Everbright Securities: Meta computing power leasing is a resource optimization method and does not change the medium-to-long-term expansion plan.
Related
Industrial Securities: The news regarding Meta's external leasing and sales of computing power should not be interpreted overly pessimistically as a sign of computing power surplus.
According to Mars Finance, on Wednesday, news that Meta plans to sell its surplus AI computing power triggered a sharp correction in global AI hardware and high-flying technology stocks. Xingye Securities believes that the news regarding Meta's external leasing and sale of computing power should not be interpreted overly pessimistically as a surplus of computing power/a comprehensive slowdown in Capex, because (1) this is not new news; there were related reports in May of this year; (2) Meta is a special case among hyperscalers; its toC-oriented business means its AI monetization capabilities mainly rely on advertising, and its exploration of cloud business can improve shareholder returns and cash flow; (3) Meta still has a computing power gap; this week, it was just reported that Google restricted its access and Meta signed an agreement with Crusoe; (4) hardware demand comes from inference, not from inflation on the training side. (Cailian Press)
Galaxy Securities: We recommend focusing on leading computing power leasing companies.
According to Mars Finance, Galaxy Securities points out that with the explosive growth in demand for AI inference computing power, the token factory business model is accelerating. Compared with traditional computing power leasing, token factories place greater emphasis on binding with downstream large-scale model inference needs. Pricing is gradually shifting from "computing power resources and duration" to "token output and unit cost." Upstream computing infrastructure can participate more in the value distribution of downstream model manufacturers, strengthening the bargaining power of the upstream computing infrastructure segment and sharing the dividends of the token economy era. Galaxy Securities believes that computing power leasing companies are currently exploring the "token factory" cooperation model with large model manufacturers, which will gradually evolve from the underlying "selling resources" model of computing power leasing to a "selling output" model. Computing power leasing companies will provide MaaS services and share the dividends of the token boom by adopting a token revenue-sharing profit model. This is expected to improve the profit margin of computing power leasing companies, and the valuation center is expected to further increase. It is recommended to pay attention to leading computing power leasing companies.
Meta's rental of computing power is similar to using existing resources to fund new ones, rather than ceasing the pursuit of high-end computing power.
According to Mars Finance, Meta is planning to launch a cloud computing business that leases computing power: one approach is to open up the model capabilities deployed on its own AI infrastructure to external customers, and the other is to lease out the more fundamental "raw computing power." However, the current situation seems more like Meta is using its existing, older computing power for cash flow recovery, rather than ceasing its pursuit of high-end computing power. In mid-to-late June, Meta was reported to have signed an agreement with Crusoe to acquire a total of approximately 1.6GW of AI computing capacity from two data centers in Texas and Missouri. Meanwhile, Meta raised its full-year capital expenditure guidance for Q1 2026 to $125-145 billion. Looking at these two events together, it seems more like a reallocation of resources across different generations and for different purposes: continuing to buy new GPUs to train cutting-edge models, but leasing out a portion of older GPUs (such as the H-series) used for high-traffic inference products and hosting external models, while not necessarily slowing down the purchase of the most scarce high-end GPUs. (Cailian Press)
Zuckerberg signals the commercialization of computing power, suggesting that Meta infrastructure may enter a dual-track model of "self-use + external leasing".
According to BlockBeats, on July 1st, Bloomberg reported that Mark Zuckerberg stated in a May shareholder call, "Almost every week, different companies come to us, either wanting us to build an API service or asking if we have computing power to sell to them, at a higher price than we paid. We haven't done that yet because we have our own need for that computing power," Zuckerberg said at the time. "But obviously, if we realize we've overbuilt, then that's an option, which partly gives us the confidence to invest in scaling up." In the rapidly evolving AI race, Zuckerberg has repeatedly stated that he believes the entire industry is facing a bottleneck in computing power, and that Meta should accumulate as much computing power as possible before deciding how to use it. Despite facing numerous complex challenges, Meta CEO Mark Zuckerberg has signaled to investors that he is willing to sell excess computing infrastructure and even launch so-called API services, allowing customers to pay based on AI usage—a business typically measured in "tokens" (i.e., the amount of data generated and consumed when a customer queries). Discussions surrounding Meta Platforms' potential commercialization of some of its AI computing power have raised concerns about a potential "oversupply" of computing power. However, analysts argue that this assessment is oversimplified. Meta currently does not have any significant surplus computing power available for sale. On one hand, it continues to expand its investment in AI infrastructure and has signed large-scale computing power agreements with companies like Crusoe. On the other hand, its existing H100/H200 resources are primarily used for internal recommendation systems and AI model training, and demand remains tight. Therefore, the so-called "selling computing power" is more of a future option for asset utilization than a current reality. Overall, it reflects a shift in AI computing power from a single scarcity to a multi-generational, tiered use and long-term capacity planning phase.
The computing power leasing concept rebounded after fluctuations, with Wangsu Science & Technology rising nearly 15%.
According to Mars Finance, the computing power leasing concept stocks rebounded in early trading, with Wangsu Science & Technology rising nearly 15%, followed by Sangfor Technologies, Cloudwalk Technology, UCloud, Aofei Data, and Xiechuang Data. Galaxy Securities believes that computing power leasing companies are currently exploring a "token factory" cooperation model with large-scale model providers, gradually evolving from the underlying "selling resources" model to a "selling output" model. Computing power leasing companies will provide MaaS services and share the benefits of the token boom through a token-sharing profit model. (Cailian Press)
CITIC Securities: Significant fluctuations will not alter the AI supercycle; emphasize the importance of domestic computing power as a "Plan B".
According to Mars Finance, CITIC Securities points out that news of Meta's plan to lease out some of its computing power has once again triggered market concerns about computing power oversupply. This, coupled with concerns about cloud vendors' cash flow pressures, the continued rise in upstream prices, the slowdown in Capex growth, and overcrowding in AI transactions in recent months, is the main reason for the significant volatility in tech stocks. In the short term, Meta's move is primarily aimed at revitalizing its existing, outdated computing power assets. Considering its continued development of advanced models and investment in next-generation computing hardware, leasing out computing power is not contradictory to further increasing its investment in computing power. Furthermore, since computing power rental fees have continued to rise recently, concerns about computing power oversupply are unfounded. This round of tech stock adjustments is more of a deleveraging and rebalancing process in the recent global liquidity tightening environment, rather than a reversal of the AI industry trend. For the medium to long term, it is crucial to pay close attention to whether the next few months will see a similar breakthrough in AI capabilities as seen with OpenClaw and Coding Agent at the beginning of the year. In addition, it has been observed that after overseas AI assets entered a phase of high crowding, high correlation, and high volatility, international funds are beginning to seek differentiated sources of return. Domestic computing power with differentiated value, dubbed "Plan B," remains resilient and is expected to attract foreign investment. With the earnings season approaching, we recommend focusing on sub-sectors with high earnings certainty and reasonable valuations: In terms of growth prospects, we recommend domestic FAB (Featured Adhesives and Materials) and equipment sectors with positive narratives, as well as the optical communication sector with relatively low valuations; in the price increase chain, segments with high AI exposure and those that have already experienced price increases have a higher probability of realizing their earnings gains, such as memory and upstream PCB industries. (Cailian Press)