On September 1, 2026, at the Hong Kong Stock Exchange, it was not the founder of Hision who rang the listing bell.
The two people ringing the bell were Gemma Dunne, the UK brand director, and Du Jiabin, a warehouse manager. Meanwhile, Xu Yangtian, who was supposed to be at the center of the stage, wore a cultural shirt with the "SHEIN" logo, stood in the corner during the photo opportunity, and did not speak. He left quickly after the event. The founder, born in 1984 in Zibo, Shandong, has rarely appeared in public over the past 20 years, with his most recent appearance being at the Guangdong High-Quality Development Conference at the beginning of the year.
There's a reason for the low-key approach. Listed at HK$48.56, the stock fell at the opening and plummeted by over 10% at one point during trading. Based on the issue price, HEYTEA's market capitalization is approximately $26.5 billion - a far cry from the $98.2 billion valuation it received during its Series D financing in 2022. In just four years, one of the world's most valuable apparel companies has shrunk by more than 70%.
A company with 273 million active users, 160 markets, and $41.8 billion in annual revenue - why is the capital market only willing to give it a quarter of its worth? Tariffs and competition are answers that everyone can see, but they are just the cover of the bill.
The real problem is written on the bill.
Yu Xiangtian in the Corner
First, look at how difficult the puzzle itself is to solve.
Xiyin's revenue growth has followed a standard downward curve: a 41.1% increase in 2023, 20.7% in 2024, 8% in 2025, and just 1.1% in the first quarter of 2026. Although revenue is still rising, the growth has almost come to a halt. In the US market, revenue fell 14% year-over-year in the first quarter of 2026, with the proportion of the company's revenue declining from 29.4% to 22.5%, slipping behind Europe (35.4%) to become the second-largest market.
The decline in profit is more direct: with a net profit of $278.9 million and a net profit margin of 8.7% in 2023, and $206.4 million and 4.9% in 2025, more than halving. In the first quarter of 2026, the company even recorded a net loss of $99 million on its books - which the company explained was due to an accounting adjustment for the change in the fair value of preferred shares, while still having an operating profit of $258 million. This explanation is technically valid, but the phrase "accounting adjustment" in the listing documents has never been able to calm the secondary market.
User data is also subtle. The number of active users increased from 186 million to 273 million, which looks good, but the amount each user spends on Xiying Flower each year decreased from $173 to $153; the annual order frequency remained at 4 times, unchanged for three years. New customers are still coming in, but each customer is buying less and not more frequently.
The capital market gives valuations based on tomorrow's growth rate, not today's profits. When the growth rate drops to zero, 98.2 billion becomes 26.5 billion.
Xiying itself, as well as most analysts, attribute the problem to two invoices.
The first one is the tax invoice.
On May 2, 2025, the United States formally terminated its duty-free policy for small packages from China valued at less than $800 (de minimis). The policy, originally intended for personal use in declaring luggage, was exploited by cross-border e-commerce companies, which turned logistics into an art form: companies like Shein and Temu would break down entire containers of goods into millions of small packages, each below the duty-free threshold, to bypass tariffs and ship directly to American consumers' doorsteps.
This is the implicit financial foundation of the "small orders, quick returns" model - direct mail is cheaper, with half the savings coming from tax exemptions.
The foundation is gone. The EU is next to dismantle: starting July 1, 2026, the EU will fully cancel the tariff exemption for parcels under 150 euros and charge a handling fee per order. This is a blow to Hi Audio - Europe is already its largest market. Two doors have been closed in twelve months, and the line in the prospectus that reads "fulfillment costs as a percentage of revenue rose from 42.1% to 47.7%" is the result of being squeezed by the door.
The second is the bill for competition. In February 2023, Temu spent heavily to purchase a Super Bowl ad spot, with the slogan "Shop like a Billionaire," making it the first time Americans remembered this Pinduoduo-affiliated app. Since then, Temu has burned through approximately $3 billion in advertising expenses over the past year, at one point becoming the largest online advertiser in the US.
The same Chinese supply chain, the same low-price mindset, and the same traffic pool from Meta and Google, TikTok Shop and Pinduoduo are facing the same challenges, with customer acquisition costs being driven up to a level that is unbearable for everyone. Xiaomi's own marketing expenses have increased by nearly 80% over the past three years, reaching $6.19 billion in 2025, accounting for 14.8% of its revenue.
By the way, the entire fast fashion industry is also slowing down: Inditex, the parent company of ZARA, saw only 0.8% growth in its 2025 fiscal year's first quarter, excluding exchange rates, and will face a 3% tariff impact on its full-year sales; H&M's sales in March 2025 fell 1% year-over-year. The tide is receding overall, and XiYin is the first to be exposed as "swimming naked" - because its business model is the most thoroughly dependent on being "cheap".
