A sharp surge in artificial intelligence spending by America’s largest technology companies has propelled U.S. equities to fresh records. Still, Goldman Sachs analysts caution that a dramatic pullback in investment could wipe out much of those gains.
In a research note released this week, strategists led by Ryan Hammond highlighted how capital expenditures from hyperscalers such as Microsoft, Amazon, Alphabet, Meta, and Oracle have fueled revenue growth across the AI supply chain — from semiconductor giants like Nvidia to cloud infrastructure and hardware providers.
Record Spending, Record Valuations
Goldman estimates that hyperscaler capital spending has already reached $368 billion in 2025, surpassing earlier forecasts. These outlays, primarily for data centers, GPUs, and cloud capacity, have translated directly into real revenues for chipmakers and service providers, powering the boom in AI-linked stocks.
Register for Tekedia Mini-MBA edition 19 (Feb 9 – May 2, 2026): big discounts for early bird.
Tekedia AI in Business Masterclass opens registrations.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab: From Technical Design to Deployment (next edition begins Jan 24 2026).
The five largest companies in the S&P 500 — Nvidia, Microsoft, Apple, Alphabet, and Amazon — currently trade at a price-to-earnings multiple of 28, lofty by historical standards but still well below the peaks of 40 in 2021 and 50 during the 2000 dot-com bubble. Goldman stresses that, unlike past speculative manias, today’s valuations are supported by tangible revenue growth.
The Risk of a Spending Slowdown
Even so, analysts warn that the market is vulnerable to a sudden shift. Should hyperscaler investment revert to 2022 levels, AI hardware and services providers could lose out on 30 percent of the projected $1 trillion in S&P 500 sales growth expected in 2026. In such an “extreme scenario,” Goldman’s model suggests the S&P 500’s overall multiple could fall by 15 to 20 percent.
“The durability of this rally is tied directly to capex momentum,” Hammond and his team wrote, noting that investor enthusiasm is highly dependent on hyperscalers maintaining their aggressive pace of investment.
Echoes of Past Tech Booms — But With a Difference
Goldman stops short of labeling the current rally a bubble. Unlike the dot-com era, the revenues generated by AI demand are immediate and measurable. Nvidia, for example, has booked record-breaking quarterly sales from data center chips, while cloud providers have reported accelerating customer adoption of AI services.
The historical parallel, however, is instructive. In the early 2000s, telecom carriers poured hundreds of billions of dollars into network buildouts, only to abruptly cut spending when revenues failed to match projections — a collapse that left equipment suppliers with excess capacity and stock prices in freefall. Similarly, the cloud-computing boom of the 2010s saw an arms race in data center investment, though in that case, demand ultimately caught up and reshaped the enterprise software industry.
Today’s AI spending wave shares elements of both: a massive upfront infrastructure buildout with uncertain long-term demand curves, but also faster near-term monetization than earlier cycles.
Momentum Still Upward — For Now
Despite concerns about sustainability, the rally continues. AI-related stocks rose 32 percent in 2024 and have added another 17 percent so far in 2025. Analysts broadly expect spending growth to slow toward the end of 2025 or into 2026, but leading tech firms have consistently revised guidance higher, suggesting the inflection point may be further away than feared.
Currently, Wall Street remains split. Bulls argue that AI is a transformative technology still in the early innings, while skeptics warn that markets are pricing in perfection. Goldman’s message to investors lands somewhere in between: the upside is real, but so are the risks.