MIXUE Group

01 — Executive summarySixty-four thousand stores, 2.3% growth, and the first profit decline since listing

MIXUE Group reported revenue of RMB 15,215.8 million for the six months ended 30 June 2026, an increase of 2.3% on the prior year. In the same six months the company added 4,202 net new stores, taking the global network to 63,987 across 17 countries, and its profit fell 14.7% to RMB 2,319.3 million. Basic earnings per share fell 16.3% to RMB 6.05. It is the first half-yearly profit decline the company has reported since its shares listed in Hong Kong in March 2025, and by the company's own admission on the results call, average turnover per store fell by a double-digit percentage.

Those two numbers — 20.7% more stores, 2.3% more revenue — describe a business model that has reached the end of its first phase. MIXUE has never been a drinks retailer. It is a vertically integrated supply chain that sells ingredients, packaging and equipment to a franchise network it does not own: 93.8% of first-half revenue was goods sold to franchisees, 3.5% was equipment, and only 2.7% was franchise and related service fees. Because the head office is paid on what franchisees buy rather than on what consumers spend, its revenue has historically been a near-mechanical function of the store count. When the store count grew 35%, revenue grew 35%. When it grew 20.7%, revenue grew 2.3%. That relationship is the single most important thing to understand about this company, and it is the subject of this report.

The question we set out to answer is narrower than whether MIXUE is a good business. It plainly is: the largest freshly-made drinks network in the world, 100% self-supplied in core ingredients, six production bases, 31 warehouses and a cold-chain system that most competitors cannot afford to replicate. The question is what the equity is worth once the store count stops being the growth driver, and how much of that has already been priced in.

Our answer, compressed into a sentence: the de-rating from a growth multiple to a value multiple is broadly complete and the balance sheet is genuinely exceptional, but the earnings base is still falling, the per-store arithmetic has not stabilised, and the company is now spending real money subsidising the franchisees who generate its revenue. At HK$170.50 the shares trade at roughly ten times trailing earnings, with RMB 21.6 billion of cash, deposits and wealth-management products against no borrowings — 39.1% of the market capitalisation. That is not an expensive price for the category leader. It is also not obviously a cheap one for a business whose revenue is flat and whose profit is shrinking.

The deceleration did not begin in 2026. The first half of 2025 was the peak of a delivery-platform subsidy war between China's three instant-retail platforms, which pushed order volumes across the entire freshly-made drinks industry to levels the underlying store economics did not support. MIXUE's revenue rose 39.3% in that half. Against that base, the first half of 2026 was always going to look weak. What matters is that it is weak on every measure, not just the growth rate: gross margin fell 120 basis points to 30.4%, selling and distribution expenses rose 22.9%, administrative expenses rose 39.4%, and the tax charge fell because profit before tax fell 14.2%. Six of the seven lines above the bottom of the income statement moved the wrong way.

Three findings follow from the primary record. Each is developed in full later in this report, and each is stated here with the number that supports it.

1. The revenue model is a store-count derivative, and the store count has run out of room

The company ended 2025 with 59,785 stores and 30 June 2026 with 63,987, covering 31 provinces and more than 300 cities in the Chinese mainland — including every county-level city in the country. Industry data from the restaurant-location provider Zhuanmen Canyan put the national tea-store count at 383,100 in the first half of 2026, down 43,200 net over twelve months, with more than 110,000 closures and a category closure rate of 16.2%. Roughly one store in six stopped trading. MIXUE is adding 4,202 net stores into that market. It is not expanding a market; it is redistributing an existing one, and the redistribution shows up as a 15% fall in revenue per store.

2. The profit decline is a cost story, not a demand story

Revenue grew 2.3% and cost of sales grew 4.1%. Gross profit fell RMB 76.7 million. Selling and distribution expenses added RMB 209.3 million and administrative expenses RMB 172.4 million. The company's explanation is that it is deliberately spending on product quality — the zhen xian chun ("real ingredients, fresh taste, simple recipes") programme that is converting ambient-temperature ingredients to chilled and frozen, plus the brand-IP content machine built around Snow King — and on staff to support store operations. Headcount rose from 9,102 to 10,374 in six months. The company plans to spend RMB 1.6 billion on the supply-chain upgrade in 2026 and RMB 1 billion subsidising franchisees' coffee-machine replacements. These are real investments. They are also, at this point, indistinguishable from price support for a franchise network under strain.

3. The balance sheet is the strongest argument in the file, and the market is not paying for it

Cash, time deposits, restricted cash and wealth-management products totalled RMB 21,640.7 million at 30 June 2026, up 8.3% in six months, with no interest-bearing bank borrowings and a gearing ratio of 17.7%. That is RMB 57.01 per share, or HK$66.7 at the current exchange rate — 39.1% of the share price. Strip the cash out and the operating business is being valued at about 6.7 times consensus FY2026 earnings. The company has just made its first distribution to shareholders, a special dividend of RMB 2.65 per share (HK$3.08304) approved on 30 September 2026 and payable on 6 November 2026. A business that pays out cash and trades at under seven times ex-cash earnings is not the profile of a company the market expects to grow quickly. It is the profile of one the market expects to shrink.

We should be explicit about where we sit relative to the street. The average target of the 21 analysts polled by S&P Global is HK$261.62, which is 53% above the current price and more than 40% above ours. Four of the five most recent published actions — Goldman Sachs at HK$285, CLSA at HK$270, Huatai at HK$266 and GF Securities at HK$255 — are all above HK$250. We are not making the case that MIXUE is a bad company. It is one of the most impressive operating machines in Asian consumer goods. We are making the narrower case that a business whose revenue has stopped growing, whose profit is falling, whose largest market has reached saturation and whose international network has shrunk for two consecutive years should not be underwritten at a multiple that assumes mid-teens growth resumes in 2027.

The report proceeds as follows. Sections 2 to 5 set out the business model, the revenue engine, the franchise contract and the eight-year financial record. Sections 6 to 8 cover the three brands. Sections 9 to 11 deal with the arithmetic of a single store and the saturation question. Sections 12 to 16 cover margin, cost, earnings quality, the balance sheet and capital allocation. Sections 17 to 19 cover the overseas business and the delivery war that distorted the 2025 base. Sections 20 to 23 cover competition, coffee, food safety and management. Sections 24 to 29 carry the valuation, the bull and bear cases, the risk matrix, the catalysts and the conclusion.

02 — Company & business modelA supply chain wearing a drinks brand, not a drinks brand with a supply chain

MIXUE Group was founded in Zhengzhou, Henan province, in 1997 by Zhang Hongchao, and opened its first franchise store in 2006. Its shares listed on the Hong Kong Stock Exchange on 3 March 2025 at HK$202.50, raising approximately HK$3,799 million net of expenses including the full exercise of the over-allotment option. The company describes itself as "a leading global freshly-made drinks company", and its products — freshly-made fruit drinks, tea drinks, ice cream, coffee and fresh beer — are typically priced around one US dollar, or approximately RMB 6, per item. Core MIXUE products sit between RMB 2 and RMB 8; Lucky Cup coffee between RMB 2 and RMB 10; FULU fresh beer at roughly RMB 6 to RMB 11 per 500ml.

That is the description the company gives. The financial statements tell a different and more interesting story. In the six months to 30 June 2026, MIXUE sold RMB 14,264.9 million of goods to its franchisees — syrups, milk, tea, coffee, fruit, grains, condiments, cups, lids, straws and packaging — plus RMB 532.7 million of equipment. It collected RMB 418.2 million in franchise and related service fees, which is 2.7% of total revenue. The consumer-facing drink is, from the income statement's point of view, an intermediate good: the thing that causes a franchisee to place the next order.

This is not a technicality. It determines almost everything about how the company behaves. Because revenue is recognised when goods are delivered to a franchisee rather than when a consumer pays, the head office is insulated from store-level demand for as long as the franchisee keeps ordering. Because the margin is earned on manufacturing and logistics rather than on the retail sale, the company's profitability depends on procurement scale, factory utilisation and route density rather than on footfall. And because the franchisee owns the store and bears the rent, the wages and the inventory risk, the group's own balance sheet carries almost no operating leverage to a decline in consumer spending — until the franchisee stops ordering, at which point the effect is abrupt.

The company's own framing is that it competes on total cost leadership built on three pillars: supply chain, store operations, and brand IP. It self-produces 100% of its core ingredients and operates six production bases — in Henan, Hainan, Guangxi, Chongqing, Anhui and Yunnan, the last of which began operations in May 2026. It runs 31 warehouses in China covering 33 provincial-level regions and more than 300 cities, plus local warehouse systems and delivery networks in nine overseas countries. It was the first company in the Chinese freshly-made drinks industry to build its own centralised factories, in 2012, and its own logistics system, in 2014. Those two decisions, taken before the category existed at scale, are the reason it can sell a cup of tea for RMB 4 and still report a 30% gross margin.

The brand is a genuine asset of a different kind. Snow King — a cartoon character with an ice-cream scepter introduced in 2018 as a "lifelong brand ambassador" — has become the only iconic IP in the Chinese freshly-made drinks industry, with an animated series, a comic series released in June 2026, merchandise, parade floats and a flagship-store format now open in 26 Chinese cities. Management treats it as a content business as much as a marketing device. It is also, as section 13 sets out, now a material cost line.

Three other structural features deserve to be stated at the outset. First, the company reports as a single operating segment: it told the exchange in Note 3 of the interim financial statements that it manages its businesses as a whole and presents no reportable segment information. Investors therefore cannot see the profitability of MIXUE separately from Lucky Cup separately from FULU, nor of China separately from overseas. Second, it has never disclosed overseas revenue, despite operating internationally since 2018. Third, it is a Chinese-incorporated issuer with a dual share structure: 150,883,058 listed H shares and 228,735,742 unlisted domestic shares, 379,618,800 in total, giving a total market capitalisation of HK$64.73 billion at the 8 October 2026 close. The controlling family retains a large majority of the unlisted shares, which is why the special dividend attracted 99.9994% of the votes cast at the extraordinary general meeting.

