Full Report

Figures converted from HKD to USD at historical FX rates (frankfurter.app). Monetary statements are shown in US$ millions; per-share figures use the matching period rate. Filing links open the native figures from which each USD value was derived.

As-reported financial statements for Meituan, FY2021–FY2025, traced to the source filings, opened with the share-price history those statements have to justify. Every linked USD figure opens the exact filing row containing the native reported value from which it was converted. Amounts in US$ millions unless noted.

Share Price — Available History Since January 2026

The stock closed at $10.32 on Jul 09, 2026 — down 18% over the window shown, trading between $8.20 and $13.49.

Source: market price feed, daily closes, Jan 2026–Jul 2026 — the feed marks this available history as partial. Price return only, excludes dividends. Prices converted from HKD to USD with date-matched or nearest-available FX.

FY2025 at a Glance

Source: FY2025 consolidated statements [1] [2] [3] [4]. Click any linked figure to open the filing page with the row highlighted.

Revenue by Operating Segment

Revenue by Operating Segment	FY2021	FY2022	FY2023	FY2024	FY2025
Core local commerce	17,525	20,617	26,488	32,222	33,508
New initiatives	5,448	7,592	8,941	11,246	13,365
Total revenues	22,973	28,209	35,429	43,468	46,873
Total revenues growth, derived	—	+22.8%	+25.6%	+22.7%	+7.8%

Source: Note 5 Segment Reporting — segment revenues (FY2021 restated to the two-segment structure in the FY2022 report) [5] [6] [7] [8]. Click any linked figure to open the filing page with the row highlighted.

Operating Profit by Segment

Operating Profit by Segment	FY2021	FY2022	FY2023	FY2024	FY2025
Core local commerce	2,413	3,784	4,954	6,749	(887)
New initiatives	(4,607)	(3,640)	(2,582)	(937)	(1,295)
Unallocated items	(772)	(891)	(655)	(1,068)	(1,035)
Total operating profit/(loss)	(2,966)	(746)	1,717	4,744	(3,217)

Source: Note 5 Segment Reporting — segment operating profit/(loss) [5] [6] [7] [8]. Click any linked figure to open the filing page with the row highlighted.

Income Statement

Income Statement	FY2021	FY2022	FY2023	FY2024	FY2025	FY2026E	FY2027E
Revenues	22,973	28,209	35,429	43,468	46,873	51,348	58,323
Cost of revenues	(17,526)	(20,289)	(22,986)	(26,757)	(32,612)	—	—
Gross profit	5,447	7,920	12,442	16,711	14,261	—	—
Selling and marketing expenses	(5,218)	(5,097)	(7,504)	(8,237)	(13,224)	—	—
Research and development expenses	(2,139)	(2,660)	(2,714)	(2,711)	(3,340)	—	—
General and administrative expenses	(1,105)	(1,253)	(1,200)	(1,381)	(1,531)	—	—
Operating profit/(loss)	(2,966)	(746)	1,717	4,744	(3,217)	—	—
Profit/(loss) before income tax	(3,022)	(866)	1,795	4,891	(3,191)	—	—
Income tax (expenses)/credits	4	9	(21)	(280)	191	—	—
Profit/(loss) for the year	(3,019)	(857)	1,774	4,611	(3,000)	—	—
Profit/(loss) attributable to equity holders of the Company	(3,019)	(857)	1,774	4,611	(3,000)	—	—
Basic earnings/(loss) per share (RMB)	(0.50)	(0.14)	0.29	0.75	(0.49)	—	—
Diluted earnings/(loss) per share (RMB)	(0.50)	(0.14)	0.27	0.73	(0.50)	(0.07)	0.52

Source: Consolidated Income Statement [1] [2] [3] [4]. Click any linked figure to open the filing page with the row highlighted.

Columns marked E are consensus analyst estimates shown alongside reported results for direct comparison; they are not company guidance.

Estimate source: Yahoo Finance analyst consensus, as of 2026-07-09. Forecasts carry no filing page links.

Balance Sheet

Balance Sheet	FY2021	FY2022	FY2023	FY2024	FY2025
Cash and cash equivalents	4,170	2,585	4,268	9,121	13,717
Short-term treasury investments	10,809	11,783	14,315	12,542	7,716
Trade receivables	230	263	351	342	427
Total current assets	18,959	18,358	23,443	27,005	28,913
Property, plant and equipment	2,926	2,847	3,326	3,894	4,972
Intangible assets	3,982	3,930	3,892	3,892	3,882
Total assets	30,864	31,355	37,514	41,764	44,568
Trade payables	1,945	2,229	2,942	3,244	4,441
Borrowings (non-current)	1,567	199	78	151	2,414
Notes payable (non-current)	3,897	4,310	4,431	4,894	6,053
Total current liabilities	8,797	9,802	12,914	13,898	15,871
Total liabilities	14,761	14,848	18,060	19,539	25,170
Equity attributable to equity holders of the Company	16,110	16,514	19,461	22,232	19,405

Source: Consolidated Statement of Financial Position [9] [10] [11] [12]. Click any linked figure to open the filing page with the row highlighted.

Cash Flow

Cash Flow	FY2021	FY2022	FY2023	FY2024	FY2025
Net cash flows from operating activities	(514)	1,464	5,188	7,358	(1,775)
Purchases and prepayments of property, plant and equipment and intangible assets	(1,156)	(735)	(881)	(1,416)	(1,705)
Net cash flows from investing activities	(7,502)	(1,887)	(3,157)	1,314	3,825
Repurchase and cancellation of ordinary shares	—	—	—	(3,359)	(47)
Net cash flows from financing activities	10,080	(1,281)	(356)	(3,916)	2,729
Net increase/(decrease) in cash and cash equivalents	2,064	(1,705)	1,674	4,756	4,779
Cash and cash equivalents at the end of the year	4,170	2,585	4,268	9,121	13,717
Free cash flow, derived	(1,670)	728	4,307	5,942	(3,480)

Source: Consolidated Statement of Cash Flows [13] [14] [15] [16]. Click any linked figure to open the filing page with the row highlighted.

Revenue by Type

Revenue by Type	FY2021	FY2022	FY2023	FY2024	FY2025
Delivery services	7,323	8,986	10,522	12,627	12,342
Commission	6,367	7,247	9,818	12,276	13,551
Online marketing services	3,730	3,946	5,187	6,340	6,670
Other services and sales (including interest revenue)	5,553	8,030	9,903	12,225	14,310
Total revenues	22,973	28,209	35,429	43,468	46,873

Source: Note 5 Segment Reporting — revenue disaggregation by service/product type [5] [6] [7] [8]. Click any linked figure to open the filing page with the row highlighted.

Long-Term Record

Fiscal year	Total revenue	Operating profit/(loss)	Profit/(loss) attributable to equity holders	Diluted EPS	Operating cash flow	Capital expenditure
FY2020	14,722	555	604	0.10	1,087	(2,029)
FY2021	22,973	(2,966)	(3,019)	(0.50)	(514)	(1,156)
FY2022	28,209	(746)	(857)	(0.14)	1,464	(735)
FY2023	35,429	1,717	1,774	0.27	5,188	(881)
FY2024	43,468	4,744	4,611	0.73	7,358	(1,416)
FY2025	46,873	(3,217)	(3,000)	(0.50)	(1,775)	(1,705)

Source: consolidated statements across filings; older years from the standardized feed [14] [1] [16] [2]. Click any linked figure to open the filing page with the row highlighted.

Traceability

258 of 258 figures on this page (100%) link to the filing page containing the native reported figure from which the USD value was converted — click a linked figure to open that source row. Unlinked figures come from standardized data feeds or pre-filing years.

• All figures are as printed in Meituan's audited consolidated financial statements, stated in RMB thousands (RMB'000). Meituan's presentation currency is RMB.

• Each of FY2021-FY2025 is cited to that year's own annual report (current-year column). The FY2020 long-term column is the comparative column of the FY2021 annual report.

• Revenue-by-segment and revenue-by-type reflect Meituan's two reportable segments (Core local commerce; New initiatives). FY2021 is shown on the restated two-segment basis presented in the FY2022 annual report; the FY2021 annual report itself used a different (three-segment) structure, so FY2021 segment figures are cited to the FY2022 report.

• The numeric data feed (data/financials/*.json) mislabels the reporting currency as 'HKD' and its USD conversions apply an HKD- USD rate to RMB figures; those USD fields are therefore not meaningful. The underlying RMB magnitudes in the feed match the filings and were used only as a cross-check.

• Debt is shown via the two non-current interest-bearing lines (Borrowings; Notes payable); Meituan also carries current portions of borrowings and notes payable disclosed on the same statement.

• 4 figure(s) differed between the data feed and the filing; the filing value is shown (see the run's financials/financials_tab.json for the audit trail).

