Citibank: Hong Kong stocks "real technology stocks" account for a small proportion, and it is difficult for the Hang Seng Index to have a significant improvement in the second half of the year.

The performance of Hong Kong stocks in the first half of the year was relatively weak, with a high of 28000 seen at the beginning of January and then a continuous decline. The main reason for this was that Hong Kong stocks are dominated by technology stocks, and the surge in memory prices has significantly increased capital expenditure for AI research and development companies. Heavyweight technology stocks in Hong Kong suffered a major decline, leading to the continuous downturn of the Hang Seng Index in the first half of the year. Citigroup's China stock strategist Liu Xianda, in an interview, pointed out that unexpected changes in interest rate expectations and regulatory measures in the mainland, coupled with downward pressure on platform stocks, make it seem like a castle in the air for the Hang Seng Index to surpass 30000 in the second half of the year. However, he does not underestimate the AI craze, but instead recommends starting to allocate funds for the next stage of AI development, not just short-term deployment within a year. At the beginning of this year, Citigroup set the year-end target for the Hang Seng Index at 30000, but by the end of the first half of the year, the target had been adjusted to 29600. Liu Xianda stated that both Hong Kong and A-shares have not performed well so far this year. The market confidence is weak due to the disparity between the anticipation of interest rate cuts at the beginning of the year and the current need to guard against interest rate hikes. The lack of significant growth in mainland consumer retail has also contributed to this weak market confidence, with the most important bearish factor for Hong Kong stocks stemming from structural issues. In the first half of the year, global capital favored AI infrastructure stocks, with concepts such as storage chips, optical communication, and PCB becoming the market leaders. In comparison, the proportion of "true technology stocks" at the forefront of technology in Hong Kong stocks is small, as the technology stocks in Hong Kong are mainly platform stocks that are more influenced by retail consumption markets such as takeaways and automobiles. Therefore, in terms of true technology stocks, Hong Kong stocks account for only 4%, while A-shares account for only 15%, much lower than the United States, mainland China, Taiwan, and South Korea, which weakens global capital's risk appetite for Hong Kong stocks. Since platform stocks still account for a significant proportion of Hong Kong stocks, it is expected that there will not be a major change in investment style in the second half of the year. Therefore, it is difficult to expect significant improvement in the Hang Seng Index in the second half of the year. When asked if it would be necessary to adjust the year-end target for the Hang Seng Index, Liu Xianda mentioned that the index forecast calculation method would consider profit growth, market capital flows, and he would continue to observe and only adjust the target as needed. He also mentioned that the Hang Seng Index company is unlikely to directly lower the weightings of platform stocks, but as the composition of the Hang Seng Index continues to expand, the addition of new stocks will help dilute the weighting of platform stocks. For example, in recent quarterly reviews, the Hang Seng Index company has added pharmaceutical, industrial, and resource stocks. Assuming that the weighting of new stock sectors gradually increases in the future, the proportion of platform stocks could be reduced from the current 32% to around 25% to 28%, which would help to diversify some of the volatility and regulatory risks. The first half of the year can be said to be a watershed moment for the AI market. After Meta announced the sale of excess computing power, concerns about AI computing power surplus deepened. After July, the AI infrastructure stocks saw a 180-degree reversal, with software stocks rebounding but AI infrastructure stocks experiencing a deep correction. Liu Xianda's analysis suggests that the AI market is far from over but will undergo phase transformations. He believes there are three stages of the AI frenzy, from the early stages of hype in 2023 to around 2025, leading to a surge in valuations of AI-related concept stocks. This year, the market is shifting towards the upstream AI infrastructure suppliers after the hype, and he expects that starting from the second half of this year, the market focus will shift from AI infrastructure investment to the implementation of AI applications, with a greater emphasis on the ability to monetize concepts. He believes that in the next one or two years, the market will no longer pay high premiums for vision alone but will pay more attention to actual order volumes and revenue growth. The prosperity of upstream computing hardware will continue, but the growth may not be as high as in the past two to three years, and there may be opportunities for prices to fall. Regarding the surge of new AI stocks this year, Liu Xianda emphasizes that the current short-term selection of new stocks is mainly based on AI concepts, and new stocks with AI and semiconductor elements will continue to be highly valued. Traditional selection criteria, such as examining the lineup of cornerstone investors and oversubscription, will remain effective. Liu Xianda says that focusing on new stocks with these elements will yield higher returns compared to just holding traditional large blue-chip stocks, with an average annual return of 6% to 9%.