Corporate profit strength in the second quarter has helped steady the stock market even as AI-related equities have seen sharp intraday and multi-day swings, according to a note from Goldman Sachs strategists.
"AI stock volatility during the past week has continued to follow the typical historical pattern following sharp Momentum rallies," strategists led by Ben Snider wrote, noting that the biggest momentum-driven rallies in recent decades have commonly been followed by consolidation periods like the recent drawdown. The team added that while earnings ultimately determine the market's direction, both historical patterns and recent investor deleveraging "suggest an improved outlook going forward."
On the earnings front, Goldman reported that 61% of S&P 500 companies had disclosed second-quarter results as of July 31, and an unusually high share - 64% - beat consensus earnings-per-share forecasts by at least one standard deviation. Excluding "other income" tied to appreciating equity investments held by mega-cap technology firms, S&P 500 EPS growth is tracking at 26% year-over-year; including those gains, the headline growth rate rises to 45%.
The median S&P 500 company is pacing to expand EPS by 12% in the quarter versus the 9% consensus estimate that prevailed heading into the reporting season. Goldman flagged that roughly one-third of that upside is attributable to AI infrastructure stocks.
Despite the robust underlying results, market responses have been uneven. Goldman strategists observed that the market's reaction to earnings beats in the Technology, Media, and Telecommunications (TMT) cohort has been muted even as the equal-weight S&P 500 has continued to climb alongside steady EPS growth. More specifically, within TMT the median stock that beat EPS lagged the S&P 500 by 192 basis points on the trading day following its report, while the median stock in other sectors that beat EPS outperformed the S&P 500 by 75 basis points.
Company-level outperformance in other sectors has helped lift forward-looking estimates. Since the start of the third quarter, the bottom-up consensus for S&P 500 2027 EPS has been revised upward by 1%, with positive revisions across most sectors, Goldman said. The strategists cautioned, however, that rising input costs remain a key risk to margins.
The note also described current dynamics in the AI investment cycle. Reports from hyperscale cloud providers signaled increasing capital expenditures and a growing need for external financing, while also showing early signs that AI spending is beginning to generate returns. For the quarter, that group recorded $182 billion of capex, just $5 billion of free cash flow, and $101 billion of debt and equity issuance.
Analyst projections now point toward more than $1 trillion in capex in 2027 - more than $100 billion above earlier estimates - and expect capex to exceed cash flow from operations from 2026 through 2028. Cloud revenue growth accelerated to 48% in the quarter, outpacing consensus expectations.
In sum, Goldman strategists portray a market where substantial earnings beats and upgraded profit forecasts are helping to offset volatility in AI-related stocks, even as heavy capex plans and margin pressures create clear areas of uncertainty.