
The semiconductors once again captured the attention of the markets this week, after the enthusiasm for new artificial intelligence developments skyrocketed the sector's quotations. The episode reminds us of how much the AI narrative continues to set the pace of global equity, far beyond the major American tech companies that until now concentrated much of the investor attention.
Behind the headline there are underlying questions: what role does each region play in the value chain of chips, whether the AI infrastructure investment cycle still has a way to go, and how to interpret the growing disparity of returns among companies. Below, four asset management firms offer complementary perspectives on these key issues for the professional investor.
The immediate trigger was the enthusiasm generated by the new artificial intelligence agent presented by Meta, which rekindled investor appetite for the entire semiconductor value chain. The movement is part of an AI infrastructure investment cycle that several managers still consider far from exhausted, although opinions differ on which regions and companies are best positioned to capture it.
The momentum itself illustrates how a single announcement can move an entire sector sensitive to spending on technological infrastructure. For investors, the challenge is not to identify the next short-term catalyst, but to understand which geographies and links in the value chain concentrate the structural growth of AI.
Yan Taw Boon, director of thematic strategies for Asia at Neuberger, points out that it is increasingly difficult to avoid the big tech companies given their weight in global indices, but argues that investors do not need to concentrate their exposure in a handful of American companies to participate in digitalization and artificial intelligence trends. In his opinion, the rise of AI infrastructures extends to sectors such as energy, construction, and industrial automation.
Boon also highlights the appeal of the "golden triangle" of semiconductors formed by Taiwan, Japan, and South Korea, where numerous medium-sized companies supply essential materials and components. In his opinion, diversifying between geographies and capitalizations allows participation in structural growth without the concentration risk of betting only on large tech companies.
Joran Mambir, product specialist at J. Safra Sarasin Sustainable AM, explains that the European strategy focuses on manufacturing in mature nodes, reflecting the region's industrial and automotive strengths rather than an ambition to compete in cutting-edge logic or memory. He recalls that Intel's project in Magdeburg, intended for advanced chips, was canceled in July 2025, leaving Europe without active production in leading-edge nodes.
However, Mambir emphasizes that Europe holds key positions in critical equipment: ASML maintains the monopoly of EUV lithography, ASM International leads atomic layer deposition, and Besi stands out in hybrid bonding. This is complemented by European leadership in analog and power semiconductors, with STMicroelectronics, Infineon, and NXP.
Therefore, he concludes that Europe's future does not lie in catching up with Taiwan in advanced nodes, but in strengthening its differential advantages.
Alex Stauffacher, equity analyst at Conviction Equities (a Vontobel boutique), along with Florence Kuang, Head of Client Portfolio Management, place the June 2026 sales wave in a known pattern: doubts about the return on AI investment, fears of an end of cycle, and Chinese competition. In their opinion, the market overestimates these risks, as investment and moderation of memory prices do not point to an oversupply.
Stauffacher and Kuang argue that the AI cycle is in an early phase compared to previous episodes, such as smartphones or the cloud, which took more than eight years to slow their growth below 20%. They point out that the capex of the hyperscalers could be around 745,000 million dollars this year, increasingly financed with debt, which they consider a moderate warning sign.
However, they warn that the supply shortage could take time to resolve, as much of the investments from TSMC, Samsung or SK Hynix will materialize beyond 2030. Therefore, they conclude that current valuations, although high, are still far from the dotcom bubble.
Alex King, investment strategy analyst, and Joshua Riefler, product reporting lead, both from Wellington Management, observe that, despite the apparent calm of the indices, volatility at the stock level has increased and correlations between companies have fallen to unusually low levels. In their opinion, concentration and exposure to individual stocks can weigh more than index volatility.
King and Riefler highlight that greater dispersion broadens opportunities for active investors, as corporate fundamentals gain weight. In addition, they point out that greater diversity in stock market leadership reduces dependence on a few stocks.
The four perspectives point to the same idea: the investment cycle in semiconductors and AI retains its course, but requires nuances. It is no longer enough to follow a handful of large American tech companies; the opportunity is spread among geographies, links in the value chain and companies with different profiles. The challenge is to combine structural conviction with discipline in selection, monitoring supply, demand and dispersion of returns.
Disclaimer: This article is for informational purposes only and does not constitute investment advice or a recommendation to buy or sell securities. Past performance does not guarantee future results. Investing in technology equities and semiconductors carries specific risks, including high volatility, sector concentration, and sensitivity to artificial intelligence infrastructure spending cycles. The comments collected come from J. Safra Sarasin Sustainable AM, Neuberger Berman, Conviction Equities (a boutique of Vontobel) and Wellington Management, and reflect the opinion of their authors at the time of writing, not necessarily that of RankiaPro.