---

Platform and Price War

Figures converted from RMB at historical FX rates — see data/company.json.fx_rates for the rate table. Ratios, margins, and multiples are unitless and unchanged.

Meituan runs China's dominant local-services and on-demand delivery platform — more than 150 million orders a day, revenue of $46.9 billion in 2025. It earned a record $4.6 billion profit in 2024, then swung to a $3.0 billion loss in 2025 when Alibaba and JD opened a subsidy war that drove even its core business into the red. The platform is dominant and funded by net cash; 2025's loss is either a temporary dip or a permanent reset.

What Meituan is

Meituan is the largest local-services platform in China: the app most people open to have a meal delivered, book a restaurant table, reserve a hotel, or buy groceries and everyday goods for same-hour delivery. Its stated mission — "we help people eat better, live better" — sits behind a business that in July 2025 was fulfilling over 150 million on-demand delivery orders a day, with more than 600 million monthly active users and over 500 million on the core app alone [1]. It is, in effect, the logistics and demand-aggregation layer for a large share of China's restaurants and neighbourhood retailers.

Sources: FY2025 Annual Report, Chairman's Statement and Financial Highlights [2]; daily-order figure from Q2 FY2025 earnings call [2].

The business is Cayman-incorporated, Beijing-headquartered, and has been listed in Hong Kong since September 2018 under a weighted-voting-rights structure. It was founded in 2010 by Wang Xing, who still runs it as Chairman and CEO.

How it makes money

Meituan reports in two segments, and the split is the key to reading the company [2].

Core Local Commerce is the profit engine: food delivery, in-store dining, hotel and travel booking, and Meituan Instashopping (quick commerce — groceries, medicine, electronics and other goods delivered on demand). It earns money four ways — delivery fees paid by users, commissions charged to merchants on each transaction, online marketing and advertising sold to merchants, and product sales from self-operated retail. In 2024 this segment produced $32.2 billion of revenue at a 20.9% operating margin [3].

New Initiatives is the investment bucket: Xiaoxiang Supermarket (self-operated grocery with nearly 1,000 distribution centres), community retail, and the overseas delivery brand Keeta (Hong Kong, Saudi Arabia, Qatar, and an evaluation of Brazil) [3]. It has always run at a loss, by design — the question each year is only how large.

Source: FY2024 Annual Report — Core Local Commerce revenue $32.2 billion; New Initiatives is total revenue of $43.5 billion less Core [3].

The shape that matters: the Core segment is large and, in normal conditions, structurally profitable; New Initiatives is a smaller, deliberately loss-making bet on future categories. For years the story was Core profits rising while New Initiatives losses narrowed. In 2025 both moved the wrong way at once.

Eight years of growth, then a swing

Revenue has compounded almost without interruption — from $8.3 billion in 2018 to $46.9 billion in 2025, roughly a five-and-a-half-fold increase [4] [5].

Source: FY2018–FY2022 from the FY2022 Annual Report five-year summary [6]; FY2023–FY2025 from the FY2024 and FY2025 Annual Reports [7] [8].

Profit has been far more volatile than revenue. Meituan lost $3.0 billion in 2021 (a year that included a regulatory antitrust fine and heavy New Initiatives spending) and $0.9 billion in 2022, turned profitable at $1.8 billion in 2023, then reached a record $4.6 billion in 2024 — a 158% jump as Core margins expanded and New Initiatives losses shrank [9]. In 2025 it swung back to a $3.0 billion loss [10].

Source: FY2022, FY2024 and FY2025 Annual Reports [11] [12] [13].

The 2025 subsidy war

The 2025 loss was not an operating stumble in the usual sense — the platform grew. Gross transaction value and order volume both rose double-digits, and annual transacting users, transaction frequency and ARPU all reached record highs [2]. What changed was competition. In early 2025 both JD and Alibaba entered on-demand food delivery aggressively, funding user and courier subsidies to buy share in a market Meituan had long led. Meituan matched the spending to defend its position, and the cost landed squarely on margins.

Total segment operating profit went from a $5.8 billion profit in 2024 to a $2.2 billion loss in 2025. The most important line: Core Local Commerce — the segment that earned $6.7 billion in 2024 at a 20.9% margin — turned to a $0.9 billion operating loss. New Initiatives losses widened to $1.3 billion, in part from stepped-up overseas investment in Keeta [2].

Source: FY2024 and FY2025 Annual Reports, Chairman's Statement / Financial Highlights [14] [15].

There is early evidence the worst has passed. In August 2025 the three platforms signalled an end to the price war under regulatory pressure, and in the first quarter of 2026 Meituan's revenue was still growing — up 5.6% to $11.6 billion — while the total segment operating loss narrowed to $0.5 billion (Core Local Commerce $0.26 billion, New Initiatives $0.27 billion), a far smaller drag than 2025's quarterly run-rate [16]. Whether normalisation restores the old margin structure or settles at a permanently lower level is unresolved; the case is most sensitive to where the Core margin settles.

The balance sheet and who owns it

Two facts bound the downside. First, Meituan absorbed a war of this scale without financial distress: it ended 2025 with $13.7 billion of cash and $7.7 billion of short-term treasury investments — about $21.4 billion of liquidity against roughly $10.3 billion of total debt, or some $11.1 billion of net cash [17]. For an investor who wants the chance of bankruptcy near zero, that buffer is the starting point of the analysis, not an afterthought.

Second, control sits with the founder. Meituan uses a dual-class structure in which each Class A share carries 10 votes and each Class B share one. As at March 2026, Wang Xing beneficially owned about 45.30% of the voting rights, and co-founder Mu Rongjun a further 5.56% — so the two founders control roughly half the vote [18]. Their economic stake is far smaller than their voting power: all Class A shares together represent only about 10% of the equity, so Wang Xing owns on the order of 8–9% of the company — a holding worth several billion dollars at today's price. That gap between control and economic ownership is a governance feature to weigh, but the alignment a value investor looks for — a founder with a large personal stake, running the business for the long term — is present. It has also been remarkably stable: at the 2018 IPO Wang Xing held about 10.4% of the equity and 47.3% of the vote [19].

The stock and what the price implies

Meituan came to market in September 2018 at up to about US$9.2 per share [20]. In the first half of 2026 the shares fell from about US$12.6 in January to a low near US$8.2 as the scale of the 2025 loss became clear, recovering to roughly US$10.3 by early July.

Source: company share price data (HKEX: 3690), converted to US dollars; indicative points across January–July 2026.

At about US$10.3, the roughly 6.08 billion shares are worth close to US$63 billion. Set against the business, that is about 1.2 times 2025 revenue and roughly 12–13 times the record $4.6 billion of profit the company earned in 2024, before the war. Consensus analyst price targets average about US$13.7, some 30% above the recent price, and the sell-side expects a return to profit over 2026–2027 as subsidies normalise.

Valuation multiples derived from reported financials and share price data; consensus targets from published analyst estimates, as reported.

The setup is a familiar one for a value or special-situation investor: a business once prized for its growth and profitability, now loss-making and out of favour, priced at roughly a dozen times its most recent normal-year earnings, sitting on net cash, with a founder still at the controls. At roughly a dozen times pre-war earnings, the price neither writes off the Core profit pool nor assumes a full return to pre-war margins; whether the 2025 subsidy war permanently reset the economics of Chinese food delivery is not yet settled in the reported numbers.

The question this report sets out to answer

The rest of the report tests both sides of that question. Meituan is China's dominant local-services platform, structurally profitable at scale until a 2025 subsidy war with Alibaba and JD turned a record profit into a loss; the question this report exists to answer is whether that collapse is a temporary, competition-driven dip in an enduring franchise — making today's depressed, net-cash-backed price a genuine margin of safety — or the first evidence that the profit pool itself is being competed away. The chapters that follow test each half of that question in turn: the durability of the core economics, the balance sheet and forward estimates in detail, the moat against two of China's best-funded rivals, the founder's record with capital, and what the price implies at each outcome.

---

Financials and Estimates

Figures converted from RMB at historical FX rates — see data/company.json.fx_rates for the rate table. Ratios, margins, and multiples are unitless and unchanged.

Meituan's three-year record splits cleanly in two: FY2022–FY2024 is a company scaling into rising profitability and heavy cash generation; FY2025 is a single year in which a subsidy war turned a record $4.6 billion profit into a $3.0 billion loss and a $7.4 billion operating cash inflow into a $1.8 billion outflow. Revenue never stopped growing. Consensus expects a return to profit by 2027, but the balance sheet — not the income statement — is what makes the wait survivable.

The three-year record

Over FY2023–FY2025 revenue rose from $35.4 billion to $46.9 billion — a 13% annual pace even after 2025's disruption — while the bottom line moved sharply: a $1.8 billion profit in 2023, a record $4.6 billion in 2024, then a $3.0 billion loss in 2025 [1].

