AI boom: Which companies benefit more?
```
With the dust settled after the third quarter earnings season, Goldman Sachs released a report stating that investment in AI infrastructure has not slowed down, but is instead accelerating upward.
Although investors have concerns about whether the credit market can absorb this investment frenzy and whether spending exceeds free cash flow, data shows that tech giants known as Hyperscalers still have enormous room for increased leverage on their balance sheets.
For investors, a key signal has emerged: AI transactions are shifting from pure infrastructure building to a stage with more marked differentiation in returns. Current capital expenditure estimates may still be underestimated, with potential upside of up to $200 billion.
At the same time, the investment logic is gradually shifting from “pick-and-shovel” players (infrastructure) towards AI platform stocks, as well as “productivity beneficiaries” that can significantly boost efficiency through AI.
Capital Expenditure Estimates Revised Sharply Upward, Return Differentiation Intensifies
The third quarter earnings season has not only catalyzed further increases in AI capital expenditure estimates, but also reignited investors’ concerns about risks lurking behind the AI investment boom.
The Goldman Sachs portfolio strategy team pointed out that market consensus for the five major AI Hyperscalers (Amazon AMZN, Google GOOGL, Meta, Microsoft MSFT, and Oracle ORCL) predicts 2026 capital expenditures rising from $467 billion at the start of earnings season (up 20% year-over-year) to $533 billion now (up 34% year-over-year).
Although “AI transactions” are currently focused mainly on infrastructure, returns within this sector are becoming more dispersed. This differentiation is driven by two main factors: first, the level of investor confidence in AI investment actually turning into revenue; second, the scale of leverage used to fund these investments. This split may provide investors with the chance to capture AI-driven returns at a more “favorable” price.
Capital Expenditure in 2026 May Have $200 Billion Upside
Although current spending figures are already eye-catching, analysts’ forecasts may still be too conservative.
Bottom-up market consensus expects annual capital expenditure growth for Hyperscalers to slow from the recent 76% to 25% by the end of 2026. This means analysts expect a significant slowdown in growth of these giants’ capital spending in 2026.
However, Goldman Sachs reviewed data from the past two years and found analyst forecasts at every stage have been not only conservative, but “overly” conservative. Based on historical analysis of tech investment cycles, Goldman Sachs believes that the current 2026 capital expenditure estimates for Hyperscalers may still have $200 billion in upward potential.
Markets are generally concerned that cash flow and balance sheet capacity will limit 2026 spending, but data disproves this view.
The vast majority of Hyperscalers’ capital expenditure so far has been funded through cash flow, but these companies also possess massive debt financing capacity. Since 2021, the combined net debt on these tech giants’ balance sheets has increased by $295 billion, but due to robust profitability, their collective net debt/EBITDA leverage is just +0.2x.
According to Goldman Sachs calculations, these five companies could add another $700 billion in net debt to their balance sheets, and even then, their net leverage would not exceed 1x. Therefore, compared to cash flow or balance sheet capacity, supply chain bottlenecks or investor appetite are more likely to be the near-term constraints on capital spending.
Leverage Pressure and the Feedback Loop
While the largest AI infrastructure companies (the aforementioned Hyperscalers) have strong balance sheets, many other listed and private companies involved in AI infrastructure face much tougher challenges.
The recent rapid growth in balance sheet debt and alternative financing has already triggered investor anxiety. Because the largest US-listed companies and smaller AI firms are linked by revenue and equity ties, this close relationship creates a feedback loop. This means that pressure in one part of the AI ecosystem (especially among smaller private firms) can easily transmit risk and impact investors throughout the entire AI sector. In fact, in the Q3 earnings season, multiple Hyperscalers reported that changes in private equity investment valuations had a major impact on their earnings.
Next-stage Winners: Platform Stocks and Productivity Beneficiaries
As enterprise adoption of AI continues to rise and concerns about the pure infrastructure sector grow, investors’ focus is shifting. Goldman Sachs highlights the following two categories of future beneficiaries:
AI platform stocks: As more companies actually adopt AI technology, these firms will directly benefit from revenue tailwinds.AI productivity beneficiaries: Goldman Sachs has screened a batch of companies with high labor costs and wage exposure, which have explicitly mentioned using AI automation to boost efficiency in recent earnings calls.
As the report soberly points out at the end, for Wall Street, these names may mean good news; but the implied “replacement” opportunities for jobs are definitely not good news for the general public.
Risk warning and disclaimerThe market is risky, and investment requires caution. This article does not constitute personal investment advice, nor does it take into account individual users’ specific investment objectives, financial situations, or needs. Users should consider whether any opinions, views, or conclusions in this article are fitting for their circumstances. Invest accordingly at your own risk. ```