The data doesn't lie. It just waits for someone to interpret it correctly.
Contrary to the euphoric projections circulating among some of Anthropic's existing investors, the company's valuation trajectory is not a signal of strength. It is a symptom of a market that has forgotten how to price risk. Six investors interviewed by the Financial Times believe that when Anthropic goes public—possibly as early as October—its valuation could exceed $2 trillion. That figure is double the $965 billion valuation from May. One investor even estimated a $3 trillion valuation based on a 30 times revenue multiple.
Let me stop here. This is not analysis. This is wishful thinking dressed in arithmetic.
Context: The Hype Cycle and the Protocol
Anthropic is an AI research company. Its primary product, Claude, is a large language model competing with OpenAI's GPT-4 and Google's Gemini. The company announced in May that its annualized revenue had surpassed $47 billion. That is a large number. It is also a number that can be easily manipulated by discounting future revenues, extending contracts, or simply reclassifying bookings. I have seen this pattern before. In 2017, I spent six weeks auditing the Waves ICO's sidechain implementation. The team claimed $100 million in token sales. The reality was a series of misconfigured cryptographic keys. The principle is the same: revenue claims are not revenue verifications.
These investors place the highest importance on revenue growth. They expect Anthropic's annualized revenue to reach between $100 billion and $120 billion by the end of the year. That implies a growth rate of over 100% in six months. Is that plausible? In a bull market, yes. Is it sustainable? The data suggests otherwise.
Hype is just volatility wearing a suit and tie.
Core: Systematic Teardown of the Valuation Assumptions
Let me dissect the assumptions underlying these projections. The investors are using a revenue multiple of 30x. That is standard for high-growth tech companies. But the comparison is flawed on three levels.
First, revenue multiples are a proxy for future cash flows. They assume that the company will maintain its growth trajectory and that the market will continue to value future earnings at the same rate. This assumption is a structural flaw. Risk is not a number; it's a structural flaw. The entire valuation framework collapses if the growth rate decelerates, if competition erodes margins, or if regulatory intervention caps revenue.
Second, the investors are ignoring the cost side. Anthropic's revenue may be $47 billion, but what is its cost of goods sold? Training large language models requires massive compute resources. The cost of inference is also high. The company's gross margins are likely lower than a typical SaaS company. A 30x revenue multiple applied to a low-margin business is a recipe for disappointment.
Third, the investors are relying on a single metric: annualized revenue. This is a lagging indicator. It tells you what happened in the past, not what will happen in the future. The protocol doesn't care about your revenue projections. The protocol cares about the underlying economics: unit economics, customer acquisition costs, churn rates, and the defensibility of the competitive moat.
Based on my experience auditing blockchain projects, I have seen this pattern repeated. A project raises a huge round based on a promise of revenue growth. The team delivers a product that generates some revenue. The investors extrapolate that revenue into infinity. Then the market corrects. The token price drops. The project fails. The same cycle is playing out here, but with a public company instead of a token.
Trust is a variable we must eliminate, not manage.
The investors are also ignoring several risks that could materially impact the valuation. Let me list them.
- Competition from low-cost models in China. Chinese AI companies are producing models that are competitive with Claude and GPT-4 at a fraction of the cost. This is not a hypothetical risk. It is already happening. The market is pricing in a premium for Anthropic's brand, but brands are not defensible when the underlying technology is a commodity.
- Conflicts with the U.S. government. The regulatory environment for AI is uncertain. The U.S. government is considering new regulations on AI models, including export controls, liability for harmful outputs, and transparency requirements. Any of these could reduce Anthropic's revenue or increase its costs.
- Companies beginning to control AI spending. The initial wave of AI adoption was driven by experimentation. Companies tried Claude, GPT-4, and other models. Now they are starting to budget for AI spending. This means that revenue growth will slow as companies optimize their usage. The low-hanging fruit has been picked.
These risks are not priced into the 30x revenue multiple. They are structural failures in the valuation model.
Let me also examine the investor's claim that Anthropic executives have not yet privately established an IPO valuation target. This is a red flag. It suggests that the company itself is uncertain about its value. The investors are projecting their own hopes onto the company. This is a classic case of the market leading the fundamentals.
I have seen this in the blockchain space. In 2021, during the NFT explosion, I wrote a 10,000-word thesis on the lack of true ownership in ERC-721 standards. The market was pricing NFTs as if they were digital property. The reality was that 80% of "decentralized" assets had single points of failure. The market was wrong. The same is true here. The market is pricing Anthropic as if it is a monopoly. It is not.
Contrarian: What the Bulls Got Right
It would be intellectually dishonest to ignore what the bulls got right. Anthropic has a strong product. Claude is widely used. The team has a strong research background. The company has a clear path to revenue growth. The investors are right to be optimistic about the long-term potential of AI.
But the blind spots are significant. The investors are ignoring the fact that the AI market is becoming commoditized. The barriers to entry are low. Anyone with a large dataset and enough compute can train a model. The moat is not technology; it is distribution. And distribution is a battlefield.
Another blind spot is the assumption that revenue growth will continue at the same rate. In the blockchain space, we have seen this pattern with DeFi protocols. A protocol generates $100 million in revenue in one year. The next year, the revenue drops to $50 million. The market is not linear. It is cyclical. The same is true for AI.
The bulls also assume that the government will not intervene. This is naive. AI is a strategic technology. Governments will regulate it. The question is not whether regulation will happen, but how much it will cost.
Risk is not a number; it's a structural flaw.
Despite these blind spots, the bulls are right that the market is underestimating the potential of AI. The technology is transformative. It will create new industries and destroy old ones. But the valuation of any single company is a function of its ability to capture that value. And that is not guaranteed.