Both invoices are genuine, but looking at just these two cannot explain a more fundamental phenomenon: why Xiyin has almost no buffer when faced with tariffs and competition. Nike will hurt from tariffs, and ZARA will hurt from tariffs, but after the pain, they are still Nike and ZARA. When Xiyin hurts, it's a direct fracture.
The difference lies in bone density, which has been spared over the past decade or so.
Xiyin's supply chain efficiency remains the industry's ceiling: new styles start with trial orders of 100 to 200 pieces, best-sellers can be restocked in as little as 5 days, while slow-sellers are immediately discontinued; inventory turnover is 36 days, compared to ZARA's 71 days, Uniqlo's 114 days, and Adidas' 164 days. This system is supported by over 1,700 self-developed software systems, connected to more than 7,500 manufacturers.
This system is called LATR - Large-Scale Automated Small-Order Rapid Response, which is presented as a technology story in the prospectus and was also described as such by CFO Gui Lei at the listing ceremony.
But the other side of efficiency is two sums of money that never need to be paid.
First, there's the issue of design fees. The essence of fast fashion is copying and modifying high-end brands, which is no secret in the industry, and ZARA also started out this way. The difference lies in the fact that ZARA has a dedicated team of designers on its payroll, for which it pays a significant amount of money; whereas, Xiyin has outsourced "design" to its supply chain - a large number of styles are developed by suppliers under the ODM model (where the supplier is responsible for design and manufacturing), and Xiyin is responsible for using algorithms to select styles and provide traffic. Normally, no one pays much attention to whose design or whose pictures they are, until someone takes it seriously.
On August 13, 2026, the High Court of London issued a 77-page ruling. The case initially seemed favorable to Xi Yin: it sued Temu for "industrially" infringing on 2,559 clothing product photos. However, the court rejected all of Xi Yin's claims - because the styles were designed by suppliers and the photos were taken by photographers, the copyrights naturally belonged to the suppliers and photographers; Xi Yin's reliance on a standard clause in the entry agreement stating "the copyright of the product images you sell to us belongs to us" was insufficient to prove the transfer of rights. The most incisive sentence in the ruling was: a standard contract cannot replace complete proof of rights.
The most dramatic piece of evidence was a "strawberry nightgown": a supplier was unable to sell it, so they put their inventory up for sale on Temu, using the same photo - a photo whose copyright was owned by a freelance photographer.
A company that claims to be "the world's top-tier fashion company" cannot even prove the copyright of its own product images. This is not an isolated case of bad luck, but rather a self-indictment of its business model: Shein's control over "fashion" stops at the distribution channel level and has never reached the creative level. The ruling also leaves a pending issue: Temu is countersuing Shein for allegedly violating competition law by using exclusive agreements to restrict over 8,000 suppliers, with the case set to continue in 2027.
Intellectual property issues are not the only problem. Uniqlo has sued it for copying the "dumpling bag"; there are around 50 intellectual property lawsuits pending in the US; in July 2025, the French regulatory agency fined it 40 million euros for "false discounts" - raising prices before offering discounts, and later fined it an additional 22.5 million euros for data compliance issues. The total fines over two years amount to nearly 500 million yuan.
The second free payment is the profit of the supply chain. Xiying has digitized the delivery time, quality, defect rate, production capacity, and quotation of each factory, and the system dynamically allocates orders - this "free management system" is touted as a digital blessing, but suppliers have a different feeling: the cost of raw materials, labor, and compliance has risen year after year, and "costs are hard to pass on upstream". A phrase circulating among factory owners in the Pearl River Delta is: What they fear is not having no orders, but having orders that don't make money.
When comparing Xi Yin with Nike and ZARA, the difference in cost structure becomes apparent. Inditex's gross margin rate has consistently been around 57% to 58%, while H&M's is approximately 53% - high gross margins come from designers, fabric research and development, and brand premiums, which are the moats built with real money every year. Xi Yin's gross margin rate exceeds 60%, reaching 67.9% in 2025, and even higher at 70.3% in the first quarter of this year.
In other words, the "dividend" for many companies going overseas is roughly equivalent to the combined effect of "not paying design fees + not paying tariffs + compressing factory profits".
These three items, one has been blocked by the court, one has been blocked by tariffs, and one is being blocked by ESG compliance audits. Three accounts are due at the same time, which is the accounting expression for "internal friction has reached a boiling point".
There is also another layer of uncertainty written on the passport of XiYin.
