Hualong Securities: Robotaxi moves from technical validation to operational validation; commercialization breaks the ice and enters a period of institutional dividends.

Hualong Securities: Robotaxi moves from technical validation to operational validation; commercialization breaks the ice and enters a period of institutional dividends. Hualong Securities issued a research report stating that on October 26, XPeng announced that it had recently been approved for the Guangzhou intelligent connected vehicle application demonstration operation qualification and can conduct L4 Robotaxi demonstration operations across all of Guangzhou. The industry has entered a period of policy dividends, leading players have already achieved positive unit economics, and attention should be paid to OEMs with in-house development capabilities, technology providers, and core links in the supply chain, maintaining the industry "recommended" rating. The main views of Hualong Securities are as follows: Robotaxi commercialization moves from pilots to a legal basis, entering a window for institutional dividend realization On August 25, 2026, the Road Traffic Safety Law of the People's Republic of China (Draft Amendment) was submitted for deliberation, adding for the first time a special chapter on "Special Provisions on Autonomous Vehicles" and clarifying the division of liability under autonomous driving functions. On July 30, 2026, the mandatory national standard Safety Requirements for the Autonomous Driving System of Intelligent Connected Vehicles was approved. It is China's first mandatory national standard for L3 and L4 autonomous driving and is scheduled to take effect in July 2027. As of the first half of 2026, China had cumulatively issued more than 20,000 intelligent connected vehicle test licenses, opened 57,000 kilometers of test roads, and total test mileage had exceeded 220 million kilometers; as of July 2026, 26 cities in China had been approved for commercial paid Robotaxi operations without safety officers. With the implementation of the above regulations and mandatory standards, the boundaries of Robotaxi liability are legally clarified, and operating entities can clearly calculate the economics of putting a single vehicle on the road, providing institutional guarantees and a fallback for large-scale deployment. Commercialization data from leading players continues to climb Robotaxi has moved from technical demonstrations to the stage of real orders, and leading domestic players accelerated fleet expansion and order acquisition in 2026. In Q2 2026, Pony.ai's Robotaxi passenger fare revenue increased by 849.3% year on year. As of H1 2026, its Robotaxi fleet reached 1,975 vehicles and is expected to reach 3,500 by the end of 2026, further raised from 3,000 at the beginning of the year. As of the end of July 2026, WeRide's global Robotaxi fleet reached 1,800 vehicles. In Q2 2026, the peak daily completed orders per domestic Robotaxi vehicle reached 28, and ride-hailing revenue increased by 140% quarter on quarter. Both Pony.ai and WeRide have cases of turning positive on a per-vehicle profit basis. Pony.ai achieved positive per-vehicle profitability in Guangzhou and Shenzhen in November 2025 and February 2026, respectively; WeRide achieved a positive per-vehicle profit model in Abu Dhabi in 2025. Robotaxi competition has moved beyond the availability stage focused on algorithm capabilities and turned to operational dimensions such as vehicle utilization, dispatch efficiency, service costs, and fleet management. Focus on three main lines: vehicle platforms + technology providers + high-level intelligent driving supply chain First, attention should be paid to OEMs with Robotaxi vehicle platforms and full-stack in-house development capabilities, which are expected to open new revenue space beyond the vehicle business through technology licensing, software services, and ecosystem cooperation. Second, L4 Robotaxi technology providers such as Pony.ai and WeRide are expected to rely on excellent model capabilities, accumulated real-vehicle data, and rich operating experience to drive fleet expansion and further validate fleet-level and enterprise-level profit models. Third, attention should be paid to core links in the high-end intelligent driving and autonomous driving supply chain. As of September 2026, Horizon Robotics' Journey series intelligent driving chips had cumulatively reached more than 15 million units in mass production and had cooperated with more than 40 automaker brands on nearly 500 designated models. With the implementation of L3/L4 national standards, suppliers of core components such as intelligent driving chips, LiDAR, domain controllers, and steer-by-wire chassis are expected to see both volume and price rise. Risk warnings: macroeconomic fluctuation risk; policy change risk; geopolitical risk; autonomous driving technology progress falling short of expectations; Robotaxi fleet expansion pace falling short of expectations; intensifying competition risk; error risk in third-party data; Hong Kong stock liquidity risk; performance of key companies falling short of expectations.