Summary of Key Points
Baidu's Q2 financial report highlighted a milestone where AI revenue accounted for 50% of its general business, suggesting a breakthrough in its transformation. However, Wall Street responded with a decline in the stock price (falling by more than 5% before the opening bell). The core issue is that the revenue generated from AI is not sufficient to offset the decline in its advertising business, and the value of AI itself as a business segment is limited (low gross margin, lack of traction in consumer-facing applications). Coupled with challenges such as talent loss, weak productization capabilities, and ongoing expenses in its chip business, Baidu is experiencing the pain of transitioning from its traditional to its new growth drivers. The new engine (AI) is just starting, while the old one (advertising) is fading, putting short-term profits under pressure and leading to skepticism from the capital market.
The Appearance of AI Revenues accounting for More than 50% vs. the Reality of Declining Profits
On the surface, the fact that AI revenue makes up 50% marks a significant milestone in Baidu's transition from an advertising company to an AI-centric enterprise. However, the profit figures are concerning:
- Both revenue and profit fell short of expectations: Total revenue was 31.3 billion yuan (290 million yuan lower than expected), and adjusted earnings per ADS were 7.22 yuan (253 million yuan lower than expected, a decrease of over 25%).
- Profits plummeted: GAAP net profit decreased by 68%, and Non-GAAP net profit decreased by 46%.
- Gross margin declined: The overall gross margin dropped from 43.9% last year to 39%, mainly due to the high costs associated with AI services, especially in computing power. In the past, advertising was a low-cost way to generate revenue; now, investing in servers, chips, and data centers is necessary to provide computing power, resulting in higher expenses for every unit of sales.
In simple terms, while AI has improved the revenue structure, it has not led to a corresponding improvement in profit margins.
The Value of AI Business: What Represents Real Growth, and What Still Needs Improvement?
Behind the 50% AI revenue figure, there are several issues:
1. Computing Power Business (AI Cloud Infrastructure): This segment is growing rapidly but has a low gross margin. Revenue was 7.3 billion yuan (+50%), with GPU cloud services increasing by 283%. However, since Baidu primarily provides computing resources for others to use their models, the high costs of hardware depreciation and procurement significantly reduce profitability.
2. AI Applications: These are crucial for generating revenue from consumer users, but they have only made minimal progress. Revenue was 2.5 billion yuan (+3%), indicating a slow growth rate. For example, Baidu's Wenxin Yiyuan (a language model) is still in the development phase and has not yet achieved significant revenue.
3. AI-Based Marketing: This should be Baidu's new core business for generating substantial profits, but it is far from reaching the scale of traditional advertising.
In conclusion, while the computing power business is keeping revenues stable, the applications and marketing strategies that could truly drive profit growth and build a competitive advantage are not yet mature enough.
The Dual Pressure of Declining Advertising Revenue and High AI Expenses
Investors have long recognized these challenges:
- Advertising Business Continues to Slump: Many firms predict a 18%-23% decline in core advertising revenue for Q2. This has been Baidu's main source of cash flow over the past 20 years, and now it is facing difficulties.
- High Capital Expenditure for Growth: Baidu spent 11.4 billion yuan on capital expenditures (such as servers and data centers), resulting in a negative free cash flow of 7.95 billion yuan (worse than last year). Investors are concerned about when these investments will pay off.
- Short-Term Profit Priority: The market values financial results; even if the AI story is promising, it cannot compensate for the loss of advertising revenue in the short term, which hinders stock price appreciation.
Three Major Barriers to Transformation
1. Talent Loss: Baidu's efforts to cultivate AI talent have not been successful, with many professionals moving to companies like ByteDance and autonomous driving startups (e.g., ByteDance's Seed and Tongyi Qianwen). The loss of senior executives (P9+ level) is around 38%, affecting the development of consumer-facing products.
2. Weak Productization Capabilities: Baidu acknowledges that its products are not on par with those from companies like ByteDance. For example, Baidu Wenku's AI features have become standard in the industry, losing their novelty. Other AI applications, such as Wenxin Yiyuan, have not yet found stable profit models.
3. Kunlun Chip: Although the Kunlun P800 chip has been mass-produced, it continues to incur expenses due to production and supply chain issues. The company's IPO target valuation is 50 billion US dollars, but the chip business relies on actual sales orders, and the capital market will not be impressed by mere potential.
The Pain of Transitioning Between Old and New Growth Drivers
Baidu is in the midst of a difficult transition from a high-gross-margin advertising company to a mid-gross-margin technology service provider. This process requires time:
- Short-term Challenges: The pressure from declining advertising revenue will persist, and AI investments will continue to consume funds, potentially leading to further profit pressures.
- Long-term Outlook: Success depends on whether AI applications (like Wenxin Yiyuan) can become successful products and whether AI-based marketing can replace traditional advertising models. Additionally, the Kunlun Chip business needs to prove its profitability.
Li Yanhong expresses confidence in long-term growth, but Wall Street demands immediate proof of short-term profits. Baidu must endure this period of instability as it transitions between its old and new growth drivers.
Finally, please note that this analysis does not constitute investment advice, and the stock market is subject to risks.
(Information sources: Baidu Q2 financial report, investment bank research reports, public media articles)