Wall Street on Chinese Stocks: From "Powerlessness" to "Opportunistic Bullishness"

With the strong rebound in the sector in July, overseas investors' pessimism toward Chinese internet and large AI model stocks has completely dissipated, marking a substantive shift from the "sense of powerlessness" in late June to "tactical bullishness."

In his latest market feedback, Bank of America (BofA) analyst Alex Liu pointed out that the three core focal points currently dominating the Chinese internet sector have become clear: the inevitable trend of commoditization in large AI models, the marginal easing of competitive pressure on ByteDance in local life services and e-commerce, and persistently weak macro consumption and summer travel demand.

Capital is re-evaluating the cost-performance ratio and fundamental delivery capabilities of major tech giants. As computing power costs rise and price wars intensify, the economic benefits at the pure model layer are being eroded. A market consensus is gradually forming: the technical difficulty of catching up in large models is not high, and the ultimate commercial winners will emerge in multi-model ecosystems, AI Agent applications, and on-device AI.

Large AI Models Mired in "Commoditization," Application Layer and On-Device AI Become Key to Breaking Through

The race for large AI models in China is experiencing intense Money Churning. Merely leading in performance by a few weeks is no longer sufficient reason to buy into a single large model stock. Capital is closely watching key models about to be released, including DeepSeek V4 Pro, Alibaba Qwen 4.0, Z.AI GLM 5.3, and MiniMax M3.1.

Among unlisted unicorns, Z.AI is regarded as a "high-quality" target due to its strong performance in users and revenue following the release of GLM 5.2. However, facing rapid iteration by peers, its ability to maintain its position in the first tier of coding capabilities is being questioned. In contrast, MiniMax is considered to lag behind the leaders in model capabilities, but its $11 billion valuation is attractive to some capital.

Among the major tech giants, after TENCENT officially released Hunyuan HY3.0 in July, market doubts regarding its progress in foundational large models largely subsided. Amid the trend of multi-model adoption and token price deflation, TENCENT is viewed as a core beneficiary due to its user leadership in the AI Agent application layer with WorkBuddy, as well as improved consumer-side economic benefits driven by the proliferation of on-device AI. However, concerns remain: the large-scale rollout of WeChat Agents may incur significant token costs, which, combined with depreciation pressures, has sparked intense debate about the risk of EPS decline in the second half of 2026 through 2027. Some capital even anticipates a year-over-year drop in earnings for 2027.

ByteDance's Offensive Slows, Trading Platforms See Tactical Bullish Window

Channel research indicates that competitive pressure on ByteDance in local life services and e-commerce is substantively decreasing. This marginal change has directly prompted capital to become more tactically bullish on listed trading platform companies.

Meituan and JD.com have become the core beneficiaries of the easing competition, with market attention significantly rebounding. Estimates show that as competition in instant retail slows, Meituan's unit economics (UE) will continue to improve into the second half of 2026, keeping bullish sentiment high; JD.com is also seeing expectations for EPS improvement, coupled with shareholder return actions such as increased buybacks, leading to a recovery in market interest. However, a common concern for both is whether the funds saved from reducing losses in food delivery and instant retail will be reinvested into new businesses with uncertain endgames, such as AI and overseas expansion. Moreover, visibility into Meituan's order volume growth remains low.

Alibaba has currently become a consensus buy, but capital no longer views it as a "buy and hold" core position. Consensus has formed around stopping losses in its instant retail business, which is expected to bring positive EPS corrections in the coming quarters; expectations for Customer Management Revenue (CMR) have been lowered. Short-term expectations for its cloud business are extremely high, with the market generally expecting its year-over-year growth in the June quarter to reach or exceed 45%, with some capital projecting cloud revenue growth of over 50% by 2027. The other side of the coin is that, facing the rise of independent AI labs, the competitiveness of Tongyi Qianwen (Qwen) and its free cash flow performance remain key areas of focus for capital.

Weak Macro Consumption Suppresses Sentiment, Summer Travel Faces Major Test

Entering July, market expectations for domestic consumption remain bearish, with growing concerns specifically about summer travel demand.

Despite a brief rebound following the end of regulatory investigations, Trip.com has not seen capital rushing to bottom-fish. Affected by weak consumer sentiment and unfavorable weather, recent summer tourism activity has been sluggish, raising market concerns about a new round of EPS downgrades. Additionally, pressure on hotel take rates is increasing, and the company faces the risk of losing market share to Meituan and ByteDance.

NetEase was a consensus long position during the sell-off in June, but the underwhelming performance of "Sea of Remnant" released in early July raised concerns about short-term EPS pressure. However, version updates for older games like "Justice Online" and the launch of unlimited servers in July are expected to support recent performance.

Other Targets

Regarding other targets, for Baidu, the focus of capital has completely shifted to its self-developed AI chip design capabilities, revenue scale, potential valuation, and spin-off timeline, with hesitation on whether to continue holding shares for the potential upside of a chip spin-off.

Bilibili has received relatively less attention, with capital waiting for TENCENT to complete its equity disposal. However, its platform value has been validated after resisting competition from ByteDance in recent years, and organic engagement growth since 2025 is creating scarce value for the company.

Kuaishou faces continuous competition from ByteDance's similar products for its "Kling" model, and lacks short-term catalysts after recent financing, resulting in little market interest. PDD and Tencent Music Entertainment (TME) are also temporarily neglected by capital due to the lack of a clear logic for bottom-fishing.

Risk Warning and Disclaimer

The market carries risks; investment requires caution. This article does not constitute personal investment advice, nor does it consider the specific investment objectives, financial status, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investors assume full responsibility for their own decisions.