Pop Mart: The Company That Sold Feelings, and the Year the Feeling Changed

Cold Open: Two Sentences, Twelve Months Apart

On a Wednesday in mid-August 2025, 王宁 Wang Ning sat in front of analysts in Hong Kong with the kind of numbers a founder gets once in a career. Half-year revenue at his toy company had more than tripled. Net profit was up nearly 400%. Somebody asked about the full-year target of RMB20 billion. Wang did not hedge. The company was on track, he said — and RMB30 billion "this year should also be quite easy."¹

The stock did what stocks do when a founder says easy. It ran as much as 14% that week to HK$319.40, the highest level since the company had listed at the end of 2020, and by the 26th of that month it printed an all-time high of HK$339.80.²³ A plastic-and-plush monster with nine serrated teeth had made 泡泡玛特 Pop Mart one of the best-performing consumer stocks on earth.

Twelve months later, almost to the day, Wang was back in the same seat with a very different sentence. Speaking after the first-half 2026 results on 20 August 2026, he told the market that Pop Mart would very likely miss the "no less than 20%" growth target it had set for itself only five months earlier, that the pressure in the second half would be greater than in the first, and that 2026 should be understood as a year of internal adjustment.⁴⁵

Between those two sentences sits everything interesting about this company.

Pop Mart is not a toy company in the way Hasbro is a toy company. It is closer to a hit-driven studio that happens to ship its content in vinyl and polyester rather than in cinemas. It buys or signs characters, wraps them in a purchase mechanic engineered to be repeated, and sells them at gross margins that would embarrass most luxury houses. When a character catches, the economics are spectacular. When a character cools, the same operating leverage runs in reverse.

The live question in September 2026 is not whether Labubu was a real phenomenon. The audited numbers settle that. The question is whether Pop Mart has built a machine that reliably manufactures the next Labubu — or whether it has been, so far, a company that got very lucky twice and is now discovering how expensive it is to run a global retail network sized for a hit that has stopped growing.

To answer that, you have to start with a variety store that didn't work.

↑ Back to Top

Origins: A Stationery Shop in Zhongguancun

The founding story of Pop Mart is unusually free of mythology, which is itself informative. There is no monsoon epiphany, no garage. There is a young man from Henan with an advertising degree who opened a shop and spent six years failing to find the product.

Wang Ning was born in 1987 and grew up in Henan province, and he founded Pop Mart in Beijing in 2010 — in 中关村 Zhongguancun, the technology district, though the business had nothing to do with technology.⁶ The original concept was a lifestyle variety store: stationery, gadgets, imported knick-knacks, small cosmetics, the sort of pleasant clutter that Japanese and Hong Kong retailers had made into a category. It was a copy of a copy. Margins were thin, differentiation was nil, and the store competed with every mall tenant selling the same imported junk.

What changed the company was a data point rather than a vision. Wang noticed that one line of merchandise — small Japanese figurines sold in sealed boxes — moved differently from everything else. Customers didn't buy one. They bought several, came back, and asked when the next series would arrive. It was the only product in the store with a repeat-purchase pattern.

So he went looking for the supply side of that behaviour. In 2016 he flew to Hong Kong to meet Kenny Wong, the designer behind a pouty, side-eyed girl character called Molly, and signed him as Pop Mart's first artist, taking exclusive commercial rights over the character.⁷ In July of that year, Pop Mart launched the Molly Zodiac blind-box series. The first drop on the 天猫 Tmall flagship store sold out in seconds.⁷

That deal contains the whole strategic idea of the company, and it is worth being precise about what it was, because it is frequently misdescribed. Wang did not create Molly. He did not draw a single line of her. What he did was recognise that an artist with a devoted but tiny following possessed something — a character with emotional pull — that he could industrialise: mass-manufacture at scale, distribute through a national retail network, refresh with new series every few months, and monetise with a purchase mechanic that turned a single sale into a collection habit.

Pop Mart, in other words, positioned itself as the aggregator and operator of designer-toy IP rather than its creator. That is a genuinely different business from Disney's, and the difference matters enormously to the risk profile. Disney owns Mickey Mouse outright and has for a century. Pop Mart's flagship characters mostly arrive through contracts with independent artists — and contracts have terms, renewal dates, and counterparties who become rich and gain leverage.

The model scaled fast because it had two cheap distribution innovations attached. The first was the retail store, small-format, mall-based, merchandised like a gallery — deliberately closer to a boutique than a toy aisle, because the price point only works if the customer believes she is buying design rather than plastic. The second was the "roboshop" — a smart vending machine, in plain terms, that could sit in a subway station or a cinema lobby and sell blind boxes with no staff at all.

The roboshop deserves a moment, because it is a genuinely elegant piece of unit economics that most Western observers skip past. A retail store in a Chinese A-mall carries rent, fit-out, and three to five staff. A roboshop carries a machine, a slot of floor space, and a restocking visit. It cannot merchandise, cannot upsell, and cannot tell a brand story — but the blind box does not need any of those things, because the product sells itself on a mechanic rather than on persuasion. Pop Mart could therefore blanket a city with distribution at a fraction of the capital intensity of a conventional retail rollout, and it could put a point of sale in transit locations where nobody would ever open a store. By the end of 2019 the company ran 114 retail stores and 825 roboshops; a year later that was 187 stores and 1,351 roboshops.⁸

There is something worth noticing about Wang's decision-making style here, and it recurs throughout the company's history. He is not, on the evidence, a visionary in the conventional founder sense. He is an unusually disciplined observer of what customers actually do, followed by a fast and total commitment of resources to whatever the observation reveals. The blind box was not a strategy he conceived; it was a behaviour he noticed in his own sales data and then chased to its source. Plush was not a category he predicted; it was a format the company tried and then poured everything into once it worked. That is a real capability — arguably a better one than vision — but it is a reactive capability, and reactive capabilities are excellent at riding waves and poor at creating them. Hold that thought; it explains a great deal about 2025 and 2026 alike.

By 2019 the company had revenue of RMB1.68 billion and, more importantly, a proven playbook: sign artist, launch series, sell through owned channels, repeat.⁸ It also had a single-character concentration problem that nobody was paying much attention to yet. Molly alone was 27.1% of revenue that year.⁸

That number is about to become the most useful piece of history in this entire story.

↑ Back to Top

Before going further into the timeline, it is worth stopping to explain the product mechanic, because outside China it is routinely described as "a toy in a box" and that description misses why the economics work.

A blind box is a sealed package containing one figurine from a series of, typically, twelve designs. You know the series. You do not know which figure you are getting. Somewhere in the series is a "secret" or "chase" variant — deliberately produced at a low ratio, often around one in a hundred and forty-four boxes.

Think of it as the retail equivalent of a slot machine with a guaranteed consolation prize. Every pull returns something you can display, so the loss is never total; but the thing you actually want is randomised, so the only way to pursue it is volume. Behavioural economists call this a variable-ratio reinforcement schedule, and it is the most powerful driver of repeat behaviour that psychology has identified. Pop Mart did not invent it — Japanese gashapon capsule machines have used it for decades — but Pop Mart industrialised it, put it in premium malls, and attached it to characters with genuine design credibility rather than licensed cartoon tie-ins.

