Goldman Sachs Strategist: Under AI Disruption, the Light Asset Aura Fades, with Capacity and Infrastructure Becoming Hard Currency

Goldman Sachs strategists say that as investors seek to avoid the disruptive risks brought by artificial intelligence (AI), stocks of companies with substantial physical production assets are performing well.

The Goldman team states that since early 2025, a basket of “capital-intensive” stocks (whose economic value mainly comes from physical assets) has outperformed “light asset” companies relying on human or digital capital by about 35%.

The report notes that investors are increasingly turning to stocks with what strategists call the “halo effect,” which have large asset bases and low obsolescence risk, mainly in sectors like utilities, basic resources, and energy.

Goldman strategist Guillaume Jaisson said, “The market is rewarding capacity, networks, infrastructure, and engineering complexity—these assets are costly to replicate and less likely to be replaced by technological advances.”

Anxiety over AI potentially disrupting business models has swept through industries from software to asset management, causing previously “sure-win” star stocks to plummet. This concern has even triggered indiscriminate selling, affecting industries like logistics that seem less directly related to AI risks.

In the short term, Hamza Lemssouguer, founder of hedge fund Arini, also said that market fears of disruption from AI are enough to raise borrowing costs for software companies, causing trouble for this heavily indebted industry.

Lemssouguer stated, “We don’t even need to wait for disruption to actually happen; problems are already emerging. The market always gets ahead. The immediate issue we’re seeing now is rising capital costs for many companies, which will eventually lead to significant defaults, shocks, and dislocations in the credit market.”

Goldman strategists also pointed out that the race for AI leadership is pushing market winners once known for light assets—the “hyperscalers” of cloud computing—toward a capital-intensive model.

It is estimated that Amazon, Microsoft, Alphabet, Meta, and Oracle will invest about $1.5 trillion in building AI infrastructure from 2023 to 2026, compared to their total prior investments of around $600 billion before 2022.

The Goldman team states that higher real yields and geopolitical-driven fiscal spending and manufacturing investments also support capital flows into capital-intensive industries.

They also noted that earnings momentum is shifting toward these companies—currently, market expectations for earnings growth and return on equity for capital-intensive firms are overall higher than for light asset companies.

Morgan Stanley strategists also pointed out that funds are gradually withdrawing from light asset industries like software. They wrote that by the end of 2025, pure long-only funds in Europe have begun reducing allocations to stocks vulnerable to AI disruption.

(Source: Cailian Press)