03 — The revenue engineWhy 97.3% of revenue is a derivative of the store count

For five years the relationship between store count and revenue at MIXUE was close to linear. The network went from 28,983 stores at the end of 2022 to 37,565 at the end of 2023, 46,479 at the end of 2024 and 59,785 at the end of 2025. Revenue went from RMB 13.58 billion to RMB 20.30 billion, RMB 24.83 billion and RMB 33.56 billion. Growth in the two series tracked each other within a few percentage points every year, because each incremental store consumes a broadly similar volume of syrup, fruit, tea and packaging per period, and the head office books that consumption as revenue.

In the year to 30 June 2026 that relationship broke. The network grew 20.7% and revenue grew 2.3%. There are three candidate explanations and they are not mutually exclusive: the incremental stores are lower-volume than the average, the existing stores are selling less because new stores are cannibalising them, or franchisees are buying less per store because consumer demand per outlet has fallen. The company's own disclosure points to all three.

The city-tier data supports the first explanation. At 30 June 2026, 58.0% of the Chinese mainland network was in third-tier cities and below, up from 57.6% a year earlier. First-tier cities were 4.9% of the network, unchanged. In absolute terms the company added 550 stores in first-tier cities and 6,786 in third-tier cities and below over the year. The stores it is adding now are in smaller, lower-income markets with lower throughput per outlet. The average ticket is similar — MIXUE's menu is priced nationally — but the number of transactions per day is lower.

The second and third explanations are harder to separate and the company does not help. It has not published average store turnover since its prospectus, in which it disclosed that average single-store terminal sales fell from RMB 1.1353 million in the first nine months of 2023 to RMB 1.0827 million in the first nine months of 2024, a decline of 4.63%, with average cups sold per store falling from 177,200 to 170,700. At the first-half 2026 results call the chief executive, Zhang Yuan, confirmed that average store turnover had fallen by a double-digit percentage, and gave two reasons: the elevated base created by the 2025 delivery-platform subsidy war, and intensifying competition. That is an unusually direct admission for a company that had, until then, declined to disclose the metric.

The competitive context makes the arithmetic unavoidable. Across the six listed Chinese freshly-made drinks groups, first-half revenue totalled RMB 36.78 billion — the sum of MIXUE's RMB 15.22 billion, Guming's RMB 7.47 billion, Chagee's RMB 6.96 billion, ChaPanda's RMB 2.66 billion, Auntea Jenny's RMB 2.59 billion and Nayuki's RMB 1.89 billion — up roughly 9% year on year. MIXUE is more than twice the size of the next player and its revenue is larger than the second and third players combined. It is also the only one of the six whose profit fell by more than a rounding error.

There is a genuine question about whether the company is now managing for revenue at all. The stated strategy for 2026 is "quality-led growth": prudent expansion, store-format upgrades, flagship stores, the chilled and frozen supply chain, and the digital rollout of smart drink dispensers, which reached more than 18,000 MIXUE stores by 30 June 2026 against 13,000 at the end of 2025. Every one of those initiatives raises cost before it raises revenue. If the strategy works, the reward is a network of stores that each sell more and a franchise system that survives; if it does not, the company will have spent a year of margin on a problem that was really about too many stores in the same street.

04 — The franchise contractWhat the franchisee buys, what they keep, and who carries the risk

MIXUE's franchise model is unusual in one respect that matters more than any other: the franchise fee is not the business. In the first half of 2026, franchise and related service fees were 2.7% of total revenue. The company states this plainly in its own disclosure, noting that "franchise and related service fees are not our primary sources of revenue" and that its role is to provide franchisees with "a competitive one-stop solution" of ingredients and equipment. The economics of the arrangement therefore turn on what the franchisee pays for goods relative to what the franchisee can sell them for.

Publicly available data on that spread is limited, because the company does not publish franchisee profitability. What is known is the cost side. A MIXUE franchise requires an upfront investment commonly reported at more than RMB 300,000 for franchise fee, equipment, decoration and deposit. Industry analysis puts the daily revenue required for a franchisee to break even at roughly RMB 3,000. The company's own prospectus disclosure implies average daily terminal sales per store of about RMB 4,400 in the first nine months of 2023 and RMB 4,184 in the first nine months of 2024. On those numbers the average franchisee was operating with a margin of perhaps 25-30% above breakeven in 2024 — not a comfortable buffer in a category with a 16.2% annual closure rate, and one that a 15% decline in store turnover would consume almost entirely.

Two developments in the last eighteen months suggest the head office knows this. The first is pricing. When the price of lemons rose sharply in 2025 — national wholesale prices up more than 60% year on year on adverse weather in the main growing regions and a global supply gap — MIXUE raised the price at which it supplies lemons to franchisees from RMB 200 per case to RMB 255 per case with effect from 30 June 2025, an increase of 27.5%, while holding consumer-facing menu prices broadly stable. The company absorbed part of the input shock but passed a substantial part of it down the chain. That is a rational decision for a business that earns its margin on the sale of the input, and it is a direct transfer of margin from franchisee to franchisor.

The second is the equipment subsidy. In 2026 the company set aside approximately RMB 1 billion to subsidise franchisees' replacement of equipment, principally the coffee machines and smart drink dispensers that underpin the fresh-ingredient pivot. On any reading this is a reversal of the usual direction of cash in a franchise system: the franchisor is funding capital expenditure on assets owned by the franchisee. Management presents it as an investment in product quality and in the coffee category. It is also, in substance, a recognition that franchisees will not make the investment themselves at current returns.

The franchisee base is still growing, which is the strongest counter-argument. Franchisee numbers rose from 23,404 at 30 June 2025 to 29,775 at 30 June 2026, and from 20,976 at the end of 2024 to 27,450 at the end of 2025. Multi-store franchisees are common, which means the average franchisee is a small business operator with more than one outlet rather than a first-time entrant. But the closure data is deteriorating at the same time. Closures rose from 1,609 in 2024 to 2,527 in 2025, an increase of 57.1%, and were 1,289 in the first half of 2026 alone — already 80% of the entire 2024 figure in half the time. Openings, meanwhile, fell from 7,721 in the first half of 2025 to 5,455 in the first half of 2026. Both sides of the churn are moving in the wrong direction.

05 — The financial recordEight years of the income statement, and the two things that never moved

MIXUE's revenue grew from RMB 10.35 billion in 2021 to RMB 33.56 billion in 2025, a compound annual growth rate of 34.2%. Net profit grew from RMB 1.91 billion to RMB 5.89 billion, a compound rate of 32.5%. Those are extraordinary numbers for a consumer business, and they were produced almost entirely by one variable: the store count, which went from roughly 20,000 to 59,785 over the same period.

What did not move, across the whole period, is the gross margin. It was 31.34% in 2021, 28.34% in 2022, 29.55% in 2023, 32.46% in 2024 and 31.14% in 2025. That is a range of about four percentage points around a mean of 30.6% over five years, in a period when revenue tripled. A company that self-produces 100% of its core ingredients, runs six factories and 31 warehouses, and tripled its procurement volume should have produced some operating leverage in its gross margin. It did not. The company's own explanation — a change in revenue mix and rising raw-material costs — is plausible, but the pattern suggests something structural: the cost savings from scale are being passed through to franchisees and consumers in the form of a lower price, which is precisely the "total cost leadership" strategy the company says it pursues. That is a defensible strategy. It does mean the equity does not have a margin-expansion story to fall back on.

The operating margin tells the same story with more noise. Reported operating income was RMB 2.44 billion in 2021, RMB 2.54 billion in 2022, RMB 3.98 billion in 2023, RMB 5.60 billion in 2024 and RMB 7.22 billion in 2025 — an operating margin of 23.6%, 18.7%, 19.6%, 22.6% and 21.5%. Again, no trend. The company converts roughly one renminbi of every five of revenue into operating profit, and has done so for five years regardless of scale.

Net profit is where the leverage appears, and for a slightly uncomfortable reason. MIXUE earns a substantial return on the cash its franchisees hand over before they receive goods. Contract liabilities — customer prepayments, in substance — were RMB 594.0 million at 30 June 2026 against RMB 473.1 million at the end of 2025, and the company holds RMB 21.6 billion of liquid resources that generate interest and investment income. Other income and gains, net, was RMB 127.5 million in the first half of 2026 and RMB 158.6 million in the first half of 2025. That is not a large number relative to profit, but it is a reminder that the reported bottom line is assisted by the float. In the first half of 2026 the assistance was less than it had been, because foreign exchange losses on foreign-currency deposits rose to RMB 153.8 million from RMB 42.2 million — a direct consequence of holding HK$3.8 billion of IPO proceeds in a company whose functional currency is the renminbi.

The first half of 2026 is the first period in the record in which every line above the tax charge moved adversely. Revenue rose 2.3%, cost of sales rose 4.1%, gross profit fell 1.6%, other income fell 19.6%, selling and distribution expenses rose 22.9%, administrative expenses rose 39.4%, research spending fell 1.6% (the only line that fell, and it is 0.3% of revenue), finance costs rose 68.1%, and profit before tax fell 14.2%. The effective tax rate rose from 21.6% to 22.1%, which is worth noting because the company benefits from a 15% preferential rate on its western-region and Hainan Free Trade Port subsidiaries and from agricultural pre-treatment exemptions. Profit for the period fell 14.7% to RMB 2,319.3 million, of which RMB 2,296.3 million was attributable to owners of the parent.

06 — Brand: MIXUEThe 59,609-store mainland network, and why density cuts both ways

The MIXUE brand is the company. At 30 June 2026 the group operated 59,609 stores in the Chinese mainland, of which all but a handful are franchised, plus a further share of the 4,378 overseas stores that trade under the same name. The mainland network covers 31 provinces, autonomous regions and municipalities and more than 300 cities, and the company states that it has achieved full coverage of county-level cities. No competitor is close: Guming had 14,351 stores, Auntea Jenny 13,155, ChaPanda 8,863, Chagee 7,639 and Nayuki 1,685.

