Why is the Three Sisters no longer delicious?

Author: Chen Dan

A doll, a piece of gold jewelry, and a cup of milk tea once supported a market value of more than HK$730 billion in the Hong Kong stock market.

In June 2025, the stock prices of Bubble Mart, Laopu Gold, and Mixue Group hit new highs. The total market value of the three companies once exceeded HK$730 billion, becoming the hottest “three consumer sisters” in the capital market. More than a year later, the stock prices of the three companies have fallen by more than half from their respective highs, with Laopu Gold falling by nearly three-quarters.

The latest bad news comes from Bubble Mart. On October 27, the company disclosed third-quarter operating data. Overall revenue fell by 45% to 50% year-on-year, and the American market dropped by as much as 70% to 75%. A year ago, LABUBU was sweeping the world, and its revenue was still growing at a rate of 245% to 250% during the same period. Growth took a turn for the worse. When the market opened the next day, Bubble Mart’s share price fell by more than 6%.

For the other two, even though profits were still growing significantly, they failed to prevent their share prices from falling.

Laopu Gold’s revenue increased by 60.3% in the first half of this year, and its adjusted net profit increased by 83.6%. However, its stock price has retreated the most among the three companies. Michelle Group’s revenue increased by 2.3% during the same period, and its net profit fell by 14.7%. It also failed to escape the downward trend of its stock price.

Trendy toys, gold jewelry, and affordable tea drinks have different growth curves, but they have encountered similar situations. Last year, the capital market was willing to pay a high price for the growth imagination brought about by blockbuster IP, high-end brands and chain expansion; now, even if there is still growth, investors are no longer willing to give the same high valuation.

At the same time, technology sectors such as AI computing power and semiconductors have become new investment hot spots in the market. Compared with consumer companies whose growth is gradually returning to normal, the demand expansion and technological breakthroughs in the AI ​​industry chain provide investors with another imagination about future growth.

Is the collective fall out of favor of the three consumer sisters due to their own performance being unable to support the high valuations in the past, or because the market has found a more attractive investment direction?

From consumption to AI, the capital market has a new favorite

In 2025, the popularity of the three consumption sisters is supported by solid demand and performance. Shen Meng, director of Chanson Capital, believes that this round of rise is driven by structural changes in domestic consumer demand. Consumer preferences such as self-pleasure, value preservation, and low prices have formed new trends, driving the rapid growth of related companies. The rise of the three companies reflects these different consumer needs.

In the first half of 2025, the international gold price hit new highs repeatedly, and long-established gold stores also set off queues. The Shanghai Yuyuan store once had a queue that lasted for 10 hours. Consumers are willing to pay a premium for ancient craftsmanship, design and branding, and also value the value support of pure gold materials. In the same year, LABUBU changed from a collection in the cabinet to a fashion accessory that young people hang on their bags, and the craze spread all the way overseas. One sells gold and the other sells dolls. They may seem unrelated, but they both make consumers willing to pay for value other than the product itself.

Image source / Unsplash

On the other side, Mixue Bingcheng has made a big business out of products that only cost a few yuan. Behind the 2 yuan ice cream and 4 yuan lemonade are the cost advantages formed by large-scale procurement, self-built production and logistics distribution. By the end of June 2025, Mixue Group will have 53,014 stores globally, of which more than half of the stores in mainland China are located in third-tier and below cities.

These three seemingly different consumption choices have all brought considerable performance growth. In 2025, Bubble Mart’s plush product revenue increased by 560.6% year-on-year to 18.708 billion yuan, and overseas revenue increased by 291.9%; Laopu Gold’s revenue in the first half of the year increased by 251%, and net profit increased by 285.8%; Mixue Group’s revenue increased by 39.3% during the same period.

The three companies respectively found consumers who were willing to pay for emotional value, brand and gold materials, and low-priced goods. This growing scarcity was an important reason why the capital market was willing to pay high valuations at that time.

In August 2025, Bubble Mart’s share price was close to HK$340, its market value exceeded HK$450 billion, and its price-to-earnings ratio exceeded 105 times. The price-to-earnings ratio of Laopu Gold once reached about 140 times. Even Mixue Group, which has a relatively low price-to-earnings ratio among the three, has a price-to-earnings ratio of more than 45 times. A price-to-earnings ratio of 100 times means that every 1 yuan of annual net profit corresponds to 100 yuan of market value. Investors are willing to pay such a price, obviously not only buying immediate profits, but also the expectation of continued growth for many years to come.

