To ride the AI wave in the long term, and without taking any single-name views, CIO likes semiconductors given their exposure to capex spending, and software due to increasing AI monetization. (UBS)
The president of South Korea—a major technology-exporting country—backtracked on imposing martial law on Wednesday just hours after his declaration. US President-elect Donald Trump’s threats of tariffs once he returns to the White House also remain a concern.
We expect tech volatility to pick up in the near term, especially in the event of disruption in the artificial intelligence (AI) supply chain. But we also think the near-term headwinds appear to be manageable at this stage, and that sound fundamentals should continue to support our positive outlook on the AI growth story.
The market impact of South Korea’s political crisis looks short-lived, with the AI supply chain more exposed to Taiwan. South Korean President Yoon Suk Yeol now faces a potential impeachment after he rescinded the martial law order under pressure from lawmakers. Emergency law declarations are not historically unusual in South Korea, and impeachments have happened before, but the market impact tends to be short-lived. Separately, the AI supply chain has disproportionately more exposure to Taiwan than South Korea despite the latter being home to key suppliers of HBM chips. We estimate that about 50-60% of big tech’s AI capex is focused on Taiwan, and about 5-10% on South Korea.
The latest tech restrictions are less severe than markets had feared, and China’s curbs on critical mineral exports should see limited impact. While South Korea did not win an exemption in HBM sales to China in the new round of export controls, the announced US restrictions overall appear less severe than previously expected. We also expect limited impact from China’s export ban of critical minerals, as AI-related chip manufacturing in the US predominantly uses more mainstream materials such as silicon and copper. Meanwhile, Trump’s tariff stance on tech imports remains to be seen, but we think several segments may secure relief or exemptions. Within semiconductors, we continue to favor AI logic chips, followed by quality foundries, memory companies, and then chip-manufacturing equipment makers.
AI demand remains robust amid improving monetization trends. Recent tech earnings showed that cloud revenue growth has continued to accelerate, with the focus in the software industry shifting away from margin pressure to improving AI monetization. We continue to see robust fundamentals for AI growth in the coming years.
So while we expect heightened volatility ahead, investors can consider utilizing structured strategies or buying the dips in quality global AI stocks. To ride the AI wave in the long term, and without taking any single-name views, we like semiconductors given their exposure to capex spending, and software due to increasing AI monetization. We remain cautious on traditional tech segments such as smartphones, PCs, and consumer electronics.
Main contributors: Solita Marcelli, Mark Haefele, Sundeep Gantori, Daisy Tseng, Jon Gordon
Original report: How to position in technology amid (geo)political uncertainty, 04 December 2024.
