Wall Street frenzy, lines drawn in Washington: Anthropic's eve of IPO

Xiaguang AI Laboratory

2026.11.20 08:37

· From Shandong

Full text 3878 words

Text | Xiaguangshe, author|Vinco, editor|Li Xiaotian

On November 18, local time, Anthropic announced that it would price its IPO Class A ordinary shares at US$83 per share, valuing the company at close to US$1.5 trillion, exceeding the previous target of US$1.4 trillion. Tonight, Beijing time, the stock will begin trading on Nasdaq under the symbol "ANT".

Two days ago, on November 16, the U.S. Department of Commerce’s Bureau of Industry and Security (BIS) issued new regulations that restrict overseas data centers from providing computing power to relevant Chinese entities and extend advanced chip export controls to four Southeast Asian countries and Brazil.

On the one hand, nearly 500 billion US dollars in subscription intentions have pushed Anthropic's issue price from the range of 70-78 US dollars to 83 US dollars. On the other hand, Washington is redrawing the licensing boundaries of advanced computing power. There is no public evidence to prove the existence of direct cause and effect between the two incidents, but their co-occurrence just constitutes the real aspect of Anthropic on the eve of its listing: At a time when the boundaries of computing power are being redrawn, why is the capital market pricing?

500 billion in headwinds: Why is Wall Street crazy about Anthropic?

Anthropic's pricing exceeds previously set caps. The company originally planned to issue 825 million Class A common shares at US$70 to US$78 per share, raising approximately US$64.4 billion based on the upper limit of the range. According to Anthropic's latest announcement, the final price is US$83 per share, approximately 6.4% higher than the upper limit (US$78); the issuance scale has been simultaneously expanded to 980 million Class A ordinary shares, of which Anthropic sold 956,351,354 shares and the selling shareholders sold 23,648,646 shares. Anthropic will not receive any income from the shares sold by the selling shareholders.

Based on 980 million shares, approximately US$81.3 billion was raised, making it one of the largest IPOs in history. The underwriters also received a 30-day over-allotment option to purchase up to an additional 147 million shares of Class A common stock.

It is reported that when Anthropic closed its order book, institutional investors' subscription intentions were approximately US$400 billion, and individual investor demand was close to US$100 billion, totaling close to US$500 billion, far exceeding the issuance scale.

For comparison, SpaceX raised US$75 billion when it went public in June this year. After exercising its over-allotment option, it raised a total of approximately US$86.25 billion, setting an IPO record at the time. Its subscription demand exceeded US$250 billion, which was oversubscribed by about 3 to 4 times, and it has always maintained a fixed issue price of US$135. By the same caliber, Anthropic's basic fundraising amount has exceeded SpaceX; more importantly, its subscription demand is close to 500 billion US dollars, which is oversubscribed by about 6 times. It has exceeded the upper limit of the pricing range and expanded the issuance scale. The intensity of the "AI asset allocation battle" can be seen.

This is not just IPO fever. After the AI ​​industry has experienced a series of recent shocks—including policy uncertainty brought about by the U.S. midterm elections, regulatory and public opinion turmoil encountered by Nvidia, and market concerns that rising interest rates may squeeze the debt repayment capabilities of data center developers and cloud service providers—Anthropic is still able to attract nearly $500 billion in subscription intentions, indicating that the public market's demand for allocation of leading AI assets remains strong. What investors are buying is a bet on a "historic moment for the artificial intelligence industry."

Looking at this release in a longer timeline, the meaning becomes clearer. From Saudi Aramco to SpaceX to Anthropic, the record for the world's largest IPO has changed hands twice in seven years. Behind it is the migration of core productivity: the objects of capital bets have changed from oil in the ground, to satellite networks in the sky, to large models in data centers. Intelligence is becoming the "oil" of the new era. The valuation logic has changed accordingly. In the past, mature giants were priced based on profits and dividends, but now they are paying in advance for long-term growth and technological paradigms. At the same time, building rockets, buying chips, and building computing power have made technology companies "heavy" again. The primary market has been unable to handle such huge capital needs. The open market has become an inevitable choice, and records have been repeatedly refreshed.

But records do not equal value. Whether this round of AI assets can convert subscription enthusiasm into sustained revenue and profits will still be answered by the financial reports in the next few years.

