Goldman Sachs Strategy Head Says AI Spending Accounts for Half of S&P 500 Profit Growth

Capital expenditures directed toward artificial intelligence are now generating roughly half of all earnings growth across the S&P 500, according to analysis from Goldman Sachs first reported by Yahoo Finance. Speaking on CNBC's Squawk on the Street, Goldman chief U.S. equity strategist Ben Snider noted that while the metric underscores the strength of the technology expansion, it also reveals that overall index profit acceleration remains heavily dependent on capital spending by a concentrated group of tech buyers.

The index has maintained momentum despite recent market fluctuations, with the SPDR S&P 500 ETF Trust (SPY) closing at 757.94 on Sept. 10 to mark an 11.13% gain for the year. Snider observed that the S&P 500 began the year trading at 22 times earnings and has since compressed to a multiple of 19 times earnings. Because overall corporate profits grew at a faster pace than share prices, equity valuations have become more grounded in realized corporate earnings rather than speculative multiple expansion.

The shifting valuation landscape comes alongside rising fixed-income yields. One year ago, the ten-year Treasury yield hovered near 4% as the S&P 500 traded around the 6600 mark. Today, the ten-year yield stands at 4.83%, placing it in the 99.6th percentile of its trailing one-year range. Snider argued that macroeconomic yield shifts present less immediate risk to equities than market volatility, provided baseline corporate profit growth maintains its trajectory.

The suppliers at the heart of the hardware cycle continue to post steep revenue gains. Nvidia reported second-quarter fiscal 2027 revenue of $96.22 billion, a 105.8% increase year over year, supported by $89.02 billion in data center sales, according to filings with the Securities and Exchange Commission. Concurrently, Broadcom reported third-quarter AI semiconductor revenue of $16.70 billion—up 221% from the prior-year period—and projected fourth-quarter sales of $21.7 billion.

On the buyer side, major technology companies are pouring hundreds of billions into data center infrastructure. Microsoft logged full-year capital expenditures of $115.95 billion and anticipates spending approximately $175 billion in capital expenditures for fiscal 2027. To keep buildout momentum strong across the ecosystem, Nvidia has extended take-or-pay commitments and selective credit enhancements to neocloud operators and AI research labs. The chipmaker has also linked up with institutional managers including Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to catalyze more than $500 billion in private infrastructure financing.

Snider highlighted that enterprise AI adoption has begun broadening beyond foundational computing infrastructure over recent months, after four years dominated primarily by data center construction. Software demand metrics support that observation: Microsoft reported that Microsoft 365 Copilot expanded beyond 30 million paid user seats, while commercial remaining performance obligations increased 84% to reach $678 billion.

Despite major equity indexes trading near historical highs, institutional investor exposure remains surprisingly subdued. Goldman's proprietary positioning indicator—measuring allocations across mutual funds, hedge funds, and other institutional accounts—has dropped to its lowest point since March. While light institutional positioning can cushion equity markets against panic selling during pullbacks, broader headwinds persist, including historical September weakness, a Cboe Volatility Index (VIX) reading of 16.46, the Federal Reserve's current 3.75% benchmark rate, and forthcoming inflation releases.

Because capital budgets are discretionary and subject to rapid adjustment, any spending slowdown by major hyperscalers could quickly reduce broader index earnings expansion. While enterprise adoption is gaining traction, most S&P 500 firms have not yet factored direct AI financial gains into reported earnings figures. Consequently, market performance over the near term will remain closely tied to the capital deployment schedules of a handful of key industry players.