Several factors point to more moderate returns in the near term, according to Goldman Sachs Research. There’s a historical pattern of seasonal weakness ahead of midterm elections, and large, recent increases in AI capex estimates and associated earnings forecasts also create a high bar for sustaining the pace of upward revisions going forward. There are signs that economic activity is slowing.
Going forward, aggregate S&P 500 earnings growth will increasingly rely on hyperscale tech companies’ ability to produce large returns on their AI investments. “Generating sufficient returns on AI investment spending will ultimately require that enterprise end users enjoy sufficient productivity gains to justify spending on AI applications,” Snider writes.
Enterprise adoption of AI remains in early stages, but Goldman Sachs Research expects the impact on productivity and earnings to become increasingly visible in coming years. Goldman Sachs Research’s forecasts embed a 0.4 percentage point boost to S&P 500 earnings growth from AI-driven productivity this year and a 1.5 percentage point boost in 2027.
It will be important for investors to diversify beyond AI-infrastructure stocks, Snider suggests, given the backdrop of narrow market breadth and elevated return differences among individual stocks. An improvement in the geopolitical outlook would likely give a bigger boost to consumer-facing sectors that are more dependent on economic growth than to AI infrastructure stocks, according to Goldman Sachs Research.
