[MSX Research Institute · Daily Observation on US Equities RWA] is a flagship daily report by MSX, the leading RWA trading platform. Leveraging our robust macro research capabilities, we capture the core dynamics of global traditional US equities, liquidity shifts, and the RWA tokenization market to help you strategically position yourself in high-quality assets.
Today's Observation
Anthropic, the parent company of Claude, has taken a key step toward its IPO. On October 26, Anthropic publicly filed its S-1 prospectus with the U.S. Securities and Exchange Commission, planning to list on the Nasdaq Global Select Market under the ticker symbol “ANT.” The company had previously submitted its filing confidentially; this public filing marks the first time its financial data, customer structure, compute investments, and governance arrangements are fully disclosed to the public market. Based on the current timeline, the company could complete its listing as early as November, though the final timing remains subject to SEC review, market conditions, and offering arrangements.
The most striking data in the prospectus is the revenue growth rate. In the first half of 2026, Anthropic generated revenue of $16.286 billion, approximately 14 times that of the same period last year. Of this, about $13.064 billion came from usage-based API and model calls, and about $3.222 billion came from subscription services. This structure indicates that Anthropic is currently more akin to an AI infrastructure and enterprise software company that charges based on computational usage.
Anthropic's net loss for the first half of the year reached $80.75 billion, of which approximately $72.4 billion stemmed from fair value changes in convertible notes, derivative liabilities, SAFE instruments, and other financing tools. As the company's valuation increased, the book value of these financing instruments—convertible into shares in the future—also rose, resulting in substantial non-cash expenses. Excluding stock-based compensation, non-cash charitable expenses, and specific litigation costs, the adjusted operating profit was approximately $1.618 billion. The operating profit for the second quarter alone reached about $974 million, marking the first profitable quarter since the filing period disclosed in the prospectus.
The significance of this prospectus goes beyond adding another major tech company. Anthropic could become the first frontier large model developer to go public, allowing investors to observe for the first time the relationship between model usage revenue, enterprise AI adoption, compute costs, and long-term infrastructure commitments.
Data per minute
• Revenue in the first half of 2026 was $16.286 billion, approximately 14 times that of the same period last year;
• Revenue based on usage amounted to approximately $13.064 billion, accounting for about 80% of the revenue for the first half of the year;
• Subscription revenue of approximately $3.222 billion, accounting for about 20%;
• Operating loss of approximately $693 million in the first half of the year;
• Adjusted operating profit of approximately $1.618 billion, with an adjusted operating margin of approximately 10%;
• Operating profit for the second quarter was approximately $974 million, marking the first quarterly operating profit since disclosure began;
• Net book loss of $80.75 billion for the first half of the year;
• As of the end of June, the company had approximately $131.7 billion in cash, cash equivalents, and other short-term liquid assets;
• As of the end of September, more than 11,900 customers contributed annual revenue of over $100,000;
• More than 1,900 customers contributed annual revenue of over $1 million;
• Approximately 250 customers contribute annual revenue of over $10 million;
• For the first nine months of 2026, the top five customers collectively contributed approximately 25% of revenue, with no single customer accounting for more than 10%;
• More than half of the Fortune 500 companies and over 30% of the Global 2000 companies have become Anthropic customers, each contributing at least $100,000 in revenue over the past 12 months;
• Approximately 48% of first-half revenue was generated through cloud marketplaces such as AWS, Google Cloud, and Microsoft Azure;
• Anthropic reported a net revenue retention rate of approximately 900%, reflecting rapid expansion in usage among existing enterprise customers; however, this metric does not include individual subscribers to Claude Pro and Max.
• The company's valuation at its private funding round in May 2026 was approximately $965 billion;
On the day the document was released, the AI and semiconductor sectors overall strengthened, with the Philadelphia Semiconductor Index rising approximately 2.9%.
MSX View
The most significant meaning of Anthropic’s IPO is that the public market can now independently price a leading large model company. Previously, investors participated in the generative AI industry indirectly through chipmakers like NVIDIA and Broadcom, cloud service providers like Amazon, Microsoft, and Google, as well as data center, power, and optical communication companies. These firms’ AI revenues are typically mixed with other business lines, making it difficult for the market to accurately determine how much end customers are paying for large models.
Anthropic's prospectus for the first time provides a relatively complete view of the revenue structure of a cutting-edge model company. Approximately 80% of the revenue in the first half of the year came from usage-based billing, meaning revenue can grow with enterprise API calls, Claude Code usage, and the number of agent tasks, without being entirely dependent on fixed-seat subscriptions.
This also explains Anthropic’s net revenue retention rate of approximately 900%. This figure does not mean the customer base has grown ninefold, but rather that the existing customer base eligible for calculation generated about nine times the quarterly revenue a year later. Companies may initially pilot Claude with a small group of developers, then expand its use to code development, customer service, data analysis, internal search, and automated workflows, leading to a rapid increase in model usage. However, as the customer base grows and enterprises complete their initial deployments, usage growth may naturally slow down. Anthropic also notes in its prospectus that future revenue growth rates may decline and become volatile.
