A $200 million difference leads to $100 billion evaporating! Who is "hijacking" Nvidia?

　　By Every Economic Daily Reporter | Yue Chupeng  Zheng Yuhang    Editor | Gao Han    

　　What happens when Nvidia's "imperfect" financial report is scrutinized by Wall Street's "perfectionism"?

　　On Wednesday (August 27) after the US market closed, Nvidia, the "world's most valuable company," delivered an impressive earnings report. However, a tiny "flaw" of just $200 million instantly pricked the market's hypersensitive nerves. Its stock price plunged more than 5% after hours, and over the next two trading days, its market value evaporated by more than $180 billions.

　　The reason lies in Nvidia's data center business revenue for the second quarter of fiscal year 2026, which came in at $41.1 billions, slightly below the market expectation of $41.3 billions. For some investors, this $200 million gap "is like scoring 98 out of 100 when everyone expected you to get a perfect score—still a disappointment."

　　The root of Wall Street's harshness toward Nvidia's performance lies in its massive $4.3 trillions market cap, which is deeply intertwined with the entire US stock market and even the new engine of the US economy—AI data center construction. Any signal deviating from expectations can be infinitely magnified and trigger a violent chain reaction.

　　Note: Non-GAAP refers to non-mandatory disclosure items, which are voluntary by management and serve as a supplement to financial statements.

　　By any objective standard, Nvidia's financial report is outstanding. Quarterly revenue reached $46.743 billions, up 56% year-on-year, slightly above the market expectation of $46.23 billions; core engine data center business revenue hit a new high at $41.1 billions, up 56% year-on-year; adjusted earnings per share were $1.05, up 54% year-on-year, also beating expectations.

　　However, these numbers still failed to satisfy Wall Street's "perfectionism." The $41.1 billions in data center business revenue did not meet analysts' expectations of $41.3 billions, and was instantly interpreted by the market as a sign that cloud computing customers are becoming cautious in AI infrastructure spending.

　　Baird investment strategy analyst Ross Mayfield bluntly stated that the market has long been accustomed to Nvidia's high growth and has placed irrational "perfect expectations" on it. Any minor flaw can be magnified and trigger an overreaction.

　　An investor named Harry, who has held Nvidia stock since 2023, told National Business Daily reporters (hereinafter referred to as "NBD reporters"): "Since Trump took office, US stocks have become more volatile, and investors are easily swayed by emotions. This drop in Nvidia's stock price has little to do with valuation and is more about an overreaction from the market." He sold all his Nvidia shares in April this year but soon regretted it and bought back in July. In his view, short-term fluctuations are inevitable, but Nvidia has built a strong moat with its computing power cards. "Buying on dips" is the investment strategy he is sticking to at the moment.

　　Another investor, Chleo, who lives in Seattle, USA, bought Nvidia when the stock price was around $70 and has held it ever since. She told NBD reporters that the impressive performance but falling stock price is mainly because the market's expectations for Nvidia are too high. "It's like scoring 98 when everyone expected 100. In over $40 billions of revenue, a $200 million gap is almost negligible." Despite the stock's volatility, she remains optimistic about Nvidia's long-term growth and plans to "buy in batches on dips."

　　Despite the stock's volatility, analysts remain bullish on Nvidia. WedBush investment bank analyst Dan Ives said in a research report sent to NBD reporters, "This financial report is very important for the entire technology sector and serves as a guide, indicating that the AI revolution is entering a new round of growth. There is only one chip company in the world driving this AI revolution, and that is Nvidia." He expects the company to reach a market cap of $5 trillions by early 2026. Several investment banks have also raised their target prices for Nvidia.

　　Some US media have pointed out that Nvidia's sustained profitability is even more important to the entire stock market than whether the Federal Reserve will cut interest rates at its next meeting. This is the deeper reason why Wall Street has such high performance requirements for it—the trend of the entire US stock market increasingly depends on Nvidia's "mood."

