My Thoughts On The Apple x Baidu Gen AI Partnership
On March 22nd, news broke that Apple (AAPL) has reportedly been in talks with Baidu (NASDAQ:BIDU) to leverage the Chinese company’s Gen AI platform for the iPhone in China. Such a deal, while still speculative, would make lots of sense for both parties. In my view, there is a close to zero probability that the Chinese government will allow Western Gen AI tools to operate in Mainland China. This thesis is anchored on the observation that data-heavy application products developed by software companies such as Meta Platforms and Google are not allowed in China. Indeed, also OpenAI’s ChatGPT is currently not allowed to operate in China. What this means for Apple is that the company will highly likely need to source Gen AI technology/ applications locally. On that note, surrendering or delaying the usage of Gen AI for iPhones sold in China won’t really be an option, in my opinion: Investors should consider that the ongoing Gen AI revolution is likely the most significant technological shift of the 21st century (personal view). In that context, if Apple were to sell iPhones in China without this technology, then Chinese handset makers–notably Huawei–would certainly step in to offer Gen AI on their products, handing a major competitive disadvantage to Apple. This would mean that Apple could lose market share in one of the company’s most important markets.
Pointing to the news report that suggested that Apple is looking for a Gen AI partner in China, I highlight that discussions are still in the very early stages. Moreover, it is unclear if Apple is also considering partnerships with other local AI firms. However, picking Baidu as the company’s core partner for developing and adopting Gen AI technology in China makes sense for Apple: I consider Baidu to be the strongest AI player in China’s internet domain. In a previous article covering Baidu, I argued that company’s LLM model, Ernie Bot, “may be as good and growing as fast as ChatGPT”, highlighting the following:
Baidu’s CTO Wang Haifeng recently disclosed that Ernie Bot has topped 100 million users. Notably, the successful passing of this milestone comes less than 5 months after the Chinese government allowed citizens’ mass access to the technology. Ernie 4.0, which is the latest version of the bot, can handle various Gen AI-supported tasks, including text generation, summarization of documents, and code generation. On that note, Baidu has reportedly been working on the technology since 2019 and should support a level of sophistication similar to OpenAI’s GPT-4, as claimed by Baidu CEO Robin Li. A similar commentary has been voiced by actual technology testers, with CNN saying: “CNN gave ERNIE and GPT a few simple tasks. The takeaway: You can’t go wrong with either” [adding that] “ERNIE beat GPT-4 on certain prompts, such as those related to current affairs. The Chinese bot knew that Taylor Swift is now a billionaire, that China had recently removed its defense minister and that “Friends” star Matthew Perry had died”.
Assuming that Apple will eventually partner with Baidu for developing and deploying Gen AI applications in China, due to reasons highlighted above, I am bullish on the related commercial implications for Baidu: On a broader level, I estimate that there are about~240-260 million iPhone users in China. Needless to say, this would imply an enormous addressable user base for Baidu. And even Baidu will likely not manage to extract a direct revenue contribution out of the partnership with Apple, the potential collaboration with would mark a critical step for the widespread adoption of Ernie LLM, as it would validate Baidu’s AI prowess and reinforce the company’s leadership position on Gen AI in the Chinese market. Moreover, Baidu would be poised to get access to lots of valuable usage/ user data, which could then be leveraged to training applications, as well as cross selling on other Baidu products (e.g., advertising on Baidu Search).
