Goldman Sachs warns of an AI bubble: When the first major player cuts spending, the entire market will reprice.

```

The AI capital expenditure boom is accumulating systemic risk. Goldman Sachs strategists warn, the AI market is now like a rubber band being stretched—market participants’ ongoing disregard for negative signals will ultimately reach a breaking point: once any major tech giant takes the lead in cutting AI spending, the entire AI sector’s valuation framework will face a comprehensive overhaul.

Rich Privorotsky, strategist at Goldman Sachs’ Global Banking and Markets division, pointed out in a research note on Tuesday that in past weeks, the market has virtually ignored all negative signals emerging from the AI capital expenditure trade. He specifically highlights a growing structural divergence: hyperscalers are ramping up spending commitments, but their stock prices have consistently underperformed; meanwhile, AI hardware stocks represented by Nvidia and TSMC have rallied against the trend. This divergence, in itself, is a sign of distorted market pricing.

The pressure on tech stocks is already showing up at the market level. South Korea’s Kospi index plunged 10% the day after hitting a record close, with Samsung Electronics and SK Hynix each falling more than 12%; Nasdaq futures fell about 2.5%, Micron Technology dropped more than 7% pre-market, and Intel fell 6.5%.

The Rubber Band Effect: The Gap Between Spending Commitments and Return Expectations

Privorotsky uses a "rubber band" analogy to describe the current internal tension in the AI market.

He notes that while hyperscalers such as Amazon, Alphabet, and Meta continue to increase AI capital investment, their stock performance keeps lagging, showing that market confidence in their investment returns is quietly wavering.

At the core of the problem is that the current pricing of the entire AI sector is based on a single assumption: as inference demand grows, capital expenditure will only ever rise and never decline. "The possibility of ‘even a slight decrease’ has not been priced in by anyone,” Privorotsky writes.

This means that once expectations reverse even slightly, the scale of market repricing will far exceed linear predictions.

Eastern Low-Cost Models: The Variable Disrupting Western Valuation Logic

Behind Goldman Sachs’ warning, a specific technological trend is rapidly advancing.

Privorotsky points out that technological advances in regions such as China and Japan are significantly lowering the operating costs of AI software, and this change has not yet been reflected in hyperscalers’ spending forecasts.

According to industry media, China's GLM-5.2 large language model was trained entirely on Huawei processors, using 100,000 Huawei chips throughout without involving Nvidia products. This case points directly to a core risk: if cutting-edge intelligence can be developed in the East at a fraction of the Western cost, then the vast AI investment by Western tech giants will face serious overinvestment risk.

"The biggest capital allocators are also the group most exposed to overinvestment risk," Privorotsky warns.

Multiple Pressures Combine, Tech Sector Valuations Under Strain

Federico Manicardi and Victoria Campos of JPMorgan’s International Market Intelligence unit also issued similar warnings about the tech sector, outlining multiple factors now weighing on tech stocks: excessive market expectations, exuberant market sentiment, lack of free cash flow, CTOs’ resistance to soaring token costs, the trend toward low-cost models, strict regulatory curbs from Washington, and increases in supply of both debt and equity.

The two strategists also particularly noted that progress in areas like orchestration, model fusion, and quantization is equally worth watching, "since these developments point to continually improving efficiency and potential headwinds for pricing power."

Who Will Hit the Brakes First

In Goldman’s analytical framework, the “breaking point” in the current AI capital expenditure cycle is likely to come from a rational awakening by one of the core spenders—when it realizes that allocating less money to AI and instead rewarding shareholders is in fact the better capital allocation choice.

Privorotsky’s warning points to a tail risk the market has not fully priced in: at a time when everyone is betting AI spending will only increase, the first giant to hit the brakes will become the fuse that triggers a market-wide repricing. When that time comes, from chip makers to cloud platforms, the entire AI value chain’s valuation logic will face renewed scrutiny.

Risk Disclosure and DisclaimerThere are risks in the market, and investments need to be made cautiously. This article does not constitute personal investment advice, nor does it take into account the special investment objectives, financial situation or needs of individual users. Users should consider whether any opinions, views or conclusions in this article fit their specific circumstances. Investing based on this is at your own risk. ```