That density is the company's principal competitive advantage and its principal constraint, and the two are the same fact seen from different angles. Density is what makes the cold chain work: a warehouse in Zhengzhou can serve enough stores on a single route to make daily refrigerated delivery economical, which is why MIXUE can supply fresh fruit to a store selling RMB 4 drinks and its competitors cannot. Density is also what caps the revenue per store, because a new outlet in a county town of 400,000 people takes transactions from the outlet that was already there.

The company is responding with format rather than count. Flagship stores — a format combining drinks, Snow King merchandise and cultural engagement — were open in 26 Chinese cities at 30 June 2026, up from 23 at the end of 2025 and from one in Zhengzhou in January 2025. The company has also begun building a small indoor Snow King theme park at its Zhengzhou headquarters. These are experiments in turning a low-ticket, high-frequency, take-away business into something with a higher average transaction value and a longer dwell time. They are also, at 26 cities out of more than 300, immaterial to the consolidated numbers today.

The operational upgrade programme is more consequential. Smart drink dispensers, which dispense measured quantities at controlled temperatures, were in more than 18,000 MIXUE stores by 30 June 2026, up from 13,000 six months earlier and from zero in October 2024. Fully automatic coffee machines, which upgrade the menu from pre-ground to freshly ground coffee, were in more than 3,000 MIXUE stores, with press reports in September 2026 putting the figure nearer 6,000. The company describes the dispensers as reducing human error, improving product standardisation and lowering food-safety risk. The subtext — that the equipment is being installed partly so that the network can be operated with less skilled labour — is consistent with the fact that headcount at head office rose 14% while store-level labour is not the company's cost to bear.

What the mainland business does not have is a pricing lever. The core menu is priced between RMB 2 and RMB 8 and the company's entire brand proposition is value for money. Management held consumer prices broadly flat through the 2025 lemon-cost shock and the 2026 ingredient-quality upgrade, absorbing cost at the gross-margin line rather than passing it to the consumer. In a market where every competitor is also discounting, that is probably the right decision. It also means the only routes to higher revenue per store are volume and mix — more transactions per day, or more of them at the higher end of the menu. Both take years.

07 — Brand: Lucky CupThe coffee brand that quietly became the third ten-thousand-store chain in China

Lucky Cup was launched in 2017 as MIXUE's freshly-made coffee brand and spent its first seven years as a footnote. In 2025 it became the most important growth vector the company has. On 24 November 2025 Lucky Cup announced that it had passed 10,000 stores, completing a target set at the start of that year roughly a month early. A year earlier it had about 4,600. That makes it the third chain in China to reach ten thousand coffee outlets, after Luckin Coffee and Cotti Coffee, and it did so at price points below the RMB 9.9 that those two use as their promotional anchor: core Lucky Cup products are priced between RMB 2 and RMB 10.

The strategic logic is straightforward and, in our view, sound. A MIXUE store already has the lease, the staff, the cold chain and the footfall. Adding a coffee machine extends the trading day from the afternoon tea occasion into the morning commute, and it uses the same cup, the same counter and the same delivery rider. The company's stated approach for 2026 is "fresh beans, fresh milk, fresh fruit, made fresh": coffee beans roasted within 60 days, fresh milk rather than powder, seasonal fruit from specific origins — Thai aromatic coconuts, Xinjiang Aksu dried apricots — and High Pressure Processing technology for fruit preparations. Signature and pour-over coffee have been introduced in pilot markets.

Two caveats belong in the record. First, the company does not disclose Lucky Cup's revenue, profit, store-level economics or capital employed. Everything above is disclosure about product and store count. The reader therefore cannot tell whether the 10,000 stores are generating incremental group margin or whether they are a way of amortising an existing supply chain over a larger order book. Second, the coffee market Lucky Cup is entering is the most brutally competitive segment of Chinese consumer goods. Luckin and Cotti both operate at or near scale with heavy discounting; the coffee price war of 2024-2025 pushed promotional prices to RMB 9.9 and below across the category. Lucky Cup's price points of RMB 2 to RMB 10 undercut that, which is consistent with MIXUE's brand but leaves very little room for error on input costs.

08 — Brand: FULU Fresh BeerRMB 297 million, 53% of a beer chain, and a third growth curve

On 1 October 2025 MIXUE announced that it would acquire 53% of FULU Fresh Beer — 鲜啤福鹿家 — for approximately RMB 297 million, through a combination of capital injection and share transfer, from Fulu Family (Zhengzhou) Enterprise Management. The transaction completed on 1 December 2025. It was the company's first material acquisition, and it gave MIXUE a third brand in a third category: fresh beer, sold on tap, priced at approximately RMB 6 to RMB 11 per 500ml.

The strategic case is the same as the one for Lucky Cup, taken one step further. Fresh beer requires exactly the capability MIXUE already has and its competitors do not: a chilled and frozen logistics network running daily routes to small outlets. Beer is a lower-frequency, higher-ticket, evening occasion than tea, which extends the trading day in the other direction. The portfolio — classic fresh beers, fruit beer, tea beer, milk beer, a five-malt classic and a zero-alcohol sparkling series launched in the first half of 2026 — is deliberately broad and cheap.

The operating numbers reported since the acquisition are impressive and should be treated with care in equal measure. Press reports in mid-2026 described FULU's store count roughly doubling within about six months of the deal to pass 3,000 outlets, and one account described the network as having expanded roughly tenfold in a year. The base is small, the ramp is off a near-standing start, and the franchise economics of a fresh-beer outlet in a Chinese city are unproven at scale. FULU also brought 36 self-operated stores onto the group balance sheet — the number rose from 18 at 30 June 2025 to 36 at 30 June 2026 — which is the only place in the accounts where the acquisition is directly visible, alongside RMB 91.9 million of goodwill and RMB 168.8 million of other intangible assets.

There is a legitimate strategic argument that a third brand dilutes rather than compounds management attention. The counter-argument is that all three brands share one procurement organisation, one cold chain and one franchise support function, so the marginal cost of the third brand is low. Our own view is that FULU is best valued as an option. It cost 1.4% of the company's liquid resources, it is not material to any consolidated line, and the company has not disclosed enough about it to underwrite anything more than that.

09 — Store economicsThe arithmetic of one shop, and why it is the whole investment case

Everything in this report eventually reduces to a single number: the revenue a MIXUE store generates per period. The company's revenue is a derivative of it, its franchisees' survival depends on it, and its own margin structure is set by the volume it needs to keep six factories and 31 warehouses running at reasonable utilisation.

Because MIXUE reports revenue at the wholesale level and store count at period end, the implied wholesale revenue per store can be calculated by dividing revenue by the average of the opening and closing store count. On that basis the figure was RMB 610,100 in FY2023, RMB 591,000 in FY2024 (down 3.1%), RMB 631,700 in FY2025 (up 6.9%), and an annualised RMB 491,800 in the first half of 2026 — down 22.1% on FY2025 and 19.4% below the 2023 level. The first half of 2025, on the same basis, was RMB 598,200 annualised, so the year-on-year decline is 17.8%.

Two adjustments matter before drawing a conclusion. The first is the base effect. The first half of 2025 was the peak of the instant-retail subsidy war, in which the three large delivery platforms subsidised consumer orders aggressively to build share in the newly competitive "instant retail" category. Freshly-made drinks were among the biggest beneficiaries because they are standardised, high-frequency and cheap. MIXUE's revenue grew 39.3% in that half against a store-count increase of 22.7%. Some of the 2025 revenue per store was subsidy, not demand. The second is that the annualisation of a half-year flatters seasonality: the first half contains the spring and early summer, which is the strong season for cold drinks in China, so doubling the first half overstates rather than understates the full-year run rate. Both adjustments cut the same way: the true underlying decline is real but smaller than 22%, and larger than zero.

The company's own disclosure gives a sense of the direction of travel over a longer horizon. Its prospectus showed average single-store terminal sales falling from RMB 1.1353 million in the first nine months of 2023 to RMB 1.0827 million in the first nine months of 2024, a decline of 4.63%, with average cups per store falling from 177,200 to 170,700. Since listing, the company has not published either metric. At the first-half 2026 results call the chief executive confirmed a double-digit percentage decline in average store turnover and attributed it to the high 2025 base and to competition. We take that statement at face value, and we note that it is the first time management has voluntarily quantified the trend.

For the franchisee, the arithmetic is tighter than the group's. On the prospectus figures the average store generated roughly RMB 4,184 of terminal sales a day in the first nine months of 2024. Industry analysis puts the daily breakeven for a MIXUE franchisee — covering rent, wages, utilities and the cost of goods bought from head office — at approximately RMB 3,000. The buffer was therefore around 28%. If store turnover has fallen by a mid-teens percentage since then, the buffer is roughly 12%, before any increase in rent or wages. That is thin. It is also the most plausible explanation for the rise in closures from 1,609 in 2024 to 2,527 in 2025, and it is why the company is now subsidising franchisee equipment rather than simply selling it to them.

10 — The saturation question63,987 stores into a market that lost 43,200 of them last year

The Chinese freshly-made drinks market stopped growing in store-count terms in 2025. Data from the restaurant-location provider Zhuanmen Canyan, cited across the Chinese trade press, put the national tea-store count at 383,100 in the first half of 2026, a net reduction of 43,200 over twelve months, with more than 110,000 individual closures and a category closure rate of 16.2%. Roughly one store in six stopped trading. The industry association's own white paper for 2026 describes the market as having moved from rapid expansion into a "stock cultivation" phase in which the competition is over efficiency rather than scale.

