Was Tencent wrongly targeted?

A debate about "who is the AI winner" is causing fundamentally solid companies to suffer unjustified sell-offs.

As investors shift their focus from steady profit growth to advanced model development and chatbot competition, Tencent's valuation has dropped to its lowest levels in recent years. Since the beginning of this year, Tencent's Hong Kong shares have fallen by 10%, and are down 21% from last year's peak.

According to Chasewind Trading Desk, Bernstein's latest research shows that Tencent's share price decline is mainly due to a dual blow: lagging AI model development progress and concerns of AI disruption in the gaming industry. Currently, Tencent is trading at 14-15 times the projected 2027 earnings, close to the trough seen during 2022-2023 when the industry faced stalled game approvals and COVID lockdowns. Analyst Robin Zhu points out that although Tencent indeed needs to catch up in chatbot development, the company continues to generate steady AI investment returns through its advertising and gaming businesses. The current valuation reflects overly pessimistic expectations.

At the same time, Alibaba has won market favor due to its AI model progress. With the launch of the Qwen agent service and the release of the Qwen3-Max-Thinking model, Alibaba holds a leading position in the domestic AI race. Alibaba Cloud's revenue grew 30-40%, plus having a global top ten AI model and a leading GPU development project in China, provides the company with long-term AI optionality. However, analysts also question when investors will demand more evidence to prove that agent AI can drive e-commerce GMV share growth and CMR growth.

Red Envelope Battle Fails to Solve Chatbot Dilemma

On the eve of the Spring Festival, Tencent, Alibaba, ByteDance, and Baidu all launched red envelope sharing campaigns, attempting to drive user growth for AI chatbots. Behind this marketing war lies a key issue: on the consumer side, AI model capabilities are only one part of the equation.

Unlike the 2015 red envelope campaign that promoted mobile payments, this time the logic chain behind incentivizing users with cash to use AI chatbots is much more indirect. Ten years ago, users accumulated cash in their WeChat Wallets, directly catalyzing the spread of online payments and laying the foundation for future transaction ecosystem growth. The closed loop of "giving users money, letting them spend within the same app" makes sense in e-commerce scenarios—for example, shopping with Tongyi Qianwen—but how red envelope funds drive users to conduct information searches or seek AI companionship is far less obvious.

QuestMobile data provides some early clues. Alibaba's milk tea subsidies pushed Qianwen's daily active users to soar to 73.5 million on February 7, then fell back to 68 million the next day. Tencent's Yuanbao app's daily active users nearly doubled to 18.3 million on February 7, then dropped to 16.7 million on February 8. This surge and plunge pattern raises questions about new user retention rates.

The deeper issue lies in user engagement. While red envelope campaigns have dramatically raised download counts and daily active users, overall engagement as measured by daily activated conversations and session length has shown limited progress. As of the end of January, high-frequency users (those using more than 10 times per day) of Tongyi Qianwen were about 500,000, while Yuanbao had about 2 million—numbers still insignificant compared to major traffic portals like WeChat, Douyin, and Xiaohongshu. Reprogramming human user behavior will take a long time.

Tencent: Underestimated Profitability

Market concerns about Tencent's AI progress are not unfounded. The company's Hunyuan Turbo S model ranked only seventh on LMarena when released in February, and the company is noticeably absent from public chatbot leaderboards. The recent Yuanbao Party campaign performed poorly, seemingly reflecting the dual-track AI development inside and outside WeChat—a messaging infrastructure for 1.4 billion users, WeChat does face a higher quality threshold before launching new features.

But this narrative pressure is already fully reflected in the valuation. According to Bernstein's projections, Tencent is currently trading at 14-15 times expected 2027 earnings, at the low end of its historical range, near the bottom seen during the severe industry headwinds of 2022-2023.

More importantly, Tencent's operating businesses continue to run steadily. In Q4, Sensor Tower data shows total mobile game revenue fell about 2% year-on-year, but taking into account Delta Force's growth, overall game growth including PC games should remain robust. The ad business also performed strongly, with Video Account revenue exceeding 10 billion RMB in Q4, up about 40% year-on-year.

Bernstein's sensitivity analysis shows that even if operating expenses increase sharply over the next two years to support AI investment, Tencent's 2027 earnings per share will still reach 33-35 RMB. Based on this, the current share price corresponds to a P/E ratio of less than 15 times. In contrast, when Google, Amazon, and Meta each take turns being the "AI laggard," the market eventually returns to fundamentals. Robin Zhu says that after a period of valuation cuts, Tencent's share price psychological model resembles that of the U.S. internet sector.

The key issue is how long investors are willing to wait. Management has hinted that Hunyuan 2.0 is the first model released after the AI development team was reorganized, and iteration speed should accelerate from now. But before that, the market may continue to focus on the "seeing is believing" story of chatbot development.

Alibaba's Bold AI Gamble

Alibaba's strategy has become clearer in recent weeks. On January 15, Qianwen launched an agent service, enabling users to perform e-commerce transactions, order food, and book travel via the app. The subsequent release of the Qwen3-Max-Thinking model scored around 40 on the Artificial Analysis intelligence index, keeping pace with Minimax and Z.ai, maintaining the company's domestic leadership position.

The company announced a 3 billion RMB subsidy program to drive user acquisition and incentivize the use of new features. If all goes to plan, incremental traffic from Qianwen and the food delivery business should help drive visits to platforms like Taobao, thereby supporting CMR growth.

But the cost cannot be ignored. In the second quarter of FY2026, Alibaba reported group capital expenditure of 31.5 billion RMB, while operating cash flow was only 10.1 billion. Bernstein expects improvement in the latter in Q3 and subsequent quarters, as spending associated with food delivery competition subsides. But if investors' tolerance for cash burn in the name of AI capital expenditure drops, this could limit Alibaba management's strategic freedom.

From a valuation perspective, Alibaba's share price has been higher than Tencent's in recent months, which Bernstein believes reflects investors using the SOTP valuation method for Alibaba Cloud. Analysts’ mental framework is still "e-commerce business of $100 per share, plus AI optionality." The core e-commerce business generates about $2.0-2.5 of quarterly NOPAT per share, providing some support for the share price.

As for how to value Alibaba Cloud and broader AI efforts, multiple angles can be debated—but in a world where Minimax and Z.ai market caps have hit $20 billion, and Baidu's semiconductor spin-off gets similar attention, having (1) a global top-ten frontier model, (2) Alibaba Cloud with 30-40% revenue growth, and (3) one of China's best GPU development projects, this combination clearly has value.

~~~~~~~~~~~~~~~~~~~~~~~~

The above excellent content is from Chasewind Trading Desk.

For more detailed analysis, including real-time commentary and frontline research, please join [Chasewind Trading Desk ▪ Annual Membership]

Risk Warning and DisclaimerThe market has risks, investment needs caution. This article does not constitute personal investment advice and does not take into account the unique investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their particular situation. Invest accordingly, and you are responsible for your own decisions.