Layer on a second mechanic: the series. A twelve-figure set creates a completion instinct. Customers who have nine figures are not buying a tenth toy; they are buying the closure of a gap. Then layer a third: scarcity by design. Limited runs, timed drops, regional exclusives, and collaborations create a secondary market, and a secondary market with rising prices converts a toy into something that at least feels like an asset.

The financial consequence of all this is straightforward. Gross margin reached 72.1% in 2025.⁹ A plastic figure that costs a few renminbi to injection-mould sells for RMB69 to RMB99 in China and considerably more abroad. Pop Mart's own filings attribute the margin expansion to two things: a higher share of overseas sales, which carry higher prices, and better procurement leverage over its contract manufacturers.⁹

There is a fourth mechanic that receives almost no attention outside China and is arguably the most commercially significant of all: Pop Mart moved the blind box online without losing what makes it work. "Pop Draw" is, in essence, a digital version of the shelf — the customer buys a draw from a displayed set on Pop Mart's own app, sees the result revealed, and then either has the physical item shipped or holds it in an account. Stripping out the physical box could easily have killed the magic; instead it removed the two frictions that limited the format, namely store hours and inventory in a specific location, while adding a genuinely new one — the ability to run a drop simultaneously to millions of members. Pop Draw generated RMB2,062.9 million in the first half of 2026 alone, up 83.3% year on year, making it larger than most listed toy companies' entire businesses.¹⁰ It is also the piece of the model Pop Mart has been slowest to export, and the reason for that gap is one of the more interesting unanswered questions about the company's international prospects.

The customer data supports the habit thesis rather than the novelty thesis, at least in China. Registered members in the Chinese mainland grew from 46.08 million at the end of 2024 to 72.58 million at the end of 2025, and members accounted for 93.7% of sales that year with a repeat purchase rate of 55.7%.⁹ By 30 June 2026 mainland registered members had reached 82.44 million and global registered members passed 100 million.¹⁰ These are not one-time tourists buying a souvenir. This is a membership base transacting repeatedly through an owned app.

Two cautions belong right here rather than in a risk section later. First, the repeat-purchase figure disclosed for the first half of 2026 — 51.6% — is measured over six months against total purchasing members, a different denominator and a shorter window than the full-year 2025 figure, so the two are not cleanly comparable and should not be read as a clean deterioration.¹⁰ Second, and more substantively: a mechanic this effective at driving repeat purchase among young consumers is exactly the kind of thing regulators eventually notice. They did. We will get there.

The mechanic explains why a hit is so profitable. It does not, by itself, explain why any particular character becomes a hit — and it certainly does not guarantee that a hit stays one. For that, look at what happened to the character that built the company.

↑ Back to Top

Molly's Lesson: What Happens After an IP Peaks

Here is the single most useful test of the "durable IP" claim that sits at the centre of every bull case on Pop Mart, and it does not require any speculation about Labubu's future. It requires only the company's own audited disclosures going back to 2019.

Molly was 27.1% of revenue in 2019, generating RMB456.0 million.⁸ In 2020, Molly's revenue fell in absolute terms to RMB356.9 million — a decline of roughly 22% — and its share of revenue halved to 14.2%.⁸ Molly recovered in later years as the overall business grew, reaching RMB705.1 million in 2021, RMB802.2 million in 2022, and RMB1,020.3 million in 2023.¹¹¹² But her share of revenue never returned to anything like 2019 levels: 15.7%, then 17.4%, then 16.2%.¹¹¹²

Now watch what happened when a bigger hit arrived. In 2025, Molly's revenue rose to RMB2,897.1 million — a fine absolute number — but her share of group revenue collapsed to 7.8%.⁹ And in the first half of 2026, Molly's revenue fell 33.6% year on year to RMB900.6 million, taking her to 5.3% of the group.¹⁰

Molly is not the only data point, and she is not even the sharpest one. PUCKY, an exclusively licensed character, was 18.7% of Pop Mart's revenue in 2019 at RMB315.3 million. In 2020 it was RMB300.0 million and 11.9%. By 2021 it was RMB182.2 million and 4.1% — an absolute decline of 42% in two years.⁸¹¹ From 2022 onward, PUCKY no longer appeared as a separately disclosed line in the IP table at all.¹²

So what does the record actually say? It says that Pop Mart characters have a demonstrated lifecycle. They rise, they peak, and then they either plateau at a materially lower share of a larger business or they decline outright. Over a seven-year window that includes two distinct management eras of the same founder-led company, no Pop Mart character has yet been shown to sustain a peak share of revenue. The company's own product-level disclosure is the evidence.

This does not reject the investment case. It narrows it. The defensible version of the claim is not "Pop Mart owns durable evergreen IP." The evidence does not support that yet, and Molly — the company's oldest and most heavily marketed character, promoted in the FY2025 report through a proliferation of sub-lines including Classic Molly, Baby Molly, Royal Molly, Space Molly and Angry Molly — is the strongest disconfirming case available, because if any character had the runway to prove evergreen status it was hers.⁹ The defensible version is narrower and more interesting: Pop Mart has demonstrated a repeatable ability to find and scale the next character before the last one rolls over. That is a claim about the company's talent pipeline and curation, not about the immortality of any given monster.

It is also a claim that can be tested. It has a specific falsifier: a year in which the fading of an incumbent character is not offset by the rise of a new one, and group revenue growth goes negative. Watch the artist-IP table in each results announcement, and specifically whether the number two and number three characters are growing fast enough in absolute renminbi to cover what the number one loses.

In the first half of 2026, that test was passed — narrowly, and only because of one character. But we are getting ahead of the story. First, the company had to survive a listing and a hangover.

↑ Back to Top

Listing, Boom, and the 2022 Reckoning

Pop Mart listed on the Main Board of the Hong Kong Stock Exchange on 11 December 2020, raising approximately HK$5.2 billion in gross proceeds at HK$38.50 per share.¹³ The retail tranche was oversubscribed hundreds of times. On debut the shares opened at HK$77.10 and closed the day up roughly 80%, valuing the company at around US$7 billion.¹⁴

It was, in retrospect, a near-perfect moment to sell equity: a novel consumer mechanic, a China-consumption narrative at its most fashionable, and a market that had not yet worked out that the company's revenue rested on a handful of characters. The company had also just opened its first overseas store, in Gangnam-gu, Seoul.⁸

Then the model met its first genuine stress test, and it did not pass cleanly.

In 2022, Pop Mart's revenue grew 2.8% to RMB4,617.3 million. Gross profit fell 3.8%. Operating profit fell 49.3%. Net profit fell 44.3% to RMB475.8 million.¹² Earnings per share dropped from RMB0.62 to RMB0.35.

The proximate cause was China's pandemic controls, which shuttered malls and crushed foot traffic through the store and roboshop network. But blaming COVID alone is too generous. The company's own 2023 filing shows that online channels also fell that year — Pop Draw revenue declined, and the Tmall flagship store fell 30.4% from RMB462.4 million in 2022 to RMB322.0 million in 2023 — which is not what you would expect if the only problem were closed shops.¹¹ Demand for the core figurine product was soft on its own terms. The blind-box novelty in China had cooled.

Two things about 2022 matter for the present.

