①Henry Wu, Chief Quantitative Strategist at Alpine Macro, stated that despite concerns over excessive valuations, rapid depreciation of chips, and threats from outside the U.S., these factors may all be overstated; ②Wu and his team expect artificial intelligence (AI) trading to overcome these concerns within the next year.
Overvaluation, the rapid depreciation of chips, and threats originating from outside the United States collectively form the basis for current concerns surrounding an artificial intelligence (AI) bubble. However, at present, several of the "loudest" reasons may have been exaggerated.
Henry Wu, Chief Quantitative Strategist at investment research firm Alpine Macro, recently stated that investors can temporarily set aside their worst fears about a potential setback in the AI boom. He pointed out that GPU depreciation, capital expenditure, and changes in “leadership” will not disrupt this feast for now, and investors should look beyond these concerns.
“Thus far, these risks are manageable and will not disrupt the development of artificial intelligence. We continue to see outstanding performance in the ‘bottleneck layer’ of the AI technology stack, which represents irreplaceable components of the supply chain with strong technological moats and extremely high market concentration,” he added.
Wu and his team focused on the following three key questions:
Is AI-related capital expenditure a risk for investors?
Will AI chips depreciate too quickly?
Is the market approaching another DeepSeek moment?
They believe the answer to all three questions is no. The strategist noted that AI-related capital expenditure has significantly increased over the past year. While this might raise some concerns about overinvestment, he argued it does not indicate that the market is nearing its peak.
“Large hyperscale companies like Google and NVIDIA are racing to establish industry standards and guide the direction of AI development. Investing in customers and suppliers ensures their technology roadmaps are more closely aligned with their ecosystems,” he added.
Wu also pointed out that, despite widespread concerns about the depreciation of AI chips, he believes such worries are exaggerated. He emphasized the growing demand both in the United States and overseas, particularly from China’s burgeoning AI market.
He noted that the utility of older AI chips will not diminish with the release of new chips, but prices will decline as supply increases. In his view, the increase in AI chip supply signifies faster-than-expected productivity growth, which investors should regard as a positive factor.
"Faster product cycles also mean accelerated productivity improvements. New chips will only bring significant enhancements," he said.
Wu's final point focused on threats from abroad, specifically mentioning DeepSeek, which disrupted the market in January. Since then, investors have been concerned that cheaper AI models might once again seize market share from U.S. investors, but Wu stated that this concern should be set aside for now.
"We believe these unexpected technological leaps are a positive factor for AI advancement, and the resulting reshuffling presents buying opportunities. These advancements will soon be imitated by competitors, thereby boosting productivity across the entire ecosystem," he added.
Finally, Wu concluded that his team expects AI-related transactions to overcome these concerns within the next year. They recommend that investors prioritize tangible investments in areas such as lithography and advanced logic manufacturing by 2026 to capitalize on AI investment opportunities.
Editor/Lambor
