U.S. earnings are entering the third-quarter reporting season with profit growth expected to remain strong on paper but increasingly dependent on a small group of AI-linked companies, according to forecasts from BofA Securities and Goldman Sachs. Wall Street expects S&P 500 earnings per share to rise 27% from a year earlier, while BofA forecasts 32% growth, or $93.50 per share.
Early results have supported those expectations. Eighty-five percent of companies that had reported had exceeded earnings-per-share estimates, above the 71% historical average. Yet the market’s response has been restrained, suggesting traders have already priced in a high level of positive results and are looking more closely at margins, spending plans and the durability of AI-driven demand.
Earnings beats are drawing a smaller market reward
BofA’s review of the second quarter offers a guide to the current environment. Seventy-seven percent of S&P 500 companies beat EPS estimates in that period, the highest share since 2021. Companies that exceeded both earnings and revenue expectations outperformed the index by an average of 0.7 percentage points on the following day.
That remains a positive reaction, but it is only half the 1.4-percentage-point average seen historically after a double beat. The penalty for disappointing results has been sharper: companies missing both sales and earnings estimates lagged the S&P 500 by 3.3 percentage points the next day, modestly worse than the typical decline.
The uneven reaction places a premium on forecasts rather than headline earnings. Companies can clear analyst estimates and still face selling pressure if they signal slower revenue growth, lower margins or rising costs. That dynamic is especially relevant for large technology companies whose valuations are tied to assumptions about AI infrastructure spending extending well beyond a single quarter.
Micron and Nvidia dominate projected profit growth
Goldman Sachs estimates that the 10 largest contributors to S&P 500 earnings growth will generate 68% of the index’s total increase in the third quarter. That is a sharp rise from 47% in the first quarter and 48% in the second.
Micron Technology alone is expected to account for 19% of S&P 500 earnings growth, followed by Nvidia at 15%. The two semiconductor companies would therefore represent roughly one-third of the index’s aggregate profit expansion. Meta is projected to contribute 7%, Alphabet 6%, and Broadcom 5%.
The concentration leaves a much less impressive picture beneath the largest AI beneficiaries. Consensus forecasts compiled by Goldman Sachs point to median third-quarter EPS growth of 9% for the typical S&P 500 company, down from 14% in the second quarter. Analysts have also lowered median profit-margin forecasts by 11 basis points since the quarter began, with every sector except information technology facing reductions.
Market breadth reflects that divide. Goldman Sachs said the S&P 500 was trading about 2% below its record high, while the median index stock remained 17% below its own peak. Its breadth measure had fallen to the lowest level since the dot-com period, indicating that index performance has depended heavily on a relatively narrow set of mega-cap names.
AI spending is lifting earnings while weighing on cash flow
The same companies helping drive profits are committing increasingly large sums to data centers, chips and networking equipment. Microsoft, Amazon, Google, Meta and Oracle are projected to increase capital expenditure by more than 90% year over year in 2026, according to the forecasts cited by BofA.
Estimates for the group’s 2026 capital spending have risen 70% over the past year. Capital expenditure is now expected to exceed 100% of their operating cash flow in 2026, a threshold that would widen the gap between reported net income and free cash flow.
That distinction is becoming central to how traders assess the AI buildout. Net income captures the current profitability of cloud and advertising businesses, while free cash flow shows how much cash remains after investment in physical infrastructure. A company can report strong earnings while directing most of its available cash into servers, chips and data centers.
The spending cycle supports semiconductor demand and has boosted earnings expectations for suppliers such as Nvidia, Micron and Broadcom. It also raises the pressure on hyperscalers to show that enterprise and consumer demand for AI products can eventually produce returns commensurate with the investment.
One-off gains add to concerns over earnings quality
BofA also identified an unusually large effect from private-investment revaluations in second-quarter index results. Reported S&P 500 EPS growth reached 50% year over year, but the figure falls to 32% when private investment valuation gains at Amazon and Google are excluded.
The adjustment does not erase the strength in corporate earnings, but it narrows the gap between reported results and underlying operating performance. It also reinforces why third-quarter guidance may carry more weight than headline EPS figures, particularly for the biggest technology platforms.
Costs are building outside technology
Raw-material costs have become a more frequent topic on S&P 500 earnings calls, reaching their highest level since mid-2022, according to BofA. U.S. retail gasoline prices rose about 30% year over year during the third quarter, creating direct pressure for transportation, retailers and industrial companies.
Labor remains the larger cost category. BofA estimates labor represents about 40% of operating costs across U.S. corporations, compared with less than 5% for energy. Wage inflation has cooled, which provides some relief for margins, but sustained energy costs could weaken consumer demand as companies pass through part of the increase.
Looking ahead, consensus forecasts call for S&P 500 EPS growth of 16% in 2027, following an expected 33% increase in 2026. BofA projects a lower 12% increase for 2027. Those estimates assume limited margin pressure outside energy, leaving little room for a prolonged rise in input costs or a material slowdown in spending.
For cryptocurrency markets, the earnings season offers a concentrated test of risk appetite rather than a direct read on digital-asset fundamentals. Results from the largest AI-linked companies, their capital-expenditure plans and signs of stress in consumer-facing sectors could shape broader trading conditions across technology equities and other higher-volatility assets.
Want to trade around concentrated AI-driven earnings? Explore Toobit’s latest market insights and strategies to sharpen your positioning.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
