【Major Bank Views】Barclays Fang Weizhang: AI Large Models and Product Iterations Rapid, No True Winners Yet

OpenClaw’s popularity triggered a nationwide “lobster farming craze,” with Chinese tech companies quickly following suit. Baring Hong Kong and China stock team leader Fang Weichang pointed out that the market was initially worried that AI would not only disrupt the entire internet but also replace existing systems. However, now tech companies are “fighting back” by launching their own AI products to enhance user experience and revenue. It is advisable to continue observing new products and AI applications.

But as AI develops rapidly and continuously introduces new features, product lifespans are becoming shorter, posing risks that need to be carefully managed. It is also expected that the hype around lobster concept stocks will be short-lived.

Currently, AI agents are called “lobsters.” Fang Weichang believes that there are no clear winners yet in large models and AI agents. He mentioned that pure large model companies are different from internet companies. The entry barrier into the large model field is not high, but excelling in this area is quite challenging. Additionally, the market’s demand for updates is high, meaning a large model can be replaced within a few months.

In terms of investment, he focuses not only on long-term value but also considers valuation and whether the company has annual growth. If there is room for valuation, he considers adding positions. Currently, he prefers hardware and raw materials.

However, he also warns that due to the unpredictable Middle East situation, the Hong Kong stock market’s direction remains unclear. Since the market is not driven by fundamental factors, any news affecting tech stocks could lead to amplified volatility.

Moreover, Fang Weichang mentioned that his funds have recently increased holdings in consumer stocks, as he believes that mainland China’s consumption adjustments have been ongoing for some time, and some companies are beginning to stabilize.