Goldman Sachs partner: "The 'AI narrative' dominates the market, but don't forget that 'the cloud giants have already allocated 100% of their cash flow to capital expenditure.'"

Goldman Sachs partner: "The 'AI narrative' dominates the market, but don't forget that 'the cloud giants have already allocated 100% of their cash flow to capital expenditure.'"

```

Goldman Sachs partners’ latest warning: AI investment narratives are once again dominating market pricing, but "cloud giants have already allocated 100% of their cash flow to capital expenditure," and market breadth continues to deteriorate, while momentum factor crowding is approaching a five-year high.

On May 3, Goldman Sachs partner Mark Wilson pointed out in the latest weekly report that despite continued disruptions from geopolitical risks, the market is once again being dominated by a single AI narrative—the Nasdaq recorded its strongest single-month performance since 2002 in April. At the same time, Microsoft, Amazon, Alphabet, and Meta, the four ultra-large-scale cloud providers, will together spend over $600 billion on capital expenditure this year, with some spending already exceeding their entire annual operating cash flow.

Wilson highlighted two core risks: First, ultra-large-scale cloud providers are very likely to spend more on capital expenditure than their operating cash flow this year; Second, market breadth continues to deteriorate, momentum factor hedge funds’ net exposure and leverage ratios are both near five-year highs, and concentration risk cannot be ignored.

He also cited the contrarian view from Goldman Sachs equity research head Jim Covello, advising investors to examine the potential gap between "AI spending and actual returns."

Capital Expenditure Surge: Stunning Numbers but Signs of Cash Flow Warning

This quarter’s big tech earnings season has once again raised the upper limit of the market’s understanding of AI investment scale.

According to Wilson, Microsoft plans to spend $120 billion on capital expenditure in the next six months alone, 45% higher than previous market estimates on an annualized rate, implying that total capital expenditure for 2025 could reach $190 billion. Amazon's capital expenditure this year will exceed $200 billion, Alphabet will spend about $185 billion, and Meta about $135 billion.

Meanwhile, on the income side, verification data is equally strong:

Google Cloud’s year-over-year revenue growth accelerated to 63%, with annualized revenue surpassing $20 billion. The backlog of orders doubled month-over-month to $460 billion, and multiple enterprise deals over $1 billion were signed;Microsoft’s AI business annualized recurring revenue reached $37 billion, a year-on-year increase of over 120%;AWS’s growth rate reached 28%, the fastest since 2021, while Amazon disclosed that its Trainium custom chip revenue commitments have exceeded $225 billion.

However, while acknowledging the historic scale of this investment cycle, Wilson explicitly issued a warning: "Do not ignore the fact – ultra-large-scale cloud providers are very likely to spend more than 100% of their operating cash flow on capital expenditure this year."

This indicates that this unprecedented AI arms race is progressing by consuming all, or even exceeding, its own ability to generate cash flow.

Market Breadth Deteriorates, Momentum Factor Crowding Nears Five-Year Extreme

The AI narrative-driven rally is showing clear signs of structural divergence.

Wilson pointed out that market breadth continues to deteriorate. At the same time, the momentum factor has surged 25% this year, and hedge funds' net exposure and overall leverage to the momentum factor are both near five-year highs.

He cited the judgment of Goldman Sachs strategist Ben Snider: The current "bad breadth" phenomenon does not indicate a market top, but rather "continued volatility in the momentum factor."

Notably, Wilson quantitatively compared the performance of momentum factors with Goldman Sachs’ AI pair portfolio and found a high degree of synchronization, with shared driving variables pointing to the relative strength of semiconductors over software, and the outperformance of AI beneficiaries over spenders.

This structure means that the current market rally is highly reliant on the continued strengthening of a few themes. Wilson specifically mentioned that over 60% of this year’s S&P 500 earnings growth forecast comes from just Micron and Nvidia.

Contrarian Examination of AI Spending vs. Returns

While affirming the persistence of the AI investment cycle, Wilson unusually introduced a "contrarian perspective" in the report.

He cited the special report "AI Spending vs. Returns" by Goldman Sachs equity research head and former semiconductor analyst Jim Covello, suggesting that investors carefully examine this question.

Wilson did not directly present Covello's conclusions in the report, but characterized them as "a contrarian viewpoint worth reading," hinting that the present market optimism about AI investment returns may be overestimated.

The background for this suggestion is: despite strong income data from cloud providers, Wilson also pointed out that ultra-large-scale cloud providers are investing 100% or more of their operating cash flow into capital expenditure. The sustainability of this model and when AI infrastructure investments can be converted into measurable business returns remain core variables that the market has yet to fully price in.

Risk Warning and DisclaimerMarkets carry risks; investment requires caution. This article does not constitute personal investment advice and does not consider the special investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Invest accordingly at your own risk. ```