Everything About AI
NextFin News - Before a single Anthropic share trades on the Nasdaq, the crowd has already priced the offering. Speculators on prediction markets are assigning 95% odds that the artificial-intelligence lab officially announces an initial public offering by the end of 2026, with the most-watched contracts pointing to a closing market capitalization of $1.75 trillion to $2 trillion — roughly double the $965 billion private valuation the company secured in May. The question is no longer whether Anthropic will list. It is whether a valuation generated by retail speculators can survive first contact with real-money institutional investors.
Anthropic, PBC filed its Form S-1 with the Securities and Exchange Commission on October 26, 2026, seeking to list Class A shares on the Nasdaq Global Select Market under the ticker ANT. Goldman Sachs and Morgan Stanley are leading a 22-bank underwriting syndicate, and the company is targeting a trading debut in the week of November 16. What sets this IPO apart from the start is the parallel price-discovery layer that has grown up around it: on Kalshi, traders define an "official IPO announcement" as the SEC declaring the S-1 effective, a priced offering, or an exchange assigning a ticker — and by that definition, the public filing has already pushed the odds near certainty.
The Numbers the Prediction Markets Are Pricing
The most liquid Anthropic contract on Polymarket, "Anthropic IPO Closing Market Cap," has drawn $2.4 million in volume since it opened on June 1. The distribution is tightly clustered around a $1.75 trillion to $2 trillion outcome: the $1.75–$2.0 trillion bucket carries 30% of the probability, the $2.0–$2.25 trillion bucket 20.3%, and the $1.5–$1.75 trillion bucket 19.6%. Only 12% of the probability sits below $1.5 trillion, and just 2% of contracts bet that no IPO happens before December 31, 2027. A separate Polymarket market gives a 72% chance of an IPO by December 31, 2026, while the timing contract puts a 56% probability on a November debut — consistent with the company's stated target.
On Kalshi, the timing is even more compressed. Traders place 95% odds that Anthropic officially announces an IPO by the end of 2026, but only 14% that trading begins before the week of November 16. On Polymarket, 69% of contracts price in a listing on a public exchange by November 20. The crowd is nearly certain the deal gets done; it is merely arguing about the exact week and the exact multiple.
The financial backdrop explains the optimism. Anthropic reported revenue of $16.29 billion in the six months ended June 30, 2026, up 1,300% from $1.16 billion a year earlier. Second-quarter revenue alone reached $11.56 billion, a 1,368% year-over-year increase. For all of 2025, revenue jumped 1,089% to $4.59 billion, after just $386 million in 2024 and $41 million in 2023. The company says its annualized revenue run rate has now surpassed $30 billion, up from approximately $9 billion at the end of 2025 — more than triple in under a year.
Customer metrics are the engine behind those numbers. More than 11,900 customers now spend $100,000 or more annually on a trailing-twelve-month basis, up from more than 950 a year earlier. Over 1,900 spend $1 million or more, and about 250 spend $10 million or more. Net revenue retention stands at approximately 900% as of September 30, 2026 — the revenue generated from the existing customer base is roughly nine times the prior-year level. More than half of the $100,000-plus customers now use both the Claude platform and a first-party application such as Claude Code or Cowork, up from about 30% at the end of 2025.
But the losses are equally staggering, and the filing is blunt about concentration risk. In the prospectus, the company warned:
We anticipate that we will continue to derive a significant portion of our revenue from a limited number of customers for the foreseeable future, and in some cases, the portion of our revenue attributable to certain customers may increase in the future.
Net loss for the first half of 2026 was $80.7 billion, though roughly $72 billion of that came from non-cash remeasurement of convertible notes and SAFEs. The 2025 net loss was about $41.9 billion, widened from $8.3 billion in 2024. Working capital stood at $131.6 billion.
Why Prediction Markets Moved Ahead of the Book-Build
The deeper story is not the IPO itself — it is that prediction markets have become a parallel price-discovery mechanism that now runs ahead of traditional IPO book-building. For decades, an IPO price was set in a closed process: the company and its underwriters canvassed institutional investors, built an order book, and printed a price the night before trading began. Retail investors and the broader market learned the valuation only after the fact. Today, a liquid, round-the-clock market in binary contracts publishes a probability-weighted valuation range weeks or months in advance, and that range becomes the anchor that every subsequent participant has to argue against.
This is a structural shift in how private companies get priced before they go public, and it is unlikely to revert. Prediction markets aggregate dispersed information continuously, they settle on observable outcomes, and they let participants express both direction and timing. Once a company as large as Anthropic — with a $965 billion private round behind it — attracts millions of dollars in contract volume, the crowd's implied valuation becomes the reference point for journalists, analysts, and even the company's own messaging. The book-build does not start from zero; it starts from the 95% and the $1.75–$2 trillion range the crowd has already printed.
There is also a self-reinforcing loop embedded in the odds. Because Kalshi's definition of an "official IPO announcement" includes the SEC declaring the S-1 effective or an exchange assigning a ticker, the act of filing publicly moved the probability materially closer to 100%. The more certain the outcome appears, the more capital flows into the "yes" contracts, which reinforces the certainty. By the time the roadshow begins, the market is no longer asking whether Anthropic will list. It is asking only at what price.
