JD.com, Inc. (NASDAQ: JD) — China’s Best-Run Retailer, Priced at Cash-and-Stakes While the War It Started Already Peaks

Report date: 2026-06-27 · Price: $25.39/ADS (2026-06-26) Currency note: JD reports in RMB (CNY); financial statements below are RMB unless marked. Valuation is anchored in USD per ADS (1 ADS = 2 Class A ordinary shares); FX ~¥7.2/US$.

⚡ Claude’s Take

This block is the author’s own independent opinion and general information, not investment advice. The analysis that follows is deliberately position-free and carries no recommendation or price target — only this block does.

Verdict: BUY / accumulate on weakness — but size it as a China-macro-beta position, not a quality-compounder. Conviction: MEDIUM. Directional value zone: an asset-backed floor near $18/ADS (net cash + listed stakes alone), a base-case fair-value band of ~$35–45/ADS on normalized post-war earnings, and a bull path to $55–70 if the ceasefire holds and the China discount partly unwinds. Accumulate in the low-to-mid $20s; don’t chase above ~$40.

The framing is deep-value / contrarian-China with a real asset floor and a freshly-inflecting operating trajectory — not momentum, and not a clean quant-value screen (the factor model shows JD as an abandoned, negative-momentum, high-China-beta proxy at its relative-strength nadir, which is exactly why the mispricing exists). Here is what the market is getting wrong: it has thrown JD to its cheapest valuation in its public history (AZI composite 4.2nd percentile; P/E ~10x; P/B ~1.16x) because in February 2025 founder Richard Liu marched the company into a food-delivery/instant-retail war that collapsed group operating income ~91% (¥39.6B → ¥3.7B) and burned ¥46.6B in the “New Businesses” line. But that loss is one discrete, self-inflicted line item layered on a healthy, compounding core — JD Retail grew operating income +25% and expanded margin to a record 4.6% (Q1’26: 5.6%) straight through the war. Strip out the ~$8.9B net cash and the ~$16.6B of marketable stakes in HK-listed JD Logistics (2618.HK) and JD Health (6618.HK) — together ~73% of the $34.8B market cap — and the market is paying roughly 1.3x EBIT for the #1 most-trusted 1P retailer in China. The war loss peaked in Q4’25 (group OI back to +¥3.8B in Q1’26), and the Chinese state is now forcing a subsidy ceasefire (“anti-involution”) — the rare instance where Beijing’s hand is a tailwind, not a hammer.

What keeps conviction at MEDIUM, not high: this is genuinely a falling knife (5-year return −16.8%/yr; trades on the KWEB correlation, not its own fundamentals), the capital is being recycled into a widening set of bets (food delivery + a ~€2.2B CECONOMY/MediaMarkt European buyout + a ~$4.5B Hong Kong build) under a founder with ~73% voting control and no board check, and no insider has bought a share on the open market at this price. The asset floor is only worth what management is willing to return. Single bullish trigger: New-Businesses loss narrows toward <¥5B/quarter while the buyback/dividend accelerates or a stake is monetized. Single bearish trigger: the ceasefire breaks (or migrates to an M&A consolidation war JD is sub-scale to fight) and the loss re-widens, or CECONOMY/HK prove to be open-ended cash sinks. Tag: “You get China’s best retailer for its cash and its stakes; the war it started is the discount, and the war just peaked.”

📈 Stock Price Action — Five-Year Event Map

Factual price history, not a recommendation. Price moves are FACT; attributed drivers are INTERPRETATION. No price target, no support/resistance.

JD’s ADS round-tripped an entire mania: from a split/dividend-adjusted all-time high of $93.92 (17-Feb-2021) to $25.39 (26-Jun-2026) — −73% off the high, sitting just above its 52-week low ($24.54) and its cheapest-ever valuation. The 52-week range is $24.54–$34.95; the lifetime low is $16.93 (Nov-2018). On the factor model this is dead-to-negative money — a −16.8%/yr 5-year return and a −79% 10-year max drawdown. The arc below is the price context the rest of this memo assumes.

Cycle narrative. (1) The Feb-2021 $93.92 ATH was the apex of the post-COVID China-internet mania — never revisited. (2) The 2021 platform crackdown de-rated the entire complex on multiple compression, not JD fundamentals. (3) 2022 added COVID-Zero and acute HFCAA delisting fear before a year-end PCAOB-access relief bounce. (4) 2023 was the worst fundamental year — Liu’s personally-driven low-price strategy and a Retail restructuring amid a weak consumer and share loss to PDD/Douyin. (5) The Sep–Oct 2024 stimulus rally (easing + trade-in subsidies booming JD’s electronics category) was JD’s last major up-move. (6) Feb-2025 launched the food-delivery war that collapsed group OI and round-tripped the 2024 rally. (7) 2026 YTD is the trough-and-recovery tension: Q4’25 peak loss, Q1’26 recovery, but a Daiwa downgrade and Beijing’s 618 warnings left JD at its cheapest-ever multiple near its 52-week low.

1. Executive Summary

JD.com is China’s largest retailer by revenue (FY2025 net revenue ¥1,309.1B / ~$184B, +13% YoY; >700M annual active customers; ~571,000 employees) and a fundamentally different animal from its marketplace peers: it is a first-party (1P) direct retailer that buys, holds, and resells inventory through an owned nationwide logistics network (>3,600 warehouses, >34M sq m), at a structurally low ~16% gross margin. Alibaba and PDD are asset-light 3P marketplaces booking high-margin commissions; JD books gross merchandise dollars. That single structural fact explains why JD’s 0.24x price/sales is partly an optical artifact, not pure value.

The investable tension is sharp and unusually clean. In February 2025, JD launched into China’s food-delivery / instant-retail war against incumbent Meituan and a counter-escalating Alibaba. The resulting subsidy war collapsed group operating income from ¥39.6B (2024) to ¥3.7B (2025), −91%, via a ¥46.6B operating loss in the “New Businesses” segment. Crucially, this is one discrete line: the core (JD Retail + JD Logistics) grew operating income ~20% to ¥56.7B during the war, with JD Retail expanding margin to a record 4.6% (its sixth straight year of expansion) and advertising growing double-digits every quarter. The market, however, responded by marking JD to its cheapest valuation ever (AZI composite 4.2nd percentile; P/E ~10x; P/B ~1.16x).

The valuation crux is a sum-of-the-parts: of the ~$34.8B market cap, ~$8.9B is net cash and ~$16.6B is JD’s economic share of its HK-listed stakes in JD Logistics (2618.HK) and JD Health (6618.HK) — together ~73% of the cap. The implied “stub” values the entire consolidated core at roughly 1.3x EBIT against JD Retail’s ¥51.4B operating income. The war loss peaked in Q4’25 (group OI recovered to +¥3.8B in Q1’26), and Chinese regulators are now actively forcing a subsidy ceasefire under an “anti-involution” campaign — the opposite posture from the punitive 2021 crackdown.

The bear case is equally real: JD bought only a distant-third ~10% order share (vs. Meituan ~55%, Alibaba ~35%) at the highest marginal cost — a textbook late-entrant capital-cycle trap that proved its logistics “moat” does not travel into on-demand delivery. The war chest is being recycled into a widening set of bets (food delivery + ~€2.2B CECONOMY/MediaMarkt + ~$4.5B Hong Kong) under a founder with ~73% voting control and no board check; the Chinese consumer is weak and deflationary; and JD trades as undifferentiated China macro beta (related stocks: KWEB, MCHI, FXI) with a −73% relative-strength reading.

The body that follows takes no position and sets no price target. It establishes: the core is a narrow-but-real, capital-heavy supply-side moat (logistics scale + authenticity reputation); the industry is structurally mixed-to-bad (mature, deflationary, recurrently self-immolating on price) with a state-imposed circuit-breaker as the one positive; the financials require segment-level and non-GAAP reading because GAAP net income is flattered by interest income on the cash pile; and the embedded expectations price JD as if the war is permanent, the consumer never recovers, and the asset backing is worth little — a set of assumptions where being wrong on even two re-rates the stock.

