Founder of Social Capital: Anthropic's IPO Safety Margin at $1 Trillion
Chamas Palihapitiya stated that Anthropic will be a significant IPO across various dimensions, but recent events will suppress pricing. He bets that the buyer's safety margin is around plus or minus $1 trillion: at a $1 trillion valuation, buyers still have substantial upside, while most sellers still win big, allowing the company to secure about $200 billion to build what it wants to build.
He stated more firmly on the All-In podcast: regulatory risks will drive the transaction price far below market expectations. The S-1 must disclose everything, making an already opaque document even harder to read; with regulatory risks on the surface, the market's only translation method is to lower seller expectations and provide buyers with a larger safety cushion.
He provided a rough conversion: even with an annualized revenue of $100 billion, a clean listing could correspond to a $2 trillion market value, but now it may drop to $1 trillion or below. Hedge funds, pensions, and long-term mutual funds have fiduciary responsibilities and will absorb all risks at lower prices. He believes this is actually good for the company; after the noise clears, it only needs to operate as a regular company and run its business.
Anthropic submitted a confidential S-1 to the SEC on June 1, with a $65 billion Series H round in May and a post-money valuation of $965 billion, with private rounds led by Altimeter, Dragoneer, Greenoaks, and Sequoia. The publicly discussed IPO target had previously reached $2 trillion, aiming to raise about $100 billion, with the timeline rumored to be pushed from October to November. The probability of an IPO this year on Polymarket has dropped from about 96% to around 76% in recent months.
Simultaneously, public discussions are also layered with export controls leading to a brief offline of leading models, lawsuits identifying supply chain risks by the Pentagon, public warnings from security researchers, a rapid increase in the proportion of open-source models used, and the governance structure of the company as a public benefit corporation with long-term interests in trust. Sellers want a $2 trillion narrative, while buyers want to factor these risks into the price.
In market mechanisms, buyers are public funds and pensions with fiduciary responsibilities, while sellers are private equity shareholders and employees whose books are close to or exceed $1 trillion. The driving events are the entry of regulatory and safety narratives into the mandatory disclosure window, not a sudden revenue interruption. Funds are flowing from "20 to 40 times revenue private equity marks" to "public offerings that must be discounted for risk." Beneficiaries are IPO buyers who can secure chips around $1 trillion and early sellers who still win relatively large based on cost bases; those under pressure are banks quoting $2 trillion and holders of secondary market premiums. If the company raises about $200 billion as he suggests, it receives construction funds but gives up valuation fantasies.
Source: Public Information
ABAB AI Insight
Chamas has built a personal brand around "storytelling IPOs" through Facebook's growth and a series of SPACs, with the market remembering him for both creating hype and leaving discounts, as seen in the post-market differentiation of Virgin Galactic, Opendoor, Clover Health, and SoFi. He is now using the same language for a reverse operation: acknowledging that Anthropic is significant, but pulling the transaction price from $2 trillion down to $1 trillion, positioning himself on the buyer's safety margin side rather than the seller's narrative side. This aligns with his summer statement that giving high terminal values to model layers is a "mathematical error," and that value lies in application and infrastructure layers—he is willing to buy the company but not pay laboratory multiples.
The capital path is that private equity marks keep risks on the balance sheet, while IPOs sell risks to public funds with fiduciary responsibilities. The $965 billion Series H round has already thickened the books of early funds and employees; $2 trillion is merely doubling the same revenue curve. Public funds are not buying the curve but the risks that must be detailed in the S-1, such as export controls, military designations, alignment failures, open-source alternatives, and the dilution of public benefit corporation boards by trusts. Discounts are not emotional; they are compliance products: the more comprehensive the risk factors are disclosed, the more the book can only use price to exchange for signatures. The $200 billion cash entering the company transforms it into an issuer bound by the Securities Act of 1933.
Analogies include Facebook's 2012 low opening after writing mobile and user metric controversies into its S-1, Uber's discount after writing its profitability path and regulations into a complex document, and tobacco companies still being able to go public after having to disclose mortality risks in their prospectus—it's possible to go public, but the multiples are rewritten by the terms. In terms of industry positioning, leading laboratories are transitioning from "who has the bigger story in private equity" to "who is willing to use price to exchange for the IPO window"; OpenAI has pushed its timeline to 2027, while Anthropic is the first to hit this door.
The structural judgment belongs to the transfer of pricing power. The mechanism is that private valuations are determined by lead investors and secondary transactions, while IPO valuations are set by fiduciaries who must read and sign off on risk factors. Kitchen-sink disclosures turn the seller's information advantage into the buyer's bargaining power. The $1 trillion is not a judgment of the company's quality but the liquidation price that public capital is willing to pay in the face of legal texts.