From 20 trillion expectations to sprinting towards 15 trillion: Anthropic IPO collides with CPI exceeding expectations, US Treasury yields nearing 5%, AI pricing maturing

On the evening of November 10, Beijing time, the U.S. October CPI rose 0.4% month-on-month, and the core CPI rose 0.3% month-on-month, both exceeding market expectations. The yield on the 10-year U.S. Treasury bond closed at 4.998%, approaching the 5% mark. On the same day, Anthropic officially launched its IPO roadshow, with the price range set at $70 to $78. Based on the post-issue share capital, the upper limit of the range corresponds to a market value of approximately $1.02 trillion. Just at the end of October, the market was still discussing a record valuation of $2 trillion.

1. CPI Released, Money Becomes More Expensive

What is truly noteworthy about the October CPI is not how astonishing the 0.4% figure is, but that it interrupts the market's smooth expectations of continued cooling inflation and a shift towards loose policy. The overall CPI rose 0.4% month-on-month, although lower than September's 0.6%, it was higher than the market expectation of 0.3%. The core CPI rose 0.3% month-on-month, also above the expected 0.2%. During the government shutdown in October 2025, no price data was collected, and this report still lacks comparable official year-on-year readings, making the impact of month-on-month pricing more concentrated.

The bond market reacted decisively. The yield on the 10-year U.S. Treasury bond rose from 4.879% the previous trading day to 4.998%, again approaching the 5% mark, and yields on 2 to 5-year bonds also rose significantly. The market then began to discuss the possibility of a significant interest rate hike in December. Regardless of where policy rates ultimately head, the most important change in this round of trading is that the market has begun to take seriously the scenario of "high rates lasting longer."

For traditional value stocks, rising interest rates may not shake the foundation of valuations. However, it is different for companies like Anthropic. Its revenue is growing rapidly today, but a significant portion of its value is built on future years of revenue and cash flow. After interest rates rise, future money discounted to today requires a larger discount, and the market will naturally demand more certain growth or friendlier prices.

This is why the CPI and Anthropic's IPO appearing on the same day is worth looking at together. One is re-establishing the price of money, and the other is re-exploring how much AI assets should sell for.

2. How the $2 Trillion Quote Was Reversed

What is most noteworthy about Anthropic's IPO is not the final issue price itself, but the rapid change in market psychological pricing within less than a month. At the end of October, after the company submitted its S-1, the market briefly discussed a listing valuation of about $2 trillion. By November 3, the reported buy-side discussion range had dropped to $1.5 trillion to $1.6 trillion. After a sharp decline in tech stocks, potential investors continued to reassess, and by around November 8, the discussion range had fallen to $1.1 trillion to $1.3 trillion. On November 10, the company officially announced 825 million Class A shares with an issue range of $70 to $78, which, based on approximately 13.087 billion shares post-issue, corresponds to a market value of about $1.02 trillion at the upper limit of $78, and about $968.4 billion at the median of $74.

The psychological price range of $1.2 trillion to $1.3 trillion from buyers and the underwriters' target of $1.5 trillion is the real negotiation space of this roadshow. Both numbers are calculated based on the post-issue share capital stated in the prospectus, excluding options and restricted stock. According to Reuters, when including employee options, restricted stock, and shares reserved for charitable donation matching plans, the fully diluted valuation at the upper limit of $78 corresponds to about $1.4 trillion, which is still slightly below some investors' expectations. Market participants say that the underwriting team is striving to push the overall valuation towards $1.5 trillion (all reports and statements are from November 10).

Rather than saying Anthropic "dropped from $2 trillion to $1 trillion" in a short time, it is more accurate to say that the market's tolerance for its valuation has undergone several rounds of downward adjustments. The $2 trillion was a high pre-IPO market expectation, while the $1.02 trillion is the upper limit given during the company's official roadshow. The calculations are not entirely the same, but placed on the same timeline, the changes are clear enough.

Using revenue multiples provides a more intuitive view. The annualized revenue of about $81 billion in October corresponds to $1.02 trillion at about 12.6 times. If calculated based on the previous market discussion valuation of $2 trillion, the multiple is close to 25 times. What the market has compressed is precisely this premium. In other words, buyers are not unwilling to give AI a premium, but are doing so more judiciously.

