Summary of Key Points
Over the past three years, the AI market has experienced three significant “style shifts”: from focusing on “conceptual stories” to investing in companies that provide the necessary “tools” for AI applications, and then to betting on practical, profit-generating AI solutions. The recent “storage crisis” in July was caused by an excessive concentration of funds in the AI sector, leading to a panic sell-off due to a large number of investors seeking to realize their profits. The long-term trend of the AI industry remains unchanged, but short-term fluctuations should be anticipated; ordinary investors need to move away from the illusion of easy wealth through random stock purchases and instead focus on selecting high-quality companies, evaluating their actual performance, and using regular investment strategies.
I. The Three Years of the AI Market: Capital’s Changing Approaches
The AI market has not followed a linear path; rather, capital has tried different approaches, akin to “searching for treasures”:
1. The First Shift: The ‘Imagination Market’ at the Beginning of 2023
After ChatGPT became popular, companies in the A-share market related to AI (including media, gaming, office software, and publishing) saw their prices soar. It was like restaurant owners hearing from a chef that a new dish would be a huge hit, so they bought flour and machinery without hesitation—capital was betting on the possibility that AI could change the world just like the internet did, fearing missing out on the opportunity.
2. The Second Shift: The ‘Tools Stock Boom’ at the End of 2023
After hearing many stories about the potential of AI, investors realized that while the “dish” (AI technologies) were not yet available, the companies selling the “tools” (such as NVIDIA’s GPUs) were making huge profits. NVIDIA’s GPU demand far exceeded supply, driving its stock price to become one of the highest in the world. A-share companies involved in manufacturing these tools also saw significant gains—this time, capital was investing in tangible orders and actual performance.
3. The Third Shift: The ‘Practical Applications Market’ Since 2025
Investors realized that they couldn’t keep buying only the tools without also developing the applications. After DeepSeek released its R1 model, it became clear that practical, low-cost AI applications were more important. Companies like Zhipu saw their stock prices soar as capital began to value those capable of generating real profits.
II. The July “Storage Crisis”: Why Such a Sudden Drop?
The sharp decline was not due to a failure of AI but rather a result of excessive concentration of funds:
- Excessive Fund Concentration: This bull market was highly focused on AI, with other industries stagnating. All funds were directed towards AI, creating a situation similar to people crowded in a narrow alley; any small movement could trigger a panic sell-off.
- Large Number of Profit-Making Investors: The storage sector had been rising for a long time, and many investors wanted to lock in their gains. Additionally, leverage funds from the Korean market were liquidating positions, and American investors were envious of the profits being made by Korean storage companies. These factors combined to cause a dramatic drop in stock prices.
- Institutional KPIs: Fund managers are evaluated every six months, and they needed to hold onto their holdings before the end of June. Once the evaluations were over in July, they started selling their positions, which coincided with individual investors chasing rising prices and then selling when prices dropped.
III. Has the AI Market Ended?
In the short term, we expect a pullback, but in the long run, the industry will continue to grow:
- Short-Term: A Normal Cooling-off: The technology sector currently lacks compelling prospects, and any rebound may be driven by individual investors buying in while institutions sell off. This year’s stock market trend is one of reaching a peak; ordinary investors should avoid trying to buy stocks at their lowest points.
- Long-Term: The Industry Is Moving Forward: Global giants (Microsoft, Google, NVIDIA) are continuing to invest in AI infrastructure, and Chinese companies are working on reducing costs and developing practical applications (such as AI-powered customer service and code writing). Stock prices reflect expectations, but the industry’s progress is real. Expectations will fluctuate, but the industry will continue to advance.
IV. What Should Ordinary Investors Do Next?
The era of easily making money by buying AI stocks is over; a new approach is needed:
1. Choose Companies That Can Survive in the Long Term: The AI sector has become widely recognized, so to achieve higher returns, investors should select companies with genuine capabilities, such as strong technical barriers (e.g., unique algorithms) and customer loyalty (e.g., being used by large enterprises). Avoid companies that are just riding on the AI trend.
2. Evaluate Financial Reports Closely: Stop relying on stories and focus on real financial data, such as actual revenue and cash flow, to distinguish between genuine AI solutions and those that are merely superficial.
3. Use Regular Investments Instead of Trying to Buy at Lowest Points: Instead of trying to buy stocks at their lowest prices, use regular investments (e.g., investing a fixed amount each month). This approach helps reduce the impact of emotional fluctuations.
In Conclusion
The AI market has not ended, but the period of easy profits has passed. The focus now shifts to understanding the true value of companies. Those who make the right choices will benefit in the future; those who miss out may be left behind. Ordinary investors need to shift from chasing trends to doing their homework and making informed decisions.