Non-farm payroll “surprise” dashes rate hike expectations, leading to heightened sector divergence in U.S. equities: the Dow Jones hits a new high while chip stocks crash.

Non-farm payrolls unexpectedly dampen rate hike expectations, but US stock market shows divergence: Dow hits new high, chip stocks crash, funds flee AI hardware for defensive sectors.

On July 2, 2026, the US June non-farm payrolls report was released. New non-farm jobs were only 57,000, far below the market expectation of 115,000; data for April and May were revised down by a combined 74,000. Although the unemployment rate unexpectedly fell from 4.3% to 4.2%, this was not due to a strengthening labor market, but rather a sharp drop in the labor participation rate from 61.8% to 61.5%, resulting in a monthly decrease of approximately 720,000 people in the labor force.

The interest rate futures market reacted swiftly: the probability of a July rate hike fell from about one-third before the report’s release to about one-fifth; CME data showed the probability of maintaining the interest rate unchanged in July rose to 82.4%; the probability of a September rate hike dropped from 64% to 52%. CICC maintained its forecast of “neither rate hike nor cut” for the year; CITIC Securities believes there is still room for downward revision of rate hike expectations.

The market quickly characterized the report as “Goldilocks” – employment cooled but did not deteriorate sharply, enough to reduce the Federal Reserve’s short-term rate hike pressure – and then began to vote with its feet. However, the market’s answer was far more complex than the simple logic of “rate cuts benefit risk assets.”

I. Dow Hits New High, Chip Stocks Crash

As of the close on July 2, the three major US stock indices showed completely opposite trends: the Dow Jones Industrial Average surged 1.14%, hitting a new historical closing high; the S&P 500 Index barely moved, rising only 0.01 points; the Nasdaq Composite Index fell 0.8%, with growth stocks under significant pressure. The divergence between the Dow and Nasdaq reflects a vigorous sector rotation. Funds collectively withdrew from AI chips, memory, and optical communication sectors, which had soared in the first half of the year, and flowed into traditional defensive stocks like McDonald’s, Coca-Cola, and Johnson & Johnson, as well as the defense and gold sectors.

Specifically: The AI hardware sector experienced a full-scale collapse. The Philadelphia Semiconductor Index plunged 5.44% in a single day, with not a single one of its 30 constituent stocks closing in positive territory. Teradyne plummeted 13%, Lam Research fell over 10%, and Applied Materials and GlobalFoundries both dropped more than 7%. The memory sector was not spared – SanDisk plunged 14%, Seagate and Western Digital fell nearly 10%, and Micron dropped 5.7%. The optical communication sector retreated across the board, with MaxLinear down 17%, Applied Optoelectronics down 12%, and Corning, Lumentum, and Marvell Technology generally falling more than 9%. Goldman Sachs’ tracked “AI beneficiary stock” portfolio fell 16% over two days, marking its worst performance on record.

Traditional defensive sectors strengthened across the board. McDonald’s surged over 4%, Coca-Cola and Johnson & Johnson rose over 3%; pharmaceutical stocks AbbVie rose nearly 4%, and Merck rose over 3%; defense stocks collectively rallied, with Lockheed Martin up 4%, and Boeing and Raytheon Technologies up over 3%; gold mining stocks Newmont Mining and Barrick Gold rose over 4%.

This was not an ordinary portfolio adjustment. Bank of America Securities’ latest fund flow report showed that US stocks recorded their first net outflow in 13 weeks ($8.5 billion), with the technology sector seeing a record $9.3 billion in weekly net outflows. Although the BofA Bull & Bear Indicator slightly retreated from 9.2 to 9.1, it remained in the extreme optimism range, with the “sell signal” continuing to be effective. REITs recorded their largest weekly net inflow since March 2024, and infrastructure attracted $1.5 billion in fund inflows. The bond market has recorded net inflows for 61 consecutive weeks. Anshul Sharma, Chief Investment Officer at Savvy Wealth, stated directly: “Funds are flowing out of these hot sectors of the past few months and rotating into other areas, but this also implies a revaluation of the AI trade itself by the market.”

II. Resonance of Triple Negative Factors, AI Narrative Faces “Stress Test”

The sector divergence on July 2 was not triggered by a single event but by the resonance of three negative factors. First: Meta “disrupts,” challenging the logic of compute power leasing. On July 1, market rumors emerged that Meta plans to commercialize its idle AI computing power, selling model access and cloud computing services externally. Previously, the market generally considered AI computing power to be a “scarce resource,” available only from a few cloud vendors. Meta’s entry means that computing power supply may significantly increase, diluting pricing power. Although JPMorgan believes this adds significant value to Meta’s AI business, it also points out that the market is re-evaluating the scarcity premium of computing power.