This is also the most valuable signal of this IPO. The market did not raise prices indefinitely due to astonishing revenue growth, but began to reassess revenue growth rates, profit margins, capital investments, and financing needs on the same table. A market that knows how to calculate can nurture a healthy bull market.

3. Revenue Soars, Bills Locked In First

Looking solely at revenue, Anthropic remains a rapidly growing company. In the third quarter, revenue was $16.96 billion, compared to about $1.38 billion in the same period last year, a year-on-year increase of about 11 times. The quarter-on-quarter growth was 47%, and the annualized revenue in October further reached about $81 billion. The demand for Claude has not diminished due to valuation discussions; enterprise clients and developers are still rapidly increasing their usage.

What has made buyers cautious are the profit statements and cost structures. In the second quarter, the company recorded a GAAP operating profit of $974 million, but in the third quarter, it turned into an operating loss of $567 million. Revenue is growing rapidly, but profits have not expanded in tandem, the most direct explanation being that the computational power required for model training and inference is indeed too large.

The key is not the short-term losses, but that a significant portion of costs must be locked in early, while revenue realization requires continuous validation. The prospectus discloses that the company has signed a large number of procurement and leasing commitments, totaling about $900 billion, of which about $447.5 billion will mature before 2030. The short-term computational expenditure in the third quarter was about $5 billion, with cumulative short-term computational contracts of about $24 billion.

In other words, Anthropic is not facing the typical software company's problem of "rapid revenue growth, delayed profits," but rather the unique economics of AI infrastructure. The stronger the model, the greater the demand, and the greater the computational consumption. Revenue growth must consistently outpace computational costs for the scale to roll smoothly. What the market needs to validate is never "Is anyone using Claude?" but whether Claude's commercialization can sustainably outpace the growth of computational bills.

4. How to Read the $900 Billion Commitment

How much can Anthropic raise in this IPO? At the upper limit of $78 for 825 million shares, it amounts to about $64.35 billion, plus about $26 billion from private placements, totaling about $90.4 billion.

Putting the $90.4 billion together with the approximately $900 billion in commitments does not lead to a simple conclusion of "fundraising cannot cover contracts." The terms, payment structures, and sources of funds for these commitments vary and cannot be treated as the same cash expenditure table. What it truly tells the market is the scale of Anthropic's future capital needs. IPO funds are important, but to support years of future computational investments, the company needs continuous revenue realization and, if necessary, to continue entering the capital market.

This is also the biggest difference between Anthropic and ordinary internet IPOs. The question for traditional software companies is whether user growth can translate into revenue. The question for cutting-edge model companies has an additional layer: while revenue is growing rapidly, they must also invest in computational power to maintain product performance, train the next generation of models, and meet customer demands. If revenue continues to grow rapidly, these investments will be absorbed by the expanding business scale. If revenue growth slows while signed commitments remain, pressure will emerge.

Buyers' desired "safety net" is also clear. It is not about whether Anthropic is making money today, but whether it can prove in the coming years that revenue growth can ultimately surpass computational costs and turn into sustainable free cash flow. A company that grows under such market pressure will, in the long run, be more solid.

5. Founders Only Give Up a Small Portion

Another detail of this IPO that can be easily exaggerated by emotions is the very low proportion of old shares sold. Of the 825 million shares issued, about 801 million are new shares, and about 23.65 million old shares are transferred, all from seven co-founders. Based on the median of $74, each founder sells about $250 million worth of shares, totaling about $1.75 billion. According to the shareholding situation disclosed in the prospectus, each person sells about 1.4%, with the vast majority of shares still retained.

This structure has tangible benefits. This is not an IPO primarily aimed at allowing early shareholders to cash out; the vast majority of the raised funds will enter the company, not the pockets of old shareholders. For an AI company that will require significant future investments, this carries substantial weight. However, it should not be interpreted as "founders using real money to prove extreme confidence in $78." They are only selling a small proportion, which can be understood as continuing to bind to the company's long-term value or as exchanging some cash in a private market with limited liquidity while retaining most of the upside potential.

What is truly worth watching is when and at what price over 98% of the shares will enter the market. The short-term issuance structure addresses the current supply, while the medium to long-term valuation will ultimately depend on whether revenue, profits, and cash flow can keep up with expectations.