MIXUE added 4,202 net stores in that half, which is 20.7% more than it had a year earlier and 9.7% of the entire net contraction of the category. Two readings are possible. The generous one is that MIXUE is taking share from weaker brands — a rational consolidation in which the low-cost operator survives and the marginal operator closes. The cautious one is that MIXUE is adding stores into a market that is shrinking around it, and that its own incremental stores are a meaningful part of the reason its existing stores sell less. Both readings are consistent with the disclosed data, and the company's own admission that average store turnover fell by a double-digit percentage while the network grew 20.7% is more consistent with the second than the first.

The forward-looking question is how much room is left. MIXUE's stated 2026 approach is "prudently expanding its store network" while "improving store performance", and its guidance for the Chinese mainland is to expand "into untapped markets and deepen its presence in existing ones". The company has already said it covers every county-level city in the country. That leaves two kinds of incremental location: additional stores in places that already have one, and additional stores in places so small that they cannot support one. Neither is a growth market in the sense the equity was underwritten for in March 2025.

It is worth being precise about what the first half of 2026 actually showed. The company opened 5,455 stores and closed 1,289. Annualised, that is 10,910 openings and 2,578 closures — a gross opening rate of about 18% of the network and a closure rate of about 4%. The network is not shrinking. It is churning faster, at a lower net rate. For a franchise system, a rising closure rate is a lagging indicator of falling franchisee returns, and the closures we are seeing in 2026 were decided by franchisees whose returns were deteriorating through 2025.

11 — Churn and the disclosure gapWhat the company stopped publishing, and why it matters

The most informative single series in this report is one that the company no longer publishes. MIXUE disclosed average single-store terminal sales and average cups sold per store in its prospectus, for the first nine months of 2023 and of 2024. It has not disclosed either since it listed. It does not publish same-store sales growth, which is the standard metric for a retail or restaurant chain and the one that would immediately answer the question this report is trying to answer. It does not disclose revenue by brand, revenue by geography, or the number of stores by brand. It reports as a single operating segment, which the exchange permits but which materially reduces the information available to investors.

This is not a criticism of the company's compliance. MIXUE discloses more than the rules require in some areas — the city-tier breakdown of the mainland network, the franchisee count, the openings-and-closures table, the smart-dispenser penetration — and it is a well-governed issuer with an Ernst & Young review of its interim statements and an audit committee of three independent non-executive directors. But the specific metrics that would allow an investor to distinguish "we are adding stores in new towns" from "we are cannibalising our own network" are absent, and they were present before the IPO. That asymmetry is worth noting. So is the fact that the store-movement table the company does publish is the closest available substitute, and that it is deteriorating.

The peer comparison in the chart above is instructive. Across the six listed groups, first-half net additions fell from 8,546 to 7,099, a 16.9% reduction. Guming — the second-largest and the best-performing operator in the group — cut its first-half additions by 37% and let its closures rise 70.8%, explicitly describing the change as a decision to protect the profitability of existing franchisees by clearing out inefficient stores. Chagee cut additions by 69% and shifted its centre of gravity abroad. Only Auntea Jenny accelerated, and it did so while extending credit terms to franchisees, with trade receivables up 108.5%.

MIXUE's own additions fell 35.7%, which is a deliberate slowing. But it is still adding more stores than the other five groups combined, and it is the only one of the six whose profit declined by double digits. The industry has concluded that the correct response to saturation is to stop adding stores. The largest player in the industry has slowed but not stopped. Whether that is discipline or inertia will be settled by the closure rate over the next four quarters.

12 — Gross margin and the fresh pivotThirty percent, and falling, for the fifth year in a row

MIXUE's gross margin has not exceeded 33% in five years. It was 31.34% in FY2021, 28.34% in FY2022, 29.55% in FY2023, 32.46% in FY2024, 31.14% in FY2025 and 30.4% in the first half of 2026, down 120 basis points year on year. The company gives two explanations for the recent decline, and both are honest: a change in revenue mix, and rising raw-material procurement costs. The first half of 2026 adds a third: a deliberate increase in cost of sales of 4.1% against revenue growth of 2.3%, which the company attributes to "strategic investments to further enhance product quality in line with our focus on real ingredients, fresh taste, and simple recipes".

That programme is the most consequential strategic decision the company has made since listing. In substance it converts the ingredient base from ambient-temperature products — powders, syrups, shelf-stable concentrates, canned fruit — to chilled and frozen equivalents: fresh milk instead of milk powder, frozen fruit instead of canned or ambient, chilled tea and condiments instead of shelf-stable. In the first half of 2026 the company expanded direct sourcing for fruit, milk, tea and coffee to the farm level, invested in ripeness management and post-harvest pre-cooling, built new production lines and retrofitted existing ones, and started up its Yunnan production base in May 2026 so that key fruit ingredients such as passionfruit can be processed close to where they are grown.

The financial consequence is that MIXUE is voluntarily giving up the cost advantage that ambient ingredients provided. Chilled and frozen logistics cost more per unit than ambient; frozen fruit costs more than canned; fresh milk costs more than powder. The company is accepting a lower gross margin in exchange for a better product, on the theory that consumers will notice, that value-for-money at a higher quality level is a stronger position than value-for-money at a lower one, and that the supply-chain capability required is one its competitors cannot replicate quickly. It is spending RMB 1.6 billion on the upgrade in 2026.

We are sympathetic to the strategy and sceptical about the timing. The sympathetic case is that the Chinese freshly-made drinks market is saturated with look-alike products, and the only durable differentiators are supply chain and brand. MIXUE has both. A company that can deliver fresh fruit to a RMB 4 drink in a county town has a moat that a company selling powder cannot cross. The sceptical case is that the upgrade is arriving at the exact moment when per-store revenue is falling, franchisee buffers are thinning and the company is subsidising franchisee equipment. Cost increases are easier to absorb when volumes are rising. MIXUE is choosing to raise its cost base in a year when its revenue is flat, and it is doing so with a three-to-five-year payback at best.

There is also a disclosure question. The company has not quantified the gross-margin impact of the fresh pivot, nor given a timeline for when it expects the investment to annualise. Consensus expects gross margin of 29.99% for FY2026 — a further 115 basis points of decline — which implies the market expects the pressure to continue through the second half rather than to reverse.

13 — Cost structure and the Snow King billEvery line grew faster than revenue except the one that matters most

The first half of 2026 is the cleanest illustration of the cost problem because the revenue line is almost flat, which removes the usual ambiguity about operating leverage. Revenue rose 2.3%. Cost of sales rose 4.1%. Selling and distribution expenses rose 22.9% to RMB 1,122.9 million, taking them from 6.1% of revenue to 7.4%. Administrative expenses rose 39.4% to RMB 610.0 million, taking them from 2.9% to 4.0%. Research and development expenses fell 1.6% to RMB 40.4 million and remained at 0.3% of revenue. Finance costs rose 68.1% to RMB 3.5 million, on additional lease liabilities for warehouses.

The company attributes the selling and distribution increase to "brand IP development initiatives and enhanced support for high-quality store operations" — in plain terms, Snow King. The Snow King content machine now includes two animated series, a three-volume comic series published in June 2026, featured merchandise including building blocks, figurines and plush toys, parade floats, festive events, pop-up stores, brand ambassadors and crossover collaborations with properties including Journey to the West and Tang dynasty sancai ceramics. The administrative increase is attributed to staff costs: headcount rose from 9,102 at the end of 2025 to 10,374 at 30 June 2026, an increase of 14% in six months, and employee benefit expenses rose from RMB 897.8 million to RMB 1,000.1 million.

The bridge above is the whole profit story in one picture. Of the RMB 399 million decline in first-half profit, RMB 382 million — 96% — came from the two operating expense lines. Gross profit contributed only RMB 77 million of the decline. Tax was a RMB 94 million offset, purely because profit fell. There is no revenue collapse, no impairment, no one-off charge, and no accounting artefact. The company spent more on marketing its mascot and on head office staff than it earned in incremental revenue.

That is not necessarily a mistake. Snow King is a genuine asset — it is the only iconic IP in the Chinese freshly-made drinks industry, it is licensed into merchandise, and the company has stated an intention to extend it into animation, film and theme parks. A content business costs money before it makes money. But the honest way to read the first half of 2026 is that MIXUE is financing a brand-and-quality transition out of its own income statement at the same time as its per-store economics are deteriorating, and that it has chosen to protect consumer prices and franchisee terms rather than its reported margin. For a company with RMB 21.6 billion of net cash, that is an affordable choice. It is not a free one.

The research and development line deserves a specific note, because it is the smallest line item in the company and the one the strategy most depends on. MIXUE spends RMB 40 million a half — 0.3% of revenue — on R&D. The company describes this as covering both application R&D for flavours and recipes and fundamental R&D for the technologies, production techniques and equipment that support its ingredients. For a business whose entire competitive claim is that its supply chain cannot be replicated, a research budget of 30 basis points of revenue is thin. It is one of the few places in this file where the rhetoric and the spending are not obviously aligned.

14 — Quality of earningsFour things in the accounts that the headline does not show

Reported profit at MIXUE is unusually clean by the standards of Chinese consumer listings. There are no material related-party transactions disclosed, no pledged assets, no contingent liabilities, no material litigation, no going-concern language and no non-GAAP reconciliation to be suspicious of, because the company does not publish an adjusted earnings figure at all. The interim statements were reviewed by Ernst & Young under HKSRE 2410. Four items nonetheless deserve a reader's attention.

First, the float. The company collects cash from franchisees before it delivers goods. Contract liabilities — prepayments received — were RMB 594.0 million at 30 June 2026 against RMB 473.1 million at 31 December 2025, a 25.6% increase in six months. That is the right direction and it is a genuinely high-quality liability, but it is also a reminder that the group's cash generation is partly a function of working capital timing rather than of profit. Trade payables fell from RMB 2,212.5 million to RMB 1,859.9 million over the same period, which is a RMB 352.6 million cash outflow — the company paid its suppliers faster. Inventory rose from RMB 3,673.0 million to RMB 3,828.2 million, a 4.2% increase against 2.3% revenue growth, which is mild but in the wrong direction for a business whose products are perishable.

