The core of the narrative of the revaluation of Chinese technology assets: Revaluing Alibaba

Author | Shen Siqi

Editor | Hard AI

The pricing logic of Chinese technology assets in the market is undergoing a dramatic change.

Chinese internet technology stocks, represented by Alibaba, are beginning to show signs of revaluation after a prolonged period of valuation stagnation. In the past week, Alibaba's stock price in Hong Kong rebounded over 13% (while the Hang Seng Index rose 5%), with an overall rebound of more than 40% from its lows.

However, the market's revaluation of Chinese technology assets may just be the beginning. When the market starts discussing Alibaba's AI and cloud computing capital expenditures (Capex), Chinese tech stocks may usher in a true AI-themed market, similar to the narrative of the US stock market's MAG7.

In the financial report that Alibaba is set to release on February 20, the scale of investment in AI and cloud computing Capex will become the focus of market attention, which means how Alibaba's management plans to use funds may influence the direction of Chinese tech stocks.

The Valuation Dilemma of Chinese Tech Stocks?

Before this rebound, Alibaba's valuation was around 10 times PE, a level that is not only far below the average 20 times PE of the US tech sector but also close to the valuation range of traditional retail companies. Some analysts even bluntly stated that the market's valuation of Alibaba is roughly based on a 10 times PE for its e-commerce business, with almost no valuation attributed to Alibaba's cloud and AI.

The market's underestimation of Alibaba mainly stems from three aspects.

First, there is a systemic underestimation of China's technological innovation capabilities. A recent report from Morgan Stanley pointed out that global investors' valuation system for Chinese tech companies still lingers in traditional internet thinking, failing to fully reflect their breakthroughs in cutting-edge technology fields.

Second, investors overly position Alibaba as an e-commerce company, neglecting its technological accumulation in cloud computing and AI. The value of Alibaba's cloud computing and AI business is almost completely ignored in the current valuation system.

Finally, the valuation continues to be under pressure due to non-fundamental factors such as geopolitical issues.

China's AI Technological Breakthroughs Bring Industry-Wide Valuation Increases

However, this underestimation of Alibaba and even Chinese technology underwent a sudden change during the Chinese Lunar New Year.

The open-source large model launched by the Chinese startup DeepSeek caused a global sensation. This model is close to international leading levels in performance, and the training and inference costs have been significantly reduced, quickly narrowing the technological gap between China and the United States in the field of artificial intelligence.

Subsequently, in the early morning of January 29, 2025, Alibaba Cloud officially released the Tongyi Qwen 2.5-Max model, which adopts a MoE (Mixture of Experts) architecture and is based on over 20 trillion tokens of pre-training data, achieving breakthrough progress in multiple core technical indicators.

In tests conducted by the authoritative evaluation platform ChatBot Arena, Tongyi Qwen 2.5-Max scored 1332 points, ranking seventh globally, particularly excelling in mathematics and programming ability tests, where it ranked first in the worldThe release of DeepSeek RI and Alibaba's Tongyi Qwen 2.5-Max has become a key turning point in reversing market perceptions.

DeepSeek's low-cost and efficient model challenges the notion that traditional AI development requires significant capital investment. Its open-source strategy in collaboration with Alibaba's Qwen allows more developers to participate in AI technology innovation, challenging the closed-source model of OpenAI and promoting the democratization of global AI technology.

The success of DeepSeek and Alibaba has triggered a reevaluation of Chinese tech stocks in the global capital markets.

Morgan Stanley pointed out that China's technological strength in the AI field is underestimated, and investors should pay attention to the innovation capabilities of Chinese tech companies.

After the Spring Festival, the technology sectors of Hong Kong and A-shares performed strongly. Chinese AI tech stocks, led by Alibaba, have seen continuous price increases, and market confidence in Chinese tech stocks has begun to recover rapidly.

Recently, foreign institutions led by Goldman Sachs, Morgan Stanley, and Bank of America have quickly changed their attitudes toward Chinese assets, believing that DeepSeek represents China's "Alibaba IPO moment." The 2014 listing of Alibaba had driven the rise of China's "new economy" sector, attracting long-term capital inflows from around the world. The rise of DeepSeek presents an opportunity for mid- to long-term value reassessment of Chinese concept tech stocks.

However, the surprises brought to the market by Chinese AI companies led by Alibaba do not stop there.

According to a report by The Information on Tuesday, Apple and Alibaba will develop AI features for Chinese iPhone users. This means that Apple recognizes Alibaba's Tongyi Qwen model, while in overseas markets, Apple's partner is OpenAI, which every AI company aspires to catch up with.

Apple's AI architecture is divided into three layers: device-side models, cloud models, and external third-party AI models.