In Shen Meng’s view, the capital market often looks for reasonable explanations for short-term stock price increases, packaging the company’s phased advantages into investment stories with long-term growth. But being given a certain concept does not mean that the company truly has the ability to continue to fulfill this expectation.

The risks began to show up early. On June 5, 2025, Hong Kong stocks’ new consumer leaders collectively pulled back, with Laopu Gold falling 9.05% in a single day and Mixue Group falling 7.72%. Just the day before, the three companies had hit record highs in share price. At that time, Wells Fargo Fund Manager Zhou Wenbo reminded that new consumer companies have attracted a lot of financial attention, but some companies are no longer cheap and there is a risk of short-term stock price adjustment.

Since then, the stock price trends of the three companies have diverged, and Bubble Mart also set a higher record in August. However, market concerns about whether high valuations can be sustained have continued to arise. Starting from the second half of 2025, AI will become a growth direction that attracts more attention in the capital market. As funds are concentrated in technology sectors such as computing power and semiconductors, the popularity of popular consumer stocks in the early public offerings has gradually cooled down.

According to Choice statistics, the number of public funds heavily invested in Bubble Mart has shrunk from 311 at the end of the second quarter of 2025, to 136 at the end of the fourth quarter of 2025, to 110 at the end of the first quarter of 2026. At the end of the second quarter of 2026, there were only 86 shares left, a cumulative reduction of more than 70%; the number of shares held dropped from 33.2482 million shares at the end of the first quarter of 2026 to 21.1468 million shares at the end of the second quarter, a decrease of approximately 36% in one quarter.

At the same time, the AI ​​industry chain has gradually become the focus of allocation of public funds. According to statistics from Guosen Securities, among the fund’s heavy holdings in the second quarter of 2026, the share of the overall stock market value of the technology sector increased by 23.6 percentage points to 60.8%; as the industry with the largest proportion within the technology sector, electronics行業The proportion of stock market value increased by 21.2 percentage points to 42.9%. The proportion of the consumer sector fell by 7.6 percentage points to 9.5%.

This change is more intuitive in the position adjustment of the same fund. In the second quarter of 2025, among the top ten holdings of the China-Canada Emerging Consumer Fund, there are also consumer companies such as Bubble Mart, Jiangxin Home Furnishing, and Gu Ming. By the first quarter of 2026, technology companies such as Cambrian and GigaDevice have entered its top ten holdings. The fund manager explained in the quarterly report that technology changing life remains the main line of industry trends throughout the year.

Shenwan Hongyuan pointed out in the research report for the second quarter of 2026 that the technology track has a capital siphoning effect on other industries. On the one hand, technology hardware has attracted allocation funds from other industries; on the other hand, technology-themed funds have attracted more investors to subscribe based on their performance.

Shen Meng believes that this is related to market liquidity and capital rotation. “The scale of market liquidity is limited, and funds are rotating between different hot sectors. Funds are chasing as safe and as high a return as possible, and the most critical indicator is policy guidance.”

But the flow of funds to AI does not mean that AI caused the decline in consumer stocks.

Investor Chen Momo believes that from a timeline perspective, the high valuations of the three consumer sisters have already caused adjustment pressure before the siphon effect of AI funds became apparent. The AI ​​market has intensified the diversion of funds and may cause consumer stocks to fall deeper and faster, but it is not the root cause of the decline.

“The move of money amplified the decline, but did not create it.” Chen Momo said that even without AI, the capital market will look for other assets with higher growth expectations, and consumer stocks may still face adjustments, but the magnitude may be smaller and the speed slower.

By the third quarter of 2026, public funds’ allocations to the electronics and communications industries have also declined, while pharmaceuticals and some consumer industries have regained their allocations. Funding preferences are still changing, but for the three consumer sisters, if they want to regain the favor of the market, the answer is how far their growth can go.

The three businesses can no longer support the original valuation?

For the same problem, the three companies faced three different tests.

The problem for Bubble Mart is how long a super IP can be popular.

In 2025, the revenue of THE MONSTERS series will be 14.16 billion yuan, a year-on-year increase of 365.7%, accounting for nearly 40% of Bubble Mart’s total revenue. LABUBU, which is popular all over the world, comes from this series.

Now, the heat is fading. In June 2025, with the official centralized replenishment, the second-hand prices of some LABUBU styles fell rapidly. In June 2026, LABUBU 4.0 “Retro Barber Shop” series was launched, and the market response was obviously not as good as before. Deutsche Bank’s channel research shows that the first-day sales performance of this series of e-commerce products is only 30% to 50% of that of the previous major IP new products in 2026, and the second-hand price of some regular models even fell below the official selling price of 159 yuan.