Second, the foreign-exchange loss. Other income and gains, net, fell 19.6% to RMB 127.5 million, primarily because foreign-exchange losses on foreign-currency deposits rose from RMB 42.2 million to RMB 153.8 million. The company holds the HK$3,799 million of IPO proceeds in Hong Kong dollars while its functional currency is the renminbi, and the renminbi appreciated through the period. This is a real economic loss and it is properly recognised, but it is also the kind of line that reverses: the same exposure that produced a RMB 154 million charge in the first half of 2026 would produce a gain if the renminbi weakened. It is not a reason to change a view on the business, and it is a reason to be careful about extrapolating the other-income line.

Third, the impairment charge. The company recognised an impairment of property, plant and equipment of RMB 5.0 million — small in absolute terms, and the first such charge in the record. It appeared alongside a reversal of RMB 0.3 million on prepayments and other receivables and a RMB 1.8 million net impairment of trade receivables. None of these is material. The significance is directional: a company that is adding 5,455 stores a half-year and simultaneously impairing its own property and equipment is telling you that some of its assets are not earning their cost of capital.

Fourth, the absence of an adjusted number. MIXUE discloses no non-IFRS earnings measure, no same-store sales growth, no like-for-like store data and no brand-level profitability. Its peers do: Guming reported a 44.4% increase in non-IFRS core profit against a reported 3.4% decline; Chagee's reported 22.7% profit increase reverses to a 23.9% decline once share-based compensation is excluded. MIXUE's decision not to publish an adjusted figure is conservative and makes comparison harder in both directions. It also means that the reported decline of 14.7% is the number, with no adjustment available to soften it.

15 — Balance sheetRMB 21.6 billion of net cash, no borrowings, and a 17.7% gearing ratio

The balance sheet is the single most reassuring document in this file, and it is the reason our rating is Neutral rather than Cautious. At 30 June 2026 MIXUE held RMB 21,640.7 million of cash and cash equivalents, time deposits, restricted cash and wealth-management products included in financial assets at fair value through profit or loss. That is up 8.3% from RMB 19,990.0 million at the end of 2025 and up 79.9% from the end of 2024. The company has no interest-bearing bank borrowings at all — the RMB 28.2 million outstanding at 31 December 2025 was repaid — and its gearing ratio, defined as total liabilities over total assets, fell from 19.6% to 17.7%.

Total assets were RMB 32,874.6 million and total liabilities RMB 5,811.3 million, giving net assets of RMB 27,063.3 million, of which RMB 26,731.0 million was attributable to owners of the parent. That is a book value of RMB 70.42 per share, or about HK$82.4 at the current exchange rate — 48% of the share price, with almost none of it in intangibles. Goodwill is RMB 91.9 million and other intangible assets RMB 168.8 million, together 0.8% of total assets. Property, plant and equipment is RMB 5,489.5 million, right-of-use assets RMB 543.4 million, and inventories RMB 3,828.2 million. This is a hard-asset balance sheet with no acquisition-related goodwill of any consequence.

The composition of the liquid resources deserves a note. Of the RMB 21.6 billion, RMB 6,027.9 million is cash and cash equivalents, RMB 5,810.2 million is current time deposits and restricted cash, RMB 1,716.6 million is non-current time deposits and restricted cash, and RMB 8,096.1 million sits in financial assets at fair value through profit or loss — substantially wealth-management products. The company states that net proceeds not immediately required are placed as short-term deposits with licensed banks or financial institutions. Wealth-management products issued by Chinese banks and trust companies are not risk-free instruments, and their classification at fair value means the carrying value moves with market conditions. In a company with a market capitalisation of RMB 55.3 billion, RMB 8.1 billion — 14.6% of market capitalisation — in Level-2 or Level-3 fair-value instruments is a governance question worth asking at the next annual general meeting.

Two further observations. First, the company's capital commitments are modest: RMB 204.5 million at 30 June 2026, down from RMB 301.8 million at the end of 2025, for building production factories and purchasing production facilities. That is a remarkably small number for a company that describes itself as being in a supply-chain investment cycle, and it suggests that the RMB 1.6 billion of 2026 upgrade spending is largely through the income statement rather than through capitalised assets. Second, there are no pledged assets, no charges over assets and no material contingent liabilities. The company told the exchange exactly that.

16 — Capital allocation & the first dividendSeventeen per cent of last year's profit, and a signal about what comes next

On 27 August 2026, alongside the interim results, the board proposed a special dividend of RMB 2.65 per share. On 11 September 2026 the company confirmed the amount, and on 30 September 2026 an extraordinary general meeting approved it: 342,789,315 votes for, 1,000 against and 900 abstentions, a 99.9994% approval on a 90.30% turnout. The record date is 12 October 2026, the payment date is 6 November 2026, and the total is approximately RMB 1,005,990,000 on the 379,618,800 shares in issue. Declared in Hong Kong dollars, the amount is HK$3.08304 per share — a yield of 1.81% at the current price, and an implied HKD/RMB rate of 0.8595.

This is the first distribution MIXUE has made since listing. The company paid nothing in FY2025 and the interim statements confirm that no dividend was paid or declared in the six months to 30 June 2026. The special dividend is therefore not a payout-ratio decision so much as a statement about the capital the business needs. On FY2025 profit of RMB 5,927.0 million it represents a 17.0% payout; on consensus FY2026 free cash flow of RMB 3.92 billion it represents 25.7%. The company has said nothing about whether a dividend will become recurring, and it has not announced a policy.

The strategic read is that management has concluded it has more cash than it can deploy at an attractive return in its existing business. That is consistent with everything else in this report: a company whose store network is close to saturation in its home market, whose overseas network has shrunk for two years, whose acquisition pipeline consists of RMB 297 million for a beer chain, and whose capital commitments are RMB 205 million. When a business with a 30% gross margin and 21.6% operating margin cannot find a productive use for RMB 21.6 billion of cash, the correct answer for shareholders is to take some of it back. The company is doing that. The open question is whether a one-off distribution of 17% of profit is the beginning of a capital-return policy or a gesture.

Our expectation, and it is an expectation rather than a disclosure, is that it becomes recurring at a modest level. Consensus already models total FY2026 dividends per share of RMB 4.00, which is RMB 2.65 of special dividend plus roughly RMB 1.35 of final dividend. If that is right, the total yield at the current price would be about 2.7%. For a company with no debt, RMB 21.6 billion of liquid resources and a business that generated RMB 1.65 billion of net liquidity in six months, that is affordable. It is also a meaningful change in the character of the equity: MIXUE is becoming a cash-return story rather than a growth story, and the market has not yet decided what multiple to put on that.

17 — Overseas I: South-East AsiaSeven years of expansion, and two years of contraction

MIXUE opened its first overseas store in Hanoi, Vietnam, in September 2018. It entered Indonesia in March 2020 and subsequently Thailand, Malaysia, Laos, Singapore, South Korea in 2022, and then Australia and the Americas. By 30 September 2024 it had 4,768 stores in South-East Asia, which was 99.5% of its overseas estate and made it, by store count and cups sold, the largest freshly-made drinks brand in the region. It entered Kazakhstan and the United States in 2025, and Mexico, Kyrgyzstan and Brazil in the first half of 2026. It now trades in 17 countries.

The store count has gone the other way. The overseas network peaked at 4,895 stores at the end of 2024, fell to 4,733 at 30 June 2025, 4,467 at the end of 2025 and 4,378 at 30 June 2026 — a contraction of 10.6% from the peak, and a fall in every single reporting period, even as the group added 17,508 stores worldwide. Overseas was 10.5% of the network at the end of 2024 and is 6.8% today.

The company's explanation is that this is deliberate. In both the FY2025 annual results and the first-half 2026 interim announcement it states that it is "optimising the operations of our existing stores" in Indonesia and Vietnam, which together were more than 70% of the overseas estate, and that "while the number of stores in these two countries declined during the Reporting Period, store operational quality improved significantly". At the FY2025 results briefing management said the new stores opened under the revised standard generate about 1.7 times the revenue of the old ones, and press reports of the FY2025 disclosure stated that same-store sales in the two markets rose more than 50% after the clean-up.

We accept the framing and note the size of the hole it leaves. If the old stores in Indonesia and Vietnam were generating roughly the group average, closing 10.6% of the overseas estate removes revenue that the company has never separately disclosed. MIXUE has operated internationally for seven years and has not once published overseas revenue, overseas profit or overseas capital employed. The reader is therefore asked to accept that a 10.6% reduction in the international network is value-accretive on the strength of a same-store-sales assertion and no financial disclosure. That may well be true. It cannot be verified.

The external evidence is mixed. Vietnamese press coverage in 2026, summarised by the Central News Agency, described MIXUE's expansion in Vietnam as having stalled for reasons that are structural rather than cyclical: franchise outlets clustered too densely during the peak expansion, creating internal competition; Vietnamese consumers shifting from price toward store experience, ingredient quality and brand image, which erodes the value of low price and standardisation as a differentiator; rising urban rents and operating costs that a low-price format cannot pass on; and the rapid emergence of local chains that adapt menus faster. The same coverage noted that the company's revenue model — selling ingredients to partners rather than earning retail margin — makes it more of a food-service supply chain than a drinks brand, which is precisely the point this report makes about the domestic business.

The overseas supply chain is being built in parallel. The company had local warehouse systems and delivery networks in eight overseas countries at the end of 2025 and nine at 30 June 2026, and extended its cold-chain network to stores in Malaysia and Vietnam during the first half. That is the necessary precondition for the fresh-ingredient programme to travel, and it is a genuine investment. It is also an admission that the ambient-temperature model that worked in Indonesia and Vietnam in 2020 does not carry the brand position the company now wants.