The first is what it revealed about the cost structure. Pop Mart's operating profit fell nearly twice as fast as gross profit, because leases, store staff, and marketing do not shrink when a hit fades. The business has high incremental margins on the way up and painful negative operating leverage on the way down. Anyone modelling a Labubu deceleration in 2026 and 2027 should have 2022 in front of them as the base case for how the P&L behaves.

The second is what management did next, and this is genuinely to their credit. Rather than defending the domestic figurine business, they pushed hard into two adjacent bets: overseas expansion, and new product formats. The company accelerated international store openings from 2022 onward, and — critically — began developing plush versions of its characters, a format that was almost irrelevant to the P&L at the time. Plush was 21.7% of revenue as recently as 2024.⁹

Recovery came fast. Revenue grew 36.5% in 2023 to RMB6,301.0 million with net profit up 128.8% to RMB1,088.8 million.¹¹ Then 2024 delivered RMB13,037.7 million, more than double the prior year, with net profit of RMB3,308.3 million.⁹

Those are excellent numbers. They are also, in hindsight, the sound of a fuse burning. Because somewhere in Thailand in 2024, an entirely different kind of event was beginning.

↑ Back to Top

Labubu: Anatomy of a Global Mania

Labubu was not a Pop Mart creation and was not, for most of its life, a hit. The character came from 龙家昇 Kasing Lung, a Hong Kong-born illustrator raised in the Netherlands, who created a Nordic-mythology-inflected tribe called The Monsters in 2015 — Labubu, Zimomo, Tycoco, Spooky, Pato — as characters in his own picture books.¹⁵ Pop Mart signed an exclusive licensing arrangement with Lung in 2019.¹⁵

For years, The Monsters was a mid-table asset. In 2020 the family generated RMB204.8 million, 8.1% of revenue; in 2021, RMB304.4 million, 6.8%; in 2022, RMB263.0 million — an absolute decline — at 5.7%.⁸¹¹¹² Pop Mart had a licence to a character that was going nowhere in particular.

What changed was a combination of geography, format, and one celebrity.

The geography was Thailand. Pop Mart opened its first Thai store at Bangkok's CentralWorld in September 2023, and Labubu found an audience there that it had never found at home.¹⁶ The format was plush: a soft, keychain-sized version designed to hang off a handbag, which converted the product from something you display on a shelf to something you wear. That is a completely different consumption occasion, with a completely different social function — it is visible, it signals taste, and it invites imitation.

The celebrity was Lisa of BLACKPINK, Thai-born and one of the most-followed people on the planet, who posted photographs with her Labubu dolls in April 2024. Rihanna, Dua Lipa, and Cher followed.¹⁶ Within months, Labubu had become a luxury-handbag accessory in a way that no Chinese consumer product had ever managed.

Look at what that did to the numbers. The Monsters went from RMB367.9 million in 2023 to RMB3,040.7 million in 2024 — a more than eightfold increase — and 23.3% of group revenue.⁹¹¹ Plush toys, effectively a rounding error two years earlier, hit RMB2,832.1 million.⁹

Then 2025 happened, and the scale becomes difficult to describe without sounding hyperbolic. The Monsters generated RMB14,161.1 million, up 365.7%, and accounted for 38.1% of a group that had itself nearly tripled.⁹ Plush toys reached RMB18,708.1 million — 50.4% of everything Pop Mart sold — up 560.6% in a single year.⁹ The Americas, a RMB802.2 million business in 2024, became a RMB6,806.2 million business, up 748.4%.⁹

Management fed the fire deliberately and skilfully. In August 2025 the company launched a mini Labubu designed to clip onto a phone — a lower price point, a new occasion, a fresh scarcity event — and third-quarter global revenue rose roughly 245% to 250% year on year.²¹⁷ The company's own 2025 report describes The Monsters joining Pop Mart's "Ten Billion Club," appearing in the Macy's Thanksgiving Day Parade, and headlining a touring "Wacky Mart" exhibition.⁹

Underneath the marketing sat a manufacturing problem that deserves more credit than it usually receives, and a strategic choice that deserves more scrutiny. Pop Mart does not own its factories; it works through contract manufacturers, which means scaling output on the order of ten times inside eighteen months required securing capacity, tooling, and quality control across a supplier base that had been sized for a mid-cap Chinese toy company. The company pulled it off — the FY2025 gross margin expansion is partly the story of that scale being turned into procurement leverage, and the cost of sales line shows goods costs rising far more slowly than revenue.⁹

But scaling supply into a scarcity-driven product is a decision, not an inevitability, and it has a cost. The entire secondary-market premium that made Labubu feel like an asset rather than a toy depended on there not being enough of them. Meeting the demand in full converted a scarce object into an available one. Pop Mart chose revenue today over mystique tomorrow — a defensible choice, and probably the right one given that scarcity rents accrue mostly to scalpers rather than to the company, but a choice whose consequences showed up in the resale market well before they showed up in the P&L.

Here is the analytical read, and it is more sober than the numbers suggest. Labubu's ascent was driven by an unusually powerful and unusually exogenous set of catalysts: a specific market where the character found organic cultural fit, a format change that altered how the product was used, and a K-pop endorsement that no company can buy or replicate on demand. Pop Mart's genuine skill was in the response — it had the supply chain to scale production tenfold, the store network to capture demand, and the merchandising instinct to keep iterating formats. That is real operational capability and it should not be dismissed.

But operational capability applied to a viral cultural moment is not the same thing as the ability to manufacture cultural moments. The bull case requires the second. The evidence to date supports mainly the first.

And viral moments, by their nature, have a shape. What goes vertical usually does not stay vertical. In March 2026, the market decided it had seen enough of the shape.

↑ Back to Top

FY2025: The Record That Broke the Stock

On 25 March 2026, Pop Mart reported the best year any Chinese consumer company has had in living memory. Revenue rose 184.7% to RMB37,120.1 million. Gross profit rose 207.4%. Operating profit rose 306.6% to RMB16,890.5 million. Net profit rose 293.3% to RMB13,012.0 million. Gross margin expanded from 66.8% to 72.1%. Basic earnings per share went from RMB2.36 to RMB9.61.⁹

The stock fell 22% that day, its worst session since listing.¹⁸¹⁹

That reaction deserves careful handling, because it is easy to write off as a market tantrum. It was not. Three specific things in the disclosure did the damage.

The first was concentration. The Monsters had gone from 23.3% of revenue to 38.1%.⁹ The company had become more dependent on one character family during a year of explosive growth, not less. Seventeen IPs cleared RMB100 million, which sounds like diversification, but seventeen characters generating a hundred million each is a rounding error against a fourteen-billion-renminbi franchise.

The second was the shape of the year. Revenue of RMB37.12 billion came in slightly under a consensus that had crept toward RMB38 billion, and the miss was concentrated in the fourth quarter, where growth decelerated sharply from the third quarter's 245%-plus pace.¹⁸ Third-party point-of-sale trackers had already flagged a sequential decline in North American sales in the December quarter.¹⁷ The market was not reading the past; it was reading the exit rate.

The third was the guidance. Wang Ning told the market Pop Mart was targeting growth of "no less than 20%" in 2026.¹⁸ Set against 184.7%, that is not conservatism; it is an admission that the company itself did not expect the run-rate to hold.