2. Business Overview

What JD does. JD.com is a “supply-chain-based technology and service provider” — in plain terms, China’s largest 1P online retailer plus a third-party logistics operator and a marketplace. It was founded by Richard Liu (Liu Qiangdong), IPO’d on NASDAQ in May 2014, and carries a primary listing on the Hong Kong Stock Exchange (9618). It is incorporated in the Cayman Islands and operates in China through a mix of wholly-foreign-owned enterprises and a small set of consolidated variable interest entities (VIEs). [FACT — 20-F FY2025; ROIC company profile]

The 1P-vs-3P identity (the single most important structural fact). JD is fundamentally a first-party retailer: it buys inventory, takes title and inventory risk, and recognizes the full gross sale as “product revenue.” This is the opposite of Alibaba (Taobao/Tmall) and PDD (Pinduoduo), which are asset-light 3P marketplaces that never touch inventory and book only commission/advertising “service revenue.” The consequences cascade through every number in this memo:

• JD’s reported revenue (~¥1.31T) is several times larger than Alibaba’s China-commerce revenue, but at a structurally low ~16% gross margin versus marketplace peers at 60–70%+.
• JD looks “cheap” on price/sales (0.24x) precisely because 1P revenue is gross merchandise dollars, not a high-margin take. P/S cheapness here is partly mechanical, not pure value. [INTERPRETATION]
• JD’s model is low-margin, high-revenue, high-asset-intensity (warehouses, ~571,000 employees including full-time warehouse and delivery staff), and working-capital-heavy (inventory). But it runs a negative cash-conversion cycle — supplier payables fund the inventory, a genuine structural advantage.

Revenue segmentation — two lenses. By revenue type: Product revenue (1P sales) splits into (i) Electronics & Home Appliances (the historic core; +7% FY2025 but −12% in Q4’25 as 2024’s government trade-in subsidy lapped) and (ii) General Merchandise (supermarket/grocery, fashion, health; +15.3% FY, five straight quarters of double-digit growth — the growth engine). Service revenue splits into (i) Marketplace & Marketing (3P commissions + the high-margin advertising business; +18.9% FY) and (ii) Logistics & other services (+27% FY). Service revenue grew +24% FY, far outpacing product revenue (+10%).

By operating segment (the lens this memo uses for profitability): JD Retail (the profit engine), JD Logistics (in-house fulfillment, separately HK-listed 2618), and New Businesses (food delivery, Jingxi lower-tier retail, international, Dada, JD Property — the loss bucket that absorbed the war).

How it makes money (the margin stack). Three profit pools, in order of margin quality: (1) advertising/marketing services — highest margin, fastest-growing, the structural profit story as “advertisers reallocate budgets to JD”; (2) 1P product gross margin — thin but improving via supply-chain scale and mix shift to general merchandise; (3) logistics/supply-chain services — modest margin, scale-dependent. The advertising overlay on the 1P base is the mechanism by which JD Retail has expanded operating margin for six consecutive years (2.7% in 2019 → 4.6% in 2025), with a stated long-term target of “high single digits.” [FACT — Q4’25 transcript, 2026-03-05]

Recurring vs. transactional. JD’s revenue is overwhelmingly transactional (per-order retail), not subscription. The closest things to recurring are JD PLUS membership (loyalty) and JDL’s enterprise supply-chain contracts. Management’s stated rationale for the entire food-delivery bet is to lift frequency — full-year shopping frequency rose +40% YoY in 2025, which management attributes partly to food delivery driving daily-use habit and cross-sell into core retail. [INTERPRETATION: the food-delivery war is best understood as a traffic/frequency-acquisition play to feed the 1P flywheel, not a standalone profit center.]

Verdict (Business Overview). JD is a capital-heavy, low-gross-margin 1P retailer with a genuinely improving high-margin advertising overlay and a maturing third-party logistics arm — a structurally lower-margin animal than asset-light marketplace peers, but with a high-quality, compounding core engine. The reported group P&L is currently disfigured by a deliberate, self-inflicted food-delivery investment; the headline “cheapness” on P/S is largely a 1P optical artifact, and the real analysis happens at the segment and SOTP level.

3. Industry Dynamics

A large, mature, slowing, deflationary market. China is the world’s largest e-commerce market, but online-retail growth has decelerated sharply post-COVID amid a weak consumer, property-driven balance-sheet caution, and outright consumer deflation — “anti-involution” (反内卷, anti-ruinous-competition) is now official policy language. JD’s own Electronics & Home Appliances category swung from a 2024/early-2025 boom (government trade-in/consumer-electronics subsidies) to −12% in Q4’25 as that stimulus lapped — a textbook policy-driven demand pull-forward now unwinding. Management expects E&HA to face a high base in H1’26 and recover in H2’26 as the trade-in program resumes, and flags a new headwind: memory-chip cost inflation pushing phone/digital prices up (could dampen volume, lift AOV). [FACT — Q4’25 transcript; CNBC 2025-07-11] The end-market is structurally low-growth and policy-whipsawed.

Competitive intensity — an oligopoly of well-capitalized giants, now in open war.

• Alibaba (Taobao/Tmall) — dominant 3P marketplace, asset-light, ROIC ~8% (2025); the most aggressive attacker in the 2025-26 instant-retail war via Ele.me and “Taobao Instant/Flash.”
• PDD/Pinduoduo — discount/social-commerce disruptor (Temu overseas); the highest-margin, most capital-efficient of the group; took massive share in price-sensitive China.
• Douyin/ByteDance — livestream/content commerce, structurally taking GMV share from shelf e-commerce.
• Meituan — the food-delivery/local-services incumbent (~55% order share), highest ROIC of the group (~18% pre-war).

The central story — the 2025-26 instant-retail / food-delivery war. JD launched into food delivery (JD Takeaway / JD 秒送 / “Seven Fresh Kitchens”) in February 2025, attacking Meituan’s turf; Alibaba counter-escalated via Ele.me/Taobao Instant. The result was a mutually destructive subsidy war (“coffee at 30 cents”). The quantified carnage is the heart of the thesis:

• Combined giants spent >¥100B (~$14B) on consumer discounts in 2025.
• JD’s New-Businesses unit lost ~¥46.6B (~$6.9B) in the war.
• Alibaba’s instant-retail losses (Q2’25–Q1’26) ran ~¥87B (~$12.9B).
• Meituan — the incumbent — swung to a ¥23.4B (~$3.4B) net loss for 2025, dragged into the red defending its own market.
• Market-share outcome (order volume, 2026): Meituan ~55%, Alibaba ~35%, JD ~10%. [FACT — CNBC 2025-07-11; Caixin 2026-03-27; BigGo Finance; greatwallstreet.net]

JD spent billions to buy a distant-third ~10% position in a business where it has no incumbency — the classic late-entrant capital-cycle trap (see Marathon read below).

The regulator-forced ceasefire (the swing factor). In July 2025, China’s SAMR demanded the giants end the subsidy war; on 17-Jun-2026 it released draft “Ten Rules for Regulating Food Delivery Platform Subsidies” — banning long-term, large-scale subsidies and coercive merchant/rider cost-sharing, and requiring 7-day advance promotion notice. Meituan, JD, and Taobao Flash all pledged support within an hour. Separately, on 11-Jun-2026 Beijing’s municipal regulator summoned the top-five platforms over unsubstantiated “¥10B subsidy” 618-festival claims. Critically, this is anti-involution / pro-rationalization regulation that protects margins — the opposite of the punitive 2021 platform crackdown (Alibaba ¥18B fine, “common prosperity”). CNBC’s 2026-06-23 framing: this round causes “less worry than 2021.” [FACT — SCMP 3357458; Caixin 2026-06-11; CNBC 2026-06-23]

Marathon capital-cycle read. Local instant-retail showed attractive incumbent returns (Meituan ~18% ROIC) → this drew in capital (JD and Alibaba both entered/escalated, each earmarking ~¥10B+). A classic supply-side surge: three deep-pocketed giants flooding capacity and subsidy into one pool simultaneously → returns collapsed for everyone, even the incumbent. The capital cycle is at peak capital-destruction; the question is whether the regulator-forced subsidy ban engineers an orderly de-escalation, or whether competition simply migrates to M&A consolidation (Alibaba’s ~$1.5B bid for Pupu; Meituan’s ~$717M Dingdong acquisition) — a ceasefire on price that is not a ceasefire on the war. [INTERPRETATION / Marathon framework]

Structural / regulatory risks. VIE/ADR delisting risk (HFCAA): JD was identified by the SEC under HFCAA on 04-May-2022; mitigated by its dual-primary HK listing (delisting from NASDAQ would not destroy the equity, as HK shares are fungible). Data-security and antitrust regimes remain live. [FACT — Paul Weiss; White & Case; 20-F]

Verdict (Industry). Structurally mixed-to-bad. Chinese e-commerce is large but mature, slow-growing, deflationary, and oligopolistic, where four-plus giants of comparable resources compete on price with weak product differentiation — the hallmark of an unattractive industry where no firm has durable pricing power and rivalry repeatedly erupts into value-destroying subsidy wars. The 2025-26 war is the proof: the industry’s own structure permitted ~$14B+ of mutual capital destruction in a single year. The one structural positive is that the Chinese state actively suppresses ruinous competition — a regulatory hand that can enforce the rationality the industry can’t self-impose. Net: a bad competitive structure with a state-imposed circuit-breaker. Not an industry that, by itself, confers durable economics.

