Goldman Sachs forecasts 27% jump in S&P 500 profits as AI spending carries the load

Goldman Sachs expects S&P 500 companies to report earnings-per-share growth of 27% year over year for the third quarter. The main engine, in the bank’s view, is corporate spending on artificial intelligence.

That would be a slowdown from the adjusted 33% growth logged in the second quarter. It would also be the third straight quarter above 25%.

What Goldman is projecting

The forecast comes from a report dated October 2, 2026, led by Goldman Sachs strategist Ben Snider. The headline figure measures earnings per share, meaning the profit each share of stock can claim.

Three consecutive quarters of growth above 25% is unusual. According to the research, streaks like this have mostly appeared when the economy was bouncing back from a crisis.

The fuel is spending on AI infrastructure, which Goldman projects will account for more than half of the index’s overall earnings growth.

The sector breakdown is lopsided. Information technology and energy together are expected to deliver nearly 80% of the S&P 500’s EPS expansion.

A very small group doing very heavy lifting

Goldman expects the top ten contributors to generate 68% of total EPS growth.

Within that group, chipmakers Nvidia and Micron are expected to account for more than one-third of the growth on their own.

The spending behind all this comes largely from hyperscalers, the giant cloud providers that rent out computing power at massive scale. Their capital expenditures are forecast to rise 116% year over year in the third quarter of 2026.

In the second quarter, hyperscaler capex growth stood at 87%.

The research also notes that AI-related spending has been responsible for nearly half of the index’s EPS growth in 2026 so far.

The bank’s bigger numbers

Goldman’s broader outlook stays upbeat. The bank projects full-year 2026 S&P 500 earnings per share of $375, which would be a 36% increase from 2025.

It has also set a 12-month target of 8,700 for the index.

The median S&P 500 company is expected to show roughly 9% EPS growth.

The risks in the fine print

The research flags several pressure points. The first is that the tailwind from AI capital spending may diminish over time.

The second risk is depreciation. When companies buy expensive equipment, they spread the cost across future years as an accounting expense, and that expense can weigh on reported profits later.

The third concern is semiconductor margins. Chipmakers have been enjoying unusually fat profits on AI hardware, and those margins could normalize.

Deceleration, not collapse

Despite those risks, Goldman’s base case is a slowdown rather than a downturn. Goldman argues that productivity gains from AI will gradually offset the fading boost from investment-led growth.

Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.