Sources: FY2025 Annual Report, Financial Summary [2] and Liquidity and Capital Resources [3].

Source: FY2025 Annual Report, Financial Summary — five-year condensed statements [4].

The table below is the statutory record for the reader's stated three-year window. The story is in the last column: revenue up 8.1% in 2025, yet gross profit down $2.4 billion and the result swinging by $7.6 billion year-on-year.

Source: FY2025 Annual Report — Financial Summary [5] and Consolidated Income Statement [6].

Where the profit went

The loss was manufactured almost entirely by competition, and it hit through two lines at once. On the cost side, cost of revenue rose from 61.6% of sales in 2024 to 69.6% in 2025 — an 8-point compression of gross margin — as delivery and rider incentives inflated the cost of each order [7]. Below gross profit, selling and marketing expense jumped from $8.2 billion to $13.2 billion — from 19.0% to 28.2% of revenue — as Meituan matched the consumer and merchant subsidies that JD and Alibaba put into the market [8]. Research and development and general expenses barely moved.

Source: FY2025 Annual Report, Management Discussion and Analysis [9]; Consolidated Income Statement [10].

Put together, reported operating profit swung from $4.7 billion (a 10.9% margin) to a $3.2 billion loss — a roughly 18-point move, split about evenly between gross-margin compression in the cost of revenue and the marketing surge [11]. The mechanism matters for the through-line: a demand collapse would show up as falling revenue; this shows up as a deliberate, subsidy-funded fight for order share while revenue kept climbing. That is a discretionary loss, which is easier to reverse than a structural one — provided the spending actually stops. The unresolved question is how much of the 8-point gross-margin hit reverts when it does, which is a matter for the unit-economics case rather than the reported ledger.

One check guards against the loss being a statutory artifact. Meituan's own non-IFRS measures — which add back share-based compensation and similar items and usually flatter the result — tell the same story: adjusted net profit swung from $5.6 billion in 2024 to a $2.4 billion adjusted loss in 2025, and adjusted EBITDA from $6.3 billion to negative $1.8 billion [12]. The loss is real cash economics, not an accounting line.

Cash tells the same story

In the two profitable years, Meituan's earnings were high quality: operating cash flow ran well ahead of reported profit — $5.2 billion against $1.8 billion in 2023, $7.4 billion against $4.6 billion in 2024 — because the platform collects from users and merchants faster than it pays, and capital intensity is low (capex under 4% of revenue) [13]. That converted into $4.3 billion and $5.9 billion of free cash flow.

FY2025 reversed this. Operating cash flow was negative $1.8 billion and free cash flow negative $3.5 billion [14]. The subsidy war did not just dent an accounting margin; it drained roughly $3.5 billion of cash out the door in twelve months.

Source: FY2025 Annual Report, Liquidity and Capital Resources [15]; FY2024 Annual Report, Liquidity and Capital Resources [16].

The clearest sign of how management reads its own position: share buybacks, which absorbed $3.4 billion of cash in 2024, were cut to $47 million in 2025 [17]. Cash was being retained rather than returned — a capital-allocation shift explored more fully in the management chapter.

The balance sheet absorbed it

This is the part that speaks to a value investor's first concern, that a holding could go to zero. Meituan's balance sheet makes that outcome unlikely. At the end of 2025 Meituan held $13.7 billion of cash and a further $7.7 billion of short-term treasury investments, $21.4 billion in all, against total borrowings of about $10.3 billion — net cash of roughly $11.1 billion even after the worst operating year in its history [18].

Source: FY2025 Annual Report, Liquidity and Capital Resources [19]; prior-year cash and borrowings from reported financials, FY2022–FY2024.

The balance sheet stayed a fortress through the war only by leaning on debt: about $3.4 billion of net cash went out, replaced with $2.7 billion of fresh borrowing, while management widened the New Initiatives loss to $1.3 billion for an unproven overseas build and cut the buyback from $3.6 billion to $0.05 billion as the stock it had called undervalued fell below $10.20. The gross cash pile held roughly flat only because that $2.7 billion financing inflow came mainly from issuing notes payable and new borrowings, which is why total debt climbed from about $7.2 billion to $10.3 billion in the year [20]. Net cash fell by roughly $3.4 billion even as the headline cash balance rose. Against that, the gearing ratio is about 53%, some 55% of that debt matures in three years or more, and none of it carries a financial covenant [21]. A covenant-free, long-dated debt stack against $21 billion of liquid assets is why a full year of cash burn barely registered as a solvency question. On the reader's own bankruptcy-risk test, the evidence points to near-zero; the main thing that would change that read is a multi-year war, not a single one.

What the estimates expect

Consensus treats 2025 as the trough. Forty analysts put FY2026 revenue near $51.4 billion (up 10.3%) and thirty-six put FY2027 near $58.3 billion (up 13.6%) — growth returning to its pre-war double-digit pace. These are consensus estimates from the analyst data feed, not company guidance, and carry no filing page.

Source: reported revenue FY2023–FY2025 [22]; FY2026–FY2027 are consensus analyst estimates (40 and 36 contributors), as reported.

Source: consensus revenue and price-target estimates (current price $10.06; mean target $13.65), as reported.

The earnings turn is the softer part of the forecast. Consensus has EPS still modestly negative in FY2026 (about $0.07 loss per share on average, with a wide $0.23-loss-to-$0.10-profit range across contributors) before recovering to roughly $0.52 in FY2027 — a return to profit, though below the $0.73 diluted peak of 2024. The revisions run two ways: FY2026 estimates have been marked up over the past month (seventeen upgrades against three cuts) as the truce held, while FY2027 estimates have been trimmed (thirteen cuts) — near-term relief, out-year caution. Sell-side positioning is constructive despite the depressed tape: of 38 ratings, 28 are buy or strong-buy and only 2 are strong-sell, and the mean target of $13.65 sits about 36% above the $10.06 price. The price marks a fallen leader; sell-side coverage stays constructive.

The numbers here establish the base the rest of the report builds on. Whether 2025 was a dip or a reset is not settled by the ledger alone — the case is most sensitive to whether the 8-point gross-margin hit reverts as subsidies fade, and to whether the FY2027 profit consensus assumes survives contact with a competitor that has not formally surrendered. Both are questions for the chapters on unit economics and competition; the financials say only that Meituan can afford to wait for the answer.

---

Order Density

Figures converted from RMB at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

Meituan's moat is real but narrow: a two-sided order-density cost advantage, not a toll bridge. Through 2025's subsidy war it kept order-volume and GTV leadership and grew users to record levels — the franchise held. Yet a well-funded JD and Alibaba erased the entire ~20.9% Core margin in a single year, so the profit was always contestable. Whether that margin reverts is most sensitive to how the two rivals now behave.

The moat as it was built

The advantage Meituan sold at its 2018 listing was not brand or app design — it was density. On-demand delivery has a self-reinforcing cost curve: more consumers draw more merchants, more merchants draw more orders, and more orders let each rider travel a shorter distance per delivery. The prospectus stated it plainly — "as the scale of our food delivery business further increases, on average, each delivery rider in our network travels shorter distance and spends less time in fulfilling an order," so that "the high order density of our delivery network enables us to efficiently manage labor cost" [1]. The daily deliveries handled by each premium rider rose 22% in a single year on that logic [2].

That flywheel showed up directly in monetization. As food-delivery scale compounded, the segment's take rate — revenue divided by transaction value — climbed from 9.0% in 2016 to 12.3% in 2017 and 12.9% in early 2018 [3]. More density did not just cut cost; it let Meituan charge more per order while still improving service.

Source: Global Offering Prospectus (2018), Our Strengths [4].

By late 2017 the network already ran on roughly 531,000 daily active riders across more than 2,800 cities and counties — a physical logistics layer no software rival could conjure overnight [5]. That density is the thing the moat analysis has to test: it is genuinely hard to replicate, but it is not free to defend.

What density bought, and what it did not

By 2024 the flywheel had matured into a Core Local Commerce segment earning a 20.9% operating margin — $6.7 billion of operating profit, up from an 18.7% margin the year before [6]. On any static read, that looks like a wide moat: a dominant platform converting scale into a high, rising margin.

The 2025 subsidy war is the stress test that separates the durable part of that moat from the contestable part. Two things happened at once, and they point in opposite directions.

The franchise held. Through the most expensive assault the sector has seen, Meituan reported record daily and monthly active users in food delivery, high retention, and — its own words — "maintained leadership in both order volume and GTV," anchored in higher average-order-value segments and user loyalty [7]. The annual report makes the same claim: it "further solidified our leading position in user structure and consumer mindshare amid a fierce competition environment" [8]. A challenger with no density of its own could not have inflicted damage without buying users away — and it did not buy them.