4. Competitive Position

The Greenwald moat-type call. JD’s only candidate competitive advantage is a supply-side / economies-of-scale advantage in logistics & supply chain (Greenwald taxonomy: cost advantage from scale plus a degree of proprietary capability), reinforced by a reputation/intangible advantage for authentic goods and reliable service — the “no counterfeits, fast and reliable delivery, real warranty” brand that is genuinely differentiated versus the counterfeit-prone marketplaces of Taobao and PDD. It is not a network-effects moat and not a consumer switching-cost moat.

The moat’s evidence. JD’s owned nationwide fulfillment network (>3,600 warehouses, >34M sq m of floor area, full-time riders, same/next-day “211” delivery, >20 automated “LangzuTech” warehouses, >130 international warehouses across 23 countries) is a real, hard-to-replicate physical asset built over ~15 years. It underpins (a) the authentic-goods reputation, (b) industry-leading delivery speed/reliability, and © JDL’s success selling logistics to external customers, now ~72% of JDL revenue — a market test that the capability has standalone value. The strongest single quantitative signal is JD Retail’s six consecutive years of operating-margin expansion (2.7% → 4.6%) while the industry fought price wars — a real, if narrow, cost/scale advantage in the 1P core. [FACT — Q4’25 transcript; ir.jd.com Q2’25]

Pressure test — why the moat is narrow and capital-heavy, not wide.

1. No consumer switching costs. Chinese shoppers multi-home aggressively across JD/Taobao/PDD/Douyin; loyalty is price- and convenience-driven. JD Plus helps at the margin but is not a lock-in — the antithesis of a captive-customer advantage.
2. Weak network effects vs. marketplace peers. JD’s two-sided 3P marketplace is structurally weaker than Alibaba’s and PDD’s; JD is supply-chain-led, not flywheel-led, with smaller merchant/buyer network density. Network effects favor the incumbents JD is not.
3. The logistics “moat” is a capital sink as much as a moat. It required massive sustained capex and ~571,000 employees; JDL’s own operating income fell 17% in FY2025 and earns sub-peer returns. A moat that needs perpetual heavy reinvestment to merely maintain position is a weak moat.
4. The smoking gun. If the logistics moat were truly wide, entering an adjacent delivery business (food) should have been cheap and high-return — JD already owns riders and last-mile density. Instead, entering food delivery destroyed 91% of group operating income in one year and bought only ~10% order share. This proves the logistics advantage does not travel: last-mile parcel scale did not confer durable advantage in on-demand restaurant delivery, where Meituan’s dense local merchant + rider network is the real moat. JD had to buy its way in with subsidies — the definitional opposite of a moat. [FACT — Q4’25 transcript; market-share data]

Direct peer comparison.

JD has the lowest gross margin (1P structural), middle-of-the-pack ROIC, and the most capital-intensive model. Its differentiation (authenticity, speed) is real but narrow and does not confer pricing power — proven by its inability to avoid being dragged into, and losing money in, every price war.

Verdict (Competitive Position). JD has a narrow, capital-heavy supply-side moat (logistics scale) plus a genuine but limited authenticity/service-reputation intangible — enough to defend and modestly improve its 1P core retail economics (six years of margin expansion is real evidence), but not a wide or travelable moat. It is closer to “the best-run, most-trusted operator in a crowded, low-margin commodity-retail industry with a capital-intensive logistics arm” than to a structurally advantaged compounder. The food-delivery debacle is the definitive falsification of any “wide logistics moat” claim. Durable advantage in 1P retail: yes, but narrow and reinvestment-hungry. Wide, pricing-power moat: no.

5. Growth History and Forward Opportunities

Historical growth. Revenue compounded from ¥745.8B (2020) to ¥1,309.1B (2025) — a ~12% five-year CAGR, decelerating from the COVID-era surge to a +13% rate in 2025 (itself flattered by the food-delivery revenue and the trade-in boom). The growth has been overwhelmingly organic, driven by category expansion (electronics → general merchandise/supermarket) and the maturing of JD Logistics’ external business, supplemented by tuck-in consolidations (Dada, Deppon).

Quality of the historical growth — mixed. The high-quality growth is in (i) general merchandise (+15.3% FY2025, five straight double-digit quarters — a genuine share-gain and diversification story into supermarket/fashion/health) and (ii) advertising/marketplace (+18.9% FY, the highest-margin line, with “advertisers reallocating budgets toward JD”). The low-quality growth is the FY2025 New-Businesses revenue (+157%), which is subsidy-bought food-delivery volume sold at a ¥46.6B loss — revenue that exists only because JD is paying to acquire it. And the 2024/early-2025 E&HA boom was policy-driven (government trade-in subsidies), now reversing (−12% Q4’25).

Forward opportunities.

• JD Retail margin runway. The single most durable growth driver is margin, not volume: management reaffirms a “high single-digit” long-term operating-margin target vs. 4.6% today (5.6% in Q1’26). Each ~100bps of Retail margin on ~¥1.1T of segment revenue is ~¥11B of incremental operating income — a multi-year, self-help earnings engine independent of macro.
• Advertising monetization. Double-digit growth every quarter; the food-delivery traffic adds an incremental ~2–3% to ad revenue. The under-monetized 1P/3P ad inventory is real optionality.
• Frequency/cross-sell from food delivery — if it earns its keep: shopping frequency +40% YoY, QAC +30%, active merchants +270%, >700M annual active customers. The bull case is that subsidy-bought traffic becomes durable habit feeding the core flywheel.
• JD Logistics external growth (~72% external revenue) and international (Joybuy/JoyExpress in Europe, Ochama, first UK automated warehouse).
• CECONOMY/MediaMarkt — a ~€2.2B European omnichannel electronics platform (~85% control, close H1’26) — a large but thin-margin, integration-heavy adjacency (see /).

Verdict (Growth). Mixed quality, with a high-quality self-help core. The most reliable forward growth is JD Retail’s margin expansion toward the high-single-digit target plus advertising monetization — both genuine, high-incremental-margin, and largely macro-independent. The volume growth is policy-whipsawed (E&HA) and partly subsidy-bought (food delivery, of uncertain durability). This is not a high-secular-growth story; it is a margin-and-mix self-help story on a low-growth, deflationary top line.

