Anthropic kicks off IPO roadshow, offering shares at $70 to $78 and targeting a trillion-dollar valuation
Key Points
- Anthropic is preparing to go public on the Nasdaq. It aims to raise up to $90.4 billion through public share sales and a private placement, with a target company value of $1.02 trillion.
- In the third quarter, the company behind Claude increased its revenue by 47 percent to almost $17 billion. It also reduced its net loss, although its growth was slower than in the previous quarter.
- The IPO faces challenges because Anthropic has promised to spend nearly $900 billion on computing capacity with partners such as Amazon and Google. Investors are also becoming more doubtful about how much money is being spent on AI.
Anthropic is meeting with investors to promote its planned IPO. Despite growing doubts on Wall Street and very high costs for its computing infrastructure, the share sale could raise up to $90.4 billion.
According to a Reuters report, AI company Anthropic plans to raise up to $64.4 billion through its public share sale. Based on the number of shares listed in its prospectus, the company, which competes with OpenAI, is aiming for a value of up to $1.02 trillion. Details of the plan were briefly sent to investment banks on Monday evening but were later withdrawn.
In addition to the public share sale, Anthropic is arranging another $26 billion through a separate private placement. Major investors in this deal include Google, Nvidia, AMD, Broadcom, Micron Technology, Microsoft, and TSMC. Together, the two deals could involve up to $90.4 billion in shares.
In its official announcement, Anthropic says the public offering includes 825 million Class A common shares. Each share will cost between $70 and $78. Anthropic itself is selling 801,351,354 shares, while current shareholders are selling another 23,648,646 shares. Anthropic will not receive any money from the shares sold by existing shareholders. The group of banks managing the offering, led by Goldman Sachs and Morgan Stanley, also has 30 days to buy up to 123.75 million additional shares. Anthropic's stock will trade on the Nasdaq under the ticker symbol "ANT."
Before the listing, potential investors had asked the company to lower its target value. An earlier figure of $1.5 trillion was considered too high. A private funding round in May valued the company, led by CEO Dario Amodei, at $965 billion. Its major existing investors include Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital. The largest IPO in the world so far this year was Elon Musk's SpaceX. It raised $75 billion in June, giving the company a value of nearly $1.75 trillion.
Billions in revenue, but data center costs are staggering
In the third quarter, Anthropic reduced its net loss and increased its revenue by 47 percent compared with the previous quarter, reaching $16.96 billion. However, this was slower growth than in the previous quarter, when revenue increased by 144 percent. Anthropic was founded in 2021 by former OpenAI researchers. The company says strong demand for its Claude AI models is driving its growth. Demand is especially strong for specialized tools used in healthcare, finance, and legal services.
But this growth comes with very high costs. According to its prospectus, Anthropic has made commitments worth nearly $900 billion to secure computing capacity from partners such as Amazon and Google. Some of the infrastructure needed in the future will depend on financing that Anthropic does not control.
The IPO, which has been delayed several times, is taking place at a time when investors are becoming more cautious. With interest rates remaining high, investors are increasingly questioning whether the current level of AI investment can continue. They are also concerned about companies funding each other through complex financial deals and the possibility that too many data centers are being built. On Thursday, the Nasdaq had its biggest one-day drop in 19 months. Strong gains by Democrats in the US midterm elections had increased fears that the government might introduce stricter rules for building AI data centers.