Artificial intelligence is set to dominate another earnings season on the American stock exchanges. This is the expectation of Goldman Sachs , which projects a 22% increase in aggregate earnings per share for the S&P 500 in the second quarter of 2026. The bank estimates that AI infrastructure companies will account for almost 60% of this profit expansion.
The concentration is striking. Just two companies, Nvidia and Micron Technology, are expected to account for about 40% of total growth, although they represent only 8% and 1% of the index, respectively.
Fueled by the demand for artificial intelligence infrastructure , both companies are coming off record results. In the last quarter, Nvidia reported record revenue of US$81.6 billion, an 85% increase compared to the same period in 2025. Micron, meanwhile, tripled its revenue year-over-year to US$23.86 billion, breaking records in all metrics — revenue, margin, earnings per share, and free cash flow.
Artificial intelligence is precisely what has maintained the bank's confidence in the performance of the American stock markets. In a report, Goldman revised its earnings forecast for the S&P 500 upwards to a 24% increase over the previous year — and raised its year-end target for the index from 7,600 to 8,000 points.
The revision came after the first-quarter earnings season led the market to raise its capital expenditure (capex) estimates for major technology companies — Amazon, Google, Meta, Microsoft, and Oracle — to $754 billion in 2026, 83% higher than last year, with projections reaching $905 billion in 2027.
According to Goldman Sachs, there is still room for surprises, since, over the past three years, market estimates have systematically underestimated the investment appetite of these companies. "The order books of hyperscalers point to a continuing imbalance between supply and demand."
On the other hand, Goldman himself recently warned that the same wave of investment that is boosting S&P 500 profits is also beginning to erode the financial returns of the largest technology companies.
In the bank's view, the sustainability of this profit cycle will increasingly depend on the ability of hyperscalers to transform investment into concrete returns, justifying the expenditure on AI applications.
For now, according to Goldman Sachs, corporate adoption is still in its early stages. The bank projects a contribution of only 0.4 percentage points to S&P 500 EPS growth from AI productivity gains in 2026 — a number that should rise to 1.5 points in 2027.
