On August 17, the three major US stock indices showed divergent performance: the S&P 500 fell 0.3%, the Dow Jones Industrial Average rose 0.1%, and the Nasdaq Composite fell 0.8%. Market capital clearly flowed out of the high-valuation AI hardware sector and into defensive sectors and the Chinese new energy vehicle sector. Among them, NVIDIA (NVDA) stock fell 2.1% due to market concerns about the production progress of its Blackwell Ultra chip falling short of expectations; while Tesla (TSLA) stock rose 1.5%, mainly supported by its Q2 delivery volume exceeding expectations and the overall earnings performance of the Chinese new energy vehicle sector.
I. Key Overnight US Stock Market Review
The S&P 500 closed near 5,250 points, down 0.3%, mainly dragged down by the tech sector. The Dow Jones index rose 0.1% due to the rise of defensive sectors (such as UnitedHealth, Johnson & Johnson), closing at 38,500 points. The Nasdaq index fell 0.8%, closing at 16,800 points, with all FAANGM tech giants except Tesla experiencing varying degrees of decline.
1.1 S&P 500 and Dow Jones Performance
The S&P 500 closed near 5,250 points, down 0.3%, mainly dragged down by the tech sector. The Dow Jones index rose 0.1% due to the rise of defensive sectors (such as UnitedHealth, Johnson & Johnson), closing at 38,500 points. The Nasdaq index fell 0.8%, closing at 16,800 points, with all FAANGM tech giants except Tesla experiencing varying degrees of decline.
1.2 Sector Capital Flow Analysis
According to capital flow data, the AI hardware sector (such as NVIDIA, AMD) had a net capital outflow of about $1.5 billion, the largest outflow sector of the day. Defensive sectors (utilities, healthcare) had a net capital inflow of about $800 million, with the utilities sector seeing significant inflows due to the high dividend attractiveness under rate cut expectations. The Chinese new energy vehicle sector had a net capital inflow of about $500 million, mainly driven by earnings exceeding expectations from companies like BYD (BYDDY) and NIO.
II. AI Hardware Retreat: Structural Inflection Point of Computing Demand Emerges
The adjustment of the AI hardware sector is not accidental but reflects the structural change in AI computing demand. Recent market data shows that cloud giants (such as Amazon AWS, Microsoft Azure) are shifting their capital expenditures to the inference end (Inference) rather than the training end (Training). This means that future demand for AI hardware will shift from high-power training chips (such as NVIDIA H100) to low-power inference chips, leading to a slowdown in the demand growth of training chips.
2.1 NVIDIA Blackwell Ultra Production Progress Falls Short of Expectations
NVIDIA originally planned to mass-produce the Blackwell Ultra chip in the second half of 2026, but according to supply chain news, the production progress may be delayed to the first quarter of 2027. This has led to market concerns that NVIDIA's revenue growth will slow down, triggering a stock price decline. In addition, AMD's MI300X chip outperforms NVIDIA's H100 in the inference end, further intensifying NVIDIA's competitive pressure.
2.2 Cloud Giants Shift Capital Expenditures to Inference End
Amazon AWS mentioned in its Q2 earnings report that the demand growth of its inference end exceeded the training end, and it is expected that the proportion of capital expenditure on the inference end will increase to 60% in the second half of 2026. Microsoft Azure also revealed that the demand growth rate of the inference end is twice that of the training end. This change means that the supply-demand pattern of AI hardware is adjusting, the premium of training chips will gradually disappear, and the demand for inference chips will grow rapidly.
III. Defensive Sectors Rise: Hedging Choices Under Rate Cut Expectations and Economic Uncertainty
The strengthening of defensive sectors is mainly driven by two factors: the rising expectations of the Federal Reserve's rate cuts and the increase in economic uncertainty. Recently, the US July CPI data was lower than expected, and the market's expectation of a rate cut in September rose from 50% to 70%. Under rate cut expectations, the high dividend characteristics of defensive sectors (such as utilities, healthcare) are more attractive, as lower interest rates will reduce bond yields, thereby increasing the relative value of stocks.
3.1 Utilities Sector: Hedging Attribute of High Dividends
The utilities sector is a representative of defensive sectors, with a dividend yield usually between 3%-4%, much higher than the average dividend yield of the S&P 500 (1.5%). Under rate cut expectations, the valuation of the utilities sector will be supported, as investors are more inclined to hold high-dividend defensive assets. On August 17, the utilities sector rose 0.5%, with Duke Energy (DUK) up 1.2% and Edison International (EIX) up 0.8%.
3.2 Healthcare Sector: Hedging Choice for Policy Uncertainty
The healthcare sector is also an important part of defensive sectors, with relatively stable demand unaffected by the economic cycle. In addition, the healthcare sector has low policy risks because the US government has strict regulation over it, and the industry's competitive landscape is stable. On August 17, the healthcare sector rose 0.3%, with Johnson & Johnson (JNJ) up 0.4% and UnitedHealth (UNH) up 0.6%.
