NextFin

Goldman’s Bell Sees More Gains for Europe’s AI Buildout Stocks

Summarized by NextFin AI
  • Goldman Sachs strategist Sharon Bell believes Europe’s AI buildout stocks have significant growth potential, as the market has expanded beyond just chip manufacturers to include various suppliers essential for AI infrastructure.
  • Goldman Sachs Research predicts AI companies could invest over $500 billion by 2026, with European AI-linked shares experiencing increased investor interest due to this anticipated spending wave.
  • The European AI market comprises diverse beneficiaries, including chip equipment makers like ASML and electrical suppliers like Schneider Electric, indicating a broad investment opportunity beyond a single category.
  • Bell's optimism is based on the structural demand for AI infrastructure, suggesting that investments in power and connectivity systems will continue to grow, making the European market more resilient than the concentrated U.S. tech sector.

NextFin News - Goldman Sachs strategist Sharon Bell is still seeing room for Europe’s AI buildout stocks to run, and the case rests on a simple market truth: the trade has broadened from a narrow chip rally into a wider infrastructure bet on who supplies the machines, power, glass, cables and software that make artificial intelligence usable at scale. Goldman Sachs Research says AI companies may invest more than $500 billion in 2026, while European AI-linked shares have already been bid up as investors keep chasing the suppliers behind that spending wave.

The point is not that Europe has suddenly produced its own version of a U.S. megacap AI platform. It is that the region has become a secondary market for the buildout itself. That matters because Europe’s listed winners are not confined to one category. They include chip equipment makers such as ASML, specialty semiconductor names such as STMicroelectronics and BE Semiconductor, and electrical and grid-linked suppliers such as Schneider Electric and Prysmian. The market has been treating those groups as different ways to own the same spending cycle.

That is why Bell’s latest stance matters. Goldman’s European equities team has been arguing that the AI trade in Europe has breadth, not just momentum. Earlier this year, the bank said European equities could attract investors worried about concentration in a handful of U.S. technology stocks, and Bell’s call suggests that view has not run its course. In a market where a single U.S. earnings miss can knock the whole theme, Europe’s scatter of beneficiaries may be more durable than the narrow leadership still dominating U.S. AI exposure.

But the more important question is whether this is just another phase of cyclical enthusiasm or the start of a structural rerating. The answer is not obvious, because the first leg of the move has been cyclical: orders, backlog and capital spending on data centers can surge, then slow when customers pause. The second leg looks more structural: AI training and inference demand are forcing long-lived investments in power, interconnect, photonics, manufacturing equipment and electrical systems that cannot be unwound quickly once built. Bell’s optimism is about that second leg.

What Exactly Is Europe’s AI Trade?

The European AI basket is less a single trade than a chain of linked exposures. At one end are the chipmakers and equipment companies that benefit when advanced semiconductors move through the fab process. At the other are the industrial groups wiring up data centers, moving power, cooling racks and handling the physical scale-up that comes with model training. That chain is the reason the same AI theme has pulled in names that would once have been seen as ordinary industrial or electrical stocks.

Reuters reported in May that one TS Lombard basket of European semiconductor supply-chain names, including ASML, Infineon and STMicroelectronics, had rallied roughly 20% since the start of April, while a separate basket of AI infrastructure names including Schneider Electric and Prysmian had gained around 22% over the same period. The split is revealing. Investors are not just buying the obvious chip tools names; they are moving down the stack into the companies that make the buildout physically possible.

That is also why the Goldman note lands differently from a generic “tech is hot” call. It points to capital expenditure as the transmission mechanism. When hyperscalers and model developers commit to more server capacity, the first-order winners are the direct suppliers. The second-order winners are the firms that sell the electrical, thermal and connectivity systems needed to keep the data center running. The third-order effect is that European earnings and order books can improve even if Europe does not host the AI platforms themselves.

One reason the market has been receptive is that the buildup already has visible scale. Goldman Sachs Research says the consensus estimate among Wall Street analysts for hyperscaler AI companies’ 2026 capital spending is $527 billion, up from $465 billion at the start of the third-quarter earnings season. Once the spending base gets that large, the trade stops behaving like a sentiment story and starts behaving like an industrial cycle.

Why The Market Keeps Repricing The Same Theme

The easy read is that investors are simply chasing whatever has worked. That is partly true, but it is not enough. The stronger explanation is that the market has repeatedly discovered the same bottleneck: AI demand is not constrained by software ambition alone, but by physical infrastructure with long lead times and limited substitutability. That includes advanced lithography, semiconductor testing, networking, grid upgrades and electrical distribution. Each bottleneck pushes investors one step farther from the obvious U.S. platform names and toward the European suppliers that sit behind them.

Goldman’s own framing supports that shift. The bank said the average stock price correlation across large public AI hyperscalers fell from 80% in June to just 20% later in the period it discussed. That matters because it shows dispersion is replacing one-factor momentum. When correlations collapse, investors are no longer buying a single AI beta; they are picking winners and losers by supply-chain exposure, margin leverage and capex sensitivity. Europe’s market gives them more places to do that.

The cyclical case is still real. Short-cycle demand for data-center equipment can cool if customers delay orders, if financing costs stay high, or if investors decide the capex boom has outrun near-term monetization. Europe’s AI-linked stocks have already seen sharp runs, and history says crowded infrastructure trades can pull back when the market briefly decides the spending curve is too steep. The 20% and 22% moves in the spring are proof that the trade is not immune to momentum unwinds.