The first two layers rely on self-developed software and hardware, while the third layer collaborates with OpenAI in overseas markets.

In the second half of last year, Apple faced severe challenges in the Chinese market. Financial reports show that China is Apple's second-largest market outside the United States; however, Apple's performance in China has continued to be under pressure: sales are expected to decline by 7.7% in fiscal year 2024, with a year-on-year drop of 11% in the fourth quarter of 2024.

Apple CEO Tim Cook directly pointed out in the recent earnings call that sales in the Chinese market in the fourth quarter of last year were affected by the iPhone's failure to provide Apple Intelligence features.

To address the issue of AI partners in China, Apple began searching for suitable AI partners in China in 2023. After evaluating several companies, including Baidu, Tencent, ByteDance, Alibaba, and DeepSeek, the technological strength and commercialization capabilities of Tongyi Qianwen ultimately stood out, opening up a vast market space covering over 200 million Chinese iOS users for the Qwen large model, a customer acquisition result that other large model companies have not necessarily achieved despite significant investment.

Additionally, it is worth noting that DeepSeek's recent popularity has also piqued Apple's interest; however, Apple has abandoned DeepSeek's model due to the team's lack of manpower and experience to support large clients like AppleAfter all, for Apple, a company that places the utmost importance on user experience, it is unacceptable for users to receive a message stating that the server is busy and to try again later after making an inquiry.

Therefore, it is only natural for Tongyi Qwen, supported by Alibaba Cloud, to become a partner of Apple. The value of cloud providers lies more in their immediate scalability and continuously updated software support rather than the cost itself.

AI brings an opportunity for Alibaba Cloud's value reassessment

As mentioned at the beginning of the article, the market values Alibaba at approximately 10 times the PE of its e-commerce business, with almost no valuation attributed to Alibaba's cloud and AI.

However, this situation may change.

Guosen Securities stated in a recent report that Chinese AI large models, represented by DeepSeek, are expected to accelerate the pace at which domestic cloud providers catch up with large models and narrow the gap in model layers. At the same time, it will accelerate the migration of domestic enterprises to the cloud, benefiting the downstream demand growth for cloud providers. This will significantly reduce the time and resource costs between the initial investment in cloud AI and the realization of applications, and under economies of scale, it is expected to further enhance the profit margins of domestic cloud providers.

In simple terms, before the Spring Festival, the gap between Chinese AI large models and those in the United States may be around 1.5 to 2 years. Although the United States has already observed a surge in demand for AI applications in cloud computing, China is still in a market similar to that of the United States two years ago, where the demand for inference computing power has not yet significantly exploded.

Google, Microsoft, Amazon, and Meta all clearly stated in their latest financial reports that strong demand in the AI field has prompted them to increase investments in data centers to meet customer computing power needs. Google and Microsoft even expressed that their cloud computing businesses are facing challenges of "capacity constraints," which has affected revenue growth. In simple terms, there is not enough computing power, which impacts the profitability of cloud computing giants. The four major cloud computing providers in North America are expected to exceed an unprecedented $320 billion in AI-related Capex by 2025 (Amazon $105 billion; Meta $60-65 billion; Google $75 billion; Microsoft $80 billion).

With the emergence of DeepSeek, the market has realized that the progress of Chinese AI large models has quickly caught up with that of the United States, and the cost reduction is even faster. The AI cloud computing business in the Chinese market is also expected to face a situation of supply not meeting demand, similar to that in the United States.

Wall Street investors are closely watching the capital expenditures of major U.S. cloud computing providers, partly because it relates to the performance of upstream hardware manufacturers, and partly because they are concerned about the growth in consumer demand for inference, which directly affects the performance of these giants.

For Chinese cloud providers represented by Alibaba Cloud, the narrative in the U.S. market is that cloud computing has shifted from a cyclical business back to a growth business.

Previously, the market mainly discussed the valuation of Alibaba's e-commerce business, the profit margins of its cloud services, and the amount of share buybacks by the group. This means that even after a significant rebound in the current stock price, AI may still represent a recovery trend for Alibaba.

When the market begins to discuss how Alibaba is spending on Capex, the narrative similar to that of the MAG7 in the U.S. stock market will truly arrive at the AI theme market. I believe that on February 20th, Alibaba's financial report will certainly make the Capex issue a focal point of market inquiries.Also, let's see how Alibaba's management expresses the matter of "spending" this time, whether they can redirect funds towards the "growth" businesses of "predictable" AI and cloud, rather than just stock buybacks.

If such a shift occurs, the market may bring Chinese AI into a new narrative environment, perhaps a new hope for the overall Chinese economy?