Claude Code is another key to understanding Anthropic’s business model. Unlike standard chatbots, programming agents can continuously read code, modify files, run tests, and complete multi-step tasks, resulting in significantly higher token and computational resource consumption per user. The prospectus shows that highly active enterprise users of Claude Code spend substantially more per month than enterprise users who do not use programming tools. This suggests that programming is not just a standalone product, but likely serves as an entry point for Anthropic to deepen enterprise customer engagement.
Financial data also shows that Anthropic's unit economics are improving. The company's operating loss for the first half of the year narrowed to $693 million, and it achieved its first-ever quarterly operating profit in the second quarter. Compared to a net loss of $80.75 billion on paper, operating profit better reflects the state of day-to-day business, as the vast majority of the net loss stems from non-cash revaluation expenses on financing instruments due to rising valuations.
However, the adjusted operating profit should not be interpreted as the company having fully resolved its profitability issues. This metric excludes stock-based compensation, non-cash charitable expenses, and certain litigation costs. For a company that needs to continuously recruit top researchers and continually train new models, stock-based compensation still represents a real dilution cost to shareholders.
The larger variable is compute commitment. Anthropic has signed large-scale, long-term infrastructure agreements with partners such as Amazon, Google, and Microsoft. Some of these agreements require the companies to commit to minimum procurement spending regardless of how much compute they ultimately use. This arrangement secures scarce GPU, TPU, network, and data center capacity in advance, ensuring Anthropic does not miss customer demand due to compute shortages, but also results in high fixed costs.
If Claude's enterprise usage continues to grow, locking in computing power in advance can become a competitive advantage; however, if demand growth slows, model prices decline, or new architectures significantly reduce inference costs, these long-term commitments may strain cash flow and capital returns.
The dual role of cloud partners is also noteworthy. AWS, Google Cloud, and Microsoft Azure not only provide Anthropic with computing power and sales channels but may also be its investors, creditors, or potential competitors. Approximately 48% of revenue in the first half of the year was generated through cloud marketplaces, indicating that these channels have significantly lowered the barrier for enterprises to purchase Claude—but also mean that Anthropic’s distribution, collections, and infrastructure are heavily dependent on a few major tech platforms.
Customer concentration is currently at a manageable but monitored level. The top five customers account for approximately 25% of revenue, with no single customer contributing more than 10%; meanwhile, over 1,900 customers have annual revenues exceeding $1 million. This indicates that Anthropic has a broad enterprise customer base, but certain large contracts still significantly impact quarterly growth.
Valuation will be the most critical issue at the IPO stage. Based on media estimates of a $1.5 trillion to $2 trillion valuation, even when annualizing revenue projected for the first half of 2026, Anthropic’s price-to-sales ratio remains high. The level of valuation the market is willing to assign depends on whether investors believe the current rates of revenue growth, customer expansion, and unit economics can be sustained, and whether future cash flows can cover long-term compute commitments. Thus, Anthropic’s IPO is testing a fundamentally new business model: Can enterprise spending on agents and model calls grow fast enough to offset the R&D, talent, and computational costs required to train the next generation of models?
This prospectus has demonstrated that enterprise AI demand is real, and Claude has moved from an experimental product to a stage of scalable deployment. The next step is to prove that the rapidly growing volume of API calls can be consistently converted into operating profits and free cash flow. If Anthropic successfully goes public and achieves a high valuation, it could establish the first public market pricing benchmark for OpenAI and other large model companies, further influencing the valuation logic of NVIDIA, cloud computing, data centers, power, and optical communications industries.
Frontier AI is transitioning from a technology race fueled by private capital to a phase subject to ongoing public market scrutiny. Anthropic’s filing of the S-1 marks the beginning of this stage.
About MSX:
MSX is a leading RWA trading platform dedicated to providing secure, efficient, and transparent access to global financial markets. As one of the earliest on-chain platforms for trading U.S. stocks, MSX has consistently remained at the forefront of the industry, driving market transformation through continuous innovation.
The platform deeply integrates blockchain technology with compliance frameworks, offering spot and derivative trading for nearly 400 tokenized stocks and Pre-IPO assets, seamlessly bridging the gap between traditional finance and the digital assets industry.
Guided by the core mission of enabling free circulation of high-quality assets, MSX has built a diversified digital financial services ecosystem encompassing U.S. stock spot and perpetual contracts, crypto-to-crypto trading, Pre-IPO investments, and MaiTong Research Institute, aiming to provide global investors with round-the-clock, high-performance access to premium assets.
Official website:https://msx.com/
Chinese TG:https://t.me/MSXOfficial
Global TG：https://t.me/MSXGlobal
Risk disclaimer: Macroeconomic conditions and U.S. stock market volatility are significant; the content of this article is for academic and research observation purposes only by MaiTong Research Institute and does not constitute any investment advice.