6. Financial Quality

Margin structure. Group gross margin is ~16.0% (FY2025), thin because ~80%+ of revenue is 1P product sales where JD takes inventory risk. Retail gross margin rose +1.1pp YoY in both Q4 and FY2025 (supply-chain leverage + mix shift to high-margin advertising/marketplace). Group operating margin, however, collapsed to ~0.3% (FY2025) from 3.4% (FY2024) — entirely the war. The quarterly path makes the discrete, peaking nature of the loss unmistakable:

The cleanest single read on the war is marketing expense: ¥40.1B (2023) → ¥48.0B (2024) → ¥84.0B (2025) — a +¥36B YoY jump (mostly consumer subsidies + rider acquisition) that maps almost 1:1 to the New-Businesses loss expansion. [FACT — 20-F]

Quality of earnings — read segments and non-GAAP, not GAAP NI. FY2025 net income attributable to shareholders was ¥19,631M against group operating income of only ¥2.8B — i.e. reported NI is ~7x operating income. The gap is non-operating: large net interest income on the ¥213B cash/short-term-investment pile, plus equity-method/investment gains, plus a ~¥7.0B “extraordinary” component. (The historical distortion was 2021’s −¥2.29 GAAP EPS, driven by the one-time Tencent share-dividend distribution to shareholders despite positive operating income.) GAAP NI is therefore a poor gauge of operating health here — it is flattered by interest income and distorted by investment marks. The better through-cycle reads are segment operating income (above) and non-GAAP NI of ¥27,032M (FY2025) vs. ¥47,827M (FY2024), −43.5% — which still shows the war roughly halved earnings. SBC is small (¥4.7B, ~0.4% of revenue) and not a QoE concern. [FACT/INTERPRETATION — 20-F; ROIC]

The core is not impaired. Stripping the New-Businesses war loss, the core (JD Retail + JD Logistics) earned ¥56.7B operating income in 2025, up ~20% from ¥47.4B in 2024, while JD Retail expanded margin to a record 4.6%. The franchise grew profit ~20% while absorbing a ¥46.6B war loss. The headline −91% group OI collapse is a single discrete line layered on a healthy, compounding core.

Free cash flow — the war’s cash signature. Operating cash flow fell from ¥58.1B (2024) to ¥19.0B (2025), −67%; FCF fell from ~¥44B to ~¥6B (¥4.8B per ROIC on a stricter capex definition), −85%+. The CFO attributes the OCF collapse to (a) the war operating losses and (b) working-capital outflows tied to the government trade-in program (a receivables build now recovering two straight quarters). The ~¥40B OCF decline approximates the ¥44B war OI swing — the cash drag is the war, and the trade-in WC effect is partly reversible timing. [FACT — FY2025 results; Q4’25 transcript]

Working capital — a structural strength. JD runs a negative cash-conversion cycle (~−25 days): it collects from customers and turns inventory faster than it pays suppliers. Supplier-payables float is an interest-free funding source that scales with GMV — a genuine balance-sheet advantage of the 1P scale model (akin to Costco/Amazon retail float).

Balance sheet — a fortress. Cash & equivalents ¥137,488M + short-term investments ¥75,744M = ¥213.2B liquid; total debt ¥107,105M (ST ¥20,924M + LT ¥86,181M); net cash ~¥63.5B. Plus ¥103,818M of LT equity investments (non-consolidated stakes), and consolidated HK-listed subsidiaries (JDL 2618, JD Health 6618, JD Industrials 7618) whose minority economics drive the ¥68.7B minority-interest line. EBITDA/interest coverage was 4.8x even on war-depressed FY2025 EBITDA (16.7x in 2024). [FACT — 20-F / ROIC]

ROIC — distorted by the war and the balance sheet. ROIC.ai reports return-on-invested-capital of 8.78% (2024) → 0.83% (2025) — depressed by both the war and a balance sheet bloated with ~¥317B of non-operating assets (cash + stakes) against a ¥293.8B equity base. A clean operating ROIC (ex-cash, ex-investments) on the negative-working-capital core was comfortably >20–30% pre-war given ¥47.4B of core OI on a small net-operating-asset base. The economics clearly improve with scale in the core; the war, not the core economics, drives the 0.83% reported figure. [INTERPRETATION]

Verdict (Financial Quality). The core is a high-quality, scale-advantaged, negative-working-capital 1P retail + captive-logistics machine that grew operating profit ~20% and expanded margin to a record 4.6% in 2025 — economics clearly improve with scale. The reported FY2025 collapse (OI −91%, FCF −85%, ROIC 8.78% → 0.83%) is almost entirely a discrete ¥46.6B New-Businesses war loss layered on a healthy core, plus a balance sheet so cash- and investment-heavy that reported returns understate true core ROIC. Earnings quality requires using segment OI and non-GAAP NI, not GAAP NI. The question is not whether the core is good (it is) — it is whether management’s war spend earns its return.

7. Capital Allocation

Buybacks — genuinely good. Under a US$5.0B authorization (effective Sep-2024, running through Aug-2027), JD repurchased and cancelled 183.2M Class A ordinary shares (=91.6M ADSs) for US$3.0B (~¥21.4B) in 2025 — ~6.3% of shares outstanding — and continued buying ~1.6% of shares (US$631M) in Q1’26. Ordinary share count fell from 3,137.7M (2023) → 2,903.4M (2024) → 2,741.8M (2025), −12.6% over two years, all cancelled (real per-share accretion). This is counter-cyclical and well-timed: management bought heavily while the stock fell to a record-cheap 4.2nd-percentile valuation.

Dividends — initiated and growing. JD initiated regular dividends in 2024; the FY2025 cash dividend was US$1.00/ADS (~US$1.4B), with per-share dividends growing (¥4.29 → ¥5.53 → ¥7.29, 2023–2025). JD now runs a committed dual return.

Capital return vs. FCF — sustainable for now, not at this ratio indefinitely. FY2025 total capital returned ~¥31.8B (¥21.4B buyback + ¥10.4B dividend) on FCF of only ~¥6B — a ~5x payout, funded from the ¥213B war chest and incremental debt (~+¥17B borrowings). JD has the balance sheet to fund years of above-FCF returns, but returning 5x FCF while burning ¥46.6B on the war is funding two competing demands from the same fortress. This works only if either (a) the war spend falls sharply in 2026 (management guides food-delivery investment down; Q1’26 New-Biz loss already narrowed ¥14.8B → ¥10.4B), restoring FCF toward the ¥40B+ pre-war level, or (b) JD monetizes investment stakes. Q1’26 actions (continued buyback + dividend) show no near-term retreat from returns.

The war as a capital-deployment decision — the central question. The food-delivery build-out was the largest capital deployment of 2025 — organic, not an acquisition: ~¥46.6B of New-Businesses operating loss (subsidies + a >150,000-rider full-time network + zero-commission merchant onboarding). This is a classic Marathon “chase a hot adjacency with subsidies” red flag. The counter (management’s framing): it is a user-frequency/engagement engine (frequency +40% YoY, QAC +30%, >700M annual customers) feeding the core retail flywheel and incremental ad revenue, and it is ROI-disciplined (loss narrowed every quarter since launch; “investment peaked in 2025,” declining in 2026 “if competition rationalizes”). The dispositive open question: does subsidy-bought frequency durably earn back ~¥45B+, or does it churn when subsidies stop? [INTERPRETATION — the central capital-allocation question of the thesis]

A widening set of bets (the empire-building caution). Layered on the war: Dada (consolidated/taken private), Deppon (further freight consolidation, Jan-2026), CECONOMY/MediaMarktSaturn (voluntary takeover at €4.60/share, ~85.2% control, ~€2.2B equity value, German FCO cleared Sep-2025, close H1’26 — a large bet on a low-margin ~1–2% EBIT European brick-and-mortar electronics retailer, adding ~€20B+ of thin-margin revenue and integration risk), and a ~US$4.5B Hong Kong build-out (Jun-2026: Kai Bo supermarket chain, JD Malls, 50% of CCB Tower as HQ, JDL hubs across 18 districts — attacking CK Hutchison’s ParkNShop/Watsons in a small ~HK$380B market). The breadth of simultaneous bets — food delivery + Jingxi + Europe + CECONOMY + HK — is the kind of multi-front deployment that, absent the share cancellations, reads as empire-building.

Capex and R&D — disciplined. Capex ~¥14.2B FY2025 (~1.1% of revenue) — the logistics network is largely built; this is maintenance + selective AI/automation, not a heavy build cycle. R&D rising but disciplined (Q4’25 +52% to ¥6.7B; JoyAI LLM).

