After the U.S. midterm elections, Wall Street began to “anti-AI”

This increase may seem counterintuitive at first glance. Cold water was poured on AI, but why did technology giants surge in value?

The reason is that after the release of the third quarterly reports of Google and Meta, the signal that free cash flow turned negative had triggered investor selling. This time, investors are betting that the Democratic-led Congress will curb the huge capital expenditures of large technology companies on AI infrastructure and help these companies “rescue” their cash flow.

If infrastructure expansion is blocked and computing power supply tightens, computing power prices are expected to rise; at the same time, the pressure on capital expenditures will be reduced, which will help ultra-large-scale cloud service providers with existing infrastructure improve their profitability.

But the same AI has two fates.

On the first trading day after the election, the AI ​​computing power chain was voted with feet by funds. CoreWeave plummeted 9.90%; Micron Technology fell 7.00%; Nvidia fell 3.17%; Oracle fell 1.95%.

Let’s break down these four. CoreWeave is an AI computing power leasing company – it barely builds its own computing power and relies on borrowed money to buy NVIDIA GPUs and then rent them out. It is the link with the highest leverage and the most sensitive to capital expenditures in the AI ​​infrastructure chain; Micron is the main player in HBM and storage, and is a direct beneficiary of AI memory demand; NVIDIA is the leader in GPUs and is a “shovel seller” in AI; Oracle is betting the future of its cloud business on AI computing power.

By concentrating on selling them, the market is pricing in a “deceleration in AI capital spending.”

Miaotou believes that as AI capital expenditures return to accounting mode, the AI ​​​​track will usher in differentiation: highly leveraged computing power lessors and data center facility chains will be under the most obvious pressure; cloud giant manufacturers will benefit from reduced capital expenditures and expected cash flow to become positive.

2. If American AI cools down, will domestic substitution win?

What does the possible slowdown in U.S. data center construction mean for Chinese technology companies?

Many investors may first think of國產替代。

After all, the United States’ restrictions on China’s advanced chips and computing power services continue. With the midterm elections over, this policy uncertainty has not gone away.

Recently, there was a case worthy of attention.

According to the Financial Times, Oracle canceled an approximately US$7 billion AI computing power leasing deal with a cloud vendor under pressure from the U.S. government. According to reports, the cloud vendor originally planned to obtain the computing power services of approximately 100,000 advanced AI chips through Oracle’s data centers in Southeast Asia.

This incident shows that non-U.S. technology companies also face policy risks when they obtain advanced computing power through overseas cloud services. In the past, even if large domestic model companies were unable to directly purchase some advanced GPUs, they could still consider leasing computing power through overseas cloud vendors.

But if this path becomes increasingly unstable, domestic companies will need to pay more attention to the supply of local computing power. From this perspective, the importance of domestic GPUs and their supporting software and hardware ecosystem may further increase.

Semiconductor equipment also deserves attention。

In April 2026, the House of Representatives and the Senate simultaneously advanced the MATCH Act, which plans to extend the scope of control from EUV lithography machines to all DUV immersion lithography machines and other equipment. Many domestic manufacturers have been listed as key regulatory facilities, and equipment exports and after-sales maintenance must apply for separate licenses. The chairman of the House of Representatives Special Committee on China proposed the SCALE bill to establish a set of “objective standards” for the export of advanced chips to China.

From Trump to Biden and back to Trump, the containment of China’s science and technology has been stepped up step by step and has never been relaxed. Now that the Democratic Party has won Congress, the legislative channel has become smoother—controls that used to rely on executive orders can now be institutionalized and long-term through legislative procedures.

Judging from the trend, domestic AI development will be further decoupled from the GPU leasing services of major overseas manufacturers; coupled with the cooling of overseas AI production expansion expectations and the rise in computing power prices, the replacement rate of domestic computing power will continue to increase. Domestic semiconductor equipment is the core part of independent control, and funds are also flowing back to the science and technology innovation and equipment sectors.

Miao Tou believes that,The long-term restrictions on China’s science and technology make domestic computing power and semiconductor equipment still an independent alternative direction worthy of focused research. However, policies can only create alternative demand. What truly determines the investment value of enterprises is technological breakthroughs, commercialization progress and profit realization.。

Moreover, domestic substitution is not the only variable facing Chinese technology companies.

In the past few years, many Chinese technology companies have also enjoyed the dividends of the expansion of US AI investment.

In particular, the performance growth of optical modules, some servers and other AI hardware companies is closely related to the capital expenditures of US technology giants.

U.S. cloud vendors will need to purchase more GPUs, optical modules, storage chips and related equipment to build more data centers. Chinese companies that enter these supply chains also gain opportunities for revenue and profit growth.

But now, as U.S. data center construction faces stricter policy scrutiny, the market is beginning to reassess the possibility of continued rapid growth in U.S. AI capital expenditures.

At present, this is more reflected in the market’s concerns about the pace of future investment, rather than the fact that US technology giants have comprehensively reduced AI investment.

If the construction of U.S. data centers does slow down in the future, the order growth of relevant Chinese suppliers may be affected.。

Therefore, for Chinese technology stocks, after the U.S. midterm elections, what is more worthy of attention is not whether the entire technology sector will benefit, but whether the growth of enterprises comes from domestic independent substitution, or whether they rely on the continued expansion of overseas AI capital expenditures. The former may receive stronger demand support, while the latter needs to be wary of the risk of slowing growth in overseas orders.

結論

So, for the technology sector, what exactly are the capital markets trading?

We believe that there are two main logics.

一是Deals cool US AI arms race. The market expects that data center construction may slow down, and technology giants are expected to ease capital expenditure and cash flow pressures. The computing power industry chain, which is highly dependent on the expansion of AI investment in North America, may face the risk of slowing order growth.

二是Long-term restrictions on technology transactions with China. The two parties in the United States have a strong consensus on restricting China’s access to advanced semiconductor technology. Independent alternatives such as domestically produced GPUs and semiconductor equipment are expected to receive stronger demand support.

Overall, the results of this election are expected to be good for domestic semiconductor substitution in the short term, but bad for the computing power industry chain that relies on the continued expansion of North American AI capital expenditures.

But investors should also be wary, as election trends tend to come and go quickly. There are still uncertainties in the regulation of U.S. data centers from political statements to policy implementation. How much AI capital expenditures will slow down, and whether domestic replacement demand can truly be converted into corporate profits, all need further verification.

Related Reading

• Nvidia live updates: Jensen Huang ousted, CFO Colette Kress to serve as interim CEO2026-11-13
• Are mangoes not afraid of red fruits?2026-11-12
• Original: After Huang Renxun, who is most likely to take over Nvidia?2026-11-12
• Samsung’s roll-screen mobile phone is exposed, creating a new form ahead of Huawei and Apple, and folding screens are expected to be eliminated2026-11-12
• Musk’s favorite car is going to be discontinued.2026-11-12