The profit did not hold. JD entered food delivery in February 2025 and, alongside Alibaba's Ele.me and Taobao Instant Commerce, funded a subsidy war reported in the billions of dollars aimed squarely at Meituan's core [9]. Meituan's own income statement absorbed the hit: total segment operating profit swung from $5.8 billion in 2024 to a $2.2 billion loss in 2025, dragging Core Local Commerce from that 20.9% margin to an operating loss [10]. The moat protected share and users; it did not protect price.

Source: FY2024 Annual Report, Company Business Highlights [11]; FY2025 margin derived from segment operating loss and revenue as reported [12].

Why the users stayed and the profit left

The mechanism matters, because it tells you what a truce can and cannot repair. Core Local Commerce revenue actually rose 4.2% in 2025, to $33.5 billion, on more transactions and more GTV — demand did not fall [13]. What fell was the profit inside each order. Delivery revenue itself declined "primarily as a result of the elevated incentives deducted from revenues," as Meituan met rivals' subsidies with its own [14]. In other words, the platform kept the customer and gave up the margin on the transaction.

That is the analytically important line, and it is where this chapter and Financials and Estimates meet: about half of 2025's operating-margin swing sat in cost of revenue — rider and delivery subsidy embedded in unit economics, not in the discretionary marketing line. A ceasefire lowers marketing spend at will; it does not automatically restore the unit-level economics if a third rival keeps buying orders below cost. The users staying is evidence the moat is real. The profit leaving so completely, so fast, is evidence it is narrow.

The Third Player's Permanence

The durability question is most sensitive to how JD and Alibaba behave: whether they treat food delivery as a land grab to be rationalized, or as a permanent business they will fund indefinitely. The corpus gives conflicting signals, and an investor should hold both.

Alibaba has committed for good. Its FY2026 chairman's letter names quick commerce "a core strategic pillar in the ongoing platform transformation of Taobao and Tmall" and "a necessary path" to its mission, framing 30-minute delivery as a standard it must meet [15]. Ele.me is already the number-two food-delivery platform at roughly a third of the market [16]. A rival of Alibaba's balance sheet, treating this as strategic rather than opportunistic, is the strongest argument that the pre-war margin does not fully return.

JD is signaling the opposite. It expects "total investment in food delivery to decrease in 2026 compared to 2025" and to "begin generating revenue through offering merchant services, achieving an orderly and rational monetization" [17]. Tellingly, JD frames its delivery push as a customer-acquisition "strategic engine for user growth" feeding its retail business, not as a standalone profit centre [18] — an entrant using subsidies to buy shoppers is one that can dial them back once the cohorts are acquired.

Meituan's own read, offered with an obvious interest, is that the war is self-limiting: management called the price-competition phase "unsustainable" and said it expected rationalization over time [19]. The early 2026 data supports that direction. In Q1 2026 industry subsidies "became more rational," competition shifted "back toward execution, user experience and operational efficiency," and the Core segment operating loss narrowed to about $0.3 billion on $8.2 billion of segment revenue [20]. Regulators have leaned the same way, prompting the market to expect narrowing losses as authorities cool the fight [21].

Sources: Competitive Set rationale [22]; Alibaba FY2026 Annual Report [23]; JD.com Q4 FY2025 transcript [24].

The read

The evidence points to a narrow but genuine moat. Order density is a real, replicable-only-at-great-cost advantage that let Meituan keep every measure of demand leadership through the worst subsidy assault in the sector's history — a software-only challenger could not have done that. What density does not protect is the margin: the 20.9% Core profit of 2024 was the equilibrium rent of a two-player market, and it went to zero the moment a well-capitalized third player chose to spend. A durable franchise and a durable margin are not the same thing here, and 2025 proved it.

The strongest fact against a clean recovery is Alibaba's stated permanence: a rival of that size treating quick commerce as core strategy can hold the equilibrium margin structurally below the pre-war level for as long as it is willing to fund it [25]. The fact that most supports recovery is JD's planned 2026 pullback and the Q1 2026 narrowing toward the very ground — execution and cost efficiency — where density gives Meituan the edge [26] [27].

What would change the read is measurable: Core Local Commerce margin recovering toward the high-teens over 2026–2027 would confirm the moat sets a high floor and the war was cyclical; a margin that stalls in low single digits while all three players remain funded would confirm a partial, permanent reset of the profit pool. One structural lever runs in Meituan's favour on that timeline — by end-2025 its autonomous delivery vehicles had completed over 5.5 million orders across more than 19 million kilometres, a route to lowering the very rider cost the war inflated [28]. The margin line over the next two years, not any single filing, is what settles it.

---

Founder and Capital

Figures converted from RMB (and, for share price and buybacks, from HK$) at historical FX rates — see data/company.json.fx_rates for the rate table. Ratios, margins, and multiples are unitless and unchanged.

Capital allocation is where a founder's discipline shows, and over the subsidy year Meituan's ran in two directions at once. The balance sheet stayed a fortress through the war only by leaning on debt: about $3.4 billion of net cash went out, replaced with $2.7 billion of fresh borrowing, while management widened the New Initiatives loss to $1.3 billion for an unproven overseas build and cut the buyback from $3.6 billion to $0.05 billion as the stock it had called undervalued fell below $10.20. The New Initiatives loss [1], the 2024 and 2025 buybacks [2] [3], and management's own "undervalued" framing [4] are on the filing record; the net-cash and borrowing movements are set out in Financials and Estimates.

The preservation reads as strength, not distress. The buyback pause and the fresh borrowing both fell in FY2025, the year a roughly $3.5 billion free-cash outflow ran through the business (Financials and Estimates); through it Meituan held about $22 billion of cash and equivalents and never raised equity [5], and liquidity had refilled to roughly $23 billion of cash and short-term investments by the first quarter of 2026 [6]. On a bankruptcy-averse reader's own test the net-cash fall is a war-year peak drain, not a solvency signal. Whether the discipline that preserved the fortress also governs how the founder now spends it is what the rest of the chapter takes up.

Wang Xing runs Meituan with 45.30% of the vote on an economic stake of roughly 8%, worth about $5.2 billion, and draws a fixed salary of $0.65 million with no bonus and no share grant to himself [7] [8]. His capital-allocation record is genuinely two-sided: a multi-year wind-down of a $13 billion-plus New Initiatives loss on one side, a $3.6 billion buyback executed near $13.9 and then paused as the stock fell below $10.3 on the other.

Ownership and alignment

The founders own the control shares outright. As at 31 December 2025 Meituan had 6,111,665,005 shares in issue, of which 579,439,171 are Class A — the high-vote class, ten votes each — and essentially all of that class sits with the two founders [9]. Wang Xing holds 515,869,783 Class A shares, which the company translates into 45.30% of the vote; Mu Rongjun's 63,283,203 carry a further 5.56% [10]. Convert every Class A share to Class B and the founders' combined economic interest is about 10.35% of the company [11] — Wang Xing's own slice is roughly 8%.

Sources: FY2025 Annual Report, Corporate Information [12] and Report of Directors [13]; stake value derived from the share count at the recent ~$10.1 quote (Platform and Price War).

For an investor who prizes skin in the game, three facts matter more than the voting arithmetic that Platform and Price War laid out. First, the stake is large in absolute terms: at the recent quote near $10.1, Wang Xing's holding is worth on the order of $5.2 billion, so his personal wealth moves with the share price a minority holder is buying. Second, he has not sold it down — the share count held through his trust and personal vehicles is unchanged in structure since the 2018 listing [14]. Third, none of it is pledged: the company states it did not pledge any assets for fundraising as at year-end, and the founders' control vehicles carry no disclosed margin borrowing [15]. A founder whose control shares are borrowed against is one forced sale away from a governance shock; that risk is absent here.

One overhang that used to sit above the register is also gone. Tencent, historically Meituan's largest outside shareholder, completed a distribution in specie of its Meituan shares to its own shareholders, and no longer appears among the substantial shareholders — the list is now the two founders, BlackRock at 6.25% of the Class B shares, and JPMorgan [16]. The one standing caveat on alignment is the wedge itself: 45.30% of the vote on ~8% of the economics means minority holders cannot outvote the founder on ordinary resolutions. That is a governance cost the reader accepts or does not; it is not a change in his incentives.

What the founder is paid

Wang Xing's FY2025 emoluments were a $0.65 million salary plus $0.03 million of pension and benefits — $0.68 million in total, with no bonus and no share-based award to himself [17]. Co-founder Mu Rongjun was paid $0.55 million on the same basis. The independent non-executive directors each received between $0.22 million and $0.36 million, most of it in shares; total directors' emoluments for the year were $2.4 million [18].

