Recently, Baidu's self-driving travel platform, Luobo Kuaipao, has been booming, driving Baidu's Hong Kong stocks to rise more than 12% on the 10th, the highest point in over a month. In terms of product structure, the operating income of 10-30 billion yuan products is respectively 401/1288/60 million yuan.

However, JPMorgan expressed its view that it is too early and not rational from a fundamental perspective for Baidu's unmanned taxi market to have such a situation. Therefore, its stock price momentum may weaken in the short term.

However, analysts further pointed out that if Luobo Kuaipao can improve its profitability, it is expected to bring a single city balance of payments in the second half of 2024.

With the introduction of the next-generation vehicles and significant changes in the proportion of vehicle safety personnel, Baidu's unmanned taxi business will achieve a balance of payments in several specific cities in the next few quarters.

However, its impact on the group's finances will remain small until this business achieves a larger scale, which may not happen until the second half of 2025 at the earliest.

Three key variables for further profit improvement

JPMorgan believes that there are three key variables to improve the profitability of Baidu's Luobo Kuaipao unmanned taxi: pricing, vehicle cost, and the proportion of vehicle safety personnel.

From the perspective of these three variables, JPMorgan conducted on-site research on Wuhan and Shanghai's unmanned taxi business and summarized the conclusions as follows:

▲ We believe that the current pricing of Baidu's unmanned taxi is not sustainable in business, because it is unlikely to produce GTV that is economically sufficient to cover costs in any operating mode (taxi, online car-hailing, unmanned taxi). We believe that the current pricing strategy is mainly to enhance consumer awareness and achieve initial participation. In our view, viable commercial pricing should be higher than taxi prices and lower than online car-hailing prices. We expect that after entering the large-scale deployment stage, Baidu will adjust its pricing strategy for unmanned taxis.

▲ At present, labor costs are three times the GTV, partly because Baidu adopts an aggressive pricing strategy, but a more important reason is that vehicle safety personnel work in shifts. We believe that this is essentially a temporary solution because the fundamental purpose of autonomous driving is to replace human drivers. According to the "Guidelines for Autonomous Driving Car Transport Safety Services (Trial)" issued by the Ministry of Transport, remote safety personnel can be used in the operation area delineated by the local regulatory authority, and each remote safety personnel must monitor no more than three unmanned taxis. We found that the unmanned taxis deployed in Wuhan use a ratio of 1:3 (each safety personnel is responsible for monitoring three cars). We believe that the ratio of 1:3 will soon be adopted in the domestic unmanned taxi industry, which may reduce labor costs by 67%.

▲ At present, hardware costs (i.e. depreciation) are one times the GTV. However, management and domestic news reports state that the cost of Baidu's next-generation taxis (Chi 06) will be 60% lower than the current generation. Therefore, using Chi 06 will reduce hardware costs by about 60%.

JPMorgan believes that although the current Shanghai unmanned taxi business is facing severe financial losses, if the three variables mentioned above are strengthened in commercial deployment, Baidu's single-city profit margin is expected to rise significantly in the second half of 2024.

The path to significant profits

JPMorgan pointed out that at present, Baidu's unmanned taxi business faces two major obstacles: the balance of payments at the unit economic efficiency level and the approval of citywide operations by regulatory authorities.

As long as these two obstacles are overcome, we believe that profits will only be affected by unit economic efficiency and the number of vehicles.

Given the highly developed supply chain of China's electric vehicles, the latter can easily achieve a large scale. Although it is difficult to predict when local authorities will approve citywide operations, we believe that this may be possible at the earliest in 2025.