Goldman Sachs: It's too early to call AI trading a bubble; investors shouldn't exit prematurely.

According to BlockBeats, on July 2nd, Goldman Sachs believes investors should not prematurely withdraw from the artificial intelligence (AI) theme. Despite the significant gains in AI-related stocks this year and rising concerns about a bubble, Ben Snider, Head of US Equity Strategy at Goldman Sachs, stated that the current market is more like an earnings-driven bull market than a speculative frenzy supported solely by valuation expansion. Snider believes AI investment opportunities remain concentrated in three main areas: AI infrastructure, power infrastructure, and recently underperforming cloud vendors and hyperscale platforms. The first includes semiconductors, servers, AI networking equipment, and data center hardware; the second includes power equipment, utilities, and energy infrastructure supporting data center expansion; and the third includes large platform companies such as Amazon, Microsoft, Meta, Alphabet, Oracle, and IBM. Goldman Sachs' core argument is that while many AI infrastructure stocks have surged, their valuations have not deviated from earnings expectations as typically seen in bubbles. Some storage, semiconductor, and data center-related stocks still reflect market skepticism rather than unanimous optimism. In other words, investors are still demanding that these companies demonstrate through orders, profit margins, and cash flow that AI capital expenditures can translate into profits. Goldman Sachs is not advocating indiscriminate buying of all AI concept stocks, but rather continuing to bet on companies that can directly benefit from AI capital expenditures and generate revenue and profit growth. They believe that AI trading has not yet entered a phase of "just telling stories, not looking at profits." As long as earnings continue to be revised upwards, AI infrastructure is likely to remain one of the most important themes in the US stock market.