Google Cloud soared 82%, and it's hard to hide hidden worries about sky-high expenses. The Big Seven's “AI faith” loosened: Apple stands alone, and Microsoft Meta faces a “capital expenditure judgment day”

The Zhitong Finance App learned that over the years, there has been an undeclared tacit agreement between US tech giants and investors: companies can spend a lot of money on artificial intelligence, and as long as revenue continues to grow, the stock market will reward them. However, this “gentleman's agreement” is suddenly falling apart. Alphabet Inc. (GOOGL.US) shares fell more than 7% on Thursday, the biggest one-day decline in more than a year, after the company raised its 2026 capital expenditure guidance to as high as $205 billion and disclosed that second-quarter free cash flow turned negative for the first time since listing in 2004.

Although Google's parent company also handed over an impressive report card of an 82% surge in cloud computing revenue, far exceeding Wall Street expectations, investors are still worried about sky-high expenses.

Jason Lemire, chief investment officer at Bold Wealth Partners, said, “The market is now highly sensitive to capital expenditure, to the point where it is even obsessed. In the past, more was better, but now less is better. Capital raising, negative cash flow, and rising debt — all are adding to the risk picture.”

This sell-off reflects a fundamental shift in the market narrative around AI and the “Big Seven US Stock Companies.” As capital expenditure continues to rise, it is becoming more difficult to please investors. Alphabet's hit is particularly noteworthy — due to the popularity of its Gemini AI service, self-developed data center chips, and rapid progress in the cloud business, it was originally regarded as the strongest beneficiary of AI among the Big Seven.

This change sets a serious tone for the next big earnings season: Microsoft and Meta Platforms will be on the list on Wednesday, while Apple and Amazon will debut on Thursday.

The index tracking the Big Seven (including Nvidia and Tesla) followed Alphabet's earnings report falling 4.8% last Thursday, the worst single-day performance since Trump's “Liberation Day” tariff statement in April 2025. The index has accumulated a 3.7% decline since 2026, and continued to soar for the previous three years. As a result, these giants, which have dominated the S&P 500 since the AI boom, are gradually ceding their leading positions to beneficiaries of their tens of billions of dollars — such as chipmakers Micron Technology and Ultrafine Semiconductors.

Microsoft was once regarded as an AI leader due to its shareholding in ChatGPT developer OpenAI, but this year it was second to last among the Big Seven, with a sharp drop of 21%. The market is worried that it will fall behind the competition, even though analysts estimate that its capital expenditure for the current calendar year has exceeded 190 billion US dollars. Meta shares fell 9.8%, and investors questioned the effectiveness of its AI investment; Amazon remained essentially flat in 2026.

According to market summary analysts' average estimates, the combined capital expenditure of Alphabet, Microsoft, Amazon, and Meta will reach about 724 billion US dollars this year, and will approach 950 billion US dollars in 2027.

Willy Lee, partner at venture capital agency Neostellar Capital, stated, “The current market environment tends to sell off capital expenditure, and Microsoft, Meta, and Amazon are joining hands with Alphabet to compete for investment. As development support continues, all of their businesses will face strict scrutiny.”

Investors' “defection” has also overshadowed the prospects of spending beneficiaries, particularly chipmakers. The Philadelphia Semiconductor Index (SOX) surged 101% in the first half of the year, but it has rebounded 17% in July, and is expected to record the worst monthly performance since June 2022 (when the market was mired in inflation and sell-off).

The 30 constituent stock indices have fluctuated sharply recently, and the 100-day volatility climbed to the highest level since the pandemic hit the market in 2020. According to the data, SOX had 17 single-day fluctuations of more than 5% during the year, the same level as in 2008; while none of the S&P 500 and the Nasdaq 100 index, which is dominated by technology stocks, fluctuated so sharply during the same period.

Bold Wealth's Lemire said, “AI winter is finally here. Current profit margins, particularly in the field of memory chips, are extremely high, but long-term maintenance is absolutely impossible. Sooner or later we will see profit compression and valuation contraction, which will have a huge impact on the market.”

Apple is in stark contrast to this. The iPhone maker avoided massive AI spending and instead chose to partner with model developers to support its services. Investors have applauded this strategy in recent weeks, driving the stock price to rise 15% in July, which is expected to be the best monthly performance in three full years. The stock rose 23% cumulatively in 2026, making it the biggest point contributor to the 8.3% increase in the S&P 500 index.

But that doesn't mean Apple is at ease. Demand for memory chips required for AI computing has soared, forcing Apple to raise prices for products such as MacBooks and iPads. How consumers react and what is the impact on profit margins is still an unresolved question.

Of course, after this round of sell-off by the Big Seven, the valuation of some individual stocks is already relatively cheap. For example, Microsoft's current price-earnings ratio is about 19 times, a significant discount compared to its 10-year average of 27 times; Meta is about 14 times, which is also 20 times lower than the ten-year average.

The problem is that competitive investment in AI computing power is changing corporate business models and introducing new risks. Alphabet's negative cash flow in the second quarter was shocking to investors for companies with rich cash flow from diversified businesses.

Brad Warden, senior portfolio manager at Nomura Asset Management, said the above factors have weakened the reference significance of historical valuations. Its funds hold Nvidia, Alphabet, Microsoft, and Amazon. Warden said, “They seem cheap now, but looking ahead to potential disruption, they are presumed guilty and need to prove their innocence. Is the current business model sustainable? Will the economic effects worsen?” He anticipates that large AI spenders will eventually reap a return on their investment, but “the key is how much pain you are willing to endure during the investment cycle and how confident you are that you can finally realize the financial return at the end of the cycle.”