Yet the stronger case is structural. There are at least three historical comparisons worth making. First, the telecom-fiber buildout of the late 1990s showed that when network demand rises, the suppliers of picks and shovels can earn for longer than the end-users of the network boom can justify. Second, the cloud-buildout years of the 2010s showed that data-center and networking capacity often expands in waves that outlast the initial product cycle. Third, the power-grid and electrification trade in Europe since the energy shock showed that when the physical system becomes the constraint, industrial beneficiaries can rerate for reasons that are not immediately reversible.

That is the key distinction Bell is betting on. A cyclical story would require the spending wave to reverse because demand normalizes. A structural story says the economy has crossed into a regime where AI compute, storage and power are now foundational inputs, and the capex keeps compounding because the installed base itself becomes the demand driver. The latter is harder to fade.

“The combination of continued corporate AI adoption and growing concerns about the AI infrastructure complex has increased recent investor focus on the next beneficiaries of the ever-expanding AI trade.”

That line captures the portfolio-level logic. If U.S. AI leadership remains concentrated, Europe becomes the alternative source of AI beta with more breadth and less single-name dependence. If that happens, the European trade can keep widening even when the headline U.S. names trade sideways.

What Would Prove Bell Wrong?

The strongest counter-thesis is that the market is overpaying for a spending cycle whose payoff is still unproven. That argument is not fringe. It is the natural objection whenever capital intensity rises faster than monetization. If hyperscalers and other buyers slow their budgets, the suppliers’ earnings leverage can flip quickly. The same names that benefit from every new order can suffer if deployment pauses or if customers decide they have already funded enough capacity for the next phase.

There is also a valuation argument. After such a strong run, Europe’s AI-linked stocks do not need a bad fundamental outcome to fall; they may only need a better-than-feared normalization. Crowded trades can reprice even without a macro shock if investors stop rewarding duration in earnings and start rewarding cash today. In that sense, the bear case is not about AI disappearing. It is about the market deciding that the next euro of capex is worth less than the last one.

That said, the counter-thesis would be stronger if the spending was visibly peaking. The falsifying signal for Bell’s structural view would be a clear slowdown in 2026 AI capex commitments, combined with a sustained drop in data-center-related order growth and a renewed collapse in supply-chain share correlations back toward the 80% level Goldman cited earlier. If that happens, the market would be telling investors that the buildout has entered late-cycle fatigue rather than early-structural expansion.

For now, the opposite is what the numbers suggest. Goldman Sachs Research points to more than $500 billion of AI investment in 2026, and its consensus figure for hyperscaler capex is $527 billion, up from $465 billion at the start of earnings season. Reuters’ spring snapshots showed European semiconductor and infrastructure baskets already outpacing the broad market, and the move was broad enough to include both chip-adjacent and electrical names. That combination argues for a theme that is bigger than a single headline and more durable than a single quarter.

Short term, the trade remains vulnerable to positioning, earnings disappointments and any sign that customers are stretching out deployments. Medium term, the winners are the companies with real leverage to recurring buildout demand: the equipment makers, grid suppliers and connectivity vendors whose order books can keep expanding as long as the compute race continues. Long term, the question is whether Europe’s capital markets are now being rerouted into a structural AI supply chain, even if the region remains a secondary venue for the platform layer itself.

The base case is more gains as long as capex stays elevated and order visibility remains strong. The upside case is a broader rerating if investors keep rotating out of concentrated U.S. megacap exposure and into Europe’s wider pick-and-shovel set. The downside case is a quick unwind if 2026 spending plans flatten and supply-chain correlations snap back higher, which would tell the market the trade is back to being just another crowded momentum bet.

Bell’s point is not that Europe owns the AI revolution. It is that Europe may own enough of the plumbing to keep profiting while the revolution is still being built.

That is the trade: not the model, but the machinery.

Explore more exclusive insights at nextfin.ai.

Insights

What are the underlying principles driving the growth of AI buildout stocks in Europe?

What historical factors contributed to the current state of Europe's AI supply chain?

How has the market perception of European AI stocks shifted in recent years?

What trends are emerging in the European AI investment landscape as of 2023?

What recent developments have affected the capital spending forecasts for AI-related companies in Europe?

How might changes in U.S. AI investment impact European AI-linked stocks?

What are the long-term implications of Europe's positioning in the global AI supply chain?

What significant challenges do European AI buildout stocks face in the next few years?

What controversies exist surrounding the valuation of European AI stocks?

How do European AI stocks compare with their U.S. counterparts in terms of market performance?

What lessons can be learned from past infrastructure booms that may inform the current AI buildout?

What indicators would suggest a slowdown in Europe's AI capital expenditure growth?

How does the diversification of European AI investments benefit investors?

What role do supply-chain bottlenecks play in shaping the future of AI investments in Europe?

What potential risks could derail the growth of Europe’s AI infrastructure stocks?

How might evolving consumer demands impact Europe's AI buildout market?

What evidence supports the argument for a structural shift in Europe's AI investment landscape?

What factors contribute to the volatility of European AI-linked stocks in the market?

How can investors identify the best opportunities within Europe’s AI buildout sector?

Search, ask or analyze anything...