Goldman Sachs: Hyperscaler AI Spending Will Keep Supporting European Stocks

Hyperscalers — the largest tech companies building cloud and AI infrastructure — will keep increasing their capital spending, and that will continue to fuel earnings growth in other sectors and regions, including Europe. That's the view of Peter Oppenheimer, chief global equity strategist at Goldman Sachs, in comments to Bloomberg Television.

Who are the hyperscalers

The term "hyperscalers" usually refers to the companies that run the world's largest cloud and data center capacity — first and foremost Microsoft, Amazon, Google (Alphabet), and Meta. Over the past several years, these companies have steadily increased capital investment in building data centers, buying AI accelerators, and securing energy infrastructure to train and run large language models. Each company's quarterly report now routinely serves as a barometer for the entire AI industry, from utilization at chipmakers' fabs to electricity demand in regions where new data center campuses are going up.

Hyperscalers' AI infrastructure spending began surging back in 2023 amid the generative AI boom and has not reversed once since then, despite periodic investor worries about an overheated market and questions about whether these investments pay off as fast as the spending itself grows.

Why the effect reaches Europe

According to Oppenheimer, rising hyperscaler spending translates into earnings growth far beyond the tech companies themselves. Part of that effect lands in Europe: local equipment makers, engineering and energy companies, and suppliers of data center components and cooling systems are winning orders from the global wave of AI investment. Goldman Sachs sees this spillover as one of the factors supporting European stock indexes, alongside traditional growth drivers.

The European beneficiaries most often cited include the Netherlands' ASML — a near-monopoly supplier of lithography equipment for manufacturing advanced chips — as well as engineering, construction, and energy companies that supply power and cooling to new data center campuses across the continent. Their revenue and production capacity utilization depend directly on how quickly US hyperscalers keep expanding their AI infrastructure.

Energy consumption is a separate issue that comes up more and more in discussions of hyperscaler capex. Growing electricity demand from AI data centers is forcing energy companies and regulators to revise grid development plans years ahead, and in some regions to discuss extending the service life of existing generating capacity. That's why analysts count not only IT contractors and chipmakers among the beneficiaries of AI capex, but the entire energy sector — from grid operators to turbine suppliers.

Are there risks to this bet?

For several quarters now, analysts have debated whether the scale of AI infrastructure spending is justified by the revenue these investments have generated so far. Skeptics point to the risk of excess capacity and to cross-financing arrangements among cloud providers, chipmakers, and AI labs. Goldman Sachs' position, as voiced by Oppenheimer, is optimistic: as long as hyperscaler investment keeps translating into earnings growth in adjacent industries and regions, the market views this capex cycle as a positive factor rather than a bubble.

What it means

Goldman Sachs' assessment backs a thesis heard more and more often in the markets: hyperscaler capital spending on AI has become a macroeconomic factor that affects corporate earnings far beyond the US and the tech industry itself.