There is a fourth thing worth flagging that received less attention: the capital position. The FY2025 balance sheet showed inventories of RMB5,472.8 million, up from RMB1,524.5 million a year earlier, against cash and equivalents of RMB13,775.1 million and total equity of RMB22,652.4 million.⁹ Building inventory into a boom is rational. Building 3.6x the inventory is a bet that the boom continues. The board also recommended a final dividend of RMB2.3817 per share, aggregating to roughly RMB3,194.0 million.⁹

What the FY2025 results genuinely established: Pop Mart's supply chain and store network can absorb a demand shock of extraordinary magnitude without breaking, and the incremental margins on a hit are as good as anything in global consumer. What they did not establish, and what the market correctly focused on: whether any of it recurs.

Five months later, the first half of the answer arrived.

↑ Back to Top

The Turn: H1 2026 and the "Business Adjustment Year"

The interim results published on 20 August 2026 are the most important document Pop Mart has released since its prospectus, because they show the machine running in both directions at once.

Before opening them, hold one earlier data point in mind, because it changes how the half reads. In its first-quarter operating update on 12 May 2026, Pop Mart disclosed that group revenue had grown 75% to 80% year on year in the three months to March, with China up 100% to 105%, Europe and other regions up 60% to 65%, the Americas up 55% to 60%, and Asia Pacific ex-China up 25% to 30%.³⁴³⁵ Every region was still growing, some of them briskly.

Set that against a first half that came in at 23.8% with two regions in outright decline, and the arithmetic is unavoidable: the second quarter did not merely decelerate, it went backwards in most of the world. Whatever broke, broke between April and June 2026. That timing matters, because it rules out the tidiest explanation — a gradual fading of a fad — in favour of something sharper: the anniversary of the Labubu comparison base arriving all at once, colliding with an overseas customer file that had no second reason to come back.

Now start with the headline, which looks fine. Revenue rose 23.8% to RMB17,172.9 million. Gross profit rose 22.6%. But operating profit rose only 11.3%, profit for the period only 8.9%, and adjusted net profit only 9.5% to RMB5,155.8 million — with the adjusted net margin falling from 33.9% to 30.0% and gross margin easing from 70.3% to 69.7%.¹⁰ Profit grew at roughly 40% of the rate of revenue. That is the 2022 operating-leverage lesson reappearing in a growth year.

Now open the segments, where the real story is.

Labubu declined for the first time. The Monsters generated RMB4,454.4 million, down 7.5% year on year, and its share of group revenue fell from 34.7% to 26.0%.¹⁰ In the seven-year record of Pop Mart's disclosed IP table, this is the first time the company's largest character family has shrunk in absolute terms while the group was growing. The pattern that consumed Molly and PUCKY has now reached the franchise that made the company famous — and it reached it within roughly eighteen months of the character going global.

Overseas revenue fell. Total international revenue dropped 11.1% to RMB4,972.4 million, the first decline since Pop Mart began reporting international markets as separate segments. Asia Pacific fell 9.7% to RMB2,575.0 million. The Americas fell 16.5% to RMB1,891.6 million. Only Europe and other regions grew, up 5.9% to RMB505.7 million.¹⁰

And within overseas, the online channel collapsed. This is the number that should hold an investor's attention. Americas online revenue fell 45.6%, from RMB1,326.6 million to roughly RMB722 million. Asia Pacific online fell 39.8%, with the Shopee channel alone down 62.1%. European online revenue fell 59.0%.¹⁰ Meanwhile overseas offline revenue grew — Americas physical channels up 19.5%, Asia Pacific offline up 16.2%.¹⁰

Read those two facts together and you get a clean diagnosis. Physical stores, where a customer walks in and browses a wall of characters, kept growing. Online, where a customer arrives with a specific search intent — almost always "Labubu" — fell off a cliff. The overseas business was not a Pop Mart business. It was a Labubu business with a Pop Mart logo on it, and when the search volume normalised, the revenue went with it.

COO 司德 Si De essentially confirmed this on the call, noting that the prior year's overseas growth had been driven mainly by Labubu and that many international customers were unfamiliar with the company's other characters.⁴ Earlier, on the Q1 update in May, he had attributed the deceleration to newly acquired overseas customers lacking familiarity with designer-toy IP and to recently hired international staff lacking operational experience.²⁰ Those are honest answers. They are also an admission that Pop Mart bought a large amount of overseas customer acquisition with a single character rather than a brand.

China, meanwhile, boomed. PRC revenue rose 47.3% to RMB12,200.5 million, taking the domestic share of the group back up from 59.7% to 71.0% — reversing three years of internationalisation in six months.¹⁰ Domestic retail store revenue rose 38.2% on only twelve net new stores, which is genuine same-store strength. Online China rose 62.7%, with the gamified Pop Draw channel up 83.3% to RMB2,062.9 million and 抖音 Douyin up 74.0%.¹⁰

And one new character carried the group. 星星人 Twinkle Twinkle — a character Pop Mart signed in 2024, which generated RMB120.8 million in that year and RMB389.4 million in the first half of 2025 — produced RMB2,650.0 million in the first half of 2026, up 580.6%, and became 15.4% of the group.⁹¹⁰ Without Twinkle Twinkle's incremental RMB2.26 billion, Pop Mart's revenue growth in the period would have been roughly 7% rather than 23.8%.

That is the succession test, and the company passed it. It is also worth stating exactly how narrowly it passed: one character, in one market, covered the entire shortfall from the previous champion. Nothing about that outcome is evidence of a diversified portfolio; it is evidence of a functioning pipeline with a sample size of one successful handoff.

Two balance-sheet items complete the picture and neither is comfortable. Inventory turnover days rose from 123 days in 2025 to 201 days in the first half of 2026, with inventories up to RMB6,101.5 million — the company attributes this to stock built ahead of overseas expansion, which is precisely the expansion that just contracted.¹⁰ And the group booked a foreign exchange loss of RMB720.3 million in the period, against a RMB120.1 million gain a year earlier, swinging other gains from a net RMB179.2 million positive to a net RMB688.8 million negative.¹⁰ Offsetting that, other income rose 350.4% to RMB302.8 million, of which government grants were RMB239.5 million, up from RMB37.8 million — a low-quality earnings contribution that is worth stripping out when assessing the underlying trend.¹⁰

Wang's framing on the call was that the extraordinary 2025 base created pressure that would be more pronounced in the second half, that the 20% full-year target was very likely out of reach, and that the company would not chase revenue that did not bring profit with it.⁴⁵²¹ Citi cut its price target on the stock to HK$198 following the release.²² The shares, which had peaked at HK$339.80 thirteen months earlier, were trading around HK$154 by early September 2026 — roughly a 55% drawdown from the high.³

The numbers now sit on the table. What they say about the people who produced them is a separate question.

↑ Back to Top

Management Under the Microscope

Any assessment of Pop Mart's management has to begin by acknowledging the obvious: Wang Ning built a company from a failing variety store into one that earned RMB13 billion in a year, and Forbes valued his fortune at US$12.6 billion in early September 2026.⁶ Founder-operators with that record earn a presumption of competence. They do not earn a presumption of accuracy.

On accuracy, the record over the last twelve months is poor, and it is poor in a specific way that investors should price.