Against a company that turned over $46.9 billion in revenue, a $0.65 million CEO salary is a rounding error — under two parts in one hundred thousand of sales. The alignment this creates is clean: Wang Xing is paid almost nothing in cash and grants himself no fresh equity, so his return on running Meituan comes through the ~8% stake he already owns, not through pay that would dilute the holders he sits alongside. The contrast with the wider employee base is worth stating plainly — company-wide share-based compensation was $0.77 billion in 2025, down from $0.98 billion in 2024 [19]. The dilution investors bear is a real, billions-a-year cost of running the platform; it just does not flow to the man at the top.

The New Initiatives loss, disciplined then reopened

The clearest window on how management spends is the New Initiatives segment — community retail, grocery, and now overseas — which has run at a loss every year and is where discretionary capital goes to be tested. Its year-by-year path is the clearest test of that discipline.

Sources: segment operating results, FY2021 [20], FY2022 [21], FY2023 [22] and FY2025 Annual Reports [23]. The 2021 figure is on the then-broader "new initiatives and others" definition.

The loss peaked at $4.9 billion in 2021, at the height of the community group-buying land grab [24]. Management then narrowed it in three consecutive steps — $3.6 billion in 2022, $2.6 billion in 2023, $0.94 billion in 2024 — each year attributed to "improving operating efficiency," especially in grocery retail [25] [26] [27]. Cumulatively the segment absorbed on the order of $13 billion over five years. That the loss came down as promised, year after year, is a promise-vs-delivery point in management's favour — the language in the calls turned into the number in the accounts.

The reopening in 2025 is a deliberate choice, not a relapse. The loss widened back to $1.3 billion, and the company is explicit that the cause is overseas investment through Keeta rather than a loss of grip on the domestic units [28]. Keeta reached positive unit economics in Hong Kong in the fourth quarter, is growing quickly in Saudi Arabia, and in the second half of 2025 launched in Qatar, Kuwait, the United Arab Emirates and Brazil [29]. This is the current use of the fortress balance sheet, and it is unproven: one market has reached break-even, the rest are early. Management's own framing for 2026 is to "prioritize operational improvement over aggressive new market expansion" while staying "financially disciplined" [30] — the same efficiency discipline that worked domestically, now pointed overseas.

Buybacks, and their timing

Meituan's other capital lever is the buyback, and here the record is less flattering. Repurchases began in January 2024, with management calling them the "preferred way to return capital right now, especially because it believes the stock is undervalued" [31]. Over 2024 the company bought back 261.4 million Class B shares for $3.6 billion at an average price of $13.87, and cancelled them [32] [33]. That buyback outran employee share issuance, so the count of shares outstanding fell.

Sources: FY2024 Annual Report, Report of Directors [34] and Share Capital note [35]; FY2025 Annual Report, Report of Directors [36].

Then the program stopped. In 2025 Meituan repurchased just 3.0 million shares for $50.3 million — barely 1% of the prior year's spend — all of it in May, at an average of $16.67 [37]. The commentary softened in step: by early 2025 management described buybacks as a tool "to offset dilution from employee stock plans" [38], and by mid-year as something it would "consider… depending on market conditions" [39]. The sequence is procyclical: the company spent $3.6 billion buying near $13.9 while calling the stock cheap, and all but withdrew as the price fell below $10.3 — cheaper still. On price alone, the 2024 buyback is underwater against the recent quote.

There is a real defence for a reader who wants bankruptcy risk near zero. The pause coincided with the worst operating year in Meituan's history, a full-year free-cash outflow of about $3.5 billion during the subsidy war (Financials and Estimates). Conserving roughly $3.6 billion of liquidity rather than spending it into a falling market is the priority a balance-sheet-first investor would choose; management held $22 billion of cash through the fight and did not have to raise equity [40]. What that defence cannot fully answer is why the buying was so heavy at $13.9 and so absent at $10.1. The read that would change is straightforward: a resumed, sizable repurchase at today's depressed price would convert the "bought high, paused low" pattern into opportunistic capital return.

Reading the record

On the alignment the reader cares about most, Meituan is close to the profile they look for: a founder still running the company he built, a multi-billion-dollar personal stake he has not sold or pledged, near-nominal cash pay, and an outside overhang cleared. The capital-allocation record beneath that is competent rather than exceptional — strong on cost discipline, where a $4.9 billion loss was walked down to $0.94 billion as promised, and weaker on buyback timing, where conviction and price moved in opposite directions. The open item is the same fortress cash that makes the 2025 loss survivable: it is now funding an overseas bet that has proven itself in one city and nowhere else, and whether that discipline travels is the thing to watch.

---

Industry Tailwinds

Figures converted from RMB at historical FX rates — see data/company.json.fx_rates for the rate table. Ratios, margins, penetration rates, and per-user counts are unitless and unchanged.

The recovery half of the case — the consensus that revenue reaccelerates to double digits and profit returns by 2027 — rests on the market itself still expanding. Meituan's own record supports that: platform transaction volume grew double digits even in the loss year, in-store orders rose over 65% in 2024, and on-demand retail over 40% in 2023. Demand is still expanding across frequency, quick commerce, and lower-tier cities. What 2025 also proved is that a growing market does not, by itself, protect the profit inside it.

The market Meituan was underwritten to grow into

The clearest independent measure of the opportunity is also the oldest one in the corpus. Meituan's 2018 listing prospectus commissioned iResearch to size the industry: China's food retail and service market was ¥8,735 billion (about $1,120 billion) in 2017 with an online penetration rate of just 13.4%, forecast to reach ¥14,132 billion (about $1,812 billion) at 29.5% online penetration by 2023 [1]. The broader consumer-service e-commerce market it sits inside was put at ¥2,705 billion (about $347 billion) in 2017, growing at a 19.8% CAGR to ¥8,011 billion (about $1,027 billion) by 2023 [2].

Sources: Global Offering Prospectus (2018), Industry Overview [3], competitive landscape [4] and business section [5].

What carries the tailwind is the low online penetration rate. A market that is only 13.4% online is one where growth comes from converting offline spending to online, not from stealing an existing online rival's share — a structurally easier tailwind, and the reason the prospectus expected the online-penetration rate to more than double in six years.

Source: Global Offering Prospectus (2018), Industry Overview [6].

Two honest limits apply to this data. It is 2018-vintage, and its forecast horizon (2023) has now passed; the corpus holds no equally rigorous, independently-sourced current market-size figure, and web research was unavailable for this run. The forward view below therefore leans on Meituan's own realized operating metrics rather than a fresh third-party TAM — which understates the independence of the evidence but not its direction.

The demand engine kept compounding

Even without a refreshed market study, the company's growth metrics show the market was still expanding right through the disruption. The prospectus itself set the pattern: transactions per user rose from 10.4 in 2015 to 12.9 in 2016 to 18.8 in 2017 — a platform whose users kept ordering more often as it added categories [7]. That same frequency-plus-breadth mechanism carried forward across the recent years, one vector at a time.

Sources: FY2023 Annual Report, Statement from the Chairman [8]; FY2024 Annual Report, Statement from the Chairman [9]; FY2025 Annual Report, Statement from the Chairman [10].

The FY2025 line speaks most directly to the through-line. In the worst operating year in the company's history, platform GTV and transaction volume still grew double digits, and annual transacting users, order frequency, and ARPU all reached historic highs [11]. The 2025 loss was not the market shrinking; the market grew. Management ended the year saying it remained "confident in the long-term growth potential and competitiveness" of the Core Local Commerce segment [12] — a claim the volume data supports even if the margin data does not.

Where the runway still sits

Three vectors carry most of the remaining upside, and all three are visible in the filings rather than asserted.

The largest is on-demand retail — quick commerce beyond food. Instashopping order volume grew over 40% in 2023, its active merchants almost 30%, and the prospectus-era "food in 30 minutes" habit has broadened into groceries, medicine, liquor, and appliances delivered on the same courier network [13]. The 2025 report shows this hardening into infrastructure — InstaMarts and Branded Flagship InstaMarts "flash warehouses," Xiaoxiang Supermarket front distribution centers accelerating city expansion in the fourth quarter, and expansion into medicine and health supply [14]. Meituan's own framing is that Instashopping users skew younger, stickier, and higher-spending than food-delivery users — a mix that lifts frequency and ticket size together [15].

The second is geographic. Meituan has been "deepening penetration into lower-tier cities" since at least 2021 [16], and in 2024 accelerated into the "county economy," streamlining onboarding for small merchants in markets where online penetration lags the tier-1 cities [17]. The third is the recovery and cross-sell of the high-margin in-store, hotel and travel category, where 2024 order volume soared over 65% and where a food-delivery user base is funneled into higher-value bookings [18].

What a tailwind does not do

A rising market is necessary for the recovery case; it is not sufficient, and two facts in the same filings bound it.