Investors as suppliers: The circular game of AI giants

After this publicized listing, where will Anthropic go? Its latest prospectus may give us a glimpse.

From the prospectus, what we can see is that an AI giant with both rapid growth and huge losses carries all the market’s expectations and enthusiasm for AI.

In the first nine months ended September 30, Anthropic achieved revenue of US$33.24 billion, an increase of approximately 1,200% over the same period last year; a net loss of US$104 billion, most of which came from changes in the fair value of convertible notes and derivative liabilities.

According to the prospectus update document disclosed by the company, as of October 31, Anthropic’s updated ARR (Run-Rate Revenue) rose to US$81 billion, an increase of US$10 billion from the end of September (this indicator is annualized based on recent consumption and subscription activities, and is not equivalent to the full-year realized revenue or revenue forecast). At the business level, the Claude series is mainly aimed at enterprise customers. The widespread popularity of the AI ​​programming tool Claude Code will drive explosive growth in 2026. Anthropic has also been competing for competitive advantages with OpenAI and Google in the enterprise market.

Currently, Anthropic's issue price of $83 corresponds to a fully diluted valuation of approximately $1.5 trillion. But there is another layer of capital binding behind the valuation. The equity value of Anthropic held by Amazon and Google exceeds US$100 billion; technology companies such as Google, Microsoft, NVIDIA, and AMD have agreed to invest a total of US$26 billion in Anthropic through private placements. At the same time, Anthropic itself also purchases high-priced computing power equipment from Nvidia, Google, Amazon, and AMD.

Are these companies Anthropic's opponents, shareholders, or suppliers? The answer is: both. The interlocking chain of interests allows Silicon Valley giants to form mutually beneficial industrial allies in competition.

Let’s look at industrial relations first. NVIDIA and AMD are both core chip suppliers and strategic investors; Google and Amazon not only provide TPU, Trainium chips and cloud infrastructure, but also hold large equity stakes, and compete directly with Anthropic in the model and cloud AI service market; Microsoft is both an investor and participates in large model ecological competition through OpenAI.

Capital injection, computing power procurement, cloud services, channel distribution and business competition are all tied into the same industrial network. Investors obtain long-term business relationships through equity, and Anthropic exchanges for chip supply, computing power scheduling, financing credit and ecological entrance. Capital here is not just a financial return tool, but also a strategic means to lock in computing resources, consolidate the supply chain, and expand ecological influence.

This structure presents distinct ecological closed-loop characteristics, which not only enhances the certainty of computing power supply and industry chain coordination, but also puts Anthropic’s bargaining power and conflicts of interest under greater pressure.

Capital is the bright line, geography is the hidden line: the other side of Anthropic’s listing

Capital, computing power and competition form the main narrative of the rivalry between Silicon Valley giants; and once this narrative is placed into the geographical coordinates of the Sino-US AI race, Anthropic’s story takes on an intriguing meaning.

On November 16, the U.S. Department of Commerce’s Bureau of Industry and Security issued new regulations “Controlling Remote Access to Advanced Semiconductors”, which included “remote access to computing power” under control: as long as overseas data centers provide computing power to relevant Chinese entities in excess of 12 million TPP – approximately 7 50 Nvidia H200s - you must apply for a license; data centers that have served relevant Chinese entities have a 90-day window to apply, except for commercial recommendation systems; the new regulations also include four Southeast Asian countries and Brazil in the scope of advanced chip export licenses, and tighten OEM compliance inspections.

It is worth noting that 750 cards hardly poses a direct constraint on the training of 10,000-card-level cutting-edge models. Although Anthropic, as a U.S. company, will not be directly impacted by this rule in the short term, the signal significance behind it cannot be ignored: the regulatory paradigm is expanding from "restricting chip exports" to "managing access to computing power." What this shift has changed is the compliance cost of global computing power, supply expectations, and the way investors price policy risks, and Anthropic is a heavy computing power purchaser.