Founder control and incentives. Dual-class (Class A = 1 vote, Class B = 20 votes). Richard Liu (founder/Chairman) holds ~73.1% of total voting power (via Max Smart Limited and Fortune Rising Holdings) on a far smaller economic stake — entrenched control with negligible minority-ADS voting influence. Liu stepped down as CEO in 2022 (Sandy Xu/Xu Ran is CEO, Ian Shan CFO) but remains Chairman and controlling shareholder and personally drove the food-delivery entry (announced the “¥10B subsidy,” set rider-hiring targets, now attends daily retail/logistics meetings several times a week). Alignment is high (huge personal stake); accountability is low (no shareholder check on capital allocation). [FACT — 20-F; Caixin 2025-05-03; 36kr]

Verdict (Capital Allocation). Mixed / jury-out, leaning disciplined-on-returns-but-aggressive-on-deployment. The shareholder-return program is genuinely good — sustained, cancellation-backed buybacks (~6%/yr) bought at a record-cheap valuation, plus a growing dividend. But JD returned ~5x FCF in 2025 while committing ~¥46.6B to a subsidy war and queuing CECONOMY + a ~$4.5B HK build, funded from the war chest and new debt. Whether this is intelligent allocation hinges entirely on whether the food-delivery spend durably earns its return. Q1’26 evidence (loss narrowing fast, spend guided down, core margins expanding) supports the disciplined read; the breadth of simultaneous bets and 73% founder voting control support the caution. Not yet a clear value-destruction verdict — the next two-to-three quarters of New-Businesses loss trajectory are dispositive.

8. Changes and Headwinds — Last Two Years

1. The founder’s operational return (the strategic-driver change). Richard Liu made a high-profile operational return starting Jan-2025 (first public appearance in 4+ years) and personally drove the food-delivery entry. The entire war is a founder-driven bet with no board check — key-person and governance risk concentrated in one person with ~73% voting power.

2. Feb-2025 — the food-delivery / instant-retail launch (the defining event). Into Meituan’s turf, ~¥10B initial subsidy, Alibaba counter-escalation → ¥46.6B FY2025 New-Businesses loss, ~10% order share. “Seven Fresh Kitchen” owned cloud-kitchens reached >50 locations by end-Feb-2026 — a differentiated 1P-style food vertical.

3. 2024 stimulus / trade-in subsidy — tailwind turned headwind. The government electronics/appliance trade-in program boomed E&HA in 2024/H1’25 (pulling demand forward), then created a high base → E&HA −12% in Q4’25. It also drove the FY2025 OCF/FCF collapse (trade-in working-capital receivable build, now recovering). H1’26 faces the toughest comps; the program resumes with recovery expected H2’26. A new headwind: memory-chip cost inflation lifting phone/digital prices and potentially denting volumes.

4. International — a second, separate cash-burn vector. CECONOMY/MediaMarkt (~€2.2B, ~85% control, close H1’26) is JD’s largest-ever overseas acquisition — a big, thin-margin European omnichannel electronics bet. Joybuy relaunched in Europe (Mar-2026) with JoyExpress own-delivery in UK/DE/FR/NL; Ochama; first overseas automated warehouse in the UK. Framed as “controlled, disciplined” investment — but it is a genuine new burn vector.

5. Hong Kong ~US$4.5B build-out (Jun-2026) — broader than logistics (Kai Bo supermarkets, JD Malls, CCB Tower HQ, 18-district JDL hubs) — a fresh large deployment layered on the war and CECONOMY.

6. Capital-return changes — dividend initiated 2024; FY2025 US$3.0B buyback (6.3%, cancelled) + US$1.0/ADS dividend; Q1’26 continued.

7. Regulatory — the state-enforced ceasefire (the swing factor, opposite of 2021). SAMR’s July-2025 demand and June-2026 draft “Ten Rules” cap subsidies; Beijing’s June-2026 618 summons pressures a shift from price war to “products, services, innovation.” This is anti-involution regulation that protects margins — and JD, welcoming it loudly, is the late-entrant that benefits most. But competition may migrate from subsidies to M&A consolidation (Alibaba-Pupu ~$1.5B, Meituan-Dingdong ~$717M) — and JD is not a named bidder, risking being left sub-scale.

Verdict (Changes & Headwinds). Net neutral-to-slightly-positive, with the sign flipping on whether the ceasefire holds. Strengthening: the food-delivery loss has demonstrably peaked (Q4’25 ¥14.8B → Q1’26 narrowing) and a state-enforced ban removes the tail of an open-ended bleed; the core kept expanding margin through the war; capital return continued uninterrupted. Weakening: the founder-driven deployment is now multi-front (war + CECONOMY + HK + Jingxi), testing the “discipline” claim; H1’26 faces brutal E&HA comps plus memory-cost inflation; and the ceasefire may convert a subsidy war into an M&A consolidation war JD is sub-scale to fight. The changes make the loss trajectory look better while adding new capital-deployment fronts; they do not resolve the core question.

9. Risk Analysis

Catastrophic-loss / total-loss assessment. A total loss is low-probability: JD is profitable on a normalized basis, holds ~$8.9B net cash plus ~$16.6B of marketable stakes, and its HK-primary listing insulates the equity from an ADR delisting. The realistic tail is not bankruptcy but a permanent low multiple — capital recycled into value-destructive adjacencies under entrenched founder control, leaving the asset floor academic and the stock a perpetual China-beta value trap. The catastrophic scenario is value erosion via capital misallocation, not insolvency.

10. Valuation Discussion

Embedded-expectations and scenario analysis only. No price target; no recommendation. Scenarios are re-rate arithmetic on normalized earnings, not forecasts.

Headline multiples — cheapest ever on the stock’s own history. At $25.39/ADS, JD’s AZI valuation-index composite sits at the 4.24th percentile of its ~10-year range — the cheapest it has ever been. P/E ~10.1x (8.7th percentile), P/B ~1.16x (1.89th), P/S ~0.24x (2.13rd). ROIC’s EV/EBITDA of 18.5x is meaningless on war-depressed FY2025 EBITDA (it was 6.4x in FY2024); EV/Sales (0.19x) and the SOTP are the cleaner reads. The P/S figure is partly a 1P optical artifact (gross merchandise dollars at ~16% GM); the P/B and the SOTP are the more reliable cheapness signals. [Own-history percentiles are context only, never a cross-sectional target.]

Sum-of-the-parts — the crux of the deep-value case. Of the ~$34.8B equity market cap:

• Net cash: ~$8.9B (¥213.2B liquid less ¥107.1B debt).
• JD Logistics (2618.HK): ~HK$71B total market cap (~$9.1B); JD owns ~64% → ~$5.8B.
• JD Health (6618.HK): ~$13.0B total; JD owns ~67.5% → ~$8.8B.
• Other stakes (JD Property, JD Industrials 7618.HK, Dada minority, strategic equity): conservatively ~$2B (likely understated — JD Property carried a large pre-IPO valuation).
• Total marketable stakes (JD’s economic share): ~$16.6B.

Net cash + stakes ≈ $25.5B ≈ ~73% of the market cap. The implied “stub” for the entire consolidated core (still including JD Retail + the consolidated JDL/JDH operations) is ~$9.4B. Against JD Retail’s segment operating income of ¥51.4B (~$7.2B), the market is paying roughly 1.3x EBIT for the #1 most-trusted 1P retailer in China — a business that grew OI +25% and expanded margin to a record. Even haircutting the stakes 30–40% for holdco discount and non-saleability, the core is priced at a mid-single-digit EBIT multiple. That is the central crux: at ~$25/ADS the market assigns roughly zero-to-low-single-digit value to the core after netting out cash and the two HK-listed stakes. [INTERPRETATION]

Normalized-earnings scenarios (anchor = non-GAAP NI, the through-cycle read; GAAP NI is interest-income-flattered). Pre-war run-rate (FY2024) was ¥47.8B; the war trough (FY2025) was ¥27.0B; consensus FY26E is ~¥30B.

