The Hong Kong stock market sees a boom in intelligent driving, let's talk about the final wave of IPOs for intelligent driving enterprises
This month, Momenta passed the hearing at the Hong Kong Stock Exchange and plans to list around July 8th. At the same time, EControl Intelligent Driving also submitted its listing application. Within a month, two intelligent driving companies are at the door of ringing the bell on the Hong Kong stock market.
On one hand, companies are queuing up to go public. On the other hand, veterans in the intelligent driving technology field who have been deeply involved for many years have switched to embodied intelligence, heading towards the next big thing.
Putting these two pictures together raises a question: After this wave, what will be left in the intelligent driving track?
The golden age of intelligent driving startups has completely passed, while its era as the infrastructure of the automotive industry is just beginning. From now on, there will be no middle - ground in the industry - either secure a core ecological position or quietly exit during the industry consolidation.
Heading for the Hong Kong Stock Market
This batch of enterprises, born during the first wave of autonomous driving startup tides from 2016 to 2018, were tied to US dollar funds from the very beginning. The VIE structure has already solidified their equity structure. Choosing the Hong Kong stock market is an inevitable result after a series of eliminations.
The A - share market naturally does not accept the listing of VIE entities. If they want to return by dismantling the structure, they need to buy out US dollar shareholders and adjust the equity structure, which requires at least a two - to three - year compliance cycle. And US dollar LPs in the exit window simply can't wait.
Although there have been successful precedents like Pony.ai and WeRide in the US stock market, the liquidity discount of Chinese concept stocks and the uncertainties in Sino - US regulations always loom large. Momenta's plan to go public in the US was once shelved, which is the most direct signal. After all, the only capital market that can accommodate the US dollar structure and connect with the mainland market is the Hong Kong stock market.
The Hong Kong stock market is not a passive receiver but actively absorbs hard - tech assets. In 2018, it relaxed the rules on different voting rights for the same shares, and in 2023, it introduced Chapter 18C. In essence, it is rewriting the valuation rules for hard - tech companies: no longer setting continuous three - year profitability as a rigid threshold, recognizing the core value of technological barriers, order scale, and mass - production ramp - up speed, and allowing enterprises to exchange growth potential for a listing ticket.
In the first five months of this year, 10 Chapter 18C enterprises have completed their listings, raising a total of over HK$25 billion, covering fields such as autonomous driving, large AI models, chips, and commercial space - this is not a scattered wave of corporate listings but the Hong Kong capital market betting on an entire package of Chinese hard - tech assets.
The deeper value lies in the industrial channel. In early June this year, Pony.ai and WeRide were simultaneously included in the Stock Connect program. Mainland funds can directly trade their stocks. Besides improving liquidity, it also opens up a path to be tied to the mainland industry: The identity of a listed company with the Stock Connect label carries a completely different weight when it comes to in - depth cooperation with automakers and negotiations on local industrial implementation.
Although people often complain about the low valuation in the Hong Kong stock market, at the current industrial stage, the value of the cross - border capital channel far exceeds the short - term valuation premium.
CATL, with hundreds of billions of cash on its books, also plans to list in Hong Kong. Even though the stock price of Seres retreated by 13% after the issuance, it still decided to list in Hong Kong. Now that automakers are going global collectively and the supply chain is expanding overseas, as core suppliers, intelligent driving companies need overseas entities to support overseas R & D, mergers and acquisitions, and local operations. The Hong Kong stock market just provides a dual - circulation platform of offshore RMB and US dollars.
However, the door of the Hong Kong stock market is not open to all enterprises. The fact that Freetech has submitted its listing application four times but still failed to pass is a clear proof. The Hong Kong stock market has only lowered the profitability threshold, and the threshold for commercialization certainty has never been relaxed: Enterprises with mass - production orders, revenue growth, and implementation scenarios can get the admission ticket, while companies that only rely on technology stories but can't produce mass - production results are still kept out.