New IP is still growing, but it has yet to fill the gap left by the overall slowdown. In the first half of 2026, Xingxingren IP revenue increased by 580.6% year-on-year. However, at the performance meeting in August, founder Wang Ning admitted that the company would most likely not be able to achieve the 20% growth target set at the beginning of the year this year.

Chen Momo believes that the market did not initially bet on LABUBU becoming an “evergreen tree”, but expected Bubble Mart to become an IP incubator that continues to produce hit products, continuously launch new IP, and distribute it through global channels. But so far, this ability to continue to replicate hot hits and use potential energy to continue expanding distribution has yet to be further verified.

The problem for Laopu Gold is whether its profits can grow sustainably and steadily.

In the first half of 2026, Laopu Gold’s revenue was 19.808 billion yuan, and its net profit attributable to the parent company was 4.267 billion yuan, a year-on-year increase of 88.2%. But when split into quarters, performance fluctuations become apparent. At the end of February this year, Laopu Gold raised product prices by about 20% to 30%, triggering a round of concentrated buying before the price increase. According to the first quarter performance forecast range and the semi-annual report, the net profit in the second quarter is approximately 467 million to 667 million yuan, which is far lower than the 3.6 billion to 3.8 billion yuan in the first quarter.

Chairman Xu Gaoming explained at the performance meeting that the international gold price fell sharply in the second quarter, coupled with the weakening consumer market, the company chose to wait and see for the time being and was not in a hurry to adjust its business strategy. He said that the average monthly profit in the second quarter still exceeded 200 million yuan, in line with the company’s expectations.

Brokerages are more cautious in their judgments. In August 2026, Kaiyuan Securities lowered its forecast of Laopu Gold’s net profit attributable to the parent company from 9.012 billion, 12.305 billion, and 15.537 billion yuan to 6.535 billion, 6.743 billion, and 8.003 billion yuan respectively from 2026 to 2028. According to the adjusted forecast, Laopu Gold’s profit growth in 2027 will be only about 3%.

The brand premium of Laopu Gold is still there, overseas revenue has maintained rapid growth, and new categories such as goldware are also contributing to the increase. However, the rapid growth in the past was partly due to rising gold prices and concentrated buying before price adjustments. Next, it needs to prove that even without these factors, repurchase, new products and channel expansion can still support performance growth.

The problem for Mixue Group is that as more and more stores are opened, profits are falling instead of increasing.

In the first half of 2026, Mixue Group’s net profit was 2.319 billion yuan, a year-on-year decrease of 14.7%. It was the first time that the half-year profit fell year-on-year after listing. As of the end of June, the group’s global stores reached 63,987, a year-on-year increase of approximately 20.7%, but revenue during the same period only increased by 2.3%.

Store openings are also slowing down. In the first half of the year, Mixue Group opened 5,455 new franchise stores, 2,266 fewer than the same period last year; a net increase of 4,166 franchise stores, down from 6,534 in the same period last year.

Picture source / Mixue Bingcheng official website

Management admitted at the performance communication meeting that in the first half of the year, the per-store sales of the group as a whole and Mixue Bingcheng’s main brand both experienced double-digit declines. Calculated based on the sales revenue of goods and equipment and the average number of franchise stores, the purchase amount of a single store from the headquarters in the first half of the year dropped by approximately 17.9% year-on-year.

This change goes to the heart of Mixue Group’s business model. The group’s revenue mainly comes from the sales of raw materials, packaging materials and equipment to franchise stores. In the past, the more stores there were, the more goods the headquarters sold; but as stores continue to be denser, single store business is more likely to be diverted, and the revenue increment brought by new stores may also gradually shrink.

Meanwhile, fees are still rising. In the first half of the year, Mixue Group’s sales and distribution expenses increased by 22.9%, and administrative expenses increased by 39.4%, both significantly faster than revenue growth. The company explained that part of the expenditure was used for product quality upgrades, brand building and store operation support.

For Michelle Group, the question has turned to “how much more money can be made per store.” Whether the headquarters’ revenue can continue to grow ultimately depends on whether franchisees can maintain profitability and whether they are willing to continue to open stores.

Shen Meng concluded that cost advantages are easily affected by upstream price fluctuations, while the differentiated advantages brought by emotional value require companies to continuously launch new products and experiences. Their competitive advantages have not disappeared. Bubble Mart still has an IP matrix, Laopu Gold still has a brand premium, and Mixue Group still has more than 60,000 stores and a mature supply chain system. What has changed is not the businesses themselves, but the capital market’s expectations for their future growth.

