Summary of Key Points
NVIDIA's financial report for the second quarter of fiscal year 2027 exceeded expectations (revenue of nearly 100 billion, gross margin of 75%), but the stock price fell by more than 3% after the report was released. However, it later reversed and rose by 4.5% after Amazon announced an additional GPU order for a total of 3 million units. The underlying logic is as follows: The market was initially concerned that "AI capital spending had peaked" (due to tight cash flows and low return on investment (ROI) among giants). Amazon's large order dispelled these short-term doubts. Additionally, NVIDIA is trying to address its long-term capital issues by promoting the concept of "AI factories" (offering complete solutions rather than just individual GPUs) and a $500 billion computing power financing platform, which aims to boost its valuation. However, this also brings new risks, as NVIDIA's fortunes are now tied to the broader macroeconomic and financial cycles.
1. Why Did the Stock Price Fall Despite Exceeding Expectations?
NVIDIA's financial numbers were impressive: revenue increased by 106% year-over-year, and the gross margin remained at 75%. The guidance for the third quarter also surpassed expectations. So, why did the stock price fall? The main reason is the market's anxiety about the sustainability of AI capital spending:
- Low Return on Investment (ROI): Goldman Sachs estimates that giants have spent $1.4 trillion on AI infrastructure, but the returns have fallen short of expectations, and the technology may become obsolete, resulting in wasted costs (for example, buying GPUs without seeing significant AI applications).
- Cash Flow Challenges: Giants like Google and Amazon have negative free cash flows; the money they earn from their cloud businesses is not enough to cover their expenses, and they even have to borrow to fund their operations. In 2026, the four major tech giants increased their debt by 79%. Investors are worried that borrowing to invest in AI may not be sustainable.
In simple terms, the good financial results reflect the past; the market is concerned that there may be no demand for GPUs in the future, so investors sold their shares.
2. Amazon's 3-Million GPU Order: A Boost to Market Confidence
Amazon's sudden announcement of an additional GPU order (from 1 million to 3 million units) effectively dispelled concerns about a peak in capital spending:
- Real-World Demand Proof: Cloud service providers are the largest buyers of GPUs, and Amazon's increased purchases indicate that the demand for AI infrastructure is still growing.
- Management Confirmation: NVIDIA stated that demand is accelerating, with growth expected to reach 70% in fiscal year 2028 (far exceeding analysts' 45% forecasts), further supporting the idea that demand has not peaked.
It's like being worried that no one will buy milk tea from a shop, only to see someone buy 100 cups at once—of course, you feel relieved, and the stock price immediately turns around.
3. The Narrative of "AI Factories": Moving Beyond Just Selling Components
NVIDIA doesn't want to just sell GPUs (like mobile phone chips); it aims to sell "AI factories" (completes solutions rather than individual components):
- Product Upgrade: The next-generation Rubin series packages GPUs, CPUs, and networking equipment into complete solutions that customers can use right away.
- Why This Strategy?
1. Preventing Replacement: Competitors like AMD and Google's TPU can match some of NVIDIA's performance, but once customers use NVIDIA's complete systems, the cost of switching is extremely high (due to integrated data center architectures and operational teams).
2. Increasing Average Order Value: Selling individual GPUs is a competitive market, while selling complete AI solutions creates a more exclusive business model, significantly increasing the value of each order.
For example, instead of just selling an engine, NVIDIA now offers a complete car—resulting in higher profits and stronger customer loyalty.
4. The $500 Billion Computing Power Financing Platform: Solving Financial Issues, but with Macroeconomic Risks
NVIDIA collaborated with six major financial institutions (including BlackRock and KKR) to create a financing platform aimed at raising $500 billion for AI infrastructure investments:
- Core Logic: This platform packages GPUs, data centers, and long-term computing power contracts into "financial assets" that can be sold to pension funds, sovereign funds, and other long-term investors. This allows cloud service providers to avoid upfront costs and repay the investment with future revenue from AI services.
- Double-Edged Sword: By financializing its assets, NVIDIA's performance is now linked to interest rates and credit spreads. For instance, if interest rates rise, financing costs increase, potentially reducing investment in AI infrastructure and increasing risks.
5. Implications for Investors:
- A-share Market Reaction: NVIDIA's performance affects the A-share market; companies in related industries (such as optical modules and fiber optics) may see short-term gains due to the connection with AI infrastructure.
- Risk Warning:
1. If NVIDIA's "AI factories" don't prove successful (e.g., if customers don't buy them), they could become another bubble, similar to the "metaverse" phenomenon.
2. The financialization of NVIDIA's business makes it more vulnerable to macroeconomic trends. A downturn in the economy or rising interest rates could lead to a sharp drop in its valuation.
In summary, NVIDIA's story continues, but investors should not be overly optimistic. While the short-term gains are significant, the key factors to watch are whether NVIDIA can actually sell its AI solutions and whether the financing platform can attract enough investment.
(The entire analysis is written in plain language to make it understandable to non-financial professionals.)