CITIC Securities: The trend of Tesla's Full Self-Driving (FSD) entering China is becoming increasingly clear, which is expected to overall benefit the intelligentization process of Chinese automotive companies.

According to the Zhitong Finance and Economics APP, CITIC Securities has released a research report stating that in the United States, Tesla's Full Self-Driving (FSD) is on the eve of formal large-scale deployment, and the next generation V12 version is expected to remove the Beta label and be officially released. If FSD officially enters China, it is expected to accelerate the intelligentization process of Chinese electric vehicles. On one hand, as an industry leader, Tesla is expected to enhance Chinese consumers' awareness of autonomous driving functions through FSD and expand the market capacity. In addition, car manufacturers that focus on advanced autonomous driving functions and use them as differentiating selling points are expected to benefit first. On the other hand, the supply chain of autonomous driving chips, domain controllers, sensors, wire-controlled chassis, and in-vehicle Ethernet chips are expected to benefit as a whole from the increase in penetration rate of advanced autonomous driving. It is recommended to pay attention to autonomous driving chip manufacturers.

In the United States, FSD technology is approaching maturity and is currently on the eve of formal large-scale deployment.

Tesla aims to achieve fully autonomous driving through FSD, but it is currently in the Beta version and is still classified as an L2+ product due to regulatory restrictions. After 3 years of testing, FSD Beta may have initially possessed L3 capabilities and is expected to achieve a high level of autonomous driving with the premise of driver takeover as a safety net.

According to Tesla's 4Q22 Shareholder Deck released on January 25, 2023, the number of FSD Beta deployed vehicles in North America (United States and Canada) has reached 400,000, with a penetration rate of over 20% compared to Tesla's cumulative sales of approximately 1.7 million vehicles in North America by the end of 2022. Since the end of 2022, Tesla has successively canceled the safety rating mechanism for FSD Beta, accelerated the feature deployment in North America, and unified the highway/city technology stack. These actions may indicate that FSD is on the eve of formal large-scale deployment.

Currently, FSD Beta has iterated to version V11.4.4, and Elon Musk stated in a reply on Twitter on June 27 that FSD V12 will remove the Beta label and be released as a stable version, with the predicted timeframe around 2024. At that time, if Musk fulfills the previously mentioned "1-month free trial" plan, the take-up rate of FSD in North America is expected to significantly increase. With the high price of $15,000 (one-time purchase price) and an estimated gross margin of around 90%, Tesla is expected to generate substantial profits and revenue.

China: The trend of FSD entering China is becoming increasingly clear, which will overall benefit the intelligentization process of Chinese car manufacturers.

Although the specific timing is unknown, the entry of FSD into China is a general trend. There are still two major hurdles before it can be fully implemented: 1) Qualification: Tesla's shadow mode is highly likely to be classified as mapping behavior. How to cooperate with first-class mapping qualification enterprises and legally and compliantly conduct road data collection in China is still under discussion. 2) Data: Road data is highly sensitive, so the data collected by Tesla in China is highly likely not to be transmitted back to the United States. It needs to establish data and supercomputing centers in China and set up an engineering team for closed-loop data. When FSD is effectively implemented in China, it is expected to accelerate the overall intelligence process of Chinese electric vehicles, while strengthening consumer education and awareness, expanding the market, and promoting the survival of the fittest among automakers. Chinese automakers that focus on advanced intelligent driving functions and use them as a unique selling point may benefit, but low- to mid-range models that are heavily impacted by price wars and have weaker self-developed capabilities may face greater pressure in the future.

Barriers: Tesla's FSD is expected to be ahead of the top players in the Chinese passenger car market by 2-3 years, mainly in the following four aspects:

1. Algorithms: Since 2020, Tesla has introduced BEV+Transformer in its perception algorithm, coupled with the concept of Occupancy Network, significantly improving the accuracy of visual perception solutions, leading the top players in China by about 2 years. In addition, Musk stated that the next generation V12 version of FSD plans to achieve end-to-end autonomous driving, which will further expand the scale of the autonomous driving model on the vehicle side, highlighting Tesla's leading position in algorithms.

2. Data Volume: Tesla's "million-car fleet + shadow mode" significantly surpasses other electric vehicle players in terms of data volume. In comparison, domestic new energy vehicle startups with strong self-developed capabilities still need to improve their sales volume and have mainly relied on a small number of test vehicles for data collection.

3. Data Loop: Tesla has built a highly automated, streamlined, and industrialized data training system, including data collection, annotation, training, and other closed-loop processes, as well as infrastructure such as the Dojo supercomputer center. This requires strong engineering practice and a wealth of experience know-how. The data loop of domestic players still needs improvement.

4. Hardware: Tesla's self-developed FSD chip and HW hardware system. The latest version HW4.0, equipped with the second-generation FSD chip, started delivery in the United States in Q1 2023. The self-developed hardware system is efficiently coupled with neural network algorithms, which is expected to help Tesla's autonomous driving solution achieve better cost reduction and efficiency improvement, while the chips and domain control of Chinese automakers are mostly outsourced.

Investment Recommendations:

In the United States, Tesla's FSD is on the eve of formal large-scale deployment, and the next generation V12 version is expected to remove the Beta label and be released as a stable version. At that time, if Musk fulfills the "1-month free trial" plan he mentioned before, the take-up rate of FSD in North America may significantly increase. It is recommended to continue monitoring the progress of FSD V12 version deployment.

In addition, Tesla is actively promoting the implementation of FSD functions in regions outside of North America, including Asia, Europe, the Middle East, and other countries. Expanding overseas markets is expected to enrich Tesla's scenario library and bring substantial revenue. It is recommended to continue tracking the progress of FSD implementation overseas. Tesla, with its algorithm, hardware, and data advantages, is expected to further expand its market.

If FSD officially enters China, it is expected to accelerate the intelligence process of Chinese electric vehicles. On one hand, as an industry leader, Tesla is expected to enhance Chinese consumers' awareness of intelligent driving functions and expand the market capacity through FSD. In addition, automakers that focus on advanced intelligent driving functions and use them as a unique selling point are expected to benefit first. On the other hand, the supply chain of autonomous driving chips, domain control, sensors, wire-controlled chassis, and in-vehicle Ethernet chips is expected to benefit as a whole from the increase in the penetration rate of advanced intelligent driving. It is recommended to pay attention to autonomous driving chip manufacturers.

Risk Factors:

FSD technology progress is slower than expected, V12 version is delayed; FSD adoption rate in North America is lower than expected; Major safety accidents occur with FSD; Progress of FSD in regions outside North America such as Europe and Asia is slower than expected; Tesla's sales volume is lower than expected; FSD is slow to land in China; FSD significantly lowers its selling price in China, squeezing the space for Chinese automakers, etc.