China is trying to prevent its booming IPO market from turning into another speculative bubble, warning investment banks to be more selective about which companies they bring public and to avoid aggressive pricing that could expose retail investors to sharp losses after listing.
The message comes after a series of extraordinary debuts on Shanghai’s technology-focused STAR Market, where some of China’s most closely watched semiconductor and robotics companies have climbed more than 400% on their first day of trading.
Chinese regulators have told bankers not to overwhelm the market with lower-quality IPO candidates and have encouraged conservative offer prices, according to the Financial Times. More than 100 Chinese companies have raised over $28 billion through IPOs so far in 2026, while the median newly listed stock has gained roughly 173% on its first trading day.
Those numbers reveal the unusual balancing act facing Beijing.
China wants its equity markets to finance semiconductor manufacturers, artificial-intelligence companies, robotics developers and other industries considered strategically important to the country’s economic future. Authorities also want households to shift more savings toward stocks after years in which property absorbed an enormous share of Chinese wealth.
But regulators do not want that policy objective to create a market where weak companies can raise money simply because investors will buy almost anything connected to AI, chips or robotics.
Recent IPOs show why that risk has become difficult to ignore.
CXMT’s 466% Debut Shows How Extreme China’s IPO Market Has Become
The most dramatic example came in July when ChangXin Memory Technologies, better known as CXMT, completed the largest semiconductor IPO ever held on mainland China.
CXMT raised 57.92 billion yuan, roughly $8.6 billion, after pricing its shares at 8.66 yuan each.
On its first trading day in Shanghai, the stock closed at 49 yuan.
That represented a 466% increase from the IPO price and pushed CXMT’s market value to roughly 3.3 trillion yuan, or nearly $490 billion, temporarily making it the most valuable company listed on a mainland Chinese exchange.
The stock generated more than 140 billion yuan of trading turnover during that single session, another indication of the extraordinary level of investor interest.
CXMT is not a random speculative startup.
It is China’s leading producer of DRAM memory chips and has become a central piece of Beijing’s effort to reduce the country’s dependence on foreign semiconductor technology.
AI demand has also produced explosive growth in the company’s business. CXMT’s first-quarter revenue surged more than sevenfold to roughly 50.8 billion yuan, while its share of the global DRAM market has been expanding.
The issue for regulators is not whether CXMT has strategic importance.
It is what a 466% first-day gain says about the IPO pricing process.
At the IPO price, investors were valuing the company at a fraction of what the secondary market was willing to pay only hours later.
Some underpricing is normal in global IPO markets. Companies and their bankers often leave room for shares to rise after listing so investors participating in the offering are rewarded.
A several-hundred-percent increase is something very different.
It can indicate that the offer price substantially underestimated demand, that speculative capital is overwhelming traditional valuation analysis, or both.
Unitree’s Rise and Fall Shows the Other Side of the Trade
Chinese humanoid-robot maker Unitree provides an even clearer warning.
Unitree shares increased more than fivefold when the company listed on Shanghai’s STAR Market in August, giving the robotics company a valuation of roughly $50 billion.
The enthusiasm reflected China’s ambition to become a global leader in humanoid robotics and Unitree’s position as one of the country’s most recognizable companies in the sector.
Investors quickly pushed the valuation far beyond the level established in the IPO.
Then the stock reversed.
Unitree shares subsequently fell roughly 45% from their post-listing high, raising concerns about speculative trading and potential losses for investors who bought after the initial surge.
That kind of price action creates exactly the problem Beijing is trying to avoid.
Aggressive first-day gains can make IPOs extremely attractive to retail investors because winning an allocation appears to offer almost guaranteed profit.
But investors purchasing shares after those gains face a completely different risk.
If a company trades at several times its IPO valuation before its underlying earnings or revenue have materially changed, future returns become increasingly dependent on sentiment remaining strong.
Once enthusiasm fades, late buyers can absorb the losses.
Chinese regulators are now reportedly raising the bar specifically for additional humanoid-robotics companies seeking listings.
Recent guidance communicated to investment banks indicates that applicants may need to demonstrate recurring revenue, progress toward reducing losses or genuinely significant technological innovation before receiving approval.
The China Securities Regulatory Commission has not publicly issued a formal rule containing those specific requirements, making the reported guidance different from a finalized regulatory change.
Still, it points toward tighter screening after Unitree’s volatile debut.
China’s IPO System Is Different From the U.S.
Part of the unusual first-day performance comes from the structure of China’s capital markets.