A Quick Comment On Baidu’s Broader Gen AI Outlook
With Q4 reporting, Baidu disclosed that the company’s revenue from AI Cloud grew by 11% YoY, to RMB5.7 billion. In a further comment, CEO Robin Li shared that out of the total revenue from AI Cloud, about 4.8% is attributable to the company’s Gen AI and foundation model. Doing the math, this would suggest that Baidu’s Gen AI business generated about RMB250-280 million of sales during the period from September through end of December 2023. While this may not be much at first glance, investors should consider that Baidu’s Ernie Bot has only launched in August 2023, and thus, the growth and commercialization of the product is still at a very early stage. On that note, it is reasonable to assume that commercial momentum of Baidu’s Gen AI business will pick up notably in 2024 and sustain high growth rates through the next few years–independent of the aforementioned collaboration with Apple. In fact, in the conference call with analysts following the Q4 results release, Baidu’s CEO Robin Li has suggested that incremental revenue coming from Gen AI will likely multiply into “several billion RMB already in 2024:
Looking into 2024, we believe this incremental revenue will multiply to several billion RMB primarily from advertising and AI cloud building. While we are beginning to commercialize Gen AI and foundation models, we see enormous monetization potential in this ground breaking technology. We invasion ERNIE as the future foundation system, serving as the foundation for millions of AI native applications developed by third-party and Baidu. This paradigm will enable us to create an ecosystem around ERNIE which opens up whereas revenue sources.
While it is hard to pinpoint the reference of “several billion” to a more precise number, I would argue that investors could likely expect a Gen AI revenue contribution in 2024 vs. 2023 of RMB 5-10 billion. Post-2024, a ~25% topline CAGR may be possible, as the Chinese Gen AI market is expanding towards a $29.5 billion SAM by 2030.
The Call For Shareholder Distributions Is Debatable
Amid slowing top-line growth and increasing profitability for Chinese internet firms, some companies like Tencent (OTCPK:TCEHY), Alibaba (BABA) and JD (JD), have turned more committed to distributing capital to shareholders, with total expected 2024 equity yields (buybacks and dividends) ranging from 5-7%. Meanwhile, Baidu has been quite reluctant to distribute cash to shareholders, despite the company’s enormous cash position of RMB 205.6 billion, paired with FY 2023 free cash flow of about RMB 25.4 billion: In FY 2023, Baidu distributed only RMB 4.8 billion to shareholders, suggesting a yield of 1.7% compared to the company’s RMB 267 billion market cap as of April 2024.
While higher capital returns to shareholders are desirable, Baidu’s conservative approach to payouts could be justifiable given its growth prospects in Generative AI. Considering the vast potential of Generative AI, alongside the uncertain landscape of competition, regulatory hurdles, and capital requirements, Baidu’s decision to retain cash may enable the company to remain more agile and responsive to opportunities as well as challenges. Therefore, the current absence of share buybacks or dividends from Baidu should not be inherently viewed as a drawback.
Investor Takeaway
Apple is reportedly in talks with Baidu to use its Gen AI platform for iPhones in China. The partnership could provide Baidu with a large user base and valuable data, boosting the company’s Gen AI business. Taking a broader view on Baidu’s Gen AI business, I highlight that the company’s CEO has suggested that Baidu’s Gen AI business generated about RMB 250-280 million of sales during the Q4 2023 period. According to further CEO commentary, I estimate that this revenue could likely multiply to about RMB 5-10 billion in FY 2024 and then run on a 25% CAGR through the next few years through 2030.
Looking ahead to FY 2024, I expect that Baidu’s total revenue will grow about 7% YoY, to RMB 145 billion, supported by a strong performance in the Cloud business, and under the assumption of a macroeconomic recovery supporting Baidu’s advertising sales at a reasonable 5% YoY growth. On profitability, I project Baidu’s adjusted net income to stay relatively unchanged vs. FY 2023, at around RMB 29 billion (I model that the higher topline in FY 2024 vs. 2023 should be largely offset by an incremental increase in R&D costs).
Despite Baidu’s healthy commercial momentum in the company’s legacy business (advertising/ search), and potentially explosive upside in the Gen AI space, Baidu shares are trading very inexpensive — pointing to an EV/EBIT of approximately 7x, compared to ~20-25x for U.S. tech businesses like Google (GOOG), Meta (META). On that note, I continue to argue that Baidu shares are trading too cheap to ignore: I have previously calculated that Baidu’s intrinsic worth could be somewhere around $179/ share; and I continue to view that estimate as reasonable. “Buy”.