Author: Chao Xiang Research
U.S. stocks weakened further on Tuesday, with the three major indices closing lower for a third consecutive session and hitting new two-week lows. Selling pressure was concentrated in chips, memory, optical communications, and AI infrastructure, with the Philadelphia Semiconductor Index, which had just re-entered a technical bull market the previous day, retreating sharply. Meanwhile, long-term U.S. Treasury yields remained elevated, oil prices continued to rise supported by Middle East risks, and market pricing for high-valuation tech assets turned cautious again. The current market is not simply trading on a downward revision of AI demand but is digesting a triple pressure from previous gains in the hardware chain, earnings expectations, and the cost of capital.
Market Performance and Key Closing Data
At the close, the S&P 500 fell 0.69% to 7,691.76 points; the Dow Jones fell 0.22% to 53,343.40 points; the Nasdaq fell 1.33% to 26,289.71 points. The Russell 2000 fell about 1.3%, indicating simultaneous pressure on small-cap stocks.
Most of the S&P's 11 sectors closed lower, with the Information Technology sector, dragged down by semiconductors, leading the decline, posting its weakest performance since July 29th. The Energy sector bucked the downtrend, supported by oil prices, and hit a recent high. The VIX rose about 0.7 points to just below 16; while the fear gauge remains low, defensive demand related to oil prices, inflation, and long-term interest rates is starting to heat up.
In the bond market, the 10-year Treasury yield was around 4.70%, and the 2-year yield around 4.17%; the 30-year Treasury yield hit a new high since 2007 during the session before retreating to around 5.28%. In commodities, WTI crude rose 0.52% to $84.94; Brent crude rose 0.17% to $91.02; COMEX gold fell 1.17% to $4,366. Cryptocurrencies saw a slight rebound, with Bitcoin around $64,500 and Ethereum around $1,900.
Chips Retreat Sharply, AI Hardware Enters an Expectations Digestion Phase
Tuesday's market narrative focused on the chip pullback. The Philadelphia Semiconductor Index closed down about 5%, having just rebounded more than 20% from its low to technically re-enter a bull market before immediately facing heavy selling pressure. This shift indicates that capital's trading in AI hardware has moved from chasing growth momentum to scrutinizing valuation levels and the pace of profit-taking.
Memory-related stocks led the decline, with SanDisk falling about 9%, Micron about 7%, and Seagate and SK Hynix ADR also showing significant weakness. Optical communications and AI infrastructure faced simultaneous pressure, with Coherent, Lumentum, CoreWeave, Fabrinet, etc., among the top decliners. Nvidia fell about 2.3%, Broadcom about 3.2%, with the correction spreading from memory to a broader range of AI hardware chain stocks.
In recent days, the market had focused on trading memory, optical communications, and semiconductors, driven by strong AI demand, tight supply, and continued acceleration in capital expenditures. However, with long-term bond yields staying high and oil prices stoking inflation concerns, the sectors with the biggest gains were the first to experience profit-taking. The chip sector still has Nvidia's earnings as a near-term catalyst, but the pre-earnings positioning has become more crowded.
Magnificent Seven Diverge, AI Trading Shifts from Rotation to Cooling Observation
The Magnificent Seven showed divergent performance. Meta had the largest decline, with Nvidia, Amazon, and Tesla weakening; Apple and Microsoft posted modest gains, while Alphabet was nearly flat. The Roundhill Magnificent Seven ETF fell about 0.7%, a smaller drop than the Philadelphia Semiconductor Index, indicating that Tuesday's capital pressure was more concentrated on the AI hardware side.
In recent periods, the market has rotated between large-cap tech stocks and the hardware chain. Large tech companies bear AI capital expenditures, while chips, memory, and optical communications directly benefit from the expansion of these expenditures. Tuesday's session showed that when pressured simultaneously by long-term rates and oil prices, both sides face re-evaluation. Large-cap tech stocks face tests on AI investment returns, while the hardware chain faces tests on recent gains and earnings expectations.
The key going forward is the direction of capital flows. If, after the chip pullback, funds flow back to large-cap tech stocks like Microsoft, Google, Amazon, and Meta, internal rotation within the AI theme could continue. If the giants also lack buying support, AI-related trading may enter a phase of cooling.
Chinese Stocks Diverge, Baidu's Earnings Weigh on Golden Dragon Index
Chinese stocks were broadly weak, with the Nasdaq Golden Dragon China Index closing down about 1%. Baidu's U.S.-listed shares fell nearly 13% post-earnings, becoming the most significant source of pressure among Chinese stocks. The company's Q2 revenue declined year-over-year, with the proportion of AI business increasing. GPU cloud maintained high growth, but the profit side was pressured by AI investments, and the market remains cautious about the transformation pace.
Alibaba's U.S.-listed shares bucked the trend, rising nearly 3%, becoming one of the few bright spots among Chinese stocks. The core of Chinese stock trading is no longer just macro sentiment; whether earnings can demonstrate that AI investments translate into revenue and profit is becoming a more important pricing factor. Baidu's movement shows that the AI narrative needs to be supported by profitability, cash flow, and core business stability.
Oil Prices and Long-Term Bonds Re-emerge as Key Anchors for U.S. Stock Pricing
The U.S.-Iran deadlock continues to elevate energy risks, with crude oil rising for a third consecutive session and Brent crude reclaiming the $91 level. Rising oil prices support the energy sector but also raise inflation expectations, squeezing the market's imagination for a Federal Reserve policy pivot.
The impact of long-term interest rates is more direct. The 30-year Treasury yield hit a new high since 2007 during the session, driven by fiscal supply, sticky inflation, and financing pressure from AI capital expenditures. AI companies need continuous investment in data centers, chips, and power infrastructure, and the government also needs continuous debt issuance. Rising funding costs affect both tech valuations and corporate financing costs.
Therefore, Tuesday's core market dynamic was not just the chip pullback, but high-valuation assets once again facing higher capital costs. As long as long-term bond yields remain high, the valuation flexibility of both AI hardware and large-cap tech stocks will be constrained.
Corporate Events: Earnings and Financing Continue to Point to Cost Pressures
Home Depot's Q2 results exceeded expectations, but housing-related demand remained weak, resulting in a limited stock reaction. U.S. consumption has not broadly decelerated, but the high-interest-rate environment continues to pressure the real estate chain and durable goods consumption.
On the AI financing front, Anthropic reportedly plans to expand its credit facility to over $100 billion to replenish funds ahead of a potential IPO. Meanwhile, market discussion is heating up around off-balance-sheet commitments and long-term capital expenditures of large tech companies. AI demand is still growing, but capital is starting to question return on investment and liability structures.
Chinese tech earnings also released similar signals. Xiaomi's Q2 revenue declined year-over-year, with memory price increases affecting smartphone shipments and gross margins; Baidu's AI cloud grew rapidly, but profits were suppressed by investments. Whether in U.S. or Chinese stocks, the market is now recalculating the same question: Can the revenue growth brought by AI cover the pressures from rising hardware costs, financing costs, and capital expenditures?
Watchpoints for Today
First, watch whether the 30-year Treasury yield can stabilize below 5.30%. If long-term bond yields continue to rise, high-valuation tech and AI hardware will remain under pressure.
Second, watch the buying support after the chip pullback. The Philadelphia Semiconductor Index retreated sharply just after re-entering a bull market. The trading volume and quality of the rebound in the coming days will determine whether this is normal profit-taking ahead of earnings or a sign of AI hardware entering a phase of cooling.
