NVIDIA plunges! Earnings beat expectations, why isn't the market buying it?

On February 27th, local time, the three major U.S. stock indices all closed lower, with the Wind U.S. Technology 7 Giants Index down 1.72%, continuing the decline from the previous trading day.

After the unexpectedly strong Q4 earnings report, NVIDIA’s stock price fell for two consecutive trading days. NVIDIA has declined a total of 6.65% this week and 7.29% in February.

Market analysts stated, “AI, which drove the market last year, is now dragging down stock prices because people are still unclear about its exact impact on the economy.”

Stock Price Diverges from Earnings Reports

The volatility of tech giants’ stock prices is mainly not due to earnings falling short of expectations.

After the U.S. stock market closed on February 25th, NVIDIA released its earnings report. The data showed that NVIDIA’s total revenue for Q4 was $68.13 billion, up 20% quarter-over-quarter and 73% year-over-year, surpassing analysts’ previous forecast of $66.2 billion. Structurally, data center revenue was $62.3 billion, compared to $35.58 billion in the same period last year; network revenue was $10.98 billion, with analysts expecting $9.02 billion; gaming revenue was $3.7 billion, with analysts expecting $4.01 billion.

Despite the quarterly earnings continuing to beat market expectations as in previous quarters, the significant increase in capital expenditures by leading tech companies has raised concerns about NVIDIA’s financial sustainability and the actual profitability of AI investments.

During NVIDIA’s earnings call, an analyst asked about this, to which CEO Jensen Huang responded that he is confident in the growth of customer cash flow and emphasized that current computing power equals revenue—whoever buys more computing power can create more value.

However, the market was not convinced. Over the two trading days following the earnings release, NVIDIA’s stock price plummeted, with a market cap evaporating over $400 billion.

Similarly, Microsoft previously announced that its fiscal second quarter (October 1, 2025, to December 31, 2025) revenue was $81.3 billion, up 17% year-over-year, with non-GAAP EPS of $4.14, up 24%. The overall earnings report showed steady growth, but Microsoft’s stock price fell 10% the next day.

Likewise, cloud infrastructure provider CoreWeave, closely linked to NVIDIA, saw its stock plunge 18.51% on February 27th due to increased net losses in its latest quarterly report and an expected capital expenditure of at least $30 billion in 2026.

Goldman Sachs warned that the growth rate of AI capital expenditures is expected to slow down in the second half of the year. This turning point could directly threaten AI infrastructure stocks that rely heavily on capital spending, potentially causing their valuation premiums to collapse.

OpenAI Secures $110 Billion in Funding

Boosting AI Infrastructure Development

Despite skepticism, tech giants continue to make large investments in AI infrastructure.

On the evening of February 27th, Beijing time, OpenAI announced it had secured $110 billion in funding, with investors including NVIDIA, Amazon, and SoftBank. OpenAI stated that this funding will be used to expand AI infrastructure to accelerate the promotion of AI benefits.

In the 2026 plans of giants like Microsoft, Google, and Amazon, record-breaking capital expenditures are planned for building new data centers and other infrastructure, further intensifying the AI “arms race.”

Google CEO Sundar Pichai said during a conference call, “Even as we increase capacity, our supply remains constrained. Clearly, our capital expenditure this year is focused on the future.”

D.A. Davidson analysts noted that these companies see the AI computing power race as the next “winner-takes-all” market, and none want to lose this competition.

AI Bubble Risks Still Exist

In response to market concerns about an AI bubble, Galaxy Securities research reports suggest that compared to the internet bubble of 2000, the current AI bubble’s risks in macroeconomic environment, valuation, and investment pace remain manageable.

The reasons include the Federal Reserve’s cumulative rate cuts of 125 basis points since September 2024, which have significantly lowered real interest rates and increased market tolerance for long-term capital expenditure and future returns uncertainty. Valuation structures show no signs of disorder; core AI companies’ P/E ratios mainly range from 35 to 45, and the current market cap expansion is accompanied by real earnings growth. In terms of investment intensity, AI investments are still in the early stages of general technology diffusion. Although tech capital expenditures as a share of GDP have increased, they are still below the peak of the internet cycle.

However, the report also mentions that with the current overlay of structural and cyclical issues in the global economy, systemic risks surpass those of the internet bubble period: “Capital and policy are passively highly concentrated. If the pace of technological implementation cannot offset macro downward pressure, it could evolve into a ‘domino effect’ of risk unwinding.”