge. JPMorgan expects limited earnings impact from this risk in 2026-27. The report’s key falsification test is whether aggregate sales and marketing expense across BABA, JD, PDD and Meituan falls year on year in the two quarters ending December 2026. If spending does not decline, JPMorgan would conclude that regulations are not changing competitive behavior and that upward earnings revisions are at risk. It therefore prefers BABA and JD over the next four quarters, while retaining a Neutral view on PDD because domestic monetization is slowing and Temu’s regulatory and profitability outcomes remain wide.

Analysis framework

JPMorgan first compares the sector’s former growth engines with current market conditions, then links the Price Conduct Rules to a potential decline in competitive spending and margin recovery. It tests that thesis through market-share trends, festival GMV and duration, trade-in subsidy comparisons, quick-commerce profit costs and company-specific monetization paths. It uses retention scenarios, cost-per-acquisition comparisons, P&L impacts and valuation multiples to frame company outcomes.

Methodology notes

• Industry AnalysisSupply-demand framework

  Assessment of whether subsidies create incremental ecommerce demand or merely transfer and pull forward demand.

  The report argues that in a mature-penetration market, subsidy spending grows faster than the demand pool, making lower promotional spending potentially positive for industry margins.

• Industry AnalysisVolume-price decomposition

  Separating reported growth affected by subsidies and festival timing from underlying share, traffic and margin trends.

  JPMorgan distinguishes policy-driven category demand from organic consumption and evaluates GMV, traffic retention, monetization and margins separately.

• Event-Driven and Behavioral FinanceEvent-driven analysis

  Analysis of the Price Conduct Rules and regulatory enforcement as catalysts changing platform behavior.

  The report treats the rules and their enforcement as the mechanism that could stop destructive spending and trigger earnings-estimate revisions.

• Other

  Quick-commerce ROI scorecard using cost per share point, cost per incremental engaged user and monetization conversion.

  The report compares campaign profit losses with market-share defense, incremental users and evidence that traffic converts into sustainable revenue or positive unit economics.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Alibaba Group Holding Limited (9988.HK / BABA)

  Preferred beneficiary of lower competitive spending and traffic monetization.

  Strengths

  0.6% software fee, Quanzhantui AI tools, 88VIP, and quick-commerce traffic growth support CMR.

  Weaknesses

  Festival-driven advertising density is exposed to weaker event demand.

  Comparison

  JPMorgan ranks Alibaba first for observable quick-commerce monetization traction.

  Risks

  Subsidy-acquired users may fail to convert into CMR.

• JD.com, Inc. (9618.HK / JD)

  Preferred beneficiary of reduced 1P discounting and renewed service and logistics pricing power.

  Strengths

  More than 1,500 warehouses, direct procurement and high-ticket-category service capabilities.

  Weaknesses

  Trade-in subsidies create difficult 2Q26 and 4Q26 comparisons; quick-commerce returns remain unproven.

  Comparison

  JPMorgan ranks JD third for current measurability of quick-commerce returns.

  Risks

  Underlying margin improvement may not offset policy-driven revenue pressure.

• PDD Holdings (PDD)

  Domestic rules constrain key promotional mechanics while Temu provides uncertain international optionality.

  Strengths

  Operating efficiency, global growth potential and lower exposure to festival demand.

  Weaknesses

  Slowing domestic monetization, limited disclosure and New Pinmu investment pressure.

  Comparison

  Less exposed than JD and Tmall to 2025 trade-in demand, but more directly constrained by price-conduct rules.

  Risks

  EU or Latin American regulation could impair Temu; localized fulfillment may reduce economics.

• Vipshop (VIPS)

  Stable niche participant less reliant on major shopping festivals.

  Strengths

  Shareholder-return commitment exceeding 75% of prior-year earnings and a resilient core business.

  Weaknesses

  Weaker revenue-growth outlook than other pure ecommerce peers.

  Comparison

  Less dependent on 618 and Double 11 than larger platforms.

  Risks

  Revenue stabilization is needed for upside optionality.

Key data

• Traditional-platform share~71% of China online physical-goods GMV in 2026EAbout 1ppt year-on-year decline versus annual losses of 3-5ppts in 2021-23.
• Comparable 618 GMV growth~1%Measured in 2026 despite an extended promotional period.
• Quick-commerce industry investmentRMB180-200bnEstimated profit consumed by Alibaba, Meituan and JD from 2Q25 through 1Q26.
• 2026 trade-in fundingRMB250bnDown from RMB300bn in 2025; JPMorgan estimates RMB725-850bn less subsidy-driven retail sales year on year.
• Alibaba FY26 China ecommerce adjusted EBITARMB107.5bnDown from RMB193.2bn in FY25; most of the RMB85.7bn decline was attributed to quick-commerce investment.
• Alibaba CMR growth12% in the Mar-2025 quarter; 8% FY26 like-for-likeDriven by take rate, software fees and higher Quanzhantui penetration.
• JD new-business operating lossRMB9.9bn in 2Q26Narrowed from RMB14.8bn in 2Q25.
• PDD total revenue growth8% in 2Q26Down sharply from more than 50% growth in 2022-23.

Impact & implications

The report argues that slower sector GMV need not imply weaker earnings if regulation reduces subsidy, merchant-concession and fulfillment spending. Alibaba is seen as best placed to monetize larger traffic through CMR and year-round tools, while JD may regain value for its service and logistics advantages despite difficult subsidy comparisons. PDD’s domestic restrictions and Temu uncertainty lead JPMorgan to retain a Neutral stance.

Risks

• Regulatory enforcement may be weak, allowing spending to be reclassified rather than reduced.
• Weaker consumer confidence could pressure discretionary categories regardless of improved industry discipline.
• Douyin could sustain rapid growth in shelf and high-ASP categories, extending incumbent share losses.
• AI agent interfaces could erode paid-ranking economics faster than expected.
• EU or Latin American action could materially reduce Temu’s international growth.
• Broader platform oversight could expand or change direction without notice.
• Tariffs, US-China tensions and ADR-related uncertainty could affect covered companies.

What to watch

• September-November 2026 results for post-subsidy order volumes and retention at Alibaba, Meituan and JD against the 45% base case.
• Double 11 duration, subsidy language, merchant terms and Alibaba’s December-quarter CMR seasonality.
• Whether PDD domestic online-marketing growth exceeds 5% for two consecutive quarters with stable adjusted operating margin.
• EU or Latin American regulatory actions affecting Temu.
• Monthly NBS appliance and communications-equipment sales versus subsidy disbursements to test JD’s underlying demand and margin thesis.
• Douyin’s high-ASP category mix and shelf GMV penetration.
• AI-agent-originated orders and the development of agent-native advertising.
• Final delivery-subsidy rules and initial penalties under the Price Conduct Rules in 2H26.
• Aggregate sales and marketing expense across BABA, JD, PDD and Meituan through December 2026.