18 — Overseas II: the new mapKazakhstan, the United States, Mexico, Kyrgyzstan and Brazil — in eighteen months

The contraction in South-East Asia has been accompanied by an unusually rapid sequence of new market entries. MIXUE entered Kazakhstan and the United States in 2025, and Mexico, Kyrgyzstan and Brazil in the first half of 2026. It now operates in 17 countries. Management's stated approach is to "deepen our presence in South-East Asia while penetrating new markets such as Central Asia and the Americas", and to formulate market entry decisions on population size, economic growth, income level, cultural characteristics and consumer preferences, refining them dynamically based on business environment and store performance in each country.

The pattern is worth stating plainly. MIXUE is simultaneously retreating from its two largest international markets and opening stores in five new ones across three continents within eighteen months. The company presents this as a portfolio approach. It is also what a business does when the original international thesis has stopped working: redeploy attention and capital to markets where the model has not yet been tested, and where the base is too small to be embarrassing. A single store in Brazil and a single store in Kyrgyzstan are not businesses; they are options, and the company is buying a lot of them.

The two entries that could matter are the United States and Mexico. The US market for freshly-made tea is real but fragmented, the labour and real-estate economics are an order of magnitude different from China's county towns, and the price point that MIXUE's model depends on — roughly one US dollar per item — is not achievable in a US city at US costs. Mixue's US stores have been reported to price above the Chinese equivalent, which breaks the brand's core promise and puts it into direct competition with established local and Chinese-American chains. Mexico and Brazil offer a closer demographic and income match, but also a franchise infrastructure, a cold chain and a regulatory environment that have to be built from nothing.

19 — The delivery-platform warThe distortion that created the 2025 base, and the re-rating it caused

The single largest exogenous event in MIXUE's short history as a public company was the instant-retail subsidy war of 2025. China's three largest delivery and instant-retail platforms competed aggressively for share in the newly contested "instant retail" category, subsidising consumer orders across food and beverage. Freshly-made drinks were among the principal beneficiaries: standardised, cheap, high-frequency, and immediately deliverable. The company's own annual report for FY2025 notes that the platforms increased subsidies, that orders migrated online, and that the subsidies raised consumers' expectations about the price-quality ratio of freshly-made drinks — a polite way of saying that the industry's customers were temporarily paying less than the products cost.

MIXUE's first-half 2025 revenue rose 39.3% against a 22.7% increase in the store count. Its full year revenue rose 35.2%. Its peers experienced the same windfall: across the six listed groups, first-half 2025 revenue growth was in the double digits almost everywhere. The category's operating margins, however, did not improve, because the platforms captured the incremental economics through commissions, delivery fees and advertising, and because the brands competed for the subsidised demand with their own discounts. Luckin Coffee's management said explicitly, on releasing its own annual results, that the rising share of delivery orders had changed its cost structure, with higher delivery costs and higher third-party platform commissions and advertising.

When the subsidies were withdrawn through late 2025 and 2026, the industry's growth rate collapsed. MIXUE's revenue growth went from 39.3% in the first half of 2025 to 2.3% in the first half of 2026. Chagee's same-store sales in China fell 16.0% in the first quarter of 2026 and 16.1% in the second. ChaPanda's active membership fell 36.2% year on year to about 35.9 million. Auntea Jenny's quarterly active membership was 1.10 million lower at mid-2026 than at the start of the year despite a 56.1% increase in selling and marketing spend.

The equity market's response has been a wholesale re-rating of the sector, and it is worth understanding the sequence because it explains where MIXUE's share price is today. In 2025, four of the six groups listed within a four-month window. MIXUE's market capitalisation passed HK$230 billion; Guming rose almost 150% above its issue price within months; Auntea Jenny's public offer was 3,600 times subscribed. By 21 September 2026, five of the six had broken issue. MIXUE's market capitalisation was a little over HK$70 billion — a fall of nearly HK$160 billion from the peak — and by 8 October 2026 it was HK$64.73 billion, down 54.7% year to date and 63.4% below its 52-week high of HK$465.80.

The important point is that the de-rating was not caused by a collapse in fundamentals. Across the six groups, first-half 2026 revenue rose about 9% and aggregate net profit exceeded RMB 5.3 billion. The de-rating was caused by the market withdrawing the growth premium it had assigned to a model that it now believes cannot be replicated indefinitely. Aggressive store expansion used to be the value driver: more stores meant more procurement scale, more supply-chain coverage, more brand presence, and therefore more franchisees. In a saturated market the same mechanism runs in reverse. That is the re-pricing MIXUE is living through, and it is the correct frame for the valuation work in section 24.

20 — The competitive mapSix listed groups, one of them twice the size of the rest, and none of them earning much

The Chinese freshly-made drinks industry now has six listed groups: MIXUE, Guming, Chagee, ChaPanda, Auntea Jenny and Nayuki. All six have reported for the six months to 30 June 2026, which makes this the first period in which a clean comparison is possible. The picture is of an industry that has stopped growing at the top and is fragmenting in the middle.

MIXUE's revenue of RMB 15.22 billion is more than twice Guming's RMB 7.47 billion, which is more than twice Chagee's RMB 6.96 billion, which is more than twice ChaPanda's RMB 2.66 billion. In profit terms the ordering is different: MIXUE earned RMB 2.30 billion of attributable profit, Guming RMB 1.57 billion, Chagee RMB 918 million, ChaPanda RMB 336 million and Auntea Jenny RMB 321 million, while Nayuki lost RMB 97 million. MIXUE is 2.0 times Guming's size by revenue and 1.5 times its size by profit.

The growth rates tell the more interesting story. Auntea Jenny grew revenue 42.4% and profit 58.3%, the best in the group, largely by adding 1,706 net stores in six months — more than it added in the whole of 2025 — while extending credit terms to franchisees, with trade receivables up 108.5%. Guming grew revenue 31.9% and, on a non-IFRS basis, core profit 44.4%, with gross margin up 190 basis points to 33.4% and every expense line growing more slowly than revenue: a textbook demonstration of operating leverage. Chagee grew revenue 3.5% but its reported profit increase of 22.7% reverses to a 23.9% decline once share-based compensation is stripped out, because its franchise revenue fell 15.4% while its self-operated revenue rose 214.7% at a cost that exceeded the incremental revenue.

MIXUE's position in that table is uncomfortable. It grew revenue 2.3% — second-worst of the six — and profit fell 14.7%, the worst of the six by a wide margin. It is the only company in the group whose selling and distribution expenses grew faster than revenue and whose administrative expenses grew faster than revenue. It is the only one of the six where both expense lines expanded simultaneously. It is, in other words, the only large player in the industry that is running its cost base up while its top line stalls.

The margin comparison is where the scale argument breaks down. Guming earned a net margin of 20.99% in the first half of 2026. MIXUE earned 15.24%. Chagee earned 13.11%, ChaPanda 12.99%, Auntea Jenny 12.41% and Nayuki lost 5.16%. MIXUE is more than twice Guming's size and earns 5.75 percentage points less on every renminbi of revenue. There are legitimate structural reasons — a national chilled and frozen logistics network is expensive, a fresh-fruit-heavy product mix is expensive, a franchise network four and a half times larger is expensive to support — and the company is deliberately buying margin in exchange for product quality. But it is worth stating without qualification: in this industry, scale does not produce margin, and the company with the smallest scale in the top three produces the best one.

Two further competitive facts deserve a place in the record. First, the industry's growth has moved abroad for those who can: Chagee's overseas stores passed 200 and its international business is now its strategic priority; ChaPanda had 72 overseas stores and Auntea Jenny 61 at 30 June 2026. MIXUE's 4,378 is not a like-for-like comparison because most of them are in South-East Asia at MIXUE price points, but it remains by a wide margin the largest international network in the group. Second, every one of the six is now selling coffee. MIXUE's Lucky Cup passed 10,000 stores in November 2025; Guming has coffee machines in about 13,500 stores, 94% coverage, with coffee at more than 20% of sales outside promotions; ChaPanda has machines in about 30% of mainland stores; Auntea Jenny has upgraded machines in more than 9,000 stores. The category has converged on the same answer to the same problem, which means the advantage will not be in having coffee but in being able to make it cheaply.

21 — Competition II: coffeeThe second front, and why the price war moved down rather than up

Coffee is the category the Chinese tea chains did not invent, did not particularly want, and could not avoid. It arrived from the other direction — from Luckin and Cotti, who spent 2023 and 2024 teaching a generation of Chinese consumers that a freshly made coffee should cost RMB 9.9, and then discovered that the price could not be put back. By the time the tea chains responded, the floor had already been set. What MIXUE did with that floor is the subject of this section, because coffee is now the second front on which the entire competitive map is being decided, and it is the one place where MIXUE's cost advantage is genuinely under test.

The vehicle is Lucky Cup (幸运咖), the coffee brand MIXUE launched in 2017 and repositioned in 2020 around a single proposition: fresh-ground coffee at a price below anything Luckin would match. It passed 10,000 stores on 24 November 2025, up from roughly 4,600 at the start of that year — a net addition of about 5,400 outlets in under twelve months. That is a faster store rollout than MIXUE's own core brand achieved in any comparable window, and it makes Lucky Cup the third ten-thousand-store coffee or tea chain in China by outlet count. The price architecture is aggressive even by MIXUE's standards: an iced americano at RMB 5, a latte at RMB 8 to 9, with promotional pricing that occasionally touches RMB 3.9. Inside MIXUE's own stores, the same coffee is sold as a "fresh coffee" line at RMB 5 to 6, dispensed from automatic machines installed in more than 3,000 outlets by mid-2026 — press estimates put the installed base closer to 6,000 — which converts an existing store's afternoon counter into a second daypart at almost no incremental rent.

The strategic logic is sound and worth stating precisely, because it is not the logic the market assumes. Coffee is a morning category; tea is an afternoon and evening category. Adding coffee to a MIXUE store does not cannibalise the tea business so much as extend the hours in which the same real estate, the same lease, the same franchisee and the same logistics network are productive. The marginal cost of the second daypart is a machine, a bag of beans and a barista module in the training manual. On that arithmetic, even a low-margin coffee sale is accretive to store-level return on invested capital, which is the metric the franchisee actually cares about. It is the same reasoning that led MIXUE into ice cream and into beer, and it is why the company treats coffee as an attachment rather than a business.