In 2024, HiFi's Executive Chairman, Tang Wei, made a widely-circulated comment at the Milken Institute Global Conference: "If you look at where we were founded and our supply chain, HiFi is a Chinese company; if you look at where our headquarters and core team are located, it is a Singaporean company; if you look at our market and value, it is an American company."
This statement was clever in its time - all three sides exchanged a glance. However, looking back at it today in the current geopolitical context, it precisely exposes the deadlock of China's Xi: none of the three completely recognize it. The US Congress has never stopped scrutinizing Chinese-background e-commerce companies, and the path to listing in New York was put on hold; after shifting to London, it was repeatedly questioned by regulators and public opinion about labor issues in its supply chain; after twists and turns, it finally returned to Hong Kong - a four-year, three-city journey, the longest IPO detour in the history of Chinese internet companies.
What may truly determine the timing of the listing is a relatively inconspicuous figure in the prospectus: $17.294 billion in convertible redeemable preferred shares that mature at the end of 2026. The net proceeds from the listing are approximately HK$13.2 billion, with 40% allocated to cloud computing, AI demand forecasting, and warehouse automation. In other words, half of this celebration is an arrangement to repay debt. The cornerstone investors include Boyu, Tencent, and Tiger Global, who subscribed to over HK$3 billion - what the old shareholders need is a liquidity exit.
Xu Yangtian held about 33% of the company's shares before its listing, with ten times the voting power through a dual-class share structure. His low-key stance, standing in the corner wearing a casual T-shirt, is a metaphor for the company's situation: being low-key to the point of near-invisibility, while also having to take the stage.
The story has come to this point, and now the initial question can be answered: why did the trillion-yuan empire depreciate?
Of that 1 trillion yuan, a considerable portion was borrowed to begin with, taking advantage of tax loopholes, ambiguous copyrights, and factory concessions. In a bull market, no one came to collect debts, and valuations served as collateral; but when the winds changed, creditors lined up at the door, and valuations were directly calculated based on the ability to repay debts. The answer given by the capital market this time is ruthless and honest: it will no longer pay a brand premium for "efficiency created through rolling up", only willing to pay for "what's left after the bills are paid".
What's left after paying off debts? The answer given by Hi Audio itself is hidden in its revenue structure: platform service revenue - the fees collected from opening up supply chain, logistics, and traffic to third-party merchants - rose from $868 million in 2023 to $4.74 billion in 2025, with the proportion increasing from 2.7% to 11.3%, and a profit margin of about twice that of the group as a whole. The profit margin for selling goods is 4.9%, while the profit margin for "selling water" (selling its own supply chain capabilities to others) is twice that. This math problem has been demonstrated by Amazon with AWS, and Alibaba is demonstrating it for the second time with its cloud business. Hi Audio wants to demonstrate it for the third time.
Charlie Munger reflected on his loss of tens of millions of dollars from investing in Alibaba in his later years, saying it is still a "damn retailer" - the retail business is one where competition never subsides and profits are as thin as a knife blade. This statement also applies to Xiying, but Xiying's situation is even more complicated: it wants to graduate from being a "damn retailer", on the premise that it first pays the tuition fees for being a student - the tuition fees for design, compliance, and branding.
The real point of upgrading is actually in Panyu. Among the more than 7,500 factories of Hiiye, the domestic portion is retreating to sampling and small orders, while large-scale orders are accelerating their relocation to Vietnam, Turkey, and Brazil. For those clothing factories that grew up with Hiiye 20 years ago, listing on the platform is not the end goal; the phrase "having orders does not equal having profits" is what truly matters. The more astute among them have already begun to focus on fabric research and development, as well as process upgrades – trying to break out of the cycle of "receiving orders, processing, and earning processing fees".
Over the past twenty years, China has manufactured the sharpest knife at sea, called efficiency. Shein has proven that this knife can cut through ZARA's price tags, and has also proven the double-edged nature of the knife: the edge facing outward is called competitiveness, and the edge facing inward is a one-time depletion of profits for the entire chain. The depletion does not disappear, it is only deferred.
From 100 billion to 25 billion, what's being erased isn't a company's market value, but the market's re-pricing of a business model: being fast doesn't mean being strong; being cheap doesn't mean having a moat; the efficiency gained from pushing others to the limit will eventually have to be paid back with interest.
Fortunately, the first step in settling accounts is to acknowledge that every entry in the ledger counts.
On September 1, Xu Yangtian stood in the corner wearing a cultural shirt - not ringing a bell, not speaking. The cultural shirt at least indicates one thing: this time, the brand on the clothes is 100% his own.