ALL SCAM! From Bubble Tea to AI: Is China’s IPO Boom Hiding a Much Bigger Financial Crisis

Hong Kong's Securities and Futures Commission, or SFC, ordered the Hong Kong Stock Exchange to immediately suspend trading in Sichuan Bai Cha by Dow Industrial at 9am on October 13th. Known publicly as Cha Panda, the tea chain built its brand around a panda mascot and fruit teas. Just two years ago, it was celebrated as the first franchise-based tea chain to go public. Now it faces an investigation into suspected manipulation of investor demand during its listing. Cha Panda issued about 147.8 million shares at HK$17.5 each on April 23, 2024. It raised roughly HK$2.5 billion. But retail investors in Hong Kong showed little interest. The public offering received just 4,593 valid applications, covering only half the shares initially set aside for them. The retail allocation was cut from 10% to 4.96%, with the remaining shares shifted to the international offering. On paper, international demand reached 1.11 times the shares available, enough to get the deal through. But ownership was highly concentrated. The five largest investors received 52.1% of the international allocation. The 10 largest took 86.75%. Retail demand was weak, institutions showed no rush to buy, and there were no major cornerstone investors committing to backing the deal. Yet the offering still raised more than HK$2.5 billion. Before trading was suspended, Cha Panda's share price had fallen to just over HK$4, about 77% below its HK$17.5 offering price. Its market value had dropped from roughly HK$25.8 billion at listing to less than HK$6 billion. The regulator's announcement outlined how the suspected scheme worked. Under undisclosed agreements between the listed company and certain groups, some IPO proceeds may have been transferred out as investments in financial products. That money may then have been secretly routed back to major investors in the offering, or people connected to them. In simple terms, money raised from the listing may have subsidized investors who helped create the appearance of strong demand. The SFC has identified two groups suspected of manipulating IPOs. The first centers on Sun Run and eight related companies. Sun is a licensed representative of Harvest Global Investments. The companies are registered in the British Virgin Islands, the Seychelles, and Samoa. The regulator issued restriction notices to six brokerage firms, blocking transactions involving up to HK$4.38 billion in the accounts. That's a record for an SFC case of this kind. Regulators suspect the group took part in fraudulent arrangements involving the IPOs of both Cha Panda and Cloudbreak Pharma. Trading in Cloudbreak Pharma was ordered suspended on September 10. The second group consists of Liu Haoren. The Reynolds-Lemkin Group, Asia Limited, and Kauffman & Company, PanAsia Limited. It's linked to the IPO of Tianzhi Dihe Tech Company. The SFC issued restriction notices to 29 brokerage firms, freezing up to about HK$285 million in related accounts. Regulators suspect both groups used similar methods. Under undisclosed agreements, some IPO proceeds may have been moved out as investments in financial products, then secretly returned to major investors or related parties. The suspected goal was to make demand for the shares look stronger than it was. SFC chief executive Julia Leung said that most market participants follow the rules, but intermediaries must stay alert to suspicious arrangements and question everything that does not make sense. She said the regulator would act without hesitation against those who undermine a fair, orderly, and healthy market. The founders of Cha Panda are a married couple, Wang Xiaokun and Liu Weihong. They opened their first milk tea shop near Wenjiang No. 2 Middle School in Chengdu in 2008. The shop was less than 215 square feet and took takeaway milk tea made from powder mainly to students. That year, packaged milk tea sales had just passed 1 billion yuan. Chains such as Jackhut, T-Storm, and Da Cash were competing at prices of 5 to 7 yuan a cup. The couple later registered the Cha Panda trademark and shifted toward freshly made drinks built around fresh fruit and Chinese tea. They built franchising in Chengdu in 2016. After adding Panda branding in 2018, they expanded nationwide. The chain quickly surpassed 3,000 stores around 2021. Shortly before listing, it brought in investors including Orchid Asia and Laurel Valley Capital, pushing its valuation to about 17.5 billion yuan, or roughly 2.5 billion U.S. dollars. Public records show the couple controlled about 82% of the company through direct holdings and companies they controlled. Their wealth once exceeded 14 billion yuan, or about 2 billion U.S. dollars. By the end of June 2026, the chain had 8,863 stores worldwide, the vast majority of which were in mainland China. 72 were abroad. They worked with over 6,000 franchisees. It appeared to be a successful franchise empire. But before the trading suspension, one person online pointed to signs of strain in Cha Panda's recent performance. I felt pretty bad when I saw this news. Cha Panda is stuck in the middle, squeezed by premium brands above it and cheaper brands below. It's a tough position. It still has more than 8,000 stores, putting it near the top of the industry, behind only Mixue and Gu Ming. It focuses on mid-priced fresh fruit tea and relies on value and delivery orders. Its prices sit between Mixue and HeyTea. It has about 3.5% of the market, ranking 5th. That looks decent, but it's struggling. Revenue rose 6.2% in the first half of this year, and profit edged up, but its gross margin fell 2.6 percentage points. It's growing on the surface, but the business underneath looked shaky. Most of its revenue comes from selling ingredients to franchisees. Franchise fee revenue fell 20%, which shows franchisees really aren't making money. The company says its lower gross margin reflects support for its coffee business, but look at the longer term, and its performance has suffered sharp drops. Net profit was cut in half the year it went public. It bounced back last year, but it still hasn't returned to its earlier level. A total of 1,218 franchisees ended their relationships with Chapanda in 2025. Its franchisee attrition rate reached 17 percent, among