"AI Rollercoaster" a Prelude to a Crash? Goldman Pushes Back: Earnings Are King—Look at Q2!

 

Ben Snider, Chief U.S. Equity Strategist at Goldman Sachs Research, argued in a recent note that investors should not mistake recent volatility in artificial intelligence equities for the onset of a broader market collapse.

 

Snider noted that while momentum-driven tech shares experienced violent swings, corporate earnings remain exceptionally strong, forward estimates continue to be revised upward, and evidence is mounting that massive AI investments are beginning to pay off.

 

For investors, Goldman contends, the key question has shifted from whether AI spending is excessive to whether earnings power can continue to support these outlays.

 

Goldman observed that the recent sell-off in AI-related stocks follows an unexpected yet familiar pattern.

 

In the note, Snider wrote that while the pullback was sharp, the heightened volatility mirrors past episodes of extreme momentum concentration.

 

Historically, such periods are followed by a phase of consolidation before the market resumes its primary uptrend.

 

The bank also highlighted that hedge funds and ETF investors have significantly reduced leverage, a de-risking move that should help dampen near-term market volatility.

 

Overall, Goldman believes market fundamentals remain vastly healthier than recent price action suggests.

 

The firm's core conclusion is that while AI-related trades may continue to undergo sharp rotations and choppiness, the broader bull market remains grounded in corporate profit growth rather than pure speculation, provided earnings do not deteriorate.

 

Earnings Serve as the Primary Driver

Goldman noted that Q2 earnings results have played a major role in underpinning market optimism.

 

As of July 31, roughly two-thirds of the S&P 500 by market capitalization had reported second-quarter results. Goldman found that 64% of reporting companies beat Wall Street earnings consensus by at least one standard deviation, making this one of the strongest reporting periods on record.

 

The bank estimates that excluding abnormally large investment gains reported by several mega-cap tech giants, S&P 500 earnings expanded 26% year-over-year. Including those gains, core earnings growth reached 45%.

 

Even median-performing companies beat expectations. Goldman estimates the median S&P 500 earnings growth rate at 12%, topping the 9% consensus estimate recorded at the start of the quarter.

 

AI Still Accounts for the Lion's Share

Snider pointed out that AI infrastructure providers continue to drive a significant portion of corporate profit growth.

 

Goldman estimates that AI infrastructure companies contributed roughly one-third of total S&P 500 earnings growth in the second quarter, a proportion that could rise above half for the remainder of 2026 and into 2027. Alphabet, Amazon, Microsoft, Nvidia, and Broadcom remain the primary drivers.

 

At the same time, Goldman noted that strong earnings are not confined exclusively to AI winners: the equal-weighted S&P 500—which strips out mega-cap tech bias—has posted steady gains as earnings expectations rise, signaling healthy underlying corporate fundamentals.

 

Wall Street Keeps Raising Forecasts

Notably, robust quarterly results have prompted analysts to adopt a more optimistic stance for the coming year.

 

Goldman highlighted that since the start of the third quarter, consensus estimates for 2027 S&P 500 earnings have been revised up by approximately 1%, with upward revisions spanning most sectors. Energy and Financials saw the largest positive adjustments.

 

However, the firm cautioned investors to keep a close eye on operating margins.

 

"Corporations have largely absorbed higher tariffs and energy costs, but analysts have begun trimming profit margin expectations for many reporting companies this quarter, signaling that input costs remain a risk even as top-line revenues hold up," the report noted.

 

AI CapEx Wave Shows No Signs of Slowing

Perhaps the single largest takeaway from the report is that hyperscalers appear more committed than ever to deploying capital.

 

Cloud revenue across Alphabet, Amazon, and Microsoft grew 48% year-over-year this quarter, accelerating from 39% in the prior quarter. Meta's revenue growth also met market expectations.

 

These operational metrics prompted analysts to raise forward CapEx forecasts.

 

Goldman now projects that capital expenditures by hyperscale data center operators will top $1 trillion by 2027—an upward revision of more than $100 billion compared to pre-earnings estimates.

 

However, outlays of this scale come at a price. Data shows that the five major hyperscalers collectively burned through $182 billion in CapEx during Q2, while generating roughly $5 billion in free cash flow over the same period.

 

To bridge this deficit, they raised a combined $101 billion in capital through debt and equity offerings during the quarter.

 

Rather than viewing this financing spree as a warning flag, Goldman interprets it as a sign of growing conviction that AI investments will generate sufficient revenue growth to sustain continued expansion.

 

Analysts project that cloud provider revenue will outpace CapEx growth over the next two years, even as capital outlays continue to climb.