Goldman Sachs Rejects "Three Major Anxieties" Over AI Bubble, Bullish on These Three Themes!

Ben Snider, Senior US Equity Strategist at Goldman Sachs Research, rejected concerns over an artificial intelligence (AI) bubble in a recent interview, stating he remains focused on three major themes within the AI trade.

Three Major Anxieties

1. Hyperscalers Suddenly Cutting Massive Spending

The market widely fears that hyperscale data center operators might abruptly scale back their massive outlays. If a company deems its return on investment unsatisfactory, it could slash spending forecasts. The belief is that if investors "reward" such capital discipline, peers might follow suit.

Snider's View: Snider believes the risk of this pivot is minimal. He noted that expanding AI infrastructure spending should continue to drive this trade.

2. Low Valuations Despite Robust Tech Earnings

Another concern is that despite explosive earnings growth in AI and tech, valuations remain quite cheap. This suggests investors may be reluctant to pay a higher premium, doubting how long the profit boom can last.

Snider's View: To Snider, this argument is counterintuitive. "Interestingly, investors view both overvaluation and undervaluation as warning signs," he said.
"I am always comforted when I hear bearish arguments," he continued. "To me, it confirms that some equity risk premium remains in the market. If a day comes when everyone thinks the outlook is bright with nothing to worry about, that will clearly signal an overvalued market."

3. Divergent Valuation Indicators

The final worry is that various valuation metrics are sending mixed signals. The Shiller P/E is near historic highs, hovering around the peaks of the dot-com bubble and 2021. Combined with the S&P 500's 20%+ gain over the past year, many wonder if upside is limited.

Snider's View: However, other metrics paint a different picture, such as the next 12-month forward P/E. Snider pointed out concisely: despite the S&P 500's massive rally over the past 12 months, its forward P/E is actually lower than it was a year ago.
This shows that the market is still driven by earnings growth, not multiple expansion. Snider stated that the most attractive parts of the market remain the top performers of recent months and years.

Three Major AI Themes

Snider highlighted the three specific sub-sectors within the AI trade he is currently focusing on:

I. AI Infrastructure Stocks

He remains focused on AI infrastructure names due to their relatively low valuations. This includes companies producing semiconductors, servers, and AI networking products.

"For most of the semiconductor industry, including memory chips, valuation multiples have barely expanded, showing the market remains skeptical of the sector," he said.

II. Power Infrastructure

Snider calls the power infrastructure theme "highly attractive," driven by ongoing AI capital expenditure and the impact of the war involving Iran on power and energy.

Energy infrastructure has been a backbone of the 2026 "HALO" (High-Asset, Low-Obsolescence) trade, as infrastructure builds are key to meeting the AI industry's endless demand for power.

III. Hyperscalers

Snider is bullish on hyperscalers investing heavily in AI builds across cloud platform data centers, such as Amazon, Microsoft, Meta, Alphabet, Oracle, and IBM.

These stocks have outperformed since late 2022, but many have lagged this year as investors rotated into stocks benefiting directly from hyperscaler spending.

"They just went through one of their worst months ever. Consequently, from an earnings perspective, their multiples are very low. On a P/E basis, they are trading at their lowest levels in a decade, very similar to trading during March 2020 or the end of the 2022 bear market," Snider added.