In the U.S., an investment bank taking a company public generally tries to price the offering close enough to market demand that investors receive an attractive first-day return without leaving an enormous amount of money on the table for the company.
Chinese regulators have historically placed greater emphasis on protecting individual investors and maintaining orderly markets.
Retail investors also play a much larger role in mainland Chinese trading than they do in many developed markets.
That creates incentives to price IPOs conservatively.
A discounted offering gives investors who receive shares at the IPO price a greater probability of earning an immediate gain.
But when that discount becomes too large, the system can produce the opposite problem: enormous opening-day rallies encourage speculation and can make new listings feel disconnected from fundamental value.
The STAR Market adds another layer.
Shanghai created the board to finance strategically important technology companies and intentionally made its listing framework more flexible than China’s traditional main boards.
The market can accept qualifying technology companies that are not yet profitable, reflecting the reality that semiconductor, biotech and other research-heavy businesses may require years of investment before producing meaningful earnings.
STAR Market shares are also not subject to normal daily price limits during their first five trading sessions. After that period, the usual daily limit is 20%.
That allows the market to find a price quickly after an IPO, but it also permits enormous first-day moves such as those seen in CXMT and Unitree.
Regulators Want Better Companies, Not Fewer Technology IPOs
Beijing’s latest stance should not necessarily be interpreted as an effort to shut down the technology IPO market.
China still needs enormous amounts of private and public capital to finance its strategic industries.
Semiconductors are an obvious example.
U.S. export restrictions have limited Chinese companies’ access to some advanced chips and manufacturing equipment, pushing Beijing to accelerate domestic production.
AI development creates similar capital requirements. Training large models, building data centers, buying chips and competing for engineering talent can consume billions of dollars before a company establishes sustainable profitability.
Robotics, advanced manufacturing, electric vehicles and biotech all require large amounts of long-term investment as well.
The CSRC has explicitly supported allowing high-quality technology companies with significant market potential to use the STAR Market even before reaching profitability.
At the same time, regulators say they want a higher threshold for what qualifies as genuine “hard technology” and stronger responsibilities for investment banks and other intermediaries bringing those companies public.
That distinction is central to Beijing’s strategy.
China does not appear to want fewer promising semiconductor or AI companies accessing capital.
It wants to prevent the popularity of those sectors from becoming a shortcut that allows companies with weak technology, poor commercialization prospects or unsustainable losses to reach public investors.
Investment Banks Could Face More Scrutiny
That puts China’s securities firms directly in the regulatory spotlight.
Investment banks act as sponsors and underwriters for IPO applicants, helping companies prepare filings, establish valuations, market offerings and satisfy listing requirements.
When regulators believe too many weak companies are entering the pipeline, bankers become one of the easiest points at which to tighten control.
China’s securities rules already require IPO intermediaries to take responsibility for the accuracy and completeness of offering documents.
Pricing rules also require institutional investors participating in bookbuilding to submit independent and reasonable valuations rather than deliberately pushing prices artificially higher or lower.
If scrutiny increases, banks could become more reluctant to sponsor borderline companies.
That could slow portions of the IPO pipeline and reduce potential underwriting fees, but it could also improve the average quality of companies ultimately reaching the public market.
The impact may be particularly noticeable in robotics and AI, where private valuations have risen quickly and companies are racing to raise larger pools of capital.
DeepSeek Is Preparing for a Potential Shanghai IPO
One of the largest tests could come from DeepSeek.
The Chinese artificial-intelligence developer has selected CITIC Securities to prepare for a potential domestic IPO on the STAR Market, according to Reuters.
The company is seeking capital for computing infrastructure, model development and talent as competition intensifies across China’s AI industry.
DeepSeek has not yet established the final size or valuation of an offering, and the IPO process is still in preparation.
Its emergence in the listing pipeline illustrates why regulators cannot simply restrict technology IPOs broadly.
DeepSeek is among China’s most strategically important AI companies, and Beijing has strong incentives to give businesses like it access to domestic capital.
But the extraordinary valuations being assigned to AI companies globally also make pricing discipline increasingly important.
If DeepSeek ultimately lists, investor demand could provide one of the clearest measures yet of how far enthusiasm for Chinese AI equities can extend.
Enflame Will Provide a More Immediate Test
Investors will not have to wait for DeepSeek to see whether the IPO frenzy continues.
Tencent-backed AI chipmaker Shanghai Enflame Technology is scheduled to begin trading on the STAR Market Friday.