Looking at the long - term perspective, intelligent driving is just the vanguard of this wave of migration. Large - model enterprises such as Zhipu and MiniMax, and chip companies such as Muxi, Biren, and Tianshu Zhixin have successively completed their listings in Hong Kong. This is a systematic migration of a whole wave of Chinese hard - tech from the US stock market to the Hong Kong stock market. Intelligent driving is just the first wave that has achieved the fastest commercialization and capitalization.
The Essence of Listing
Listing is not the end but getting the admission ticket for the next round of competition. Two industrial inflection points have locked in the general trend for the next two or three years.
The first inflection point: On the eve of the explosion in penetration rate, the entire industry has stepped into the "valley of death" short of money. Official data from the Ministry of Industry and Information Technology shows that from January to February 2026, the penetration rate of new passenger cars with L2 - level combined driving assistance functions was 69.15%. Intelligent driving is no longer just a gimmick for high - end options but is starting to spread rapidly to the mass - market price segment.
The price of L2+ has dropped from 200,000 yuan to 150,000 yuan. In another year or two, it will be out of date for a 100,000 - yuan family car not to have basic intelligent driving.
However, the period of penetration rate growth is exactly when the entire industry spends the most and has the tightest cash flow. On the chip side, Horizon Journey 6 and Black Sesame Wudang C1200 have just entered the mass - production ramp - up stage. A single advanced - process tape - out costs hundreds of millions, and then there are tool - chain construction, developer ecosystem building, and vehicle - model adaptation, all of which are extremely costly.
The pressure on solution providers is more direct. Momenta's prospectus shows that it has a cumulative total of 170 vehicle - model orders, 68 models in mass production, and over 680,000 vehicles equipped globally, which are impressive figures. But behind this is a huge amount of delivery prepayment - the payment cycle in the automotive industry is generally 6 to 12 months. If the vehicle - equipped volume doubles, the prepayment will also double. The larger the scale, the bigger the cash - flow hole.
Coincidentally, the primary market is almost out of supply. US dollar funds are in the exit cycle, and RMB funds have shifted their attention to large models and embodied intelligence. No one is willing to accompany intelligent driving companies through the difficult days of mass - production ramp - up.
Therefore, this wave of listing tides is not an expansion for the sake of improvement but a necessary life - saving fund. Those that can't get it will probably fall before the full - scale explosion of the penetration rate.
The second inflection point: The game between self - research and external supply has reached a crossroads of defining boundaries. After Huawei's automotive BU was split into Yinwang Technology, its quasi - Tier 0.5 position has become more stable; BYD's Tian Shen Zhi Yan full - stack self - research has been fully implemented in its vehicles, with full control over perception hardware, planning and control algorithms, and underlying chips; The intelligent driving teams of new - force automakers such as NIO, XPeng, and Li Auto generally have over a thousand people, and they are pursuing further full - scale self - control of core algorithms and data closed - loop.
Once the self - research systems of leading automakers are fully operational, the incremental market that the remaining solution providers can tap into will more rely on mid - tier automakers and overseas markets.
Now rushing to list is not for aggressive expansion but to get a survival admission ticket first. With the credit endorsement of a listed company, small and medium - sized automakers will dare to give you orders without worrying about your sudden collapse and supply interruption. With the ability to continuously raise funds in the public market, you can withstand price wars and not be eliminated in advance.
Convergence of the Endgame
After getting the admission ticket, the industry will accelerate its differentiation, and three main lines are clearly visible.
Among passenger - car solution providers, several well - known ones at present - Yinwang, Horizon, Momenta, DeepRoute.ai, QCraft, Zhuoyue, those that are already listed or on the way to listing, naturally have capital and credit advantages. They have more confidence in competing for orders, launching price wars, and expanding delivery teams. Those that are not listed can't raise money in the primary market, can't scale up, and can't reduce costs. The gap will only widen.
In two or three years, those that are not listed will either be acquired by automakers and leading companies or completely exit the passenger - car market and turn to mines and ports for survival.
