Source: New Intelligence Report
This morning, countless people's social media circles and groups were flooded with "crash" and "collapse."
People were rushing to tell each other: last night, the US stock market encountered an epic bloodbath.
In just 40 minutes, the market value of the S&P 500 index evaporated by $2.5 trillion!
This is one of the fastest and most brutal moments of wealth loss in the history of the US stock market.
Furthermore, Anthropic's IPO is also in jeopardy!
Now, Anthropic's IPO valuation has dropped from $2 trillion to $1.5 trillion, and it may even be further adjusted down to $1 trillion, or even the rumored $500 billion!
It is simply a "knee-capping."
Worse still, with the US technology stocks evaporating $2.5 trillion overnight, whether Anthropic can go public is now a question.
Epic Collapse:
US Stocks Plummet Across the Board, Big Tech Giants Suffer a "Bloodbath"
Last night on Wall Street, the only way to describe it is "brutal."
All three major US stock indices fell sharply—
The Nasdaq plunged over 1400 points, a decrease of 5.52%.
The S&P 500 evaporated $2.9 trillion in a single day.
The Dow Jones Industrial Average plunged 1833.06 points, with a decline of 3.44%.
The sell-off in tech stocks was like a flood bursting its banks, with the total market value of the "Big Tech Seven" evaporating over $1 trillion!
Specifically for individual stocks, once-illustrious favorites fell from grace.
This morning, the storm crossed the Pacific directly, with South Korea's KOSPI opening down nearly 9%, triggering a circuit breaker, pausing all trading.
Blood-red screens everywhere, mud and sand mixed together.
Just two days ago, the US stock market was celebrating the results of the midterm elections, with the Dow surging 1500 points; merely 24 hours later, the Dow reversed and plunged 1833 points, wiping out all gains in an instant—it's too dramatic.
NVIDIA plummeted 9.26%, evaporating over $440 billion in a single day, nearly reducing the annual increase to zero;
AMD plummeted 15%, Micron fell 17%, SanDisk dropped 22%;
Qualcomm fell nearly 9.5%.
The high cost of storage chips is severely eroding profits.
The two giants of computing power leasing suffered a severe bloodbath: CoreWeave plummeted 10%, down 76% from last year's peak; Nebius dropped 15%, halving 58% from this year's peak, with even energy stocks like GE Vernova and Constellation Energy, responsible for power supply, facing fierce selling.
Tesla, under Elon Musk, plunged nearly 11%, and SpaceX fell over 9%, as concerns about the expansion of AI training centers quickly spread.
AI concept stocks were detonated one by one!
The direct catalyst was the result of the US midterm elections—Democrats secured both the Senate and House. What Wall Street fears the most is that AI infrastructure may be "choked."
It's crucial to know that the reason the US stock market has consistently hit new highs over the past year is due to an all-in approach to "AI computing power chains" and "data center construction."
If this logic is disrupted, the entire industry chain will be overturned!
Anthropic's Shocking Thunder: $895 Billion Debt Black Hole, Valuation Crashes from $2 Trillion to "Knee-Cap"
Originally, Anthropic was exalted by Wall Street, seen as the ballast of global tech stocks in 2026.
A few days ago, the market was still clamoring for its IPO valuation expectations of $2 trillion.
However, the latest disclosed data left major Wall Street investment banks stunned:
Anthropic is burdened with irrevocable commitments for computing power and infrastructure expenditures amounting to as much as $895 billion!
What does this mean?
Among the tech giants in the U.S., apart from Google's parent company Alphabet's $1.06 trillion commitment, Anthropic's bill directly dwarfs Meta ($866 billion), crushing the entire Silicon Valley.
This means that even if the company makes no money tomorrow, it must staunchly pay this exorbitant bill. Among them:
Broadcom: $161 billion;
Google: $121 billion, plus up to $150 billion in AI chip procurement commitments;
SpaceX: $121 billion;
Amazon: at least $110 billion.
Even more deadly is the repayment timetable.
In just the next 12 months, Anthropic must cough up a full $70 billion in cash; by the end of 2030, it must shell out $448 billion.
Can it earn that much money?
Despite a 47% year-over-year increase in revenue in Q3, reaching $17 billion, the bottomless pit of large model R&D and substantial depreciation means its net losses are expanding at astronomical rates.
Annual income in the billions, debts nearing trillions, and a $70 billion repayment due in one year.
Once the news broke, the capital market flipped the table completely.
Investment institutions collectively turned against it, crazily urging reforms.
A few days ago, the prospect of a $2 trillion valuation for its IPO was quickly slashed in half to $1.5 trillion, $1 trillion, and there were even rumors that someone in the private trading was offering a "knee-cap" low price of $500 billion.
When Wall Street found out that even the most elite AI creation models were reliant on "trillion-dollar paper notes," everything collapsed.
Anthropic Faces a Huge Crisis
Can AI models earn back the money burned?
According to data from the prediction market platform Kalshi, the probability of Anthropic going public in 2026 has dropped from nearly 90% before the midterm elections to less than 60%.
A former Morgan Stanley employee and Silicon Valley partner bluntly stated: the drop in valuation is insignificant because this game may already be over.
He pointed out that the entire existence of Anthropic (and similar AI giants) is built on an "extreme asset-liability mismatch" financing model.
This is essentially a "reflexive trap of the confidence game."