　　Now, Nvidia's market cap is close to $4.3 trillions, accounting for 8% of the total market cap of the S&P 500 Index. According to independent investment research firm Leuthold Group, this proportion has surpassed any company they have tracked in the past 35 years. In the tech-heavy Nasdaq 100 Index, Nvidia's market cap accounts for as much as 14.43%, even surpassing Cisco at the peak of the Internet bubble.

　　This unprecedented concentration has had a huge impact on the entire US stock market. Because index funds and ETFs, which are passive funds, manage vast wealth, they "passively" continue to buy Nvidia based on market cap weighting. Morningstar data shows that the number of ETFs providing leveraged exposure to Nvidia is now equal to the number of ETFs tracking the S&P 500 Index, which has a total market cap of $52 trillions. Among them, the larger GraniteShares 2x Long Nvidia Daily ETF has accumulated assets of $4.56 billions since its launch in December 2022. Huge amounts of capital are linked to Nvidia's stock price, and a single move can affect the whole market.

　　Apollo Global Management chief economist Torsten Slok said that in the first half of this year, as much as 35% of the S&P 500 Index's market cap growth came from Nvidia alone. According to SimCorp's calculations based on the Axioma US Equity Factor Risk Model, if Nvidia's stock price falls by 25%, the S&P 500 Index could fall by 4.4% as a result.

　　JPMorgan data shows that retail investors currently account for about 18% of US stock trading volume, nearly double that of a decade ago. Most of these funds have flowed into the stock market through index funds, with the most popular being those tracking the S&P 500 Index. Slok believes that buying S&P 500 index funds often gives people the impression of buying 500 different stocks and thus achieving diversification. But in reality, the increasing concentration of the S&P 500 Index has become a major problem.

　　As Nvidia's market cap continues to break records, some Wall Street investors can't help but compare it to Cisco in the early days of the Internet bubble. However, in terms of profitability, Nvidia far outperforms Cisco back then. Data shows that Cisco's five-year average net profit margin from 1996 to 2000 was about 17.2%, while as of January 2025, Nvidia's five-year average net profit margin is as high as 40.34%.

　　The stock market is the "barometer" of the economy. As the index weight becomes increasingly concentrated in AI companies like Nvidia, the driving force of the US economy is also undergoing a historic shift: the "cash-burning" frenzy triggered by AI data center construction is replacing traditional consumer spending as the biggest engine of economic growth.

　　Estimates by analysts at Renaissance Macro Research show that since 2025, the contribution of AI data center spending to US GDP growth has surpassed consumer spending for the first time in history. Some analyses even point out that without the construction boom of AI data centers, GDP might have already shrunk amid an uncertain macroeconomic environment. Therefore, data center spending may have delayed the arrival of an economic recession.

　　It is clear that the US economy is now deeply bound to Nvidia. This is because most AI-related spending flows to Nvidia. Nvidia CFO Colette Kress said on August 28 that in the latest fiscal quarter, "large cloud service providers" accounted for about 50% of the company's data center revenue, while the data center business accounted for 88% of Nvidia's total revenue.

　　Alphabet, Microsoft, Meta, and Amazon have also announced that they will collectively invest $400 billions in capital expenditures this year, most of which will go to AI infrastructure construction. Nvidia's revenue for the second quarter of fiscal 2026 is highly dependent on two mysterious customers. "Customer A" accounts for 23% of total revenue, and "Customer B" accounts for 16%, with the two together accounting for 39%. This proportion is significantly higher than the 25% (14% and 11%, respectively) in the same period last year. Nvidia expects total AI infrastructure spending to reach $3-4 trillions by the end of this decade.

　　Morningstar senior equity analyst Brian Colello told NBD reporters that currently, Nvidia's GPUs and AI infrastructure are still in short supply. Unlike the Internet bubble era, Nvidia's customers today are the world's largest tech companies, which are still sparing no effort to expand their infrastructure.

　　Although compared with the Internet bubble period, Nvidia has a more solid profit base and a stronger customer base, this highly concentrated revenue structure itself constitutes the biggest risk in the market. When the fate of a company is so closely intertwined with the pulse of the entire market, or even the economic growth of a country, no investor can ignore the chill at the top. Nvidia's growth continues, but the "perfectionist" shackles on it are also becoming heavier and heavier.