the highest in the industry. Revenue from royalties and franchise fees fell 19.5 percent from a year earlier. Franchisees paid real money for the right to use a brand whose market appeal may have been built partly on artificial demand. Supplier payment schedules, equipment purchases, and franchise fee refunds could all face ripple effects. Trading in Cha Bai Dao's Hong Kong shares was suspended on October 13. There are several short-term effects. First, ordinary investors and potential franchisees will want to wait and see. Being a listed company no longer carries the same assurance, so don't rush in. Second, existing franchisees will lose confidence. Some will have to recalculate how long it will take to cover their investment. Headquarters will also probably cut back on marketing subsidies and discounts on supplies. Third, stores in smaller cities and crowded shopping districts face pressure from Yishui and Guoming. Those already barely breaking even will be more likely to close when their lease expires. That means the impact of the suspension will reach beyond the stock market and quickly spread through the franchise network. In the short term, the suspension puts pressure on both funding and confidence across the business. If manipulation is established, lasting damage to the brand could undermine the franchise network. Small business owners and retail investors often end up bearing the losses. A line from the film, Let the Bullets Fly, captures that imbalance. The wealthy get all their money back, while ordinary people's money is split 30-70 between those running the scheme. Luckin Coffee inflated sales and revenue to boost its share price. The suspected arrangement at Cha Panda involves a different method, routing outside money through a circle of transactions to support demand for its own shares, drawing retail investors' money into transfers that benefit connected parties. Posts on X suggest Cha Panda's sudden suspension may be connected to recent rumors involving former Chongqing mayor Huang Qifan. The posts say his companies were key vehicles in the efforts to turn Chongqing's state assets into investment capital. They describe Liao Qingxuan, a former head of Chongqing's state assets regulator and chairman of Southwest Securities, and Wu Jian as key associates who later came under investigation. The posts also allege that records involving Fang Xinghai, the former vice chairman of China's security regulator responsible for overseas listings, point to Huang's son. They accuse him of using his father's influence in financial regulation and local political and business circles. Operating behind the scenes in Hong Kong and Canada, he supposedly shared in proceeds from as many as 29 Hong Kong IPOs. More evidence is needed to establish whether these accounts have any basis. Cha Panda highlights questionable practices in Hong Kong fundraising by traditional consumer businesses. Another wave of risk involves technology companies with higher valuations, bigger promises and faster cash burn. Moonshot AI is a prominent example. Bloomberg reported on October 14 that Bank of America had withdrawn from talks about a possible underwriting role in Moonshot AI's Hong Kong IPO. The bank had discussed serving as an overall coordinator but never received a formal appointment. Days earlier, Bloomberg reported that Moonshot AI was valued at about 50 billion US dollars after its latest funding round. It planned to list in Hong Kong as early as the first quarter of 2027 and was preparing to begin preliminary investor presentations. But the company, promoted as a rising star in China's AI industry, already faces several controversies ahead of its listing. In September, Anthropic publicly accused Moonshot AI of secretly forwarding requests meant for Kimi to Claude, then presenting Claude's responses to users. In one case described by Anthropic, nearly 300,000 customer requests were sent to Claude over 10 days, using 5,380 suspected fake accounts. Anthropic also said, some responses were saved and used to extract information about the model's reasoning to train the company's own systems. Its report named seven Chinese companies accused of large-scale misuse of Claude. Moonshot AI, DeepSeek, Alibaba, ZAI, SenseTime, Minimax, and Xiaomi. Meanwhile, Reuters reported on October 11th that Moonshot founders Yang Zhiling and Zhang Yutao remained involved in an arbitration dispute in Hong Kong with shareholders of Recurrent AI. The shareholders accused them of making false statements, improperly transferring proprietary data, and diverting benefits while setting up Moonshot. They are seeking compensation equal to the value of a 10.5% stake in the company. Biopharmaceutical listings also slowed sharply in 2026. Huihe Medical's application for Shanghai's Star Market was accepted in June and withdrawn in September, less than three months later. The company had also been selected for an on-site inspection. Success Biotech ended its Star Market IPO in October, marking its third failed listing attempt in six years. Tiankang Pharmaceutical, Yan'an Pharmaceutical, and Boron Pharmaceutical also withdrew applications to the Beijing Stock Exchange. Mainland-listed HuaHeng Biotechnology, Northland Biotechnology, and Boron Pharmaceutical cancelled plans to issue H shares in Hong Kong. Boron's controlling shareholder was released on bail pending investigation into suspected securities market manipulation. HuaHeng also became involved in legal troubles linked to its controlling shareholder. In mid-October, Dongcheng Pharmaceutical cancelled plans to spin off Lan attempting for a Hong Kong listing. Its failure to complete a qualifying IPO, as agreed, also triggered an obligation to buy back minority shareholder stakes. These companies are seeking funding through capital markets. Commercial space takes the model further, as the Chinese Communist Party packages state-led industrial projects as market investment opportunities. More than a dozen companies, including Landspace, CAS Space, Galactic Energy and Space Pioneer are lining up to list on Shanghai's Star Market or the Hong Kong Stock Exchange. The listing wave functions as a state-directed funding mechanism. Beijing sets entry rules around national priorities, maintains state dominance in ownership, and retains