Embedded expectations — what the price underwrites. At ~10x trailing P/E, ~1.3x EV/Retail-EBIT, and a stub worth less than net cash + stakes, the market is pricing some combination of: (1) the food-delivery war is permanent; (2) the Chinese consumer never recovers; (3) the stakes are worth far less than market and/or capital is trapped (recycled into more hot adjacencies, never returned); (4) a China-ADR/VIE/HFCAA de-rating justifies a permanent low multiple regardless of fundamentals; (5) the founder misallocates the war chest indefinitely. The market could be wrong because the war loss already peaked (Q4’25 −¥4.55B → Q1’26 +¥3.8B), the core grew OI +25% through the war, the regulator is actively forcing a ceasefire, and the asset backing is real and HK-listed (the HFCAA tail is hedged by the 9618 primary listing). If even two of (1)–(5) are wrong, the multiple is mispriced low — that is the variant-perception fulcrum. [INTERPRETATION]

11. Variant Perception

Consensus — a two-layered disconnect. Sell-side is nominally bullish (37 analysts, “Strong Buy,” average PT ~$40.55 / ~+60%), underwriting the “cheap, war-losses-peaked, earnings-inflecting” thesis. But the tape/buy-side treats JD as a China-internet value trap — a margin-impaired, capital-immolating, founder-controlled ADR that trades as undifferentiated China macro beta. The disconnect (Strong Buy ratings vs. a stock −21% over 1 year, −16.8%/yr over 5 years, at its cheapest-ever multiple) is the setup. Daiwa’s June-2026 downgrade to Hold captures the marginal sell-side capitulation toward the bearish tape.

Strongest bull case. (1) Cheapest-ever on every metric with a hard asset floor (~73% of the cap in net cash + listed stakes) — you get China’s #1 trusted 1P retailer for ~1.3x EBIT. (2) The war has peaked (Q4’25 trough → Q1’26 +¥3.8B; mgmt says investment “peaked in 2025”). (3) Regulator-forced ceasefire is a tailwind (opposite of 2021). (4) Core compounding + frequency gains are real (frequency +40%, QAC +30%, double-digit ad growth). (5) Aggressive buyback at the trough (~6.3% retired & cancelled in 2025), per-share accretive — management voting with the balance sheet.

Strongest bear case. (1) Food delivery = permanent margin sink — JD bought a distant-third ~10% share at the highest marginal cost (late-entrant capital-cycle trap); subsidy-bought frequency may churn when subsidies stop. (2) Weak consumer/deflation is structural; E&HA already rolled −12%. (3) Alibaba + Meituan are structurally stronger in local services; JD’s logistics moat provably did not travel, and the war may migrate to M&A consolidation. (4) Capital is trapped/recycled into hot adjacencies (food delivery + CECONOMY + $4.5B HK) under a founder with 73% votes and zero accountability; no insider open-market buys despite the cheapest-ever price. (5) VIE/ADR delisting + geopolitical tail justifies a permanent discount; JD is undifferentiated China macro beta (KWEB 0.95 similarity).

The assumptions that matter most, with falsification tests.

1. The war is ending, not permanent. Bull falsified if New-Businesses loss re-widens in any 2026 quarter or stays >¥10B/quarter through 2026. Bear falsified if it narrows to <¥5B/quarter and management holds the “peaked” line (Q2–Q3’26 confirm Q1’26 wasn’t a one-off).
2. The core keeps compounding. Bull falsified if Retail OI margin stalls/reverses below 4.6% or ad growth decelerates to single digits. Bear falsified if margin keeps climbing toward the high-single-digit target with double-digit ad growth (a 7th straight year of expansion).
3. The asset backing accrues to shareholders. Bull falsified if capital is visibly recycled into more loss-making adjacencies with no monetization. Bear falsified if JD monetizes a stake or sustains/raises capital return toward FCF-funded levels as the war fades.
4. China discount/HFCAA is an overhang, not a thesis-killer. Bull falsified by a fresh PCAOB-access breakdown or forced-delisting headline. Bear falsified if the HK-primary listing keeps insulating the equity and the discount narrows on a sector re-rate.

Factor-positioning input. JD is not trading on idiosyncratics — it is a high-China-beta macro proxy. The All-Factors model shows Country:China beta +1.60 (R² 0.61, the dominant exposure), with related stocks entirely China-internet ETFs (KWEB 0.95, MCHI, FXI, YINN) plus BABA (0.94); idiosyncratic vol is only ~25%/yr. Style loadings: Momentum −0.70, USDollar −0.75, with Value only +0.06 (it does not screen as a clean quant-value name despite cheap multiples — a beaten-down China/growth name, not a value darling). Relative strength rs_peak −73.0 (near its nadir); 5-year return −16.8%/yr; 10-year max drawdown −79%. This is the empirical signature of an abandoned, negative-momentum, high-China-beta falling-knife at its momentum nadir — and that is precisely the source of the mispricing: consensus prices JD as undifferentiated China beta while the idiosyncratic SOTP/core-quality argument is ignored. A durable re-rate likely needs either a China-macro risk-on regime shift or an idiosyncratic catalyst (war-ceasefire confirmation, stake monetization) strong enough to break the KWEB correlation. [INTERPRETATION — regime-caveated, not a price call]

12. Fact vs. Interpretation Table

13. Open Questions

1. Does the ceasefire hold, or migrate to M&A consolidation? Alibaba/Meituan are buying dark-store networks (Pupu/Dingdong); JD is not a named bidder. Does JD get left sub-scale at ~10%, forcing renewed spend?
2. Food-delivery path to breakeven — management asserts improving unit economics but gives no unit data or breakeven timeline. Can ~10% share against a 55%-incumbent ever reach acceptable economics?
3. Can the stakes be monetized to JD shareholders, or is value trapped (recycled into new bets)? This determines whether the SOTP floor is real or academic.
4. CECONOMY consolidation timing and margin/integration drag — bringing ~€20B+ of thin-margin EU brick-and-mortar on-balance-sheet (close expected H1’26).
5. H1’26 E&HA — how much do memory-chip cost inflation + the trade-in high base dent the core category before the H2 recovery?
6. Exact JD ownership % of JDL (62% vs 68% swings the stake ~$0.6B) and the under-researched value of JD Property/JD Industrials/Dada “other stakes.”

14. What Must Be True

Bull case — what must be true. The food-delivery loss must keep narrowing toward breakeven (New-Businesses loss <¥5B/quarter through 2026) as the regulator-enforced ceasefire holds; JD Retail must keep expanding margin toward its high-single-digit target with double-digit advertising growth (a 7th straight year); and management must either monetize a stake or return the war chest (sustained/rising buyback) rather than recycling it into the next hot adjacency — converting the SOTP asset floor from academic to real. Falsification test (bull): any 2026 quarter in which the New-Businesses loss re-widens, OR JD Retail margin stalls below 4.6%, OR a fresh large cash-burn deployment (beyond CECONOMY/HK) is announced with no acceleration of capital return.

Bear case — what must be true. Food delivery must prove a permanent margin sink — ~10% subsidy-bought share that never reaches viable unit economics against a 55% incumbent, with frequency gains churning when subsidies stop; the war must migrate to an M&A consolidation fight JD is too sub-scale to win; and the capital must stay trapped — recycled into CECONOMY + HK + international under unaccountable founder control — so the asset floor never accrues to minority holders, leaving JD a perpetual China-beta value trap. Falsification test (bear): two consecutive 2026 quarters of New-Businesses loss narrowing to <¥5B with the “investment peaked” line intact, OR a visible stake monetization / step-up in capital return, OR a sustained sector re-rate that breaks JD’s KWEB correlation on idiosyncratic news.

15. Source Appendix

See the Source Appendix below for the full citation list. Primary sources include: JD FY2025 Form 20-F (filed 2026-04-16, SEC EDGAR CIK 1549802); JD Q4/FY2025 earnings release and call transcript (2026-03-05); JD Q1’26 results (2026-05-12, 6-K); ROIC.ai financial statements and ratios; AZI valuation-index and price history; FactorsToday factor model; and public sources cited inline (CNBC, Caixin, SCMP, S&P Global, companiesmarketcap.com, stockanalysis.com) accessed 2026-06-27.

The – institutional analysis carries no investment recommendation and no price target; the sole exception is the labeled “Claude.s Take” opening block, which is the author.s own independent opinion.

APPENDIX A — Standard Diligence Questionnaire

Supplemental to the research memo. Fact / Interpretation / Assumption labels applied where material. Currency RMB unless marked; valuation in USD per ADS (1 ADS = 2 ordinary).