In August 2025, Wang told analysts RMB20 billion was on track and RMB30 billion "should also be quite easy."¹ The company finished 2025 at RMB37.1 billion, so that call was not merely met but exceeded — which is the point. The guidance was directionally right and rhetorically loose, and it was loose in the direction that flatters the stock. In March 2026, the same executive guided to "no less than 20%" growth.¹⁸ By 20 August 2026 — five months later, and roughly twelve months after "quite easy" — he was telling the market that the 20% figure was very likely unreachable and that this was a year of internal adjustment.⁴⁵

This is a same-management, same-narrative reversal inside a twelve-month window, and it is not attributable to an external shock. There was no tariff wall that closed overnight, no regulatory ban, no recall. What happened is that a character stopped growing at the rate it had been growing, which is the single most foreseeable risk in this business and the one Pop Mart's own IP table had been warning about since 2020.

Two mitigating observations are fair. First, when the miss arrived, management did not hide behind aggregates: the interim announcement discloses channel-level online declines by region, the IP table shows the negative print for The Monsters in plain sight, and the inventory-days deterioration is stated explicitly.¹⁰ That is better disclosure than many companies offer in a bad half. Second, the explanations offered — high base, unfamiliarity of new overseas customers with non-Labubu IP, inexperienced newly-hired international staff, cross-border logistics and inventory strain from a rapidly built global network — are specific and operationally plausible rather than evasive.⁴²⁰

Two aggravating observations are also fair. The COO's explanation that overseas growth "had been driven mainly by Labubu" was true in 2025 as well as in 2026; it was knowable at the time the 20% target was set. And CFO Yang Jingbing had already flagged in May 2026 that raw material costs were running 3 to 5 percentage points higher year on year and would compress gross margin.²⁰ The ingredients of the miss were visible to management before the target was reaffirmed.

The People in the Room

It is worth pausing on who actually speaks for this company, because Pop Mart's public voice is unusually distributed for a founder-controlled business, and the division of labour is revealing.

Wang Ning is the narrative. His register at results briefings runs to the expansive and the conceptual — the company as an IP business rather than a toy business, the ambition to build something Disney-shaped, the confidence that a number is "quite easy." He does not, on the public record, get into channel-level operating detail. When the story is good, this is an asset: founders who talk in decades rather than quarters attract long-duration shareholders. When the story turns, the same register becomes a liability, because there is no track record of calibrated, boring guidance to fall back on. The market has nothing to anchor to except the founder's mood.

司德 Si De, the chief operating officer, has been the one delivering the diagnoses, and his have consistently been the more useful statements. He joined the company in 2019, before the listing and before Labubu, and it was Si De who supplied the two most analytically honest sentences of the past year: that the overseas growth had been driven mainly by Labubu, and that newly hired international staff lacked operational experience while newly acquired international customers lacked familiarity with the wider IP portfolio.⁴²⁰ Neither statement flatters the company. Both are specific enough to be checked.

CFO Yang Jingbing — an accountant by training, with a finance background in Beijing property development before joining the toy business — has been the earliest and most accurate of the three. He flagged the input-cost problem in May 2026, quantifying raw material inflation at 3 to 5 percentage points and guiding gross margin down, well before the interim results confirmed it.²⁰ That is what functional CFO communication looks like.

The pattern across the three is consistent and worth naming plainly: the operators have been broadly reliable, and the founder has not. An investor reading Pop Mart's disclosures would have been better served by the COO's and CFO's remarks than by the chairman's, in both directions — they were less exuberant in 2025 and more precise in 2026. That is not a governance failure, but it is a useful weighting rule for anyone reading the next set of results.

Then there is capital allocation, where the behaviour is more interesting than the rhetoric.

In the first half of 2026, Pop Mart repurchased 11,220,000 shares for approximately HK$1,744,493,320 and cancelled them. The disclosed monthly detail shows purchases in January at HK$177.70–194.90, in March at HK$141.40–157.80, and in April at HK$140.90–150.20.¹⁰ The company paid its FY2025 final dividend of roughly RMB3,194.0 million in May 2026, declared no interim dividend for the first half of 2026, and announced an intention to repurchase between RMB2 billion and RMB5 billion of stock over the following six months.⁹¹⁰²³ Taken together, dividends plus buybacks for the year are on track to approximate the entirety of the prior year's earnings.²¹

How should a sceptic read that? Charitably: a company with RMB12,442.1 million of cash, no meaningful leverage, and a shrinking need for growth capex is returning capital rather than empire-building, which is the correct instinct.¹⁰ Less charitably: a substantial share of the buyback was executed at prices well above where the stock now trades, and the switch from a paid interim dividend to a buyback intention with a wide RMB2–5 billion range preserves flexibility for management while sounding more supportive than a fixed commitment would be. It is not obviously disciplined and it is not obviously undisciplined; it is a company trying to put a floor under a falling share price with the tool that is easiest to reverse.

What would genuinely test management credibility from here is narrower and more checkable than any of this. It is whether the second half of 2026 lands within the range implied by "greater pressure than the first half," or whether it lands materially below — and whether the next target the company sets is one it can actually hit. On the evidence of the last twelve months, the appropriate posture toward Pop Mart guidance is to treat it as a description of management's hopes rather than as a forecast.

↑ Back to Top

Beyond the Toy Aisle: Parks, Film, and the Disney Question

The strategic answer Pop Mart offers to the concentration problem is, in essence: stop being a toy company. Become an IP company — one where a character generates revenue across toys, parks, film, food, accessories, and licensing, so that no single product cycle determines the outcome. The comparison management and the market reach for is Disney, and occasionally サンリオ Sanrio.

The comparison is worth taking seriously, and also worth stress-testing, because it is the load-bearing claim in every long-duration bull case on this stock.

The physical expression of the strategy is POP LAND, a theme park in Beijing that opened in September 2023.⁹ Pop Mart began equipment upgrades in April 2025, and on 25 April 2026 unveiled a substantial expansion: three new zones, five large-scale amusement facilities, and a Labubu Forest area themed around The Monsters and DIMOO, which opened on 30 April.⁹²⁴ The park reported visitor traffic up 70% year on year, with non-family visitors at 59% of footfall and non-local tourists at 58% — meaning the park is drawing genuine destination traffic rather than neighbourhood repeat visits.²⁴ A Phase II expansion is planned for 2027, and the company's VP and general manager Hu Jian described the park as a live platform to test how IP translates into physical experiences and consumer spending.²⁴

The narrative expression is film. On 18–19 March 2026, Pop Mart announced a partnership with Sony Pictures to co-develop a Labubu feature, a live-action and CGI hybrid to be directed and co-written by Paul King — the filmmaker behind Paddington and Wonka — with Kasing Lung as executive producer.²⁵

And there is a long tail of format experiments: POP BAKERY, a dessert brand that ran over ten themed pop-ups across mainland cities; popop, an accessories concept with its first stores in Beijing and Shanghai; touring IP exhibitions; and a music output from the park team, with a single written for Twinkle Twinkle reportedly drawing over 15 million views.⁹

Now the discipline. Certification is not commercialisation, and a signed development deal is not a released film. Pop Mart's own disclosure is that revenue from POP LAND, POP BAKERY and popop is not separately broken out; it sits inside a "wholesales and others" line for the PRC that totalled RMB553.1 million in the first half of 2026 — around 3% of group revenue, and that line includes genuine wholesale as well.¹⁰ After nearly three years of operation, the park is not yet a material revenue contributor, and by the company's own account only about one-third of it was open to the public during 2025, with two-thirds under renovation as of April 2026.⁹²⁴

There is also a history worth holding in view. The original 2020 concepts for the park included roller coasters that were shelved on feasibility grounds, and the park opened in 2023 at roughly half its originally envisioned scope.²⁴ That is not a scandal — theme park development is hard and descoping is normal — but it is the relevant base rate for how quickly this company converts an ambitious physical concept into an operating asset. A Phase II slated for 2027 is, on that record, a 2028-or-later revenue event.