The competition has converged on the identical answer, which is the problem. Guming has installed coffee machines in roughly 13,500 stores — about 94% coverage — with coffee now accounting for more than 20% of sales outside promotional periods. ChaPanda has machines in around 30% of its mainland estate. Auntea Jenny has upgraded machines in more than 9,000 stores. Every one of the six listed groups has reached the same conclusion, and they have reached it simultaneously. When a capability becomes universal, it stops being an advantage and becomes a cost of staying in the game; the only question that survives is who can deliver it cheapest. That is a question MIXUE usually wins, but not on this particular battlefield.

The reason is the input. MIXUE's structural advantage in tea rests on vertical integration in a fresh-fruit supply chain it largely controls — its own lemon orchards, its own processing, its own chilled logistics. Coffee beans are a globally traded commodity with a small number of very large roasters and a price set in New York and London. MIXUE buys beans at scale, but so does Luckin, and Luckin buys more of them. The company has no proprietary advantage in roasting, no unique origin, and no ability to move the bean price by the force of its own procurement. In coffee, MIXUE is a price-taker on the input and a price-setter on the output, and that is the least favourable position in the value chain.

There is a second, subtler exposure. Lucky Cup is a standalone brand with its own store network, its own franchisees and its own unit economics, and its price points are lower than MIXUE's tea average. Lower ticket means a lower gross-profit pool per store to cover the same rent, the same labour and the same franchisee return expectation. The breakeven sales threshold for a Lucky Cup outlet is therefore higher in volume terms than for a MIXUE outlet, and the franchisee base is younger and less capitalised. If the coffee price war intensifies — and there is no evidence it is ending — the marginal Lucky Cup franchisee is the most fragile economic unit in the entire group. That is not a forecast; it is an asymmetry worth pricing, and the market has not yet had to price it because Lucky Cup's losses are not disclosed separately inside a single operating segment.

We are not bearish on coffee. The category is growing faster than tea, the daypart extension is real, and 10,000 stores is a genuine achievement that gives MIXUE a national coffee footprint overnight. But the coffee business is best understood as a defensive move that protects store traffic rather than an offensive move that creates a new profit pool. It keeps MIXUE relevant in the morning; it does not add a second engine. Investors who are buying MIXUE for the coffee story are buying the least differentiated part of the company, and paying for it with the margin compression the coffee rollout itself helps to cause.

22 — Food safety and the licence to operateSixty-four thousand kitchens, one brand, and a residual liability the accounts cannot hold

Of MIXUE's 63,987 stores, 63,951 are franchised. Thirty-six are operated by the company. That single statistic defines the food-safety problem: the company earns its revenue from a supply chain it controls and its reputation from 63,951 kitchens it does not. Every franchisee is the licensed food-business operator at their own premises, but the brand that absorbs the consequences of a mistake is the parent's. There is no contractual arrangement that transfers a viral video back to the franchisee.

The recorded complaints are substantial and growing. On the Black Cat consumer-complaint platform (黑猫投诉), MIXUE's cumulative complaint count reached 11,216 by 29 March 2026, of which more than 70% relate to food safety — foreign objects, spoilage, hygiene in preparation. That is a cumulative figure across a network of this size and it must be read against roughly 5 billion cups a year, so the rate is low in absolute terms. But the direction and the composition matter. Complaints of this type are the leading indicator of the incidents that occasionally escape the platform and reach national attention.

Those incidents have a pattern. In 2022 a Changchun outlet was cited for expired ingredients; in 2023 an Xiamen store was exposed using expired materials; in 2024 a Beijing outlet featured in the annual consumer-rights broadcast over relabelled and expired ingredients, and the company issued a public apology and closed the store. In August 2026 a Nantong outlet was filmed with a member of staff smoking at the till; the store was closed and fined RMB 4,000. The financial penalty in every case is trivial — four thousand renminbi against a company with RMB 21.6 billion of liquid resources. The risk is not the fine. The risk is that the value proposition MIXUE sells is cheap and acceptable, and a single vivid incident attacks the second half of that promise, which is the half that is hardest to rebuild.

The company's mitigation is more serious than its critics allow. More than 18,000 stores now run smart dispensing equipment — up from 13,000 at the end of 2025 — that enforces recipe, portion and, critically, batch dating at the point of preparation, removing a large share of the human discretion that produced the historical incidents. The network is covered by in-store cameras, mystery-shopper audits and a franchisee training and certification regime, and food safety is embedded in the operational scorecard that governs supply terms. This is a genuine and expensive control system, and it is why the incident rate has not scaled linearly with the store count.

But the structural arithmetic is unforgiving. At 63,987 stores, an incident rate of one in ten thousand per store per year still produces six incidents a year that could each, on the wrong day, become a national story. The company is running a reputation business with a statistical tail it cannot eliminate, only shrink. And the mechanism that makes MIXUE's economics work — the franchise model — is precisely the mechanism that puts the kitchens outside its direct control. The company can monitor, train and penalise; it cannot be in the room. For an investor, this is the single largest un-hedgeable risk in the equity, and it belongs in the position-sizing decision rather than in a footnote.

23 — Management and the Zhang Yuan transitionA finance-trained outsider takes the wheel at the hardest moment in the company's history

MIXUE is a founder business. It began in 1997 as a shaved-ice stall in Zhengzhou run by Zhang Hongchao, and it grew into the world's largest freshly-made drinks network under the joint direction of the two Zhang brothers — Hongchao as chairman and Hongfu as chief executive. That structure survived the IPO in March 2025 and the first year of public-market life. It ended, in operational terms, on 24 March 2026, when Zhang Hongfu stepped down as chief executive and moved to a co-chairman role, and Zhang Yuan (张渊) was appointed chief executive.

Zhang Yuan's profile is the opposite of the founder's. Born around 1990, he holds a master's degree in finance from Tsinghua University and spent his pre-MIXUE career at Bank of America Securities and then Hillhouse Capital, before joining the company in 2023 as chief financial officer. He is, therefore, a capital-markets professional running an operations business — the first non-founder to hold the chief executive role, and the first occupant of that seat whose instincts were formed in valuation and financing rather than in store rollout and supply chains.

The timing is what makes the appointment interesting. MIXUE is entering the most difficult phase of its existence: the domestic store network is approaching saturation, same-store sales are falling, gross margin has compressed for five consecutive years, and the growth that remains is capital-intensive, low-margin and located in unfamiliar jurisdictions from Kazakhstan to Brazil. These are precisely the problems a finance-trained chief executive is hired to solve — capital allocation, cost discipline, portfolio triage — and they are also precisely the problems in which a founder's operational instinct is most valuable. The company has bet that the next phase is a spreadsheet problem rather than a store problem. That bet may be right; it is not obviously right, and it has not yet been tested by a full year of results.

What the transition has not changed is control. The company remains family-controlled, and the two brothers retain the chairmanship between them. The share register reflects the structure: 150,883,058 H shares are listed and freely traded, while 228,735,742 domestic shares remain unlisted, giving a total of 379,618,800 shares of which a clear majority is in family and insider hands. Minority holders own a claim on the cash flows; they do not own the decision rights. For a company that has just begun returning capital, that concentration is a double-edged inheritance — it insulates management from short-term market pressure, and it means the dividend policy is a family decision rather than a board negotiation.

The capital-allocation record to date is, on balance, disciplined. The March 2025 IPO raised approximately HK$3,799 million net and the company has carried no interest-bearing borrowings since listing. It has committed RMB 1.6 billion to a supply-chain upgrade and RMB 1 billion to a franchisee equipment subsidy in 2026 — both investments in the core network rather than diversifications away from it — and it declared its first distribution in 2026, a special dividend of RMB 2.65 per share. Headcount, however, tells a less comfortable story: it rose from 9,102 to 10,374, an increase of 14%, while revenue grew 2.3%. Some of that is the cost of building the next phase; some of it is the cost of a larger corporate centre. The market will want to see the ratio move the other way before it credits the new management with cost discipline.

24 — ValuationA no-growth price on a business that is still growing, and what the cash is worth

MIXUE's shares closed at HK$170.50 on 8 October 2026, giving a market capitalisation of HK$64.73 billion and a trailing price-to-earnings ratio of 10.2×. The stock is down 54.7% in the year to date and has fallen from a 52-week high of HK$465.80 to sit at the bottom of that range. Measured from the peak market capitalisation of more than HK$230 billion reached in the months after the March 2025 listing, the shares have lost roughly 72% of their value in about eighteen months. This is not a stock that has been gently de-rated. It is a stock that has been repriced from a growth multiple to a value multiple, and the question is whether it has overshot.

The starting point for any answer is that the market is now valuing MIXUE as a business that will not grow. Consensus revenue for FY2026 is RMB 33.50 billion, which is a decline of 0.2% on FY2025 — a flat year — and consensus earnings per share of RMB 13.29 implies a 15.1% fall in profit. On those numbers, the forward multiple is 10.2×, essentially identical to the trailing multiple, which is what happens when a company is priced for terminal stagnation. The street's FY2027 estimates are more constructive — revenue RMB 36.81 billion, up 9.9%, and EPS of RMB 15.17, up 14.1% — but the current price assigns them little weight. A stock on 10× forward earnings is a stock the market expects to disappoint.

The balance sheet changes the arithmetic materially, and it is the part of the MIXUE story that the market appears to have mislaid. At 30 June 2026 the company held RMB 21.64 billion of liquid resources — cash, current and non-current deposits and wealth-management products — against no interest-bearing borrowings whatsoever. Gearing was 17.7%, down from 19.6% at the end of 2025. Net cash therefore approximates the entire liquid balance, which at our reference rate of 0.855 renminbi to the Hong Kong dollar is roughly HK$25.3 billion, or HK$66.7 per share. That is 39.1% of the current share price sitting in cash. Strip it out and the operating business — the network of 63,987 stores, six production bases, the brand and the franchise system — is being valued at about HK$103.8 per share, or roughly 6.7× consensus FY2026 earnings. Investors are being asked to pay less than seven times earnings for the largest drinks network in the world, with the cash thrown in.