First, the macro drag is real and persistent. Some version of "despite challenges from the macro environment" appears in every recent annual report; in 2022 the Core segment grew 17.6% "despite negative impacts from the macro environment" [19]. That phrasing cuts both ways: the platform kept growing through weak Chinese consumption, but the demand tailwind has been running into a consumption headwind for years, and value-oriented formats like Pin Hao Fan exist precisely because consumers are trading down.

Second, a growing market did not defend the profit pool in 2025. Platform volume grew double digits while the Core segment swung from a $6.7 billion operating profit to a $0.9 billion loss. The tailwind and the loss coexisted in the same twelve months. That separates two questions the recovery case tends to blur: whether the market grows (the evidence says yes) and whether Meituan keeps the profit on that growth (unresolved, and the subject of Order Density). A larger addressable market raises the ceiling on the recovery; it says nothing about the floor, which competition sets.

The read that follows from the evidence: the industry tailwind is genuine and multi-vector — frequency, on-demand retail, lower-tier penetration, in-store recovery — and it is the demand-side support under the double-digit forward revenue consensus. The main fact against leaning on it is that the hard independent sizing is 2018-vintage and that 2025 showed volume growth and profit collapse can happen at once. What would sharpen the read: sustained double-digit on-demand-retail GTV growth in the 2026 quarterly disclosures, continued gains in order frequency and ARPU, and — the piece the tailwind cannot supply — evidence that the growth is again converting to segment profit as subsidies normalize.

---

What's Priced In

Figures converted from Chinese renminbi, and Hong Kong dollar share prices, at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

At about US$10.3 the market pays roughly seven times the Core Local Commerce engine's pre-war operating profit and gets net cash worth about a fifth of the market value on top, but that pre-war margin was competed to an operating loss in a single year, so whether the price is cheap depends on where the Core margin resettles once subsidies normalize.

That seven-times multiple is struck on the FY2024 pre-war Core operating profit of US$6.7 billion, while the enterprise value it is measured against is current — so it is a multiple of last year's profit engine, not of today's earnings. On the FY2025 basis the Core segment swung to a US$0.9 billion operating loss, which leaves the same multiple undefined [1]; the figure is a bet on where the steady-state margin resettles, not a claim of cheapness on current earnings. The moat chapter showed demand holding through the war while the entire pre-war margin was competed away in a single year (Order Density). The net cash of about US$11.1 billion, near a fifth of the US$63 billion market value, sits on top of the operating business for nothing [2].

The stakes turn on where the Core Local Commerce operating margin settles once subsidies normalize. The table below holds today's US$63 billion market value fixed and varies only the steady-state profit.

Source: author's scenarios holding market value and net cash fixed; the reset and reversion profit figures are illustrative, the partial-recovery figure is the 2027 consensus, as reported.

The arithmetic frames the debate rather than settling it. If the subsidy war permanently reset the economics — Core margins stuck in low single digits — then at 35-to-45 times earnings today's price is expensive, and the sales multiple's apparent cheapness is a mirage. If 2025 was a discretionary dip and Core margins revert toward their pre-war high teens, the stock is on roughly 10 to 13 times a net-cash-backed, dominant platform and the upside is large. The consensus sits between, and the price sits with the consensus: a partial recovery, roughly fairly valued, with the two tails wide open.

Because Meituan's profit is temporarily below its own potential — a subsidy war turned a record 2024 into a 2025 loss (Financials and Estimates) — a trailing price-to-earnings multiple is close to meaningless. The useful lenses are the ones that see through a depressed year: what the enterprise value is as a multiple of sales, what it costs against a normalized profit, and how sensitive that answer is to where core margins settle.

The price, stripped of cash

Meituan had 6,111,665,005 shares in issue at the end of 2025 [3]. At the July 2026 price of about US$10.3 that is a market capitalization near US$63 billion. Against it sits a genuine cash pile: US$13.7 billion of cash and equivalents plus US$7.7 billion of short-term treasury investments — some US$21.4 billion of liquidity — against roughly US$10.3 billion of total debt, leaving about US$11.1 billion of net cash [4].

That net cash is about US$1.96 of the US$10.3 share price — roughly 19% of the market value is money in the bank, not the operating business. The remaining US$8.4 per share, an enterprise value near US$47 billion, is what a buyer actually pays for the platform.

Source: share count from FY2025 Annual Report [5]; net cash derived from cash, treasury and debt in the FY2025 Annual Report [6]; share price as reported.

The cash matters twice. It removes bankruptcy from the list of things that can go wrong (Financials and Estimates), and it means the operating business is being valued more cheaply than the headline price suggests.

One times sales, for a platform

An enterprise value near US$47 billion is about 1.0 times 2025 revenue of US$46.9 billion [7]. On the consensus revenue path — US$51.7 billion in 2026 and US$58.8 billion in 2027 — it falls to about 0.9 and 0.8 times forward sales.

Source: EV derived from market value less net cash; revenue from FY2025 Annual Report [8] and consensus estimates, as reported.

For a business that at its 2024 peak converted about 10% of revenue into net profit, one times sales is not a demanding multiple — provided the profit comes back. The number does the work of a warning as much as an invitation: at 1.0 times sales, a business earning a 10% net margin trades near 10 times earnings, but a business earning 3% trades near 33 times. The sales multiple looks cheap only under the recovery it cannot itself confirm.

Value Against Normalized Profit

Against a normalized profit, the stock is neither the bargain nor the trap the sales multiple alternately implies. Two anchors bound the range. Meituan earned US$4.6 billion in 2024, a record [9]; at today's market value that is about 12.7 times peak earnings, or 10.3 times on an enterprise basis. But the sell-side does not expect a return to that peak. Consensus has Meituan back to a profit of roughly US$3.2 billion by 2027 — real money, but about 31% below the 2024 record.

Source: FY2024 profit from FY2024 Annual Report [10]; FY2025 loss from FY2025 Annual Report [11]; 2027 consensus estimate, as reported.

On the consensus 2027 number, the enterprise value is about 14.8 times earnings and the market value about 18.2 times. That is the crux of the multiple: the "return to profit" the sell-side models is a return to roughly 70% of the old peak, and the price already reflects it. Buying the 2027 recovery at 15 times enterprise value is paying a fair, not cheap, multiple for a partial rebound.

Source: derived from reported financials — market value less US$11.1 billion net cash [12], revenue and profit from the FY2024 and FY2025 Annual Reports [13] [14], and consensus estimates.

The most telling line in that table is the last. The Core Local Commerce segment earned US$6.7 billion of operating profit at a 20.9% margin in 2024, before the war [15]. Against that pre-war run-rate, today's enterprise value is about 7 times the operating profit of the profit engine alone — with the loss-making New Initiatives bucket, the overseas Keeta build, and the balance-sheet cash thrown in for nothing. Seven times is cheap for a dominant platform — if the 20.9% margin is a level the business returns to. That margin is the disputed number, and it is settled by operating results, not by any valuation multiple.

Outside anchors: a market that cannot agree

The sell-side dispersion tells the same story. The mean analyst price target is about US$13.7 and the median US$13.9 — some 32% above the recent US$10.3 — but the range runs from US$7.2 to US$17.8, a high roughly 2.4 times the low. Recommendations skew positive: 6 strong-buy and 22 buy ratings against 8 holds and 2 strong-sells. A spread that wide on a single name is itself a reading: the analyst community is not debating a few percentage points of growth, it is split between the dip and the reset.

Source: consensus analyst price targets and recent price, as reported.

One further anchor sits inside the company. Through 2024, Meituan repurchased US$3.6 billion of its own stock at an average of US$13.9, in buybacks it framed as creating value for shareholders [16]. That is management's revealed view of fair value — about a third above today's quote. The signal is weak on its own, because the buying was procyclical: heavy near US$13.9, all but absent below US$10 (Founder and Capital). A resumed, sizable repurchase at today's price would turn that internal mark into a genuine valuation signal; its continued absence leaves it a caveat.

Where this leaves the price

The honest read is a measured one. At about US$10.3 Meituan trades at roughly one times sales, about 15 times the consensus enterprise earnings of a partial recovery, with a fifth of the market value backed by net cash. On the base case the stock is close to fairly valued; the asymmetry — cheap if core margins revert, expensive if they reset — is real but unresolved by any multiple. The strongest fact against the bull case is the one the earnings lens exposes: even a successful recovery, as the sell-side models it, leaves 2027 profit about a third below the 2024 record, so "return to profit" is not "return to normal." The read would change with the realized Core Local Commerce operating margin as subsidies normalize. Valuation sets the range; the realized margin locates the business within it.