Enflame raised 6.12 billion yuan, approximately $912 million, by selling 43 million shares at 142.18 yuan each.
The offering values the company at roughly 61.2 billion yuan, or about $9.1 billion.
Only approximately 4.16% of Enflame’s shares will initially be freely tradable, creating a relatively small public float that could contribute to volatile trading if demand is strong.
The company’s financial profile also captures the type of trade-off regulators are trying to manage.
Enflame expects revenue between 2.3 billion yuan and 3 billion yuan during the first nine months of 2026, representing growth of roughly 326% to 455%.
But it still expects a net loss of approximately 700 million to 860 million yuan.
Tencent will remain Enflame’s largest shareholder with a stake of nearly 18% after the IPO. Tencent also represented almost 84% of Enflame’s 2025 revenue, creating significant customer-concentration risk.
Enflame therefore has rapid growth and strategic value, but also losses and heavy dependence on a single major customer.
How investors price those risks after trading begins could show whether the market is becoming more selective.
Beijing Wants More Long-Term Capital in Stocks
China is also trying to change who owns its equities.
Regulators have repeatedly called for more insurance companies, pension funds, state-backed institutions and other long-term investors to put money into the stock market.
The policy is often described as encouraging “patient capital” — investors willing to hold companies through long development cycles rather than trade around short-term price movements.
The CSRC reiterated that objective earlier this year as part of its longer-term capital-market planning.
Regulators have also experimented with providing larger IPO allocations to institutions willing to accept longer lock-up periods, particularly for unprofitable technology companies.
The approach could reduce the amount of stock immediately available for speculative trading and align large investors more closely with a company’s longer-term performance.
But lockups introduce another risk.
When large quantities of previously restricted shares eventually become tradable, the increase in available supply can pressure the stock.
Moore Threads recently offered a real-world example.
Shares of the Chinese GPU developer dropped sharply after a lock-up expiration released millions of additional shares into the market, even though early investors remained substantially above the IPO price.
That is one reason the structure of shareholder ownership before and after an IPO matters almost as much as the initial valuation.
Why Beijing Wants Households in Stocks
The IPO boom also fits into a much larger economic transition.
For decades, Chinese households treated residential property as one of their primary stores of wealth.
The country’s extended real-estate downturn has damaged that model.
Home prices have fallen in many cities, developers have defaulted, construction has slowed and households have become more reluctant to commit additional savings to property.
Beijing would benefit if at least some of that capital moved into productive businesses through equity markets instead.
A stronger stock market could help companies finance expansion while giving households another potential source of investment returns.
But that strategy depends heavily on confidence.
If retail investors repeatedly buy newly listed companies after enormous first-day rallies and then suffer steep losses, households may become less willing to move savings into equities.
That helps explain why regulators appear willing to tolerate significant IPO underpricing while simultaneously trying to prevent lower-quality issuers from exploiting strong demand.
The policy objective is not simply to maximize the number of listings.
It is to create enough successful investments to convince households that stocks are a viable long-term alternative to property.
What Investors Should Watch Next
Friday’s Enflame debut will provide the first immediate test of whether regulators’ tougher message is affecting investor behavior.
Another triple-digit opening gain would suggest speculative appetite remains extremely strong.
A more moderate move could indicate investors are beginning to differentiate between strategic importance and valuation.
Unitree deserves attention as well.
Its roughly 45% decline from post-IPO highs has become an example regulators can point to when warning about the risks of chasing newly listed technology stocks after enormous first-day gains.
The pipeline itself will be another signal.
If robotics startups begin delaying listings or investment banks withdraw weaker candidates, Beijing’s informal guidance may already be changing behavior before any new formal regulation is published.
CXMT’s performance will also matter because its $8.6 billion IPO remains the defining transaction of China’s 2026 equity boom.
The company’s long-term ability to justify a valuation created by its 466% debut will depend on revenue growth, profitability, technological progress and its ability to compete with global memory leaders including Samsung Electronics, SK Hynix and Micron.
China’s regulators face a difficult trade-off.
Too much restriction could starve strategically important technology companies of capital and undermine Beijing’s goal of creating stronger domestic competitors in chips, AI and robotics.
Too little control could allow a flood of weak companies to exploit investor excitement and eventually produce losses severe enough to damage confidence in the entire market.
With more than $28 billion already raised and the median Chinese IPO gaining 173% on its first day this year, Beijing is no longer dealing with a shortage of investor enthusiasm.
Its challenge now is making sure that enthusiasm funds companies capable of eventually justifying it.