6. The Pricing Method of AI is Switching, Anthropic Will Affect Pricing Across the Entire Industry Chain

In the past two years, AI assets have been the easiest to command a premium, due to high demand, rapid iteration, and broad space. As long as revenue growth continues to exceed expectations, the market is willing to discount future profits to today.

Now the environment has changed. Interest rates are approaching 5% again, AI infrastructure investments are entering the hundreds of billions of dollars, and model companies' commitments to computational power are measured in hundreds of billions. The market naturally begins to ask more practical questions: Can revenue growth cover capital investments? When can investments turn into free cash flow? Anthropic happens to concentrate these questions in one company. It has strong revenue growth and enormous capital needs. It proves that AI applications have real commercial demand and shows the market that the commercialization of cutting-edge model companies does not mean the capital expenditure cycle has ended.

Therefore, the real change in this IPO is not "AI dropping from $2 trillion to $1 trillion," but rather the shift in AI valuation from "how big can it be in the future" to "how much must be invested to achieve such size, and how much of that investment can ultimately be recouped." This is also the significance of looking at the CPI and Anthropic's IPO together: CPI sets the price of money, and the IPO sets the premium the market is willing to give for growth. One manages costs, while the other needs to prove; both are pushing AI transactions towards a more solid direction.

If Anthropic can successfully complete its issuance and stabilize around one trillion dollars, the signal is not "the end of the AI bubble." On the contrary, this indicates that capital is still willing to give high valuations to cutting-edge AI companies, but with clearer conditions attached: revenue growth must be fast enough, commercialization must continue to advance, and capital investment must ultimately turn into cash flow. If the valuation continues to be adjusted downward during the roadshow, or if there is a significant drop after the listing, the impact will not be limited to just Anthropic. The market will reassess the entire chain related to AI capital expenditures, including GPUs, cloud computing, data centers, Neocloud, and the long-term procurement contracts surrounding these companies.

In the past, the market could continuously raise prices upstream based on "growing AI demand"; when the commercialization of model companies begins to be constrained by funding costs, the growth of upstream orders must also undergo new scrutiny. This does not mean that Nvidia, cloud vendors, or data center assets will decline, but rather that their valuations will need to answer a more practical question in the future: can the cash flow ultimately generated by the AI industry outperform the capital invested across the entire industry chain? Companies that can answer this question well will secure a stronger position in the next round of pricing.

VII. In Conclusion: AI Hasn't Changed, but the Pricing Method Has

The most noteworthy aspect of this IPO is its timing: on one hand, the October CPI exceeded expectations, and the yield on 10-year U.S. Treasury bonds approached 5%. On the other hand, one of the most watched AI companies in the world officially entered the pricing stage. The long-term expectations that could previously be masked by low interest rates and high-risk appetite must now face the scrutiny of higher funding costs.

Therefore, what Anthropic faces is not "whether AI still has a story," but rather how much the AI story can sell for. If the company can secure stable institutional demand around one trillion dollars and continue to prove through revenue growth that computing power investment ultimately translates into profits and cash flow, this round of valuation contraction resembles a healthy repricing. AI has not ended; the market simply demands a thicker margin of safety, and a thicker margin of safety is precisely the foundation for a long-term market to go far.

If the offering price still cannot attract enough buyers, or if it continues to drop after listing, the market will need to reassess not only Anthropic but also the pricing habits established in the entire AI industry chain over the past two years. The long-term logic of AI has not changed due to a single CPI or a single IPO, but the way the capital market prices the AI story is changing. In the past, everyone was more concerned about "how big the future can grow"; now the question to answer is another: how much to invest today for this growth, and when will this money come back? Companies that can answer this question well will be the true protagonists in the next market cycle.

Disclaimer: The content of this article is for general information and market commentary purposes only, compiled based on publicly available information as of the time mentioned in the text. Relevant market data, expectations, and probabilities may change with market conditions. The views and investment strategies cited from third-party institutions, analysts, or other individuals represent the opinions of the relevant third parties at a specific time and do not represent the views or recommendations of BIT. This article does not constitute investment advice, investment research, solicitation, or recommendation for any securities, investment products, or trading strategies, nor should it be used as the basis for any investment decision. Financial markets carry risks, and securities prices and market performance may fluctuate; past performance and historical market trends do not represent or guarantee future results. Investors should independently assess relevant risks based on their own circumstances and seek professional advice when necessary.