For most investors, the next S&P 500 earnings season will probably feel a lot like a test of whether a handful of companies can continue carrying the story.
Goldman Sachs expects another strong quarter, but its forecast suggests the benefits of the AI spending surge are still heavily concentrated.
Nevertheless, shareholders could see significantly different results depending on who owns the companies that build, supply, or directly monetize AI infrastructure.
It’s important to note that the broader market isn’t moving in lockstep. Some companies are still facing sluggish growth, tighter margins, and higher costs, even as AI-linked businesses continue to pull the index higher.
That creates a more complicated setup than a simple “strong earnings” headline suggests.
The real question now is whether AI spending can begin to create a broader earnings lift across the market, or if the next season will reinforce how dependent the S&P 500 has become on a relatively narrow group of winners.
Goldman Sachs expects another strong earnings season, with AI doing more of the work
Goldman Sachs expects the S&P 500 to deliver another robust quarter of bottom-line growth, with AI investment accounting for an increasingly large share of the gains.
Consensus forecasts call for Q3 S&P 500 earnings per share to rise 27% from a year earlier, according to an Oct. 2 report led by Goldman strategist Ben Snider, as reported by Seeking Alpha.
That would still be a very strong result, even if it marks a slowdown from the roughly 33% growth recorded in Q2, after accounting distortions are stripped out.
Goldman also expects most companies to beat consensus estimates again.
AI infrastructure companies are expected to account for more than 50% of the S&P 500’s earnings growth this quarter, while information technology and energy together are projected to generate nearly 80% of the total increase.
AI is driving the index, but the average company tells a different story
Goldman’s headline forecast looks strong, but perhaps the more revealing number is 68%. That’s how much of the S&P 500 earnings growth the firm expects in Q3 from just 10 companies to generate.
Micron (MU) and Nvidia (NVDA) alone are projected to account for more than one-third. Micron has already set the tone, delivering year-over-year earnings growth of 1,003% and beating expectations.
Information technology and energy together are expected to produce nearly 80% of total index earnings growth.
More Wall Street:
- Wall Street’s AI trade faces its biggest valuation test
- The next Wall Street shift is already underway
- Wall Street sends strong 4-word verdict on the stock market
That concentration changes how I would read the season.
The S&P 500 may deliver 27% EPS growth even while the median company grows earnings just 9%, with revenue growth slowing and median net margins slipping to 14.7% from 15.1%.
Independent data strengthens Goldman’s broader optimism without clipping away at that divide. FactSet currently sees S&P 500 earnings rising 29.5%, up from a 26.7% estimate at the start of the quarter.
Even more unusually, analysts actually raised quarterly earnings estimates by 1.4%; historically, they tend to cut them. Seventy-two companies have also issued positive EPS guidance, far above the five-year average of 42 as reported by FactSet.
So the question now is largely about who is producing those profits. With the median S&P 500 stock sitting well behind its record while the index remains near its high, another AI-heavy earnings beat might reinforce rather than resolve the market’s unusually narrow leadership.
AI spending is no longer the story: returns are
Naturally, for investors, the big question is whether AI spending can produce enough sales, bottom-line expansion, and strong productivity gains to justify the next investment leg.
Goldman Sachs expects hyperscaler capital spending to jump to 116% year over year in Q3, after already climbing 87% in the second. The firm also thinks spending could grow more than 50% again in 2027 and ultimately exceed the roughly $1.1 trillion currently embedded in consensus estimates.
That’s an enormous capital commitment, which raises the bar for what investors should demand in return.
The encouraging sign is cloud revenue.
Goldman expects combined growth at Amazon (AMZN), Alphabet (GOOGL), Microsoft (MSFT), and Oracle (ORCL) to accelerate to 55% from 48%. If that high bar is achieved, it would show that at least part of the spending boom is translating into real business demand.
The weaker point is productivity.
About half of S&P 500 companies discussed AI in connection with efficiency last quarter, yet only 2% earnings impact. That gap between enthusiasm and proof is still too wide.
There is also less help from accounting this quarter. Goldman does not expect a repeat of the roughly $150 billion in Q2 mega-cap technology gains tied to rising investment values.
If revenue, margins, and productivity keep improving, the market can justify more investment. If they do not, the same spending that powered earnings growth could become the source of the next valuation problem.
It’s important to note that the S&P 500 is not especially cheap, but its valuation has become less stretched as earnings estimates have risen.
FactSet puts the index at about 19 times forward 12-month earnings, below its five-year average of 19.8 times and just under its 10-year average of 19.1 times.
Related: Nvidia’s AI boom revives a Warren Buffett warning for investors