The growth rate of Baidu's advertising and cloud business previously slowed slightly. Looking back at 2024Q1, Baidu achieved quarterly revenue of 31.5 billion yuan, a year-on-year increase of 1%. In Baidu's core business, online marketing advertising revenue was about 17 billion yuan, an increase of 3% over the previous year, and the growth rate has slowed. Smart cloud and other business revenue was 6.8 billion, up 6% year over year. We determined that the tightening of IT spending by downstream companies affected cloud demand. The company's non-GAAP net profit for the quarter was 7.01 billion yuan, and the non-GAAP net profit margin was about 22%.

Short-term outlook, advertising growth may be pressured by a high base. Looking ahead to 2024Q2, on the one hand, last year's advertising base was high; on the other hand, the current desire for downstream marketing is still limited, and the revenue side is expected to continue to be under pressure.

Cloud businesses have limited benefits from AI in the short term, but AI growth is a core focus for the next 5-10 years. In the short term, the cloud business will continue to benefit from the incremental demand brought about by AI training. In the medium to long term, an application ecosystem based on the Wenxin Big Model is gradually being established. Looking forward to the future, in order to reduce inference delays and ecological collaboration, customers based on the Wenxin Big Model will open up more room for growth for Baidu Cloud.

The advancement of AI capabilities and ecosystems is accelerating. At the Baidu WAVESUMMIT Deep Learning Developers Conference held on June 28, Baidu CTO Wang Haifeng announced that currently the number of users of the Wenxin Big Model has exceeded 0.3 billion, and the number of daily calls has reached 0.5 billion. As a comparison, Li Yanhong said in April of this year that Wenxin said that the number of users exceeded 0.2 billion, and the average daily API calls also exceeded 0.2 billion. It can be seen that Baidu's big model ecosystem is progressing rapidly. It is expected that with the launch of free versions of some models and a sharp drop in inference costs, the number of users and API calls will continue to increase. In terms of ability, the algorithm training efficiency of Wenxin's big model has increased to 5.1 times the original, the average weekly training efficiency rate has reached 98.8%, the inference performance has increased 105 times, and the cost of inference has been reduced to 1% of the original.

Baidu had a first-mover advantage in the launch of Robotaxi. In July 2024, orders for Radish Express began rapidly in Wuhan. The peak number of single-day cycling orders exceeded 20, which is close to the level of taxi drivers in a single day. As of April 19, the total number of orders had exceeded 6 million orders.

From the perspective of first-mover advantage, Baidu is in a leading position in the country's Robotaxi operation layout. From the perspective of network effects, Baidu itself has a strong Internet user ecosystem. Next, through efforts on the vehicle supply side, it is conducive to the formation of a “user-vehicle” operational-level network effect.

In terms of profit, unit volume increases and vehicle cost reduction are core factors, and Baidu is making efforts through research and development of new intergenerational models and improving operational efficiency. We expect the Wuhan region to achieve UE balance in the near future, and we hope to replicate it successfully in other regions in the future.

Maintain an “Overweight” rating. We expect the company's 2024-2026 revenue to be 140.1/152.7/166.3 billion yuan, an increase of 4%/9%/9% year-on-year, and non-GAAP net profit of 26.7/27.9/30.8 billion yuan, a year-on-year change of -7%/+4.5%/+10.1%. Maintain an “Overweight” rating.

Risk warning: the risk that advertising demand falls short of expectations, the risk that mass production implementation of the autonomous driving business falls short of expectations, and the risk that AIGC application progress falls short of expectations.