The film has an even longer fuse and a harder economics question attached. A studio co-development deal typically leaves the toy company with modest economics on the film itself; the value accrues if the film extends the character's cultural life and lifts merchandise. That is a real prize. It is also a prize that arrives, at the earliest, several years after the point at which The Monsters has already begun declining — and Hollywood's record of converting toy properties into durable franchises is mixed enough that no confident forecast is available.

There is a better benchmark than Disney for what Pop Mart is attempting, and it is worth naming because it sets a more realistic bar. サンリオ Sanrio built a multi-decade business on Hello Kitty by licensing the character into thousands of third-party products rather than manufacturing them, converting a design asset into a royalty stream that requires almost no working capital and carries almost no inventory risk. That is the high-margin, low-capital end state of an IP company — and it is precisely the opposite of the position Pop Mart occupies today, with RMB6.1 billion of finished goods on its balance sheet and a global lease portfolio.¹⁰

The distinction matters because it clarifies what "becoming an IP company" would actually require. It is not building a park or making a film. It is reaching the point where a meaningful share of profit arrives as licensing income from other people's products — where Pop Mart's characters appear on someone else's stationery, someone else's apparel, someone else's confectionery, and Pop Mart collects a percentage without carrying the goods.

The company has begun this. Income from IP licence fees and other services rose from RMB25.8 million in the first half of 2025 to RMB60.6 million in the first half of 2026.¹⁰ That is real, it is growing fast, and it is roughly a third of one percent of revenue. It is also, arguably, the single most informative small number in the whole disclosure: it is the clearest available measure of how far the company has travelled from selling toys with characters on them toward owning characters that other people pay to use. Sanrio took decades to get there. Pop Mart is a few years in.

The honest framing: the IP-company strategy is the correct strategic response to the concentration problem, it is being pursued with real capital and credible partners, and there is currently no financial evidence that it works at scale. It should be treated as unpriced optionality with a multi-year gestation, not as a pillar of the near-term case. The KPI that would change that assessment is the "wholesales and others" line growing to a genuinely material share of group revenue with the company choosing to break it out separately — a decision management would only make if the numbers flattered them.

↑ Back to Top

Myth vs. Reality

Four consensus narratives surround this company. Each contains something true and something that does not survive contact with the disclosure.

Myth: "Pop Mart is a gambling business dressed as a toy company, and regulation will kill it."

The mechanic is genuinely engineered for repetition, and Chinese regulators have engaged with it directly. The State Administration for Market Regulation issued trial blind-box guidelines in June 2023, banning certain contents and prohibiting sales to children under eight, with vendors urged to prevent minors from becoming addicted.²⁶ On 20 June 2025, 人民日报 People's Daily ran a commentary calling blind boxes and mystery cards "business traps" that use unpredictable rewards to drive repeat purchase, and urged stronger protections for minors. It did not name Pop Mart. The stock fell 8.8% across two sessions, erasing roughly HK$30.9 billion of market value.²⁷

The reality is more mundane than either the bulls or the bears want. Pop Mart's core customer is not a child; it is an adult, disproportionately female, in her twenties or thirties, buying an emotional object and a social signal. An under-eight prohibition is close to irrelevant to the revenue base. The regulatory risk is real but it is a tail risk — the scenario that matters is not the 2023 guideline, it is a future rule that constrains the randomised-outcome mechanic for adults, or that caps drop pricing. Nothing of that kind has been enacted. Investors should hold this as a live overhang with unknowable timing rather than as an imminent event, and note that the company's response has been to position itself inside the regulatory process: its own filings state that it initiated the formulation of the first national standard for the pop toy industry.¹⁰

Myth: "Labubu was a bubble and it has burst."

Half true, and the half that is true is about the secondary market rather than the primary one. As Pop Mart scaled production, resale premiums compressed sharply; by mid-2026 discounted listings were flooding second-hand platforms within hours of a release, and StockX data indicated only around 23% of Labubu SKUs released since 2022 had held or increased their value after launch.²⁸ Counterfeiting became a genuine problem — US Customs and Border Protection seized over 11,000 fake collectibles worth more than US$500,000 at Seattle airport in September 2025, UK authorities seized £3.5 million of fakes, and the US Consumer Product Safety Commission issued a warning in August 2025 about lookalike dolls presenting choking and toxicity risks — prompting Pop Mart to escalate anti-counterfeiting enforcement.²⁸²⁹

But the primary business did not burst. The Monsters still generated RMB4,454.4 million in the first half of 2026 — more than the entire company earned in revenue in any year before 2021.¹⁰¹¹ What died was the speculative overlay: the belief that a plastic monster was an appreciating asset. Losing that is bad for volumes at the margin and unambiguously good for brand durability, because a customer base composed of flippers is worth far less than one composed of collectors.

Myth: "China is the mature market and overseas is the growth engine."

This was the consensus through all of 2025, and the first half of 2026 inverted it completely. China grew 47.3% while overseas shrank 11.1%; China's share of the group rose from 59.7% to 71.0%.¹⁰ The domestic business is not the ex-growth legacy asset; it is currently the only large part of the company that is compounding, powered by an 82-million-member base and gamified owned channels that no overseas market has replicated.

Myth: "Pop Mart has a wide moat because it owns its IP."

This is the most consequential myth and it needs the most careful handling, so it gets its own section.

↑ Back to Top

War Game: The Competitive Board

Put yourself on the other side of the table. You are 名创优品 Miniso, or a well-funded Chinese startup, or a Japanese incumbent, and you have watched Pop Mart earn a 72% gross margin. What stops you?

Less than the bulls think, and more than the bears do.

What does not stop you: the product. A blind box is not technically hard. Injection moulding and plush manufacturing are commodity capabilities available from the same Chinese and Vietnamese contract manufacturers Pop Mart uses. There is no patent thicket, no process secret, no regulatory approval. Anyone can be in this business next quarter.

What does not stop you: retail. Mall space is available to anyone who can pay rent. Roboshops are vending machines.

What partially stops you: the artist relationships and the curation. This is where the actual advantage sits. Pop Mart has exclusive commercial rights to characters that matter and a design organisation that has now identified several winners. That is what Hamilton Helmer would call Cornered Resource — control of a coveted asset on preferential terms — combined with a degree of Process Power in the incubate-launch-iterate loop.