---

After the Trough

Figures converted from Chinese renminbi at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

The subsidy war that turned a record 2024 into a 2025 loss reached its worst point in the third quarter of 2025, when Core Local Commerce ran a $1.8 billion operating loss — a negative 20.9% margin on a segment that ran at a 20.9% profit margin in 2024. It has narrowed every quarter since. By the first quarter of 2026 the Core loss was $0.3 billion, management describes subsidies as "more rational," and the cash pile has refilled. The de-escalation is real; the recovery is not yet complete.

Where earlier chapters established that the 2025 collapse was a discretionary fight for price rather than a loss of customers (Order Density) and that the price already embeds a partial recovery (What's Priced In), this chapter tracks the series that speaks most directly to which reading is winning: the Core Local Commerce operating margin, quarter by quarter, as the war escalated and then cooled.

The quarterly arc

Core Local Commerce is the profit engine — food delivery, in-store, hotel and travel — and its operating margin is where the war shows up first. In 2024 it ran at 20.9% for the full year [1]. JD entered food delivery in February 2025, Alibaba's Ele.me escalated, and the margin followed a clean path down and back up.

Source: quarterly results announcements, segment information — Q2 2025 [2]; Q3 2025 [3]; FY2025 annual report [4]; Q1 2026 [5].

The shape matters more than any single point. The first quarter of 2025 still carried the pre-war economics — a $1.7 billion Core operating profit on $8.2 billion of revenue [6]. By the second quarter the margin had thinned to 5.7% [7]. The third quarter was the trough: a $1.8 billion Core operating loss, the deepest of the war, as all three platforms spent through the summer peak season [8]. The fourth quarter loss narrowed to about $1.3 billion, and the first quarter of 2026 to $0.3 billion [9]. The margin has narrowed for two consecutive quarters off the third-quarter 2025 trough — a recovery of 17.7 points across a three-quarter window, on roughly flat revenue — the strongest evidence yet that 2025 was a dip rather than a reset, but the line is still below zero, and the 2024 profit engine has not been rebuilt.

The latest quarter, in the numbers

The first quarter of 2026 is the freshest read, and the one that carries the recovery signal.

Source: Q1 2026 Results Announcement, financial summary and management discussion [10].

Revenue grew 5.6% year over year to $11.6 billion even as the company held subsidies tighter, and the reported net result was a $0.9 billion loss against a $1.3 billion profit a year earlier [11]. Total segment operating profit was negative $0.5 billion, split between a $0.3 billion Core loss and a New Initiatives loss that narrowed to $0.3 billion; adjusted net profit was negative $0.6 billion [12]. The Core operating margin of negative 3.2% compares with positive 21.1% a year earlier — the year-over-year gap is still large — but against the negative 20.9% trough two quarters before, the sequential move is what changed [13].

The table below sets the war window out in full, so the two comparisons — down 24 points year over year, up 18 points off the trough — sit side by side.

Source: quarterly results announcements and FY2025 annual report, segment information [14] [15] [16] [17].

Revenue held roughly flat across the five quarters — the platform kept its orders through the war, consistent with the demand-side durability shown earlier (Order Density). What moved was the margin, and it has moved back toward zero rather than settling at the bottom.

Management's read, and the condition attached

Management frames the first quarter of 2026 as the war cooling. The results announcement attributes the quarter's performance to "the moderation of intense industry competition" and "a more disciplined approach to subsidies" [18]. On the earnings call, management said industry subsidies "became more rational" during the quarter, shifting competition back toward execution, user experience and operational efficiency — the ground on which Meituan believes it holds an edge [19]. The same call noted that the "irrational subsidy moderated compared with the last quarter" while food delivery still added a large number of new users [20].

The forward statement is where the calibration sits, because management attached a condition and a warning to it. It expects "meaningful unit economics improvement in Q2 if competition stays rational" — the recovery is conditioned on the truce holding, not assumed — and it cautioned that second-half order growth could turn negative year over year on tougher comparisons, with delivery costs seasonally higher in the third and fourth quarters [21]. That is the honest counter to the recovery line: the improvement so far runs on a rival's restraint, and Meituan's own guide is for slower volume growth ahead. A single re-escalation by JD or Alibaba would show up in the next quarter's margin the way the last one did.

The cash cost, and the refill

The war was expensive in cash, and the balance sheet shows both the drain and the refill. At the third-quarter trough, cash and treasury investments had fallen to $18.2 billion — $12.7 billion of cash plus $5.4 billion of treasury investments — as the company funded a $2.1 billion adjusted net loss and a $1.9 billion adjusted EBITDA loss in the quarter [22]. By the end of the first quarter of 2026, liquidity had rebuilt to $23.0 billion — $14.9 billion of cash plus $8.1 billion of treasury investments — as the losses narrowed [23]. The fortress balance sheet documented earlier (Founder and Capital) absorbed the worst year without strain, and it is refilling as the operating losses shrink — which is why the war was survivable at this depth but would become a genuine constraint only if the trough recurred and persisted.

What would settle it

The dip-versus-reset question is now most sensitive to one observable series — the Core operating margin — and the next two quarters carry it. If the Core Local Commerce operating margin continues its climb and crosses back above zero in the second or third quarter of 2026 while revenue holds, the 2025 loss reads as a discretionary, competition-driven dip that is unwinding — the value in the shares would sit toward the middle-to-cheap end of the scenarios in What's Priced In. If instead the margin stalls in negative territory, or a re-escalation drives it back down, the reset tail stays live and the stock is dear on any normalized-earnings measure.

Three datapoints resolve it, in order of weight. First, the Core operating margin trajectory — the direct measure, which has narrowed for two consecutive quarters off the third-quarter 2025 trough, 17.7 points across a three-quarter window and still below zero. Second, whether JD's guided-down 2026 investment and Alibaba's posture actually hold through the second half, against management's own warning that order growth could turn negative. Third, whether the refilling balance sheet is redeployed — a resumed buyback at the depressed price would be management's own signal on where it thinks value sits (Founder and Capital). The first of those prints in August 2026, with the half-year results.

---

Regulatory Overhang

Figures converted from Chinese renminbi, and Hong Kong dollar share prices, at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

The report so far weighs one force on Meituan's profit pool: competition. A second bears on the same margin — the Chinese state. It has already cost the company once, a US$0.44 billion antitrust fine that helped push 2021 to a loss, and it now sits on both sides of the recovery. Regulators forced the 2025 subsidy truce that is helping profits heal, while a new rider social-insurance obligation threatens a permanent, non-competitive step-up in delivery costs. The case is most sensitive to where the steady-state Core margin settles, and that margin is, in part, a policy variable.

The 2021 precedent

The last time Meituan reported a full-year loss before 2025, a regulator was part of the reason. In April 2021 the State Administration for Market Regulation opened an anti-monopoly investigation, and in October 2021 it imposed a fine of ¥3,442 million (about US$0.44 billion) for abuse of market dominance — the "choosing one of two" merchant-exclusivity practice common across China's platforms [1]. The penalty landed in a year the company already reported a US$3.0 billion operating loss, and management named the fine explicitly as a driver [2].

Two things carry forward from that episode. The first is posture: Beijing has shown it will act directly against Meituan's economics when it judges the platform's conduct anti-competitive, and the amount — roughly 3% of that year's revenue — was set by rule, not negotiation. The second is that the specific practice sanctioned, merchant exclusivity, is one of the levers a density-based network could otherwise use to defend share (Order Density); it is now legally off the table. The fine itself was a one-time cash cost long since paid. Its lasting effect is the boundary it drew around how the moat may be defended.

A cost floor under delivery

The live regulatory question is not conduct but labor. Meituan's delivery network runs on more than three million active riders, historically engaged as flexible gig workers rather than employees [3]. Through 2024 the company's obligation to them was largely commercial: it required its delivery partners to buy employer's-liability and accident insurance, and left retirement provision to the riders themselves [4].

That changed in 2025. Under regulatory encouragement to extend protections to workers in "new forms of employment," Meituan launched what it calls the first industry-wide pension program covering all types of couriers, and extended occupational-injury insurance across 17 provinces and cities [5]. It began funding pension and injury insurance for roughly 820,000 long-term riders from the second quarter of 2025, then expanded the subsidy nationwide later in the year [6]. This is not a soft ESG footnote: management's own cost bridge names "enriched benefits for couriers" among the reasons cost of revenue rose to 69.6% of sales in 2025 from 61.6% a year earlier — the same line that absorbed the subsidy war [7].

The distinction that governs how large this can get is between a subsidy and a mandate. What Meituan runs today is a subsidy toward flexible-employment schemes — the company pays part of a voluntary contribution while riders keep their gig status. The tail risk is reclassification: a regulatory move to treat stable riders as employees would trigger full employer social-insurance contributions on the rider base, a cost that recurs every year and does not unwind when subsidies do. Meituan does not disclose its total courier cost, but delivery-services revenue — the closest disclosed proxy for the scale of the rider-cost base — ran at about US$12.3 billion in 2025, roughly flat on 2024 [8]. Against that base, the sensitivity is straightforward. For comparison, Core Local Commerce operating profit was US$6.7 billion before the war and swung to a US$0.9 billion loss in 2025 [9].