General

What thoughtful questions have other investors asked about this company? The dominant question is whether the February-2025 food-delivery war is a value-destructive ego move by a returned founder or a defensible frequency/traffic play that feeds the core retail flywheel — i.e., does the ~¥46.6B New-Businesses loss earn its return? Secondary questions: (1) Is the SOTP asset floor real, or is capital permanently trapped/recycled? (2) Can JD ever reach viable unit economics at ~10% delivery share against Meituan’s 55%? (3) Does the HK-primary listing fully neutralize HFCAA delisting risk? (4) Is the cheapest-ever multiple a value opportunity or a permanent China-beta de-rating? (Interpretation)

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? A self-inflicted low. Group operating income collapsed −91% (¥39.6B → ¥3.7B) in 2025 on the delivery war; non-GAAP NI fell −44% to ¥27.0B vs. ¥47.8B pre-war (2024). The core (JD Retail + Logistics) is at a high (OI +20% to ¥56.7B, record Retail margin). So group earnings are at a war-depressed trough while the underlying core is at a cyclical high. (Fact/Interpretation)

Driven by the external environment or internal actions? Predominantly internal — the loss is a deliberate, founder-driven capital deployment into food delivery, not a demand shock. External overlays: a weak/deflationary Chinese consumer and the lapping of 2024 government trade-in stimulus (E&HA −12% in Q4’25). (Interpretation)

How stable are revenues? Revenue itself is stable-to-growing (¥745.8B → ¥1,309.1B over 2020–25; +13% in 2025), but its quality is mixed: general merchandise and advertising are durable; E&HA is policy-whipsawed; New-Businesses revenue is subsidy-bought. (Fact)

Outlook for products/services? Core retail margin has a multi-year self-help runway (4.6% → high-single-digit target); advertising grows double-digits; food delivery’s profitability is unproven. E&HA faces a tough H1’26 (trade-in comps + memory-chip cost inflation), recovering H2’26. (Interpretation)

How big will this market be — growing, shrinking, domestic or international? Chinese e-commerce is the world’s largest but mature, slow-growing, and deflationary. JD is expanding internationally (Europe via CECONOMY/Joybuy/Ochama; Hong Kong) — adding TAM but also thin-margin integration risk. (Interpretation)

Business Quality & Competitive Moat

Is the industry getting more or less competitive? More — the 2025-26 instant-retail war drew three giants into mutual capital destruction (~$14B+ combined 2025 losses). The one offset is the state-enforced “anti-involution” ceasefire that may force rationalization. (Fact/Interpretation)

How profitable is the business (ROIC, ROE)? Reported ROIC fell 8.78% (2024) → 0.83% (2025), depressed by the war and a balance sheet bloated with ~¥317B non-operating assets. Clean operating ROIC on the negative-working-capital core was comfortably >20–30% pre-war. ROE/GAAP figures are distorted by interest income and the cash/stake pile — read segment OI, not consolidated returns. (Interpretation)

How profitable is the industry — competitors, barriers? Low-to-moderate and unstable. Peer pre-war ROIC: Meituan ~18%, JD ~10%, Alibaba ~8%; PDD most capital-efficient. Barriers to entry in 1P logistics are high (capital), but barriers in marketplace/delivery are low enough that well-capitalized entrants repeatedly ignite price wars. (Fact/Interpretation)

Can the business be easily understood? Mostly — it is a 1P retailer + logistics operator + marketplace, complicated by consolidated HK-listed subsidiaries (JDL/JDH/JD Industrials), a VIE structure (small, ~7% of revenue), and a war-distorted P&L requiring segment/non-GAAP reading. (Interpretation)

Can it be undermined by foreign low-cost labor? Not directly — it is a domestic-China retail/logistics network; the competitive threat is domestic (Alibaba/PDD/Meituan/Douyin), not offshore labor arbitrage. (Interpretation)

Do brands matter? Yes — JD’s brand is the moat-adjacent intangible: “authentic goods, fast/reliable delivery, real warranty,” differentiated from counterfeit-prone marketplaces. (Fact/Interpretation)

What is the nature of competition? Price/subsidy-led, with weak product differentiation — periodic value-destructive subsidy wars. (Fact)

Customers’ switching costs? Low — Chinese consumers multi-home aggressively; JD Plus is loyalty, not lock-in. (Interpretation)

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? Yes — the market value of consolidated HK-listed stakes (JDL 2618 ~$5.8B, JD Health 6618 ~$8.8B JD-share) far exceeds carrying value; JD Property’s pre-IPO valuation and the logistics real estate may be understated. (Interpretation)

Off-balance-sheet liabilities? Operating/finance leases are capitalized; ~¥33B of total capital leases on-balance. No material flagged off-balance-sheet exposure beyond standard VIE contractual arrangements. (Fact)

How conservative is the accounting? Reasonable on the operating side (SBC small at ~0.4% of revenue; clear segment disclosure), but GAAP NI is flattered by interest income and swung by investment marks — non-GAAP NI and segment OI are the cleaner reads. (Interpretation)

How CapEx-hungry is the business? Moderate now — capex ~¥14.2B (~1.1% of revenue); the logistics network is largely built (maintenance + selective AI/automation). The cash hunger is in subsidies/M&A (food delivery, CECONOMY, HK), not capex. (Fact/Interpretation)

Capital Allocation & Management

How much FCF, and how is it used? FCF fell to ~¥6B (2025) from ~¥44B (2024) on the war; pre-war ~¥40B+ run-rate. Uses: ~¥21.4B buyback + ¥10.4B dividend (~5x FCF, funded from the war chest + debt) plus the ~¥46.6B war spend. Philosophy: committed dual shareholder return and aggressive reinvestment, funded by the fortress balance sheet. (Fact)

Significant acquisitions recently? CECONOMY/MediaMarkt (~€2.2B, ~85% control, close H1’26); Dada (take-private); Deppon (further consolidation Jan-2026); ~$4.5B Hong Kong build-out (organic). (Fact)

Buying back shares? Yes — US$3.0B / 6.3% of shares in 2025, all cancelled; +US$631M in Q1’26; ordinary share count −12.6% over two years. (Fact)

Issuing large amounts of new shares to insiders? No — SBC is small (~0.4% of revenue) and net share count is falling. Insiders receive RSU/option awards (code-M vests) but no large dilutive issuance. (Fact)

Compensation policy / incentives? Dual-class (Class B = 20 votes); founder Liu ~73% voting power. Standard executive equity awards; no disclosed ROIC-linked metric surfaced. Alignment is high (founder’s huge stake) but accountability is low (no shareholder check). (Fact/Interpretation)

Motivations of management? Founder Richard Liu returned operationally in 2025 and personally drove the delivery war — a long-term, control-driven strategic bet. Alignment via ownership is strong; the risk is unaccountable empire-building. (Interpretation)

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? An ADR (NASDAQ: JD; 1 ADS = 2 Class A ordinary shares), with a primary HK listing (9618). Not an MLP; no K-1 (issues 1099/standard ADR tax treatment). Cayman incorporation, VIE structure (~7% of revenue). (Fact)

Dividend policy? Regular annual cash dividend initiated 2024; FY2025 US$1.00/ADS (~US$1.4B), growing. (Fact)

How profitable is the business? War-depressed at the group level (non-GAAP NI ¥27.0B, ~2.1% net margin); the core is healthy (Retail OI margin record 4.6%, Q1’26 5.6%). (Fact)

Is net income diverging from cash from operations? Yes, sharply in 2025 — OCF ¥19.0B vs. GAAP NI ¥19.6B looks aligned, but both are far below the ¥58.1B OCF / ¥41.4B NI of 2024; the war hit both. GAAP NI also diverges upward from operating income (~7x) due to interest/investment income. Read segment OI. (Fact/Interpretation)

Risks & Downside

What factors would cause the stock to decline? Re-escalation of the delivery war (loss re-widens); M&A consolidation leaving JD sub-scale; further consumer/deflation weakness; capital visibly trapped/misallocated (CECONOMY/HK drag); HFCAA/geopolitical de-rating; RMB depreciation. (Interpretation)