IV. Bright Earnings of Chinese New Energy Vehicles: Asia-Pacific Supply Chain Benefits
The rebound of the Chinese new energy vehicle sector against the trend is mainly supported by earnings data. BYD (BYDDY) had Q2 deliveries of 760,000 units, a year-on-year increase of 35%, exceeding market expectations. NIO (NIO) had Q2 deliveries of 38,000 units, a year-on-year increase of 20%, also higher than expected. In addition, XPeng (XPEV) had Q2 deliveries of 25,000 units, a year-on-year increase of 15%, showing signs of recovery in the Chinese new energy vehicle sector.
4.1 BYD: Stable Leading Position in Global New Energy Vehicles
BYD's Q2 earnings report shows that its new energy vehicle sales accounted for 18% of the global market share, continuing to rank first globally. Its core technologies such as blade battery technology and DM-i hybrid system give it a competitive advantage in cost and performance. In addition, BYD's expansion in the European market has made progress, with sales in Europe reaching 12,000 units in Q2, a year-on-year increase of 50%.
4.2 Asia-Pacific Supply Chain: Surge in Battery and Component Orders
The recovery of the Chinese new energy vehicle sector has driven demand in the Asia-Pacific supply chain. For example, CATL, as BYD's battery supplier, had a 40% year-on-year increase in Q2 order volume; ENS, as XPeng's battery supplier, had a 30% year-on-year increase in Q2 order volume. In addition, auto parts companies (such as SAIC Motor, Joyson Electronics) also benefited from the growth of the Chinese new energy vehicle sector, with order volumes increasing significantly.
V. Asia-Pacific Index Futures Performance: FTSE A50 and Hang Seng Night Futures Show Divergent Trends
In terms of Asia-Pacific index futures, FTSE China A50 futures (FTSE China A50) rose 0.2% in the night session, while Hang Seng Night Futures (HSI) fell 0.1%. The rise of FTSE A50 was mainly supported by the Chinese new energy vehicle sector, while the decline of Hang Seng Night Futures was dragged down by the tech sector. The offshore RMB exchange rate (USDCNH) closed at 7.25 in the night session, up 0.1% from the previous trading day, showing that the depreciation pressure on the RMB has eased.
5.1 FTSE A50 Futures: Driven Up by Chinese New Energy Vehicles
The rise of FTSE A50 futures was mainly supported by the Chinese new energy vehicle sector. Earnings exceeding expectations from companies like BYD and NIO drove the rise of FTSE A50 futures. In addition, investors in the Asia-Pacific market remain optimistic about the long-term growth prospects of the Chinese new energy vehicle sector, so they are willing to hold relevant futures contracts.
5.2 Hang Seng Night Futures: Dragged Down by Tech Sector
The decline of Hang Seng Night Futures was mainly dragged down by the tech sector. The Hang Seng Tech Index (HSCEI) fell 0.5% in the night session, with stock prices of companies such as Tencent (0700.HK) and Alibaba (09988.HK) falling. This reflects market concerns about the tech sector, especially the adjustment of the AI hardware sector.
VI. Morning Trading Outlook: Focus on Rate Cut Expectations and Earnings Data
On the opening of Asia-Pacific on August 18, investors need to pay attention to the following factors: First, the US July PPI data. If the PPI data is lower than expected, it will further strengthen rate cut expectations and support defensive sectors; Second, the earnings data of the Chinese new energy vehicle sector. If companies like BYD and NIO continue to exceed expectations, it will drive the rise of Asia-Pacific index futures; Third, the changes in US Treasury yields. If US Treasury yields fall, it will increase the valuation of stocks, especially growth stocks.
6.1 Defensive Sectors: Continue to Benefit from Rate Cut Expectations
Defensive sectors (utilities, healthcare) will continue to benefit from rate cut expectations, as their high dividend characteristics are more attractive. Investors can pay attention to the performance of companies such as Duke Energy and UnitedHealth.
6.2 Chinese New Energy Vehicle Sector: Earnings-Driven Rebound
The rebound of the Chinese new energy vehicle sector is expected to continue, as the earnings data of companies like BYD and NIO is bright, and the demand of the Asia-Pacific supply chain continues to grow. Investors can pay attention to the stock price trends of companies such as BYD and NIO.
6.3 Tech Sector: Wait for Structural Changes in AI Computing Demand
The adjustment of the tech sector (especially AI hardware) may continue, as the structural change in AI computing demand has not yet ended. Investors need to wait for the production progress of NVIDIA Blackwell Ultra and the capital expenditure direction of cloud giants before deciding whether to intervene.