Source: illustrative sensitivity derived from FY2025 delivery-services revenue (~US$12.3bn) and FY2024 Core Local Commerce operating profit (US$6.7bn); Meituan does not disclose total courier cost or a social-insurance cost estimate [10] [11].

A step-up equal to even a tenth of the delivery-revenue base — a plausible order of magnitude for full pension and medical contributions on a stable rider cohort — is close to a fifth of the pre-war Core profit pool, and it does not depend on any competitor's behaviour. This is what the labor question adds to the report: What's Priced In showed the equity value is most sensitive to where the steady-state Core margin settles, and reclassification is a way that margin could be lowered permanently even if the subsidy war never returns. The counter-fact is real and worth equal weight: today the program is a subsidy Meituan chooses to pace, its cost is already inside the 2025 numbers rather than lurking ahead of them, and China has so far preserved the flexible "new employment forms" category precisely because forcing platform-wide employment would strain the gig-economy model nationwide. The reclassification tail is a risk, not a base case.

The same hand on both sides

The state that threatens the cost floor is also, on the present evidence, what capped the war's downside. Through mid-2025 regulators pressed the platforms over the subsidy fight, and in August 2025 Alibaba, Meituan and JD publicly pledged to end the price war after Beijing's warnings [12]. Meituan's own account is that it is "cooperating with regulators investigating the market" — the intervention that helped turn the trough [13]. The three-quarter margin recovery that the bull case leans on (After the Trough) is therefore partly regulatory in origin, not purely a market outcome.

That symmetry is the useful way to hold the regulatory variable. It is not a one-directional negative. The same authority raises the floor under rider costs and lowers the ceiling on how far rivals can bid delivery below cost.

Sources: FY2021 Annual Report [14]; FY2025 Annual Report [15]; news summary [16].

What would change the read

On the evidence in the corpus, regulation is a genuine but currently bounded second threat to the profit pool. The rider program is a subsidy the company is pacing, its cost is already embedded in the 2025 margin, and the state's recent net effect — through the subsidy truce — has been to support the recovery rather than deepen the loss. What would move this from a bounded cost to a structural one is a specific, observable shift: any move from voluntary "subsidy" language to a mandate treating stable riders as employees. Three markers make that visible. The first is the wording of national rules on "new employment forms" — whether they harden from encouragement into a contribution requirement. The second is the courier-benefit component inside cost of revenue, which the company has begun to flag and could be pressed to quantify. The third is whether regulators keep enforcing the delivery truce, or let the platforms return to bidding orders below cost — the same lever, pointed the other way. None of these prints on a fixed date, but each is checkable in the next several filings, and together they determine whether the state is a cost the model absorbs or one that resets the Core margin.

---

Dip or Reset

Figures converted from Chinese renminbi, and Hong Kong dollar share prices, at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

The earlier chapters lay out a dominant, net-cash-backed platform whose 2025 loss is narrowing but not yet reversed. This chapter sets the bull and bear readings against the same shared facts, then reduces the disagreement to a short list of falsifiable datapoints. The margin of safety sits in the balance sheet and the demand base; the case is most sensitive to where the Core Local Commerce operating margin settles as subsidies normalize.

The same facts, two readings

The disagreement is not about what happened. Both sides accept the same record. The Core Local Commerce operating margin bottomed at negative 20.9% in the third quarter of 2025 — a US$1.8 billion operating loss — and then narrowed for two consecutive quarters to negative 3.2% by the first quarter of 2026, a 17.7-point recovery off the trough across a three-quarter window that is still below zero [1] [2]. Group revenue held through the fight while total segment operating profit swung from a US$5.8 billion profit in 2024 to a US$2.2 billion loss in 2025 [3]. Management now calls subsidies "more rational" but conditions further recovery on the truce holding, and warns second-half order growth could turn negative year over year [4]. JD guides its 2026 food-delivery investment down while Alibaba calls quick commerce a "core strategic pillar" of Taobao and Tmall [5] [6]. Behind it all sits US$11.1 billion of net cash that funded the war without an equity raise [7]. What splits the readings is what comes next.

The table sets the two cases against each shared fact, with the evidence that would settle each one.

Sources: quarterly segment tables [8] [9]; FY2025 Annual Report [10]; Q1 2026 earnings call [11]; JD.com Q4 2025 call [12]; Alibaba FY2026 Annual Report [13]; net cash from the FY2025 Annual Report [14]; consensus 2027 estimate as reported.

Two of these rows carry most of the weight. The margin trajectory is the direct measure of the question, and it has narrowed for two consecutive quarters off the Q3 2025 trough (After the Trough). Against it stands the single strongest fact in the bear's hand: Alibaba's stated permanence. A rival of its balance sheet, calling 30-minute delivery a strategic necessity rather than an experiment, can hold the equilibrium margin structurally below the pre-war level for as long as it chooses to fund it [15]. The recovery so far runs on JD's restraint, not on Meituan's own pricing power — and management has attached its own warning to that, expecting further improvement only "if competition stays rational" [16].

The margin of safety, and where it stops

For an investor who prizes a large margin of safety and a near-zero chance of bankruptcy, the protection here is unusually concrete, and it is worth separating from the disputed part of the case.

Sources: net cash US$11.1bn from the FY2025 Annual Report [17] against market value (What's Priced In); founder economics [18]; latest Core margin [19]; 2027 consensus as reported.

Three things are protected. The balance sheet is the first: some US$11.1 billion of net cash — about a fifth of the market value — sat on the books after the worst operating year in the company's history, funded through the fight with new borrowing rather than an equity raise [20]. Bankruptcy is not on the list of things that can go wrong. The second is alignment: the founder holds 45.30% of the vote on roughly 8% of the economics, has not sold the stake or pledged it, and draws near-nominal cash pay — so a forced-sale governance shock is absent [21] [22]. The third is the franchise itself: demand held through the war, so the market's fear is priced onto one part of the business — the profitability of core delivery — not onto the platform's users or its share.

Where the protection stops is the same place the disagreement lives. The case is most sensitive to where the Core margin settles, and the margin of safety in the balance sheet does not answer it. At about US$10.4 the market pays roughly seven times the Core Local Commerce engine's pre-war operating profit and gets net cash worth about a fifth of the market value on top, but that pre-war margin was competed to an operating loss in a single year, so whether the price is cheap depends on where the Core margin resettles once subsidies normalize (What's Priced In). Two of the safety props also carry a caveat the reader should hold. The buyback that would most directly convert the fortress cash into per-share value was paused through the depressed year — just US$0.05 billion repurchased in 2025 against US$3.3 billion the year before — so management is conserving the cushion rather than pressing the value case at today's price [23]. And part of the cash is already committed: the New Initiatives loss widened back to US$1.3 billion in 2025 to fund the overseas Keeta build, which has proven positive unit economics in one city and remains early everywhere else [24].

What to watch

The disagreement resolves into a small set of observable datapoints, each falsifiable and each tied to a filing line and a window. They are ordered by how directly they move the read.

Sources: watch thresholds derived from the segment tables, cash-flow statement and management guidance cited above — Q1 2026 segment table [25], Q1 2026 outlook [26], and the buyback and New Initiatives disclosures [27] [28].

The first line does most of the work. The Core operating margin is the direct measure of dip-versus-reset, it has narrowed for two consecutive quarters off the trough, and the next reading prints with the half-year results in August 2026. If it crosses back above zero while revenue holds, the reset tail narrows and the value sits toward the middle-to-cheap end of the scenarios in What's Priced In. If it stalls in negative territory — or a rival re-escalates and drives it back down — the reset case stays live and the shares are dear on any normalized-earnings measure.

Where the evidence leaves it

The record supports a measured reading rather than a call. The franchise came through the war intact, the balance sheet absorbed the worst year without strain, and the founder's incentives point the same way as a minority holder's — so the downside is bounded by real net cash and a durable user base, which is the protection a margin-of-safety investor is looking for. What that protection cannot supply is the answer to the disputed number: whether the Core Local Commerce margin reverts toward its pre-war high teens or settles permanently lower. The strongest single fact against reversion is Alibaba's stated commitment to quick commerce as core strategy, with a mandated rise in rider social-insurance costs a second policy-side lever that could reset the margin (Regulatory Overhang); the strongest fact for it is two consecutive quarters of narrowing losses off the Q3 2025 trough and a rival, JD, guiding its own spend down. The evidence does not yet decide between them, and the honest position is to name the datapoints that will — the Core margin's path, order-volume growth, and the two rivals' spend through the second half of 2026 — rather than to force the question closed a quarter early.