The L4 track will completely split into two completely different paths. Everyone in the industry knows the high - cost nature of the Robotaxi path. Operating in one city requires dozens or even hundreds of test vehicles, plus road - right applications, operation teams, and policy communication, costing billions every year. The profit inflection point is still far away. In the end, only companies like Pony.ai and WeRide, which are listed in both the US and Hong Kong and have the ability to continuously raise funds in the public market, can hold on until the day when the cost per kilometer is balanced.
On the other hand, autonomous driving in commercial vehicles follows a completely practical route. Mines, ports, and trunk logistics have closed - loop scenarios, fast implementation, and high unit value. Customers are willing to pay real money for cost reduction and efficiency improvement. In the mine scenario, three companies, Xidi, EControl, and Tage, have emerged, each binding with leading mining enterprises, with orders and revenues secured.
However, the ceiling is also visible. The number of mines and the number of mining trucks in the country are limited, which can't support multiple high - valuation companies. In the end, each company will hold on to its core customers and survive in a small - but - beautiful way. Cross - scenario expansion is extremely difficult - for a company in the mine field to enter the port field, it has to build the scheduling logic, customer channels, and operation system from scratch, which is no different from starting a new business.
More importantly, after listing, the logic of storytelling must be completely switched.
In the primary market, people look at algorithm advancement, test mileage, and scenario coverage, telling a technological narrative of full - scenario autonomous driving. After listing, they have to be responsible for the financial statements. The capital market only looks at revenue per vehicle, operating cost, and pay - back period, telling a business narrative of operational efficiency. Shifting from technological idealism to business accounting is much more difficult than conquering algorithms for many founding teams with a technical background. A number of companies will die on the way of this narrative switch.
In the hardware field, the second round of price wars is about to start. The first round was about using primary - market financing to burn money to compete for orders, sacrificing profits for market share, competing on who can raise more money and withstand losses. The second round is that listed companies use mass - production scale and continuous financing ability to fight a cost war, competing on who has higher supply - chain efficiency and stronger scale effects.
After Hesai and Sute (assuming it's a company name, if there's a specific English name, it should be replaced) went public, the unit price of lidars dropped from tens of thousands to thousands of yuan. After Horizon and Black Sesame listed, the larger the chip shipment volume, the lower the unit cost, and the price can continue to drop. This is good news for downstream solution providers, as the whole solution can be applied to lower - priced vehicle models. But for small and medium - sized sensor and chip companies that are not listed, this is a disaster - without mass - production scale and continuous funds, they can't withstand long - term low prices. They will either be incorporated by leading companies or completely exit the mainstream market.
Ultimately, intelligent driving hardware will quickly move towards standardization and commoditization, and the profit margin will continue to narrow. The industry value will gradually shift to the software and data end.
After this wave of listing tides, what's left in the intelligent driving industry is not a mess after the bubble bursts but a real industry that has returned from the startup boom to the normal industrial state.
Ten years ago, when the first batch of entrepreneurs entered the field, everyone was telling stories of subverting the automotive industry, and the valuations soared with the stories. It was a typical boom track. After ten years, the technological routes have converged, the commercialization paths are clear, and the bubble has gradually receded. People finally realize that no matter how high - tech autonomous driving is, it will ultimately become part of the automotive industry, like an engine, chassis, or in - vehicle system. It's not flashy but indispensable.
The listed companies have survived and become part of the infrastructure. In the future, they will follow the cycle of the automotive industry and make money from scale, efficiency, and engineering. Those that didn't make it will either be incorporated into large - company systems or their teams will disband and rush to the next big thing.
Talent and capital are always flowing. The intelligent driving story is over, and the next narrative is about embodied intelligence and general artificial intelligence. Just like the transition from PC to the Internet and then to the mobile Internet, wave after wave, there will always be new booms attracting the top talents and the hottest money.
However, this is not a bad thing. When an industry is no longer crowded with startup companies telling stories and no longer occupies the technology headlines by hyping valuations every day, it is when it truly starts to precipitate and create industrial value.