But test the strength of that cornered resource honestly. Pop Mart's biggest character was licensed from an independent artist, not owned outright. Kasing Lung is now globally famous and, one presumes, considerably wealthier and better advised than he was in 2019; the FY2025 filing discloses that design and licensing costs rose from RMB506.8 million to RMB950.5 million, attributed partly to higher IP licensing fees.⁹ Cornered resources whose owners gain bargaining power tend to become less cornered over time. Pop Mart has been building in-house design capability since its PDC in-house team launched original IPs in 2020, and artist IPs rose to 90.0% of revenue in 2025 — but "artist IPs" includes both owned and licensed characters, and the company does not disclose the split.⁸⁹

What partially stops you: brand and membership. Pop Mart's 82.44 million mainland registered members generating 92.9% of mainland revenue represents genuine Switching Cost — not the contractual kind, but the collection kind. A customer nine figures into a series does not casually restart with a competitor's series.¹⁰ This is a real, durable, China-specific asset. It is also, notably, an asset the company has not successfully exported: the overseas online collapse is the direct evidence that international customers were transacting with a character, not with a brand.

Now the actual competitor. TOP TOY, Miniso's collectible-toy arm, launched in December 2020 — the same month Pop Mart listed. Its revenue went from RMB1,461 million in 2023 to RMB1,909 million in 2024 to RMB3,587 million in 2025, up 87.9%.³⁰ It ended 2025 with 334 stores, holds roughly 4.8% share of the Chinese collectible-toy market on Frost & Sullivan data, and has filed for a Hong Kong listing.³⁰ Its growth continued into 2026 — up 51.4% in the March quarter — and it opened its first US stores in Times Square.³¹³²

Three things about TOP TOY sharpen the competitive read, and they cut in both directions.

First, TOP TOY is decelerating too. Miniso disclosed 32.7% growth for TOP TOY in the first half of 2026, guided to roughly flat performance in the second half, and to low-double-digit growth for the full year.³² Whatever is cooling in the Chinese pop-toy market is not a Pop Mart-specific problem.

Second, TOP TOY's model is meaningfully different and structurally weaker on economics. It leans on licensed and externally sourced product; proprietary IP was only about 10% of its sales in the first half of 2026, and its gross margin has held around 32% across three years — less than half Pop Mart's.³⁰³² That gap is the clearest quantitative evidence that Pop Mart's owned-and-exclusive-IP position is worth something real.

Third, and least comfortable for Pop Mart: TOP TOY is proving that proprietary IP can be built from a standing start. Its flagship character Nommi passed RMB300 million in cumulative GMV, and YOYO — launched in June 2024 — cleared RMB100 million globally within six months and did RMB165 million in the March 2026 quarter alone.³¹³² Those are Twinkle-Twinkle-shaped trajectories, achieved by a competitor with a third of Pop Mart's gross margin.

Running Porter's five forces across this board produces a mixed verdict. Rivalry is intensifying, with a well-capitalised number two about to access public equity markets. Barriers to entry are low in capability and moderate in brand. Buyer power is individually negligible but collectively enormous, because the buyer is a fashion-sensitive young consumer with no switching cost beyond her own collection. Supplier power is low on manufacturing and rising on artists. Substitutes are the broadest threat: Pop Mart does not compete only against other toys but against every discretionary purchase that delivers a small hit of identity and delight — sneakers, cosmetics, 谷子经济 anime merchandise, concert tickets, mobile games.

The correct conclusion is not that Pop Mart has no moat. It is that the moat is narrower and shallower than a 72% gross margin implies. It consists of owned-IP economics that competitors demonstrably cannot match, plus a Chinese membership and channel asset that is genuinely hard to replicate, plus a demonstrated curation capability with a short track record. It does not consist of a structural barrier. Anyone framing Pop Mart as uncatchable is not reading the disclosure of the company chasing it.

↑ Back to Top

The Bull Case and the Bear Case

The bull case, stated at its strongest.

Pop Mart is an IP portfolio company that the market persistently misprices as a fashion cycle. The evidence for the portfolio thesis is the first half of 2026 itself: the flagship character declined 7.5% and the group still grew 23.8%, because a character signed in 2024 delivered RMB2.65 billion.¹⁰ That is the succession mechanism working in real time, under observation, in the worst possible conditions.

Underneath it sits a Chinese business compounding at 47% with 38% growth in store revenue on twelve net new locations — that is productivity, not square footage.¹⁰ It sits on 82 million mainland members transacting through owned channels, including a gamified Pop Draw format that grew 83.3% and has no Western equivalent.¹⁰ It sits on gross margins near 70%, no meaningful debt, RMB12.4 billion of cash, and management returning approximately all of prior-year earnings to shareholders.¹⁰

And the optionality is genuinely unpriced: a theme park drawing destination traffic and 70% attendance growth, a Sony feature with a proven family-franchise director, and a global brand-awareness position that a decade of marketing spend could not have bought.²⁴²⁵ On Helmer's framework the company plausibly holds Cornered Resource, emerging Branding, real Switching Costs in China through collection completion, and Scale Economies in a supply chain that has proven it can flex tenfold.

The bull's core claim: 2026 is a digestion year in which an unsustainable Labubu comparison washes out, after which a broader, more balanced portfolio with a bigger store base and a larger member file compounds from a reset level.

The bear case, stated at its strongest.

Pop Mart is a hit-driven business with the cost structure of a retailer, and the market has consistently paid a platform multiple for a studio's earnings volatility.

The historical record is unkind to the durability claim. Molly peaked at 27.1% of revenue in 2019 and fell in absolute terms the following year; PUCKY went from 18.7% of revenue to 4.1% in two years and out of the disclosure table entirely; The Monsters has now posted its first decline, having peaked at 38.1% just one year after reaching 23.3%.⁸⁹¹⁰¹¹ The pattern is consistent and has now repeated three times.

The overseas story is worse than a slowdown; it is a failed brand transfer. Online revenue fell 45.6% in the Americas, 59.0% in Europe and 39.8% in Asia Pacific in a single half.¹⁰ Pop Mart's own COO said the overseas base was Labubu-driven and unfamiliar with other characters.⁴ The company spent 2025 building a store network, hiring sales staff from 6,219 to 9,734, and expanding leases — lease-related expenses rose 43.3% and employee benefit expenses 45.7% — sized for a demand curve that has since inverted.¹⁰ Inventory days went from 123 to 201.¹⁰ That is the 2022 setup with a global cost base attached.

The valuation implication is the crux. At roughly HK$203 billion of market capitalisation in early September 2026 against RMB13.0 billion of FY2025 net profit, the stock is not obviously expensive on trailing earnings.³⁹ But trailing earnings reflect a peak-Labubu year that management has explicitly said will not repeat. The bear's question is not "what multiple of last year" but "what is the normalised earnings power of a business whose largest franchise is declining, whose international online channel has halved, and whose margins are compressing" — and that number is not disclosed by anybody, including the company.

Layer on the governance overlay a short-seller would write. Pop Mart is a Cayman-incorporated entity operating through consolidated affiliated structures, with GWF Holding Limited as ultimate holding company and a founder who is simultaneously chairman, CEO, and dominant shareholder.¹² Guidance has been revised sharply within twelve months. The interim dividend was skipped in favour of a buyback with a range wide enough to mean RMB2 billion or RMB5 billion. Other income was flattered by a sevenfold increase in government grants. Inventory is building into a demand decline. None of these individually is damning; collectively they are the checklist an activist would open with.