“Although their performance is different, investors’ expectations for their future are basically the same, which is not enough to support the current valuation.” Shen Meng said.

AI took over the premium,

I also took the same test question

So, will the valuation of AI assets, which have become the new favorite of the capital market, be more confident?

So far, the answers are inconsistent. In the AI ​​industry chain, some companies have turned demand into profits, while others are still in the high-investment stage, and the capital market’s pricing for them has also shown clear differentiation.

Let’s look at the big model companies first.

In January 2026, Zhipu and MiniMax were listed on the Hong Kong stock market. In the first half of this year, Zhipu’s revenue was 954 million yuan, a year-on-year increase of 399.7%, and its adjusted net loss was 1.964 billion yuan, which expanded year-on-year; MiniMax’s revenue was approximately US$117 million, a year-on-year increase of 283%, and its adjusted net loss was US$293 million, a year-on-year expansion of 111.2%.

Both companies’ revenue increased several times, but their adjusted net losses were still more than twice their revenue during the same period. Model training, inference and R&D require continuous investment, and the rapidly growing revenue has not yet been converted into profitability. The attitude of the capital market has also changed. After the two companies went public, their stock prices rose sharply; entering 2026, they experienced deep retracements one after another. As of October 20, the share prices of Zhipu and MiniMax have fallen by more than 80% from their respective historical highs.

In contrast, funds prefer AI hardware companies that have realized profits.

Image source / Unsplash

Li Xiangjie, fund manager of China Asset Management, previously pointed out that the current investment focus is on upstream hardware, while downstream applications are still in the differentiation and observation stage. Shen Meng also believes that the popularity of big model stories is fading, and in the short term the capital market will pay more attention to AI hardware companies whose performance has been realized.

Zhongji InnoLight is the representative. In the first half of 2026, the company’s revenue was 41.778 billion yuan, a year-on-year increase of 182.49%; the net profit attributable to the parent company was 13.651 billion yuan, a year-on-year increase of 241.70%. Major AI customers continue to expand data centers, driving growth in shipments of 800G and 1.6T high-speed optical modules.

The demand side is also still expanding. Gartner predicted in September this year that global AI spending will reach US$2.67 trillion in 2026, a year-on-year increase of 49.5%, with AI infrastructure being the largest spending direction.

Some AI hardware companies are in a stage of rapid growth in both revenue and profit, but this does not mean that valuations can rise indefinitely. Once growth slows down, they will still be recalculated. This is exactly what the three consumer sisters are experiencing.

Institutional adjustments have already released some signals.

According to statistics from Huafu Securities, in the third quarter of 2026, the allocation ratio of active equity funds to the electronics and communications industries continued to fall, falling by 6.81 and 1.69 percentage points respectively from the previous quarter. The proportion of electronic allocations fell to about 36%, and communications dropped to about 15%.

Zhongji InnoLight is still the largest holding of public funds, but the number of funds holding it and the number of shares held have decreased compared with the end of the second quarter. Electronics and communications are still important allocation directions for public funds, and institutions have not completely withdrawn from AI. However, as stock prices rise and positions become increasingly concentrated, some funds have begun to reduce their positions and re-weigh future returns and valuation risks.

In Shen Meng’s view, consumer demand still lacks obvious support in the short term, and the technology sector still has a growth narrative. Although some funds have begun to allocate consumer stocks after the correction, they may not be able to reap returns soon.

When most institutions were reducing their holdings of Bubble Mart, Duan Yongping chose to buy.

In August 2025, Duan Yongping still said he “can’t understand”, but after Bubble Mart released its 2025 annual report, he changed his position. Starting in April 2026, he began to make arrangements by selling put options (selling Put). Subsequently, through multiple rounds of holding increases, he became the second largest shareholder after Wang Ning. In the third quarter, after Bubble Mart’s revenue was nearly halved, Duan Yongping still said that he “will buy some more”, with the upper limit being 10% of the total equity.

Public funds are reducing their holdings, while Duan Yongping is adding to their holdings. This seems to be two opposite judgments. In Chen Momo’s view, different investors may have different holding periods. Buying a company’s stocks is mainly based on the judgment of trading opportunities. The key is to see whether there is room for profit between the current stock price and the expected value.

With the same performance and the same risks, some people are worried about slowing growth, while others believe that the stock price has fully digested this concern. How to price a company has always been an eternal problem in the capital market.

*The title picture comes from Unsplash.