Simply put: AI companies need to continuously purchase computing power from super-scale data centers (like Amazon, Google); to pay for it, they must constantly finance and issue new shares at extremely high valuations. This model can only be sustained as long as the stock price keeps soaring.
However, once the IPO pricing makes an error or the stock market environment deteriorates, a death spiral will begin.
If there is no highly valued equity, no money can be borrowed; without credit, the computing power bills cannot be paid; if the computing power bills cannot be paid, the company cannot survive because that nearly trillion-dollar irrevocable commitment is a deadly negative asset mortgage.
"If credit rating agencies don't grant investment-grade bond ratings for a $2 trillion market cap, they absolutely won't give a $1 trillion junk bond rating."
In this situation, Anthropic's smooth IPO faces an extremely grim life-and-death exam.
If it forcefully conducts roadshows, it may encounter the embarrassing situation of no one showing interest; if it delays, the $70 billion bill due next year is about to be overdue.
Black Swan Forces AI Infrastructure to Hit the Brakes
The result of the US midterm elections directly pulled the plug on the global AI computing power supply chain.
The Democrats not only secured both the Senate and the House but also unexpectedly won the gubernatorial election in Texas, the home base for AI computing power.
The newly elected governor clearly stated during the campaign: a temporary ban on high-energy-consuming data centers will be implemented, and tax breaks for data center operators will be abolished.
In the past two years, global tech giants have been building factories here, with Google's, Meta's, Oracle's, and OpenAI's super computing centers clustered together. As AI models grow bigger, their electricity consumption is becoming increasingly alarming.
Now, the reckoning has arrived.
The market has suddenly realized: the biggest bottleneck for AI is not algorithms, but electricity; and the biggest obstacle to electricity is policy.
The reason the US stock market can hit historic highs this year is due to the unlimited investment expectations around "giants crazily buying chips, purchasing electricity, and building data centers." Now expectations have been directly halved, leading to a violent stomp across the entire AI infrastructure chain.
This time, the market is not cutting "with shovels," but specifically targeting "selling shovels."
In the past, capital believed that as long as chips could be made and data centers could be built, returns would eventually materialize. Now, with the giants burdened by nearly a trillion dollars in bills and data center approvals being policy-blocked, Wall Street has begun to tally up:
These hundreds of billions slammed into the ground, apart from flashy PPTs, when will genuine commercial returns be seen?
The blood-red screens everywhere are the cruel reply Wall Street is giving.
Proposed Permanent Ban on "Super Intelligence," Violators Face Up to 20 Years
Moreover, the US may comprehensively prohibit super intelligence, with violators facing up to 20 years!
Less than 24 hours after the election results were released, Arizona Democratic Representative Yassamin Ansari publicly declared war on the entire AI industry:
The top agenda for Democrats on their first day in power should be to pause advanced AI development and completely ban super intelligence.
This tweet quickly surged past 14 million views, and Elon Musk immediately retweeted it, mocking: "It's a beautiful new world."
Unexpectedly, with a new majority in Congress, Ansari directly confronted Musk, leaving a fiery warning: "See you in Congress soon."
The panic in Silicon Valley and Wall Street is due to the fact that Ansari is not merely making empty threats.
The "Ban on Super Artificial Intelligence Act" (S.5493/H.R.10538), which she co-sponsored, has long moved beyond mere proposals.
As early as this September, this act was officially submitted to Congress by Senator Sanders and Representative Casar.
Looking through this legislative text, it contains three iron-fisted clauses that could uproot Silicon Valley:
The first blow superficially governs the "future," but actually cuts into the "present."
The act stipulates that any training computing power exceeding 10²⁵ floating-point operations triggers a "mandatory pause order"—training, fine-tuning, or even launching on the market is strictly prohibited without approval from the newly established federal "AI department."
It's important to note that as early as three years ago, the first generation of GPT-4 had already crossed this threshold!
This means that the next-generation flagship models that are currently running in various Silicon Valley labs are all already treading on the illegal line from the moment the data center powers on.
The second blow completely seals the ceiling on "super intelligence."
As long as AI’s cognitive abilities surpass human levels in most domains, or there exists a risk of losing control, it will be permanently classified as a no-go zone—cutting off the entire chain from coding, investment, to open-source deployment.
The third blow treats large model development as "a felony."
Once the line is crossed, involved tech companies will be directly ordered by the federal government to disband; engineers and executives may face up to 20 years of federal imprisonment!
The entire X exploded in an instant.
Prominent tech investor Jen Zhu commented, "Why not just ban linear algebra, probability statistics, and vector calculus as well!"
Wall Street traders were furious, criticizing politicians for single-handedly crashing the entire US stock market with a single tweet.
Macro institutions also warned that hitting the brakes now is tantamount to "self-destruction."
For the capital market, this act directly cuts off the source of demand for the entire AI hardware supply chain.
If next-generation large models cannot be trained, then there is no need for so many GPUs.
If there is no need for GPUs, then there is no need for so much HBM and storage; if chips are not bought, then there is naturally no need to build data centers.
It's crucial to recognize that AI is the only reason the US GDP can still maintain positive growth.
Reference Material:
https://finance.yahoo.com/news/anthropic-ipo-odds-drop-compute-commitments
https://www.bloomberg.com/news/articles/anthropic-valuation-tumbles-pre-ipo
https://x.com/i/trending/2118481444595236976
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到[email protected]，本平台相关工作人员将会进行核查。