That is the bull case in a single sentence, and it is why we are not sellers. But a low multiple is not by itself an argument, and we are not buyers at this price either. The reason is that the operating business is, on the current trajectory, shrinking: per-store revenue annualised to RMB 491,800 in the first half of 2026, down 22.1% from FY2025 and 19.4% from FY2023. A business whose revenue per unit is falling at that rate does not deserve a growth multiple, and the market is right to have taken one away. What it does deserve is a fair multiple on a stabilised earnings base, and the disagreement is over whether RMB 5 billion of net income is a floor or a waypoint on the way down.

We model three scenarios. In the bear case, domestic same-store sales continue to fall, closures keep pace with openings, and net income settles at RMB 4.5 billion — EPS of roughly RMB 12.00 — which we capitalise at 8×, the multiple a no-growth consumer franchise deserves. That gives HK$115, only a third below the current price, because the cash floor and the earnings base both provide support. In the base case, the domestic network stabilises around 60,000 stores, overseas adds modestly, margin compression bottoms in 2027, and earnings recover to the consensus RMB 15.17 by FY2027. At 10.5× — a discount to Guming's multiple and a premium to nothing in particular — that is HK$185. In the bull case, the international network re-accelerates, coffee and beer contribute, the margin trough proves to be a trough, and earnings reach RMB 17.50 by FY2027, capitalised at 13×, for HK$265. Weighted 30/50/20 across bear, base and bull, the probability-adjusted value is HK$180.

The sensitivity of that base case to its two inputs is worth making explicit, because it shows how much of the outcome rests on the exit multiple rather than on the earnings. Holding the FY2027 earnings estimate at consensus, every turn of multiple is worth roughly HK$17.7 of value per share. A move from 8× to 13× — the full width of our scenario range — is a swing of HK$88, or more than half the current share price, without a single renminbi of earnings changing. In a stock with this much cash on the balance sheet and this much uncertainty in the terminal margin, the multiple, not the earnings, is the variable that will determine the return. Investors should be explicit that they are making a re-rating bet, not an earnings bet.

The final piece is how the components of value stack against the price. Of the HK$170.50 the market pays, HK$66.7 is cash, HK$93.1 is the capitalised value of the current earnings stream at 10×, and HK$10.7 is what remains as the market's entire estimate of future growth — the coffee rollout, the overseas expansion, the beer business and the brand's option value, all of it, for about 6% of the share price. The market is not merely sceptical about MIXUE's growth; it has assigned it approximately nothing. That is an unusually asymmetric setup. If the growth is worth anything at all, the stock is mispriced to the upside; if it is worth less than nothing, the cash and the earnings still put a floor under the downside.

Our call is Neutral, with a 12-month target of HK$185 — the base case, which happens to sit almost exactly on the probability-weighted fair value. The stock offers roughly 8.5% upside to target against 33% downside to the bear case, a payoff ratio of approximately one to four that does not compensate for the risks we have catalogued, and the arithmetic is made worse by the absence of any near-term catalyst: the margin trough is a 2027 event, the overseas inflection is unproven, and the next full-year print will still be lapping the inflated delivery-subsidy base of 2025. Against a street consensus target of HK$261.62 — implying 53% upside, with all 21 covering analysts rated Buy — we are materially more cautious, and the reason is simply that we do not believe the market is wrong to have removed the growth multiple until the company demonstrates, with disclosed same-store sales, that the decline has stopped.

25 — The bull caseThe seven arguments a buyer has to believe

We do not hold a bearish view of MIXUE; we hold a neutral one. The distinction matters, because the bull case here is not a story about a great company being cheap for no reason — it is a coherent set of arguments, and any investor considering a position should be able to state them and then decide whether the price already reflects them. These are the seven we consider strongest.

26 — The bear caseThe seven arguments a seller has to believe

The bear case is not a short thesis; MIXUE has no debt, no liquidity problem and a cash pile that would take years of losses to erode. It is a case about a business whose growth engine has stalled at exactly the moment the market stopped paying for growth, and about a management transition that has not yet been tested. These are the seven arguments we consider strongest, and we have weighted the last of them heavily in our Neutral rating.

27 — Risk matrixTwelve risks, ranked by probability times impact

The register below ranks the risks we consider material to the investment case. Probability and impact are scored from one to five, where one is remote or immaterial and five is near-certain or severe. The score is the product. A score of 16 or above means the risk is both likely and consequential and should be reflected in the price; a score below 8 means it is a tail to be monitored rather than a thesis. The distribution is unusually concentrated: four risks carry a score of 15 or higher, and three of those four concern the same underlying variable — the direction of revenue per store in mainland China.

28 — CatalystsWhat we are watching, and when

29 — ConclusionNeutral at HK$185, on a good business and a price that already assumes the worst

MIXUE is not a broken company. It is the largest freshly-made drinks network in the world, with 63,987 stores, six production bases, a vertically integrated fresh-fruit supply chain, no debt and RMB 21.6 billion of liquid resources. It earns a 30% gross margin at price points no competitor in the top six can profitably match, it grew revenue 35% in FY2025, and it has just begun returning capital to shareholders. Any account of the business that does not begin here is not an account of the business.

But the growth engine that produced those numbers has stalled. Revenue grew 2.3% in the first half of 2026, profit fell 14.7%, gross margin compressed for a fifth consecutive period, and revenue per store — the metric that determines whether a franchisee renews or closes — fell 22.1% against the prior full year. The domestic market is shrinking faster than MIXUE is growing into it, closures are rising faster than openings, and the company has responded by withdrawing the one disclosure that would let investors judge whether the decline is a cyclical hangover from the 2025 delivery-subsidy war or a structural descent into saturation.

The valuation is where the argument gets interesting. At HK$170.50 the market capitalisation is HK$64.73 billion, of which HK$25.3 billion — 39.1% — is net cash. Strip the cash out and the operating business is valued at roughly 6.7 times consensus FY2026 earnings. The market has assigned approximately 6% of the share price to all future growth: the coffee rollout, the beer business, the international expansion, the brand itself. That is a remarkably pessimistic reading, and it is the reason we are not sellers.

It is also not enough to make us buyers. The payoff is asymmetric in the wrong direction: roughly 8.5% upside to our base case of HK$185 against 32.6% downside to a bear case of HK$115, a ratio of about one to four that does not compensate for a food-safety tail the company cannot hedge, a management transition not yet tested by a full year, and a disclosure gap that prevents investors from verifying the single assumption the whole thesis rests on. Against a street consensus target of HK$261.62 and 21 Buy ratings out of 21, we are materially more cautious — not because we think the analysts are wrong about the business, but because we think the market is right to withhold a growth multiple until the company shows, with numbers it currently refuses to publish, that the decline has stopped.

30 — Appendix: financialsSummary tables from the filings

All figures are drawn from MIXUE Group's filings and results announcements as identified in the sources section. Amounts are in millions of renminbi unless stated otherwise, and the reporting currency is the renminbi. Financial years end 31 December. The company reports as a single operating segment; there is no product- or geography-level profitability disclosure.

31 — SourcesPrimary and secondary references

This report is built from primary disclosure wherever possible. The following documents and sources were used. Where a figure is a Farstar estimate rather than a reported number, it is labelled as such in the text.

Primary — company filings and announcements

MIXUE Group Inc. (2097.HK) prospectus and listing documents, February 2025, including the historical financial information, the franchise model description and the risk factors. MIXUE Group Inc. allotment results announcement, 3 March 2025. MIXUE Group Inc. annual results announcement for the year ended 31 December 2025, March 2026, including the consolidated financial statements, the segment note and the management discussion. MIXUE Group Inc. interim results announcement for the six months ended 30 June 2026, 27 August 2026, including the condensed consolidated financial statements, the profit bridge, the balance sheet and the store-count disclosures. MIXUE Group Inc. announcements regarding the acquisition of a 53% interest in FULU Fresh Beer, October and December 2025. MIXUE Group Inc. circular and announcements regarding the special dividend of RMB 2.65 per share, including the extraordinary general meeting results of 30 September 2026 and the record and payment dates of 12 October and 6 November 2026. MIXUE Group Inc. environmental, social and governance report, 2025. MIXUE Group Inc. monthly and periodic store-count and operational updates published on the Hong Kong Stock Exchange news service.

Competitor filings

Guming Holdings Limited interim results for the six months ended 30 June 2026. Chagee Holdings Limited interim results for the six months ended 30 June 2026. Sichuan Baicha Baidao Industrial Co., Ltd. (ChaPanda) interim results for the six months ended 30 June 2026. Auntea Jenny (Shanghai) Industrial Co., Ltd. interim results for the six months ended 30 June 2026. Nayuki Holdings Limited interim results for the six months ended 30 June 2026.

Industry and market data

China Chain Store & Franchise Association and national catering-association statistics on the freshly-made drinks market and the national store count, 2024 to 2026. Black Cat consumer-complaint platform (黑猫投诉) cumulative complaint data for MIXUE, as at 29 March 2026. Hong Kong Stock Exchange market data and closing prices for 2097.HK, 8 October 2026, including the 52-week range, market capitalisation and share count. Sell-side consensus estimates for MIXUE Group, aggregated from published broker research, as at 8 October 2026, including the 21-analyst Buy consensus and the HK$261.62 consensus target price. Press reporting and regulatory notices relating to food-safety incidents at MIXUE outlets, 2022 to 2026, including the 2024 consumer-rights broadcast and the August 2026 Nantong enforcement action.

32 — DisclosureConflicts, limitations, and revision policy