Michael Burry remains skeptical of Anthropic's IPO filing

A skeptical look at the S-1

Anthropic recently filed its Form S-1, the official document required for an initial public offering (IPO). This filing provides a detailed look at the company's finances and business structure for potential investors. However, Michael Burry, the investor famous for his role in the housing bubble collapse, says the document failed to change his negative outlook on the company.

Burry told his Substack subscribers that while the filing contains a lot of information, it does not provide any new evidence that the company is a winner. He argued that the document is outdated because it only covers information through June 30. According to Burry, the developer landscape has shifted significantly in the four months since that date.

The investor is known for his contrarian approach and deep skepticism regarding the current AI boom. He has previously raised concerns about aggressive accounting and hidden liabilities within the sector. His critique of Anthropic suggests that the company's public debut may not be as smooth as the market expects.

Signs of slowing revenue growth

One of the primary reasons for Burry's skepticism involves the pace of Anthropic's revenue growth. The company reported that its third-quarter revenue was about $16.9 billion, which was 86 times the $197 million it generated in the fourth quarter of 2024. While this is a large number, the growth rate between quarters is beginning to show signs of cooling.

The data shows that Anthropic's revenue increased by 144% from the first quarter to the second quarter of this year. However, that growth rate dropped to around 46% between the second and third quarters. Burry noted that this slowing is not necessarily due to a decline in 'tokenmaxxing,' which is the practice of maximizing AI token consumption.

Burry pointed out that he did not expect the company to mention the decline in tokenmaxxing in their filing. This trend has fallen out of favor recently because of the high costs associated with it. The shift suggests that users are becoming more mindful of how they use AI resources, which could impact the bottom line for providers.

Efficiency as a headwind

A major theme in Burry's analysis is the idea that AI models, such as Anthropic's Claude, are becoming more efficient. He argues that as these models improve, they are being used more effectively by developers and companies. While this is a technical success, Burry believes it will eventually work to the detriment of the growth rates for the companies selling the models.

If a company can achieve the same results using fewer resources or fewer tokens, the total amount of money spent on those services may decrease. This creates a paradox where the product gets better, but the revenue generated from it might shrink. Burry is watching this trend closely as it could fundamentally change the economics of the AI industry.

The investor also highlighted the commoditization of tokens as a risk. He acknowledged reports that large technology companies are moving away from third-party models like Anthropic's to build their own. This shift toward internal tools could leave independent AI startups with a smaller share of the market.

Fragile debt and chip hurdles

Beyond the specific numbers of Anthropic, Burry is looking at the broader structural risks in the AI sector. He expressed concern regarding the complexity and fragility of the debt financing that many AI deals currently rely on. He believes the underlying fundamentals of these deals are more important than the headlines surrounding individual companies.

Burry also pointed to physical and logistical hurdles in the hardware space. He noted the difficulties involved in installing new graphics chips, which can shore up demand for older chips. This dynamic could create a complicated and unpredictable market for the hardware that powers the AI revolution.

Additionally, he warned of a potential crash in memory chip prices once the current shortage eases. He is also watching the rise of small language models (SLMs), which require less computing power than larger models. These smaller models could reduce the overall need for the massive infrastructure that many investors are currently funding.

The road to the IPO

Anthropic has not yet responded to requests for comment regarding Burry's specific criticisms. The company remains focused on its path to becoming a public entity, which would be one of the largest IPOs in history. However, Burry's comments serve as a reminder of the scrutiny the company will face from sophisticated investors who look past the hype.

The investor's focus on 'fundamentals' suggests that the next phase of the AI market will be defined by profitability and efficiency rather than just raw scale. He is urging his followers to look at the circular financing deals and the actual costs of doing business in the AI space. This perspective may influence how institutional money flows into the sector in the coming months.

As the IPO approaches, the market will be looking for clarity on these issues. Whether Anthropic can prove its growth is sustainable despite the cooling of token consumption will be a key factor. For now, Burry remains unconvinced that the current filing provides a clear picture of a winning business.