According to media disclosures, Southeast Asia has been a key area for Chinese-funded computing power to go global in the past few years: Among the approved data center projects in Thailand, the proportion of Chinese-funded backgrounds or mainly serving customers of large Chinese companies is said to be as high as 60 to 70%; in the computing power hub of Johor, Malaysia, Chinese cloud manufacturers and Chinese-funded third-party data centers are also the main players. Malaysia has expressed its intention to legislate to require data centers to disclose operators, electricity consumption and GPU import quantities. These rules do not directly target Anthropic, but they have actually increased the compliance costs of the global computing power market.

Nvidia-related turmoil is a manifestation of risks in the same context.

The resignation of Nvidia CEO Jensen Huang last week triggered market concerns about the stability of core links in the AI ​​supply chain, and also brought the political sensitivity of the computing industry to the forefront. Judging from the disclosed IPO information, Anthropic’s computing power strategy itself is leaning towards “eliminating single reliance”. The company plans to integrate four types of chip resources: Amazon Trainium, Google TPU, Nvidia GPU and AMD GPU into a unified and flexibly allocable computing power pool, and use software technology to allocate tasks based on the performance advantages of each chip in different scenarios.

This layout is interpreted by some analysts as a means for large AI companies to enhance their bargaining power and reduce the risk of supply fluctuations; however, the price is that the cost structure will be more affected by the price of multi-source computing power and the compliance environment, and NVIDIA's advantages in performance, network interconnection, and software ecology will still be difficult to completely replace in the short term.

Therefore, Anthropic’s issuance price of US$83 is a high-priced vote for the capital market’s top AI assets. Nearly US$500 billion in subscription intentions and a valuation approaching US$1.5 trillion jointly show that the public market is still willing to pay a premium for the AI ​​growth story. On the eve of the listing, Dario Amodei mentioned in an interview the restraint of "not letting a company's valuation be even 1 dollar higher than its actual value." This also added an extra layer of tension to the IPO.

But the geopolitics of computing power offers another line. Chip export licenses, overseas data center computing power caps in China, tape-out compliance inspections, and four congressional bills are turning computing power from a commercial cost into a strategic variable. Anthropic’s valuation narrative is based on computing power procurement, model training and enterprise customer growth.

The market will give new answers next. The soon-to-be-announced trading performance on the first day of listing and whether the underwriters exercise their over-allotment option are short-term observations; the longer-term observation point is whether Anthropic can efficiently transform huge computing power expenditures into sustained growth in revenue within the boundaries of computing power geopolitics and supply chain.

Respect for the original work, if you are qualified, you should be appreciated.

According to the real-name requirement of the Cybersecurity Law, please bind your mobile phone number before posting a comment.

Express

08:38

CITIC Securities: The value of real estate-related equity assets may bottom out and rise in 2027

08:37

Financial Street Securities: The company’s chairman is currently under investigation and has been subject to lien measures

Agency: Global artificial intelligence spending is expected to exceed US$4 trillion in 2027

08:35

The Nikkei 225 Index fell 3% on the day

08:25

Tailor-made financing solutions, banks increase support for the construction of "six networks"

08:20

Apple head of camera and photo software departs

08:05

Nvidia closes in on acquiring diffusion AI startup Inception for more than $500 million

08:04

EU approves Hungary's 5.4 billion euro defense loan plan

08:01

Wang Yi introduces the situation of the meeting between the heads of state of China, Russia and the United States

07:59

MSCI will add Anthropic to its large-cap index on December 1

07:39

Anthropic CEO: The company hopes to obtain an investment-grade credit rating in the future

07:36

Anthropic CEO responds to US$900 billion AI investment commitment: It is not large compared to existing cash reserves and revenue

07:34

SpaceX reportedly plans to issue at least $32 billion in AI chip-related debt

07:32

Anthropic rings the Nasdaq opening bell, setting a record for the largest IPO in history

07:20

U.S. stock market closing comments: The three major stock indexes all fell more than 2% after opening high and moving low. Anthropic closed up 17% in its debut.

07:14

Trump says he won't 'rush' into action against Iran before New Year

07:08

The net value of the fund reveals the direction of position adjustment, and institutions explore the certainty of the next stage.

2026-11-19 23:11

Director Zhu Lidan sues Wahaha Commercial Co., Ltd.

2026-11-19 23:02

Domestic commodity futures markets mostly closed higher at night, with black products leading the gains.

2026-11-19 23:01

International Energy Agency: Middle East conflict may reshape global energy landscape

Scan to download App