Takeaway: Accountability Call
Anthropic's $2 trillion valuation is a story of hype. The investors are betting on a future that may not materialize. The risks are real. The competition is real. The costs are real. The valuation is a wish, not a fact.
The blockchain industry has taught me that the best way to evaluate a project is to look at the code, not the marketing. The same principle applies here. Look at the financials. Look at the unit economics. Look at the competitive landscape. Do not rely on a multiple of revenue.
Hype is just volatility wearing a suit and tie.
The next time you hear a valuation projection, ask yourself: what is the underlying structural integrity? If the answer is a revenue multiple, you are not analyzing; you are guessing.
Additional Analysis: The Blockchain Parallel
Let me draw a parallel to the blockchain industry. In 2020, during the DeFi Summer, I deeply analyzed the lending logic of Compound Finance. I spent three months tracing the interest rate accumulation algorithms. I discovered a potential edge case in the liquidation threshold calculation that could be exploited under high volatility. The market was pricing Compound as if it was a safe, high-growth protocol. The reality was that the protocol had a structural flaw that could be exploited.
The same is true for Anthropic. The market is pricing it as a safe, high-growth company. The reality is that the company has structural flaws that could be exploited by competitors, regulators, or market dynamics.
The protocol doesn't care about your revenue projections.
The Fallacy of the 30x Multiple
Let me deconstruct the 30x revenue multiple. The multiple is derived from comparable companies. But the comparables are not truly comparable. Anthropic is a high-growth, high-risk company. The only comparable companies are other AI companies, and they are all trading at similar multiples. This is circular logic. The multiple is based on the market's assumption that the high growth will continue. But the market is often wrong.
Consider the case of Zoom. During the pandemic, Zoom's stock traded at over 100x revenue. The market assumed that the growth would continue. It did not. The stock crashed. The multiple compressed. The same pattern is likely for Anthropic.
The Role of Investor Psychology
The investors are not rational. They are driven by the fear of missing out. They see the success of other AI companies and want to be part of the next big thing. This is the same psychology that drives the crypto market. I have seen it in every bull market. The investors ignore the risks because they are focused on the potential reward.
Hype is just volatility wearing a suit and tie.
The Structural Flaw in the Business Model
Anthropic's business model is based on selling access to its API. This is a low-margin business. The cost of inference is high, and the competition is intense. The company's moat is not the technology; it is the brand. But brands are ephemeral. They can be damaged by a single scandal or a competitor's breakthrough.
The Government Risk
Let me expand on the government risk. The U.S. government is considering new regulations on AI. The European Union has already passed the AI Act. These regulations will impose costs on companies like Anthropic. They will also limit the company's ability to scale. The investors are not pricing this risk into their valuation.
The Competition from China
Chinese AI companies are making rapid progress. They have access to large datasets and cheap compute. They are also operating in a regulatory environment that is more favorable to AI development. The investors are underestimating the threat from China.
The Structural Flaw in the Valuation
The valuation is based on a single metric: revenue. But revenue is not the same as profit. The company may be unprofitable. It may be burning cash. The valuation does not account for the cost of capital. The investors are ignoring the balance sheet.
The Analogous Crypto Case: Solana
In 2021, Solana was trading at a valuation of over $100 billion. The market believed that it would become the next Ethereum. The reality was that the network had structural flaws that led to multiple outages. The price crashed. The same pattern is playing out with Anthropic.
Risk is not a number; it's a structural flaw.
The Importance of First-Principles Analysis
The only way to evaluate a company like Anthropic is to use first-principles analysis. Look at the underlying technology. Look at the market dynamics. Look at the competitive landscape. Do not rely on a multiple of revenue. That is a lazy heuristic.
The Call for Accountability
I call on the investors to be more rigorous in their analysis. Do not rely on hype. Do not rely on emotional language. Use data. Use code. Use analysis.
Trust is a variable we must eliminate, not manage.
The investors are betting on a future that may not materialize. The risks are real. The competition is real. The costs are real. The valuation is a wish, not a fact.
Final Thoughts
Anthropic's $2 trillion valuation is a symptom of a market that has lost its discipline. The blockchain industry has taught me that the best way to evaluate a project is to look at the code, not the marketing. The same principle applies here.
The protocol doesn't care about your revenue projections.
I will continue to write about the structural flaws in the market. The hype will fade. The data will remain. And the investors who ignored the data will be left holding the bag.
Appendix: Technical Analysis of the Revenue Multiple
Let me provide a technical analysis of the 30x revenue multiple. The multiple is derived from the formula: Valuation = Revenue * Multiple. The multiple is a function of expected growth, risk, and discount rate. If the growth rate is 100% and the risk is low, the multiple can be 30x. But if the growth rate slows to 50% and the risk increases, the multiple should contract to 15x. That would cut the valuation in half.
The investors are assuming that the growth rate will remain high. This is a dangerous assumption. The law of large numbers dictates that growth rates slow as companies get larger. Anthropic's revenue is already $47 billion. To double that to $100 billion, the company would need to add $53 billion in new revenue. That is a massive amount. It is not impossible, but it is highly unlikely.
The Implications for the Blockchain Industry
The Anthropic case has implications for the blockchain industry. The same valuation techniques are used in crypto. The same hype cycles exist. The same risks are ignored. The blockchain industry should learn from this case. The next time you see a project with a high valuation based on a revenue multiple, be skeptical. Look at the underlying structure. Look at the code. Look at the data.
Hype is just volatility wearing a suit and tie.
Conclusion
Anthropic's $2 trillion valuation is a structural flaw in the market's pricing mechanism. The investors are ignoring the risks. The competition is real. The government is coming. The costs are high. The valuation is a wish, not a fact.
I will continue to be the cold dissector. I will continue to expose the structural flaws. The data will speak for itself.