Analogy to the 2020-21 Chinese Internet, HSBC says: Another new trend emerges, Asian AI stocks peak
HSBC believes that, similar to the performance of Chinese internet stocks after the peak in 2020-2021, Asian AI stocks are also undergoing a "lagging adjustment". The timing of when the massive investments in AI will pay off remains unknown, and the market's optimism towards the AI sector is changing, gradually lowering the target prices and profit expectations for these stocks
On the 20th, HSBC released its latest research report, stating that Asian AI stocks are undergoing an adjustment similar to the Chinese internet stocks in 2020-2021.
According to HSBC's report, since July 2024, the prices of AI-related stocks in Asia have dropped significantly, with their custom Asian AI index falling by 15% from its peak on July 11. This pullback has become the most severe sell-off the sector has experienced since the release of ChatGPT in 2022.
The report mentioned that as the prices of AI-related stocks rapidly decline, the market's optimism towards the AI sector is changing, gradually lowering the target prices and profit expectations of these stocks. HSBC analysts believe that Asian AI stocks may have peaked and are entering a new adjustment phase.
The report specifically highlighted three major challenges facing the current Asian AI market: weak demand, regulatory pressure, and reduced profit expectations, with "reduced profit expectations" being a factor that has only recently started to take effect.
Reduced profit expectations, similar to Chinese internet stocks in 2020-2021
HSBC pointed out in the report that signals of reduced profit expectations and target price adjustments have emerged.
The report stated that the recent financial reports of Mag7 and the subsequent stock price trends fully demonstrate a shift in investor sentiment. While most financial reports did not have significant surprises, the stock prices of these companies reacted negatively after the reports were released. This impact has spread to the Asian tech industry.
Although most AI stocks still maintain optimistic ratings, with 89% of AI stocks currently rated as "buy," analysts are beginning to adopt a cautious stance on their profit outlook.
The target price increases seen earlier this year have passed. Since July, target prices have fallen by 2%. As some target prices are benchmarked against the entire industry, we suspect that these figures may see a significant decline in the coming weeks.
HSBC pointed out that similar to the performance of Chinese internet stocks after the peak in 2020-2021, Asian AI stocks are also undergoing a "lagging adjustment."
Profit forecasts have also peaked. Profit forecasts have been slightly reduced. This trend (analysts lowering forecasts two months after stock prices peak) is very similar to what we saw in Chinese internet stocks in 2020-2021.
This indicates that further adjustment in the AI sector may only be a matter of time.
When will the massive investments in AI pay off remains unknown
HSBC believes that although tech giants are increasing their investments in AI infrastructure, the market is skeptical about how these investments will translate into actual returns.
NVIDIA CEO Jensen Huang previously predicted that global AI infrastructure investments will reach $1 trillion in the next 4 to 5 years, but the market demand has not yet caught up with such massive investments.
Large tech companies continue to heavily invest in strengthening infrastructure, far exceeding the actual demand that has yet to materialize. According to their latest data, in the second quarter of 2024, the average capital expenditure of Mag7 increased by 50% year-on-year Some people want to know how these investments can be monetized? Is there enough demand to ensure such a huge investment?
AI Regulatory Risks Highlighted
Finally, HSBC believes that the future regulatory risks that AI may face mainly lie in the power consumption of AI data centers.
Europe and other regions have successively introduced stricter regulations on data privacy and the AI industry. The United States has also launched an antitrust investigation against NVIDIA. The report also points out that the huge power consumption of AI data centers may bring more regulatory risks in the future.
For example, the power consumption of one search by ChatGPT is 10 times that of Google search.
In addition, the research report points out that the excessive concentration of funds in the AI field has also raised concerns in the market. HSBC stated that due to high valuations and risk exposure, the company will maintain a underweight position in the Chinese and Taiwanese markets