strategic control. The remaining shares are then sold to funds, institutions, and individual investors. The result is a form of fiscal cost shifting. Spending on industrial policy moves off the government's balance sheet and reappears as private investors' losses, rather than creating productive capital. Under the U.S. and British systems, an IPO allows a private company to sell shares to the public under disclosure rules. Buyers gain real insights to the company's financial returns and a voice in its governance. The CCP uses the same term to present state projects as ordinary market activity. The Star Market has even created a special listing route for commercial space companies. It waives profit and revenue thresholds and instead requires technical milestones. Taking on national missions or participating in major state-led space projects is explicitly listed as a reason for priority support. No Western exchange writes service to a national mission into its listing rules. The funding source matters. Instead of providing long-term subsidies directly through the budget, the government lets companies with unproven business models enter the stock market. Risk moves from the government's balance sheet to funds, institutions, and individual investors. The state can still steer strategy through state-backed funds, policy eligibility, and industry access. Public shareholders bear the costs of falling share prices, technical failures, and dilution from new share issues. Rocket development consumes enormous amounts of capital. A single technical failure can wipe out years of investment. The Shanghai Stock Exchange opened its fifth set of listing standards specifically because many of these companies cannot meet traditional revenue and profit requirements. Policy can lower listing barriers. It cannot reduce the actual likelihood of rocket explosions, weak orders, failed technology, or depleted cash reserves. Financial risks under the CCP have spread from isolated problems into threats to the wider system. The concern goes beyond any single failure or trading suspension. The system has long-spread risks across the public while concentrating gains in private hands. Ordinary people's savings and investors' capital keep absorbing the risks as local government debt grows. Effective market discipline and independent oversight remain lacking. For more than a decade, local governments have borrowed heavily through local government financing vehicles, or LGFVs. The money supported infrastructure expansion and a fiscal model built around land revenue. Total debt has long exceeded publicly disclosed figures. As the property market weakened, revenue from land sales fell sharply. Many financing vehicles now depend on new borrowing to repay old debt and on implicit government guarantees. A default on local infrastructure debt could hit related investments held by banks, trusts, insurers, and retail investors. The CCP often uses administrative measures to contain these failures. Banks are told to renew loans, state-owned companies to take over troubled assets, and local governments to resolve their debts. These steps push risk into the future without covering losses already embedded in the underlying assets. The quality of bank assets is also fragile. Reported bad loan ratios remain relatively low, but loan extensions, debt refinancing and off-balance sheet transfers often conceal the true picture. Small and mid-sized banks, especially rural commercial banks, are particularly exposed. They depend heavily on local property markets and government financing vehicles. Local fiscal trouble can quickly weaken their capital buffers and available cash. If depositors lose confidence, the risk of bank runs can spread from individual institutions to the wider system. Capital markets themselves are also amplifying risk. The IPO manipulation cases involving Chapanda, Cloudbreak Pharma, and Tianzhu Dihe, the compliance disputes surrounding Moonshot AI, and the wave of withdrawn biopharmaceutical listings all reveal problems in share issuance and fundraising. These include coordinated share subscriptions, money routed back to investors, and artificial demand. They also expose delayed oversight and weak market discipline. The CCP uses capital markets to finance industrial policy. Through the Star Market and Hong Kong listings, projects with unproven business models are presented as public investment opportunities. Risks that should fall on government finances shift to funds, institutions, and retail investors. The state retains control, while public shareholders absorb losses. This pattern appears repeatedly in heavily promoted sectors such as commercial space, AI, and biopharmaceuticals. It shifts fiscal costs to investors rather than creating productive capital. Investors keep encountering manipulated IPO demand, inflated financial figures, and sudden problems involving controlling shareholders. That erodes trust and confidence. Capital outflows, weak valuations, and newly listed shares falling below their offering prices become more common. Hong Kong's standing as an international financial center has already suffered. Mainland markets remain caught in a cycle of policy intervention and market deception. An aging population, difficulty finding work among young people, and weaker household spending power further erode the financial resources needed to support rapid growth and high valuations. CCP officials repeatedly promised to prevent and resolve financial risks, but those promises have had little effect without ending improper transfers of benefits between political power and capital, and allowing failed businesses to exit the market. Risks keep accumulating and being pushed into the future. Increasing government intervention and more ordinary people's money are then needed to cover the losses. Chapanda's suspension offers a view into a bureaucracy accustomed to using deception to maintain the appearance of stability and public money to absorb losses. The deeper crisis is its inability to generate sustainable, real returns. It keeps operating by shifting risk elsewhere. If an external shock coincides with a collapse in domestic confidence, risks long held in check could spread quickly through the financial system, government finances, and the broader economy.