Risk of a catastrophic loss? Low on a fundamental basis — net cash ~$8.9B + ~$16.6B stakes, normalized profitability, HK-listing insulation. The realistic tail is a permanent low multiple (value erosion via misallocation), not insolvency. (Interpretation)

Chance of a total loss? Very low — would require simultaneous forced ADR delisting and VIE-contract invalidation and capital destruction; the HK-primary listing and small VIE footprint make this remote. (Interpretation)

Recent News & Events

Has the business environment changed recently? Yes — (1) the delivery-war loss peaked Q4’25 and Q1’26 inflected to profit; (2) Chinese regulators are forcing a subsidy ceasefire (“Ten Rules,” June 2026); (3) Daiwa downgraded JD to Hold (June 2026); (4) Beijing summoned the giants over 618 tactics. (Fact)

Significant acquisitions? CECONOMY (~€2.2B), Deppon, Dada; ~$4.5B HK build. (Fact)

Change in accounting policies? None material flagged. (Fact)

Recent changes — new markets, facilities, management? International push (Europe/HK); founder Liu’s operational return (2025); food-delivery business built from scratch with a >150,000-rider network; Seven Fresh Kitchens >50 locations. CEO Sandy Xu (since 2023), CFO Ian Shan. (Fact)

APPENDIX B — Source Appendix

All sources accessed 2026-06-27 unless noted. Primary sources first. Each non-obvious memo fact traces to an entry here.

Primary filings & company disclosures

• JD.com FY2025 Form 20-F, filed 2026-04-16, SEC EDGAR CIK 0001549802 — https://www.sec.gov/Archives/edgar/data/1549802/000119312526157870/d53690d20f.htm (segment revenue/OI, VIE structure & ~7% revenue share, HFCAA disclosure, dual-class voting, Richard Liu ~73.1% voting power, marketing expense ¥84.0B, SBC allocation, capital leases, dividend/buyback).
• JD.com Q4 & FY2025 results release, GlobeNewswire, 2026-03-05 — segment operating income (JD Retail ¥51.4B; JD Logistics ¥5.27B; New Businesses −¥46.64B), Retail margin 4.6%, buyback US$3.0B/6.3%, dividend US$1.0/ADS.
• JD.com Q4/FY2025 earnings-call transcript, 2026-03-05 (via ROIC.ai get_earnings_call_transcript / get_latest_earnings_call) — CEO Sandy Xu / CFO Ian Shan: “investment in food delivery has peaked in 2025”; frequency +40% YoY; QAC +30%; advertising reallocation; “high single-digit” Retail margin target; regulator “catalyst not constraint.”
• JD.com Q1’26 results, GlobeNewswire/Nasdaq, 2026-05-12 — https://www.globenewswire.com/news-release/2026/05/12/3292587/0/en/jd-com-announces-first-quarter-2026-results.html — revenue ¥316B (+5%), non-GAAP NI ¥7.4B, JD Retail OI +16.5% to ¥15B (record 5.6% margin), New-Biz loss narrowed.
• JD.com IR — CECONOMY voluntary public takeover offer — https://ir.jd.com/news-releases/news-release-details/jdcom-announces-decision-make-voluntary-public-takeover-offer (€4.60/share, ~85.2% control, ~€2.2B; German FCO cleared 2025-09-18; close H1-2026).
• JD.com Q2 2025 interim results, ir.jd.com; jdcorporateblog.com — JD Logistics: >3,600 warehouses, >34M sq m, ~72% external revenue, >130 international warehouses across 23 countries.
• SEC EDGAR filing corpus (CIK 1549802): 5× 20-F, 96× 6-K, Form 3 (12) / Form 4 (10) over ~5 years; insider Form 4s = code-M RSU/option vests only, zero open-market purchases.

Quantitative data sources

• ROIC.ai MCP — income statement, balance sheet, cash flow (annual + quarterly), profitability/per-share ratios, enterprise value (multi-year RMB statements; net cash −¥63.5B; segment trend; EV ¥248.4B). Third-party aggregated; reconciled to the 20-F.
• AZI valuation_index (azitrading.com) — own-history percentiles (2026-06-26): composite 4.24th pctile; P/E 10.1x/8.7th; P/B 1.16x/1.89th; P/S 0.24x/2.13rd; TTM EPS $2.51/ADS, BVPS $21.95.
• AZI price history CSV (azitrading.com/controls/download-data.php?t=JD) — split/dividend-adjusted OHLCV; ATH $93.92 (2021-02-17), current $25.39, 52wk $24.54–$34.95, lifetime low $16.93.
• FactorsToday (factorstoday.com/api) — stock-loadings (Country:China +1.60, Momentum −0.70, Value +0.06), leaderboard (y1 −21%, y5 −16.8%/yr, 10yr maxDD −79%), related-stocks (KWEB 0.95, MCHI, FXI, BABA 0.94), stock-info (beta 0.93, rs_peak −73.0).
• companiesmarketcap.com — JD Logistics 2618.HK ~HK$71B; JD Health 6618.HK ~$13.0B (stake values).
• stockanalysis.com/stocks/jd/forecast — 37 analysts, Strong Buy, avg PT $40.55; FY26E EPS ¥12.47/ord (+23.6%).

Industry, competitive & regulatory (public press / research)

• CNBC, “Coffee at 30 cents escalates into billion-dollar burn for JD.com, Meituan, and Alibaba,” 2025-07-11 — https://www.cnbc.com/2025/07/11/china-instant-commerce-price-war-billions-subsidies-jd-alibaba-meituan-ele-me.html
• Caixin Global, “Meituan Swings to $3 Billion Loss as Delivery Price War Bites,” 2026-03-27 — https://www.caixinglobal.com/2026-03-27/meituan-swings-to-3-billion-loss-as-delivery-price-war-bites-102427802.html
• BigGo Finance, “China’s Food Delivery Subsidy Crackdown” — war-loss tallies (Alibaba ~¥87B, JD new-biz ~¥46.6B, Meituan ¥23.4B).
• Caixin Global, “Beijing Summons E-Commerce Giants Over Excessive Competition,” 2026-06-11 — https://www.caixinglobal.com/2026-06-11/beijing-summons-e-commerce-giants-over-excessive-competition-102453188.html
• SCMP, “As 618 shopping frenzy grows, Beijing warns e-commerce giants over price-war tactics,” 2026-06 — https://www.scmp.com/tech/article/3356744 ; SAMR “Ten Rules” draft — https://www.scmp.com/tech/article/3357458
• CNBC, “Beijing is summoning executives again… less worry than in 2021,” 2026-06-23.
• Caixin, “Chinese Regulators Demand Tech Giants End ‘Instant Retail’ Subsidy War,” 2025-07-19 — https://www.caixinglobal.com/2025-07-19/chinese-regulators-demand-tech-giants-end-instant-retail-subsidy-war-102342818.html
• S&P Global Ratings — “the worst of the subsidy war may be over” (June 2026).
• M&A consolidation: Alibaba ~$1.5B Pupu (Bloomberg via Benzinga); Meituan ~$717M Dingdong (Dingdong 6-K SPA 2026-02-05).
• Richard Liu operational return: Caixin 2025-05-03; SCMP “billionaire founder delivers food”; 36kr “Liu Qiangdong Returns to the Front Line.”
• HK ~US$4.5B build-out: Bamboo Works / Benzinga, 2026-06-15 (Kai Bo, JD Mall Wan Chai, CCB Tower HQ, JDL 18-district hubs).
• Daiwa Capital downgrade Buy→Hold (analyst John Choi), Benzinga, 2026-06-23.
• Peer ROIC comparison: kavout.com; stockdoctor.substack.com (Meituan ~18%, Alibaba ~8%, JD ~10%); order-share data (Meituan ~55% / Alibaba ~35% / JD ~10%), greatwallstreet.net.
• VIE/HFCAA framework: Paul Weiss; White & Case; JD identified under HFCAA 2022-05-04; PCAOB regained China/HK inspection access Dec-2022.

Analytical frameworks

• Analytical frameworks applied: Greenwald & Kahn, Competition Demystified (moat taxonomy, barriers-to-entry, ROIC/share-stability tests); Chancellor / Marathon, Capital Returns (supply-side capital-cycle; high returns attract capital → mean reversion).