Goldman Sachs strategist: The trading sector of AI has been "sold to the extreme," but there are still positive aspects after the Democratic Party's big win
The Block
45m ago
Ai Focus
Rich Pivrotzky, head of equity trading in Europe at Goldman Sachs, stated that after a round of selling, the valuation of AI related assets has significantly corrected, and there is a need for hedge positions to be covered. He believes that although a Democratic victory could lead to stricter regulation of data centers in some regions, companies may accelerate their capital expenditures and transactions as a result, and the value of already established computing power assets could also increase.
Helpful
No.Help

After a round of selling and the implementation of elections, the position structure has been cleared, and bears need to cover their positions. The resilience of the growth sector still exists.

In recent weeks, the market sentiment towards the AI sector has clearly turned negative. The results of the mid-term elections have led some investors to believe that the worst-case scenario they feared for this sector has now materialized.

However, a strategist at Goldman Sachs holds a different view. The valuations of related targets have already corrected, and there is a need to close short positions; the final outcome of the election may actually accelerate capital expenditure rather than suppress it.

Rich Priwrotzki, the head of equity trading in Goldman Sachs Europe ( Rich Privorotsky ), believes that the AI sector has been “sold to extreme”. His view for fund managers is that the fundamentals are still strong, the uncertainty surrounding elections has subsided, and the portfolio composition is much clearer now.

Before the voting on Tuesday, the question that investors were most concerned about was whether a Democratic victory would curb the wave of data center expansion. The market paid particular attention to the Texas governor election, as approximately a quarter of all data center planning projects in the United States are located in that state. The defeat of the current Republican governor, Greg Abbott, to Gina Inohosa means that the local regulatory environment may become stricter.

However, Privorotsky's logic is as follows: Corporate decision-makers may decide to advance their capital investment to 2027, taking advantage of the current approval processes before any adjustments to regulatory frameworks are made to complete strategic transactions. By 2029, when a new president takes office, tax policies may change, and administrative regulations as well as merger and acquisition reviews are likely to become stricter. Major hyperscale cloud providers will choose to act swiftly to implement their plans.

He also put forward a view: if the expansion of computing power in the future is restricted, the value of the existing computing power that has already been built and put into operation will increase accordingly. Theoretically, this will be beneficial to large-scale cloud providers such as Amazon, Microsoft, and Alphabet which have already deployed a large amount of computing power.

In his trading notes sent to clients on Wednesday, Privorotsky also acknowledged that delays or suspensions of some data center projects would drag down what investors refer to as the "shovel-selling" sector, which includes manufacturers of chips, cooling systems, power equipment, and energy service providers.

However, these sectors have already experienced corrections before. For example, the Philadelphia Semiconductor Index ETF ( SOX ) has fallen by 20% in the past month.

This research report uses AMD's impressive third-quarter report as evidence that the current AI capital expenditure cycle still possesses resilience. The guidance for the fourth quarter is very strong, with an expected year-on-year growth rate increase of another 27%. Privorotsky believes that limited expansion on the supply side may not eliminate demand; instead, it could lead to higher prices lasting longer.

If the construction of data centers is delayed, large-scale cloud providers may experience a situation where orders exceed demand. However, Privorotsky believes that this has already been reflected in the currently low valuations. For emerging cloud providers such as Nebius and CoreWeave, the investment conclusions are more complex and depend to a large extent on whether long-term contracts for computing power have been signed and whether the computing power has already been established.

Finally, Privorotsky believes that once the elections are held and uncertainty decreases, it will be difficult for high volatility, represented by the panic index VIX, to continue. If volatility falls rapidly as he predicts, and since volatility is negatively correlated with stock market indices, then the overall market index is expected to rise; AI and AI related stocks account for nearly half of the weight in these indices.

Credit attribution:Global Market Report cites trading notes sent to clients by Rich Priwrotzki, the head of Goldman Sachs' European equity trading department.

Tip
$0
Like
0
Save
0
Views 24
WalletJYS reminds readers to view blockchain rationally, stay aware of risks, and beware of virtual token issuance and speculation. All content on this site represents market information or related viewpoints only and does not constitute any form of investment advice. If you find sensitive content, please click“Report”,and we will handle it promptly。
Submit
Comment 0
Hot
Latest
No comments yet. Be the first!
Related
Tesla's discontinued Model S exterior design patent approved, with protection period extending to 2041
Tesla has been granted a design patent that closely matches Model S, with protection extending until 2041. The article states that this patent was approved several months after the production of the model was discontinued. Tesla had previously stopped accepting new orders for Model S and Model X, and has launched special editions of Signature Edition for these two models.
The Block
·2026-11-05 16:37:17
8
Baidu Wenxin releases the open-source ERNIE 5.5 Flash Base high-efficiency base model, with a total of 280B parameters
Baidu Wenxin Large Model releases ERNIE-5.5-Flash-Base, making available the pre-trained weights from ERNIE up to 5.5 times Flash. This model is a sparse MoE language model with a total of 280B parameters, and each token activation has 14B parameters. It was pre-trained from scratch on 33.5T of tokens memory, with a maximum training context of 128K. Fine-tuning and inference are supported with a context size of 1M tokens.
The Block
·2026-11-05 16:37:15
6
Microsoft Successively Cancels Nscale: First Cuts a $3 Billion Contract in the UK, Then Abandons an Expansion Order in Texas
British startup Nscale faces another cancellation by Microsoft on the eve of its IPO push. Following the cancellation of a $3 billion lease in the UK at the beginning of the year, Microsoft has now also abandoned the option to expand its data center in Texas. Nscale claims that its backlog of $103 billion in contracts is not affected, but the company still faces pressures related to financing, construction delays, and IPO.
Wallstreetcn
·2026-11-05 16:29:44
13
Asian AI stocks plummet, losing $550 billion; Bitcoin rises 4%
Asian stock markets faced large-scale selling on Thursday, with South Korea, Taiwan, and Japan collectively losing nearly $550 billion in market value, with semiconductor stocks leading the decline. Investors are concerned that after the Democratic Party in the United States wins the mid-term elections, it may tighten regulations on AI companies and data centers, thereby slowing down AI spending. Meanwhile, Bitcoin rose by more than 4%, reaching around $81,000 at one point.
Coinpedia
·2026-11-05 16:17:39
25
CZ Warning: Trust Wallet Users, Beware of Fake Websites! There Are Also Rumors of a Vulnerability Worth Approximately $1.1 Million
Chao Changpeng, co-founder of Binance, warns users about a phishing website that impersonates Trust Wallet, which induces visitors to connect to encrypted wallets. Meanwhile, another report suggests that a vulnerability in devices running iOS 18.5 may have resulted in losses of approximately $1.1 million.
crypto.news
·2026-11-05 16:17:36
22
View More