Goldman Sachs expects the boost that large technology companies' AI investment spending provides to the economy to reach a turning point in the fourth quarter of 2026, as the growth rate of such spending relative to U.S. GDP approaches its anticipated peak.
The forecast, led by chief economist Jan Hatzius, raises a core question: after the pace of infrastructure investment expansion slows, can AI-related companies' earnings growth support current valuations?
Goldman Sachs estimates that the year-over-year increment in hyperscaler investment as a share of U.S. GDP is expected to hit a high of approximately 1.5% in the fourth quarter of 2026.
It should be clarified that this forecast points to a peak in the growth slope of investment relative to the economy, and does not mean that total AI spending will decline.
The report cites estimates that leading cloud and technology companies including Amazon.com (AMZN.US), Microsoft (MSFT.US), Alphabet (GOOGL.US), Meta Platforms, Inc. (META.US) and Oracle (ORCL.US) will spend between $725 billion and $800 billion on capital expenditures in 2026, with funds continuously flowing into data centers, semiconductors and computing infrastructure.
This growth inflection point may reshape the contribution pattern of AI investment to economic growth.
Companies may still continue to increase investment, but the marginal economic boost from year-over-year expansion of new investment will gradually diminish.
Goldman Sachs also modeled how stock market gains could feed back into the investment cycle: under an optimistic scenario assuming AI sector stocks rise 55% and other stocks rise 20%, household wealth appreciation could drive consumption.
The above gains are only scenario analysis assumptions, not Goldman Sachs' baseline market forecast, and are intended to demonstrate the linkage between equity valuations, household wealth and consumer demand.
Now, tech giants are about to disclose their financial results, and the market will face a reality check.
Microsoft and Alphabet will release earnings on October 28 U.S. Eastern Time, with Amazon.com following on October 29.
The market will focus on capital expenditure guidance, cloud revenue growth, AI service demand and free cash flow to judge whether massive infrastructure investment can deliver sufficient financial returns.
Goldman Sachs has another estimate: hyperscalers need approximately $300 billion in AI-related revenue annually to cover AI investments and achieve break-even.
This figure highlights the extremely high revenue threshold the industry must meet to fulfill its infrastructure commitments.
Companies deeply tied to AI infrastructure and services, such as NVIDIA (NVDA.US), Microsoft and Amazon.com, are particularly affected by this investment outlook.
If corporate earnings fall short of expectations, a slowdown in capital expenditure growth could put pressure on currently elevated valuations.
To reiterate, Goldman Sachs' core judgment is that the year-over-year increment in hyperscaler investment relative to U.S. GDP will peak in the fourth quarter of 2026; it does not mean that total AI infrastructure spending will necessarily decline.
Comments