The second-order consequence is the one most investors are not discussing: prediction markets may compress the IPO pop. The first-day surge that early employees, venture backers, and pre-IPO shareholders bank on depends on the offering price being set below where the public is willing to trade. But when a $1.75–$2 trillion range is already consensus among speculators, underwriters face pressure to price near the top of what the crowd will bear — leaving less money on the table for the opening bell. The real trade, in that world, migrates from the IPO itself to the secondary lockup expiry, when insiders can finally sell and the prediction-market thesis meets actual supply.
The Counter-Thesis: A $150 Billion Valuation Against a $2 Trillion Consensus
The strongest challenge to the crowd's verdict comes from independent research firm New Constructs, which in October valued Anthropic at $150 billion — one-thirteenth of the modal prediction-market outcome — and called the potential offering the "most ridiculous IPO of 2026." The firm argued that a $2 trillion market capitalization would require profits roughly equal to twice Nvidia's trailing four-quarter total, and wrote:
While Anthropic offers more to society than WeWork ever did, at a $2 trillion valuation, its IPO presents far bigger risks and is positioned to be a far bigger rip off of U.S. capital markets.
New Constructs added that the offering's purpose is not to create wealth for public-market investors but to provide liquidity for the company's Wall Street backers. Anthropic did not respond to requests for comment.
The arithmetic behind that skepticism is concrete. Dr. Chan Ahn, a former Goldman Sachs and JPMorgan executive, estimated that supporting a $2 trillion valuation at a 10% cost of equity, a 25% free-cash-flow margin, and a 25x terminal multiple would require roughly $725 billion in annual revenue by 2036. At a 13% discount rate, the required revenue rises to about $950 billion. Anthropic's current run rate is above $30 billion. The gap between $30 billion and $725 billion is not a rounding error — it is a decade of uninterrupted hypergrowth with no competitive or technological interruption.
The compute bill is the other side of the ledger. Anthropic has committed approximately $518 billion to cloud and infrastructure buildout, with about 80% of those commitments noncancelable. Broadcom is providing up to $42 billion in potential financing, Google up to $20 billion, and Amazon up to $15 billion; a separate financing package announced alongside the filing totals $77 billion. The company is also moving into its own data-center model through "Anthropic Direct," under which third-party partners secure the real estate and power while Anthropic leases the compute — rent payments of $125.2 billion over a five-year lease, excluding interest. In the first half of 2026, compute and infrastructure expenses grew 300% even as second-quarter revenue grew 1,368%. Growth is outpacing the bill — for now.
The bear case, stated fairly, is this: Anthropic is selling intelligence at a price that does not yet cover the cost of producing it, and the customers paying today include the same hyperscalers — Amazon, Google, and their ecosystem — who are simultaneously building competing models and financing the very competitor that threatens them. If open-source models compress pricing power, or if the hyperscalers throttle their own spending once their in-house alternatives mature, the 900% retention rate that underpins the $2 trillion range could decay faster than the prediction markets have priced in.
What Would Break the Thesis
The prediction-market consensus is a bet on durability: that Anthropic's growth is structural, not cyclical. The clearest falsifying signal is the gross-margin trajectory in the company's first public quarterly report. Usage-based revenue — 80% of first-half revenue, metered in tokens and priced by model — must begin to outpace compute costs if the company is to reach sustained positive operating cash flow. If gross margin fails to expand and operating losses remain in the tens of billions despite revenue scaling past $30 billion, the structural-growth thesis is wrong regardless of the retention rate.
A second signal is net revenue retention. The approximately 900% figure implies existing customers are expanding their spend at an extraordinary pace. If that number falls toward 300% or below — still exceptional by software standards, but no longer compounding at the current rate — the revenue path to $725 billion by 2036 breaks down.
There is also a timing risk the crowd may have underpriced. Kalshi assigns only 14% odds to trading before the week of November 16, and the SEC review process for a filing of this size can stretch well past the company's target. A delay does not change the long-term outcome, but it reprices the short-term contracts and pushes the IPO closer to a macro window that may be less forgiving.
What Comes Next
Short term (the listing itself): The base case is a November debut near the $1.75–$2 trillion range the prediction markets have already printed, with a muted first-day pop because the crowd has arbitraged the surprise. The upside case is a hotter-than-expected order book that pushes the closing print toward $2.25 trillion, where 13% of Polymarket probability currently sits. The downside case is an SEC delay or a broader AI de-rating that forces pricing below $1.5 trillion — the two buckets that together hold 12% of the probability today.
Medium term (the first two to four quarters public): Watch the gross-margin bridge and the customer-concentration disclosure. The company already warns that a limited number of customers drive a significant share of revenue; if the top-customer share rises while growth slows, the multiple will compress even if revenue keeps climbing. The hyperscaler customers — Amazon, Google, and their partners — are at once the company's biggest buyers, its largest shareholders, and its most capable future competitors. That triangle cannot hold forever without one side breaking.
Long term (the structural question): If Anthropic sustains positive operating cash flow while keeping net revenue retention above 300%, the $2 trillion range becomes a floor rather than a ceiling, and prediction markets will have proven themselves a more accurate pre-IPO pricing mechanism than traditional book-building. If margins stall and retention decays, the $150 billion valuation looks less like bearish hyperbole and more like a warning that was ignored.
The final irony of Anthropic's IPO is that the most honest price discovery is happening not in the roadshow rooms of its lead underwriters, but in the binary contracts of speculators who have already voted. The public market will get its chance to agree or disagree — and for the first time, it will know exactly what it is disagreeing with before the bell rings.
Explore more exclusive insights at nextfin.ai.