Second: AI companies are starting to “calculate.” Anthropic is reportedly in discussions with Samsung to develop its own AI chips and is considering mass production using a 2-nanometer process. The market quickly realized that AI companies are starting to think about how to improve the return on investment for every dollar of capital expenditure, rather than just continuing to buy more GPUs. Once the shift from “competing on investment” to “competing on efficiency” becomes an industry consensus, the valuation logic for the entire AI hardware industry chain will be rewritten. BNP Paribas analysts noted that market focus is shifting from “Token Maxxing” to “Token Optimization.”

Third: Non-farm data is just a “catalyst,” not the “cause.” The non-farm data itself is positive – reduced rate hike expectations, eased liquidity pressure. However, the market’s choice to sell AI chip stocks amidst “good news” indicates that the real problem is not macroeconomic, but internal to the industry. JPMorgan warned in a report released on July 3 that a dangerous signal, reminiscent of the dot-com bubble burst, is appearing in the US stock market: AI hardware stocks have soared, while tech giants making massive investments in AI capital expenditure have clearly fallen behind.

III. The First Half Was Too Strong, a Correction Was Inevitable

Looking back, this adjustment was not unexpected. In the first half of 2026, the Nasdaq 100 Index’s approximately 20% gain was contributed almost entirely by 10 stocks. Memory chip manufacturer Micron Technology became the biggest winner, with its stock price soaring 4 times, contributing 26% of the Nasdaq 100 Index’s gains and 17% of the S&P 500 Index’s gains from a single stock. The Philadelphia Semiconductor Index surged 81% in the entire second quarter, marking its largest quarterly gain in history.

But the feast was not shared by everyone. AI chip leader NVIDIA’s year-to-date gain was only about 7%, ranking last among semiconductor index constituents. The “Magnificent Seven” (Mag7) also showed severe internal divergence – Microsoft led the decline with over 22% drop, and Tesla and Meta experienced varying degrees of decline. Goldman Sachs derivatives experts pointed out that investors are underweighting US tech stocks, especially the “Mag7.” BofA clients have sold tech stocks for four consecutive weeks, with individual stocks recording $9.9 billion in outflows, the fourth-largest weekly withdrawal since 2008. When a few stocks contribute the majority of an index’s gains, any pullback in these stocks will send chills throughout the entire market.

IV. How Professional Investors Should Respond? — From “Sector Bets” to “Portfolio Management”

The core contradiction in the current market is: rate cut expectations are positive, but structural changes are occurring within the AI industry — compute power is moving from “extreme scarcity” to “efficiency optimization,” and AI companies are shifting from “buying GPUs at any cost” to “meticulously calculating returns.” For professional investors, this means:

First, concentration risk needs active management. The excess returns in the AI sector in the first half were accompanied by extreme trading congestion. When sector rotation occurs, concentrated holdings in a single sector face significant volatility shocks. How to retain long-term AI allocation while managing short-term volatility through tools is one of the core issues of current portfolio management.

Second, the allocation value of the Hong Kong stock market is worth attention. Amidst the sharp volatility in the US AI sector, Hong Kong stocks have shown a certain degree of “resilience.” On July 2, the Hang Seng Index held above 23,000 points, rising 0.76%. Although southbound funds showed net outflows, stocks like UBTech, Zhipu AI, and Alibaba still saw significant net purchases. For professional investors, participating in Hong Kong IPO anchor or cornerstone placements – exchanging liquidity for discounts, and time for space – offers an allocation path distinct from direct secondary market purchases in the current market environment.

Third, volatility itself is when risk management tools come into play. In an environment of sharp fluctuations, customized tools such as OTC options for individual stocks or indices can help investors manage downside risk while retaining upside potential. These tools have a high threshold and are typically only available to professional investors, but in extreme market conditions, they are among the few means for fine-grained risk management.

V. Conclusion

The June non-farm payroll report provided market participants with some relief – rate hikes are not as imminent. However, the US stock market’s response was far more complex than the simple logic of “bad news is good news”: the Dow hit a new high, while chip stocks crashed. This is not simply “bad news is good news” at play, but rather funds voting with their feet and redefining the winners and losers of the AI era.

The core engine driving the US stock market in the first half – AI hardware – is undergoing a collective valuation reassessment. Traditional defensive sectors, after a long period of being overlooked, have regained favor with investors. How long this rotation will last depends on whether the fundamentals of the AI industry can deliver a convincing performance in the second half. For professional investors, in this environment where “everything is expensive and everything is volatile,” what is truly scarce is not the ability to judge bull and bear markets, but a set of tools that can span markets and cover multiple asset classes – from Hong Kong and US stock trading to IPO placements, from OTC derivatives hedging to flexible allocation of underlying assets. These tools collectively form an asset allocation system that can be finely managed even in turbulent cycles.

This article is for reference by professional investors only and does not constitute any investment advice. The market is risky, and investment requires caution.

[De Shang Qi Dian Technology]