As Anthropic prepares to price what could become one of the most consequential public offerings in a generation, NYU Stern finance professor Aswath Damodaran offers a measured and methodical reality check in his October 22, 2026 Substack essay1. The piece does not tell investors whether to buy. It does something more useful: it explains why so few of them are even trying to value the company in the first place.
The Life Cycle Diagnosis
Damodaran situates Anthropic and OpenAI clearly within a corporate life cycle framework. Both companies are reporting sky-high revenue growth alongside operating losses and deeply negative free cash flows, the defining characteristics of early-stage, growth-phase firms. What makes them unusual is scale. With Anthropic disclosing a Q3 2026 revenue figure of $16.8 billion and an annualized revenue run rate of $67 billion, these are not garage-stage startups. They are, as Damodaran observes, "growing at scale." That combination, early-stage economics at mega-cap ambition, is precisely what makes conventional valuation so difficult.
Pricing Is Not Valuation
The most consequential distinction in the piece is between pricing and valuation. Damodaran argues that intrinsic valuation, estimating future cash flows, growth rates, and discount rates, requires inputs that simply do not exist in stable form for Anthropic today. Business models are still in flux. Market size is contested. There is little financial history to draw on. In the absence of those inputs, he notes, investors fall back on pricing: "attaching a number to an asset or business, based upon what others are paying for similar assets or business." This is not a failure of rigour. It is the rational response to radical uncertainty.
The ARR Problem
Damodaran is candid about the limitations of Annualized Revenue Run Rate, the metric Anthropic and OpenAI are both leaning on heavily. ARR is not standardized. Choices about whether to use gross or net revenues, and whether to annualize from daily, weekly, or monthly figures, can produce materially different numbers. OpenAI's shift toward net revenue reporting, for instance, dropped its ARR from $70 billion to $50 billion. Anthropic continues to use gross revenues. Damodaran frames the precedent for this kind of metric evolution with a clean historical analogy: dot-com companies in the late 1990s were once priced on eyeballs, the time website visitors spent on pages, a metric that "had some basis," he argues, even if it felt speculative at the time. ARR is the current era's version of that creative but imprecise proxy.
The Peer Group Problem
For pricing to function, it needs a peer group. Anthropic's IPO roadshow faces a structural problem: there are almost no publicly traded comparables. Gemini is embedded inside Alphabet. xAI is part of SpaceX. Claude's closest mirror, OpenAI, is still private. Damodaran sketches the creative workarounds investors will use: pairing Anthropic with OpenAI in private-to-private comparisons, or expanding the comparison to the Mag Seven on the basis that investors who seek trillion-dollar market cap companies will compare them against each other, regardless of business model differences. Neither approach is ideal. Both are inevitable.
The Implicit Assumptions in Every Multiple
The most clarifying section of Damodaran's analysis is his insistence that pricing does not escape the assumptions that make valuation so difficult. It merely embeds them silently. An enterprise-value-to-revenue multiple for Anthropic implicitly assumes something about revenue growth trajectories, gross margin evolution, operating leverage from compute cost efficiencies, and risk convergence. "To make a case for Anthropic at a $1.5 or $2 trillion pricing," Damodaran writes, "you will have to make defensible arguments for sky-high revenue growth in the early years, a surge in operating margins to 30% or higher, with favorable unit economics and economies of scale providing tailwinds." The multiple looks clean. The story underneath it is not.
5 Key Takeaways for Advisors and Investors
- Pricing is not valuation. For companies like Anthropic, comparable pricing through peer multiples is the practical tool, but it carries all the same assumptions as a discounted cash flow model, just less visibly.
- ARR is a useful but gameable metric. Gross versus net revenue choices and annualization periods can shift the number dramatically. Advisors should treat ARR figures from different companies as non-comparable until methodology is disclosed.
- The absence of a peer group is a real risk. Without publicly traded LLM-focused comparables, pricing anchors for the Anthropic IPO will rely on creative and imprecise analogies to the Mag Seven or SpaceX.
- Operating margin trajectory is the central variable. A move from current losses to 30%-plus operating margins is the linchpin of the bull case. Advisors should probe whether that path is credible before accepting any pricing multiple as justified.
- Wait for the prospectus. Damodaran himself defers his intrinsic valuation to when the full prospectus is filed. That is sound counsel for advisors advising clients on IPO participation.
Footnote:
1 Damodaran, Aswath. "From Eyeballs to ARR: Buzzy Pricing Metrics and the AI Trade!" Aswath Damodaran's Substack, 22 Oct. 2026, https://aswathdamodaran.substack.com/p/from-eyeballs-to-arr-buzzy-pricing.