Weighing it. The bear case has the better evidence on the specific question of IP durability, and it is the company's own data that supplies it. The bull case has the better evidence on the question of institutional capability — the China engine, the member file, the supply chain, and one clean succession. The synthesis is that this is neither a broken business nor a compounder: it is a genuinely profitable, cash-generative, well-run hit factory whose central uncertainty — the rate at which it produces hits relative to the rate at which they decay — is empirically unresolved with a sample size of two-and-a-half characters. Anyone claiming confidence in either direction is claiming knowledge that does not exist in the public record.

↑ Back to Top

Risk Radar and the Second Layer

Only the risks with a live business mechanism are worth listing, and four qualify.

Demand cyclicality with negative operating leverage. This is not a hypothetical; it happened in 2022, when a 2.8% revenue increase produced a 49.3% operating profit decline.¹² The company now carries a far larger lease and payroll base — 105 more net stores year on year and 9,734 sales personnel versus 6,219 — which means the sensitivity is higher than it was, not lower.¹⁰

Inventory and working capital. Inventory turnover days nearly doubled to 201 in the first half of 2026, with RMB6.1 billion of finished goods on the balance sheet, and the company recorded an impairment of inventories of RMB22.8 million in the period against RMB4.3 million a year earlier — small in absolute terms, but the direction is the point.¹⁰ If overseas demand does not recover in the second half, the risk is markdowns, which land directly on the gross margin that carries the entire equity story.

Trade and supply chain. Pop Mart manufactures predominantly in China and sells increasingly into the United States, which makes it structurally exposed to tariff policy. The company has been shifting production, with its Vietnam facility reported at around 10% of total output and additional procurement from contract manufacturers in Vietnam and Malaysia for lower-tier SKUs bound for Western markets, alongside selective US price increases.³³ Input costs are a separate live pressure: management flagged raw material costs running 3 to 5 percentage points higher year on year in May 2026, and the interim report attributes part of the gross margin decline to higher raw material prices.¹⁰²⁰

Currency. The RMB720.3 million exchange loss in the first half of 2026 is not a rounding error against RMB5.1 billion of net profit.¹⁰ A company earning in dollars, euros, baht and won while reporting in renminbi has taken on a translation exposure it did not have three years ago, and it is not disclosed as hedged.

Three lighter second-layer observations, each worth a sentence rather than a section. The FY2025 balance sheet shows deferred income tax assets jumping from RMB147.0 million to RMB1,753.6 million, a large enough swing to be worth understanding from the annual report detail.⁹ The company reported no leverage stress — net current assets of RMB17.4 billion at 30 June 2026 — so refinancing risk is not currently a live concern.¹⁰ And the audit committee comprises two independent non-executive directors and one non-executive director, chaired by Wu Liansheng, with Ngan King Leung Gary holding an accountancy qualification — a conventional structure with no disclosed auditor concerns in the period.¹⁰

What is deliberately not on this list: AI disruption, which has no obvious transmission mechanism to a physical collectibles business, and cybersecurity, which is a generic risk with no company-specific evidence of elevated exposure.

↑ Back to Top

The Three Numbers That Matter

Most of the metrics Pop Mart discloses are interesting. Three are decisive, and an investor tracking only these would understand this company better than one tracking twenty.

One: the revenue of the number-two and number-three artist IPs, in absolute renminbi. Not the leader — the leader's decay is now the base case. The entire investment question is whether the successors are scaling fast enough to cover it. In the first half of 2026, Twinkle Twinkle's incremental RMB2.26 billion more than covered The Monsters' RMB360 million decline.¹⁰ The half in which that arithmetic stops working is the half the thesis breaks. Read the IP table first, every time.

Two: overseas revenue growth, split between online and offline. This is the cleanest available proxy for whether Pop Mart is exporting a brand or exporting a character. Offline growth with online decline — the pattern in the first half of 2026 — means customers browse the brand but search for one product.¹⁰ A return to online growth in the Americas and Asia Pacific, without a new viral catalyst, would be the first hard evidence that international customers have adopted the portfolio rather than a single monster.

Three: gross margin alongside inventory turnover days. These two travel together and jointly reveal whether the company is selling at full price or clearing stock. Margin easing to 69.7% while inventory days climb to 201 is an early-warning configuration.¹⁰ Margin holding while days normalise would indicate demand absorbed the build; margin falling while days stay elevated would indicate markdowns, and would mean the earnings base is lower than it currently appears.

Everything else — store count, member count, park attendance — is context. These three are the scoreboard.

↑ Back to Top

Where the Story Stands

There is a version of this story that ends with a company that learned to manufacture cultural relevance on a schedule, and a version that ends with a company that caught two waves and spent the intervening years building infrastructure for a third that never came.

As of September 2026, the evidence does not permit choosing between them. What it does permit is a more precise statement of what has actually been proven. Pop Mart has proven it can find characters that resonate, industrialise them faster than anyone else in its category, and earn extraordinary margins doing it. It has proven, three times now, that its characters decay. It has proven, once, that it can produce a successor large enough to absorb that decay. And it has proven that a global store network built on the strength of a single character is an expensive thing to own when that character stops growing.

Wang Ning's own language has completed a full arc — from "quite easy" to "internal adjustment" — inside twelve months, without a single external shock to explain the gap. That is the most reliable signal in the entire file, and it is not a signal about Labubu. It is a signal about how much confidence to place in this management team's forecasts, which is: considerably less than the market placed in them in August 2025.

What makes the case genuinely hard, rather than merely uncertain, is that the two possible futures produce identical evidence in the short run. A company with a working hit pipeline and a company that got lucky twice both look exactly like this in the year after a peak: one character declining, one character surging, a domestic base holding, an overseas experiment retrenching, and management asking for patience. There is no test available today that separates them. The separation happens over the next two or three character cycles, and it happens in the artist-IP table rather than in anything management says at a briefing.

The interesting thing about a hit factory is that it always looks like a platform on the way up and always looks like a fad on the way down, and the truth is neither. It is a business whose value depends entirely on a hit rate that nobody, including the people running it, can yet estimate with confidence. Pop Mart has one clean succession on the board. The second half of 2026, and the character that follows Twinkle Twinkle, will start to tell us whether that was a system or a coincidence.

Pop Mart International Group Limited

Table of Contents

Pop Mart: The Company That Sold Feelings, and the Year the Feeling Changed

Cold Open: Two Sentences, Twelve Months Apart

Origins: A Stationery Shop in Zhongguancun

The Blind Box Machine: How Pop Mart Actually Makes Money

Molly's Lesson: What Happens After an IP Peaks

Listing, Boom, and the 2022 Reckoning

Labubu: Anatomy of a Global Mania

FY2025: The Record That Broke the Stock

The Turn: H1 2026 and the "Business Adjustment Year"

Management Under the Microscope

The People in the Room

Beyond the Toy Aisle: Parks, Film, and the Disney Question

Myth vs. Reality

War Game: The Competitive Board

The Bull Case and the Bear Case

Risk Radar and the Second Layer

The Three Numbers That Matter

Where the Story Stands

References