OpenAI remains at least six months away from going public, potentially longer. However, as the company progresses toward an initial public offering (IPO), investment banks seeking a role are reportedly contacting public market investors to gauge sentiment regarding OpenAI's post-listing prospects, according to a source familiar with the outreach.
Market sentiment appears mixed. Investors anticipate the IPO could raise tens or even hundreds of billions of dollars, potentially surpassing the scale of Saudi Aramco's 2019 offering, which raised over $25 billion. Despite recognizing OpenAI's leading position in the AI competitive landscape, many public market investors maintain a degree of skepticism. Interviews with eleven investors revealed that the majority currently do not hold equity in OpenAI.
Several investors are adopting a cautious stance toward the IPO. One key concern is the lack of sufficient detailed financial performance data available before the company files its registration statement. Another significant worry is OpenAI's long-term profitability. The developer of ChatGPT is projected to continue incurring substantial costs, with profitability not expected until at least 2030.
The company is finalizing a new funding round at an $850 billion valuation, with investors including NVIDIA, Amazon, and SoftBank. This valuation is already considered high, and investment banks might price the IPO even higher. At $850 billion, OpenAI's valuation would be 28 times its projected 2026 revenue. In contrast, NVIDIA's forward price-to-sales multiple is approximately 12.
Bob Lang, Founder of Explosive Options, stated, "I do believe OpenAI is a great company with a strong moat, but I don't see the valuation on day one offering value to investors." He indicated he would likely not invest in a public OpenAI, especially if its IPO valuation multiple exceeds that of NVIDIA. "The beneficiaries will be the current investors, like the cloud giants and early backers, who will get a chance to sell down some of their stakes," he added.
Some investors have expressed eagerness to short OpenAI stock, believing public markets will not tolerate its prolonged path to profitability. Jim Chanos, a prominent short-seller and investment manager, suggested that NVIDIA sets a valuation benchmark for AI firms like OpenAI. "NVIDIA is largely a monopolist, with explosive growth, enormous margins, and huge cash flow. So why would you pay a higher multiple for OpenAI?" he questioned.
Chanos noted that there is currently insufficient financial information to analyze OpenAI thoroughly. However, once it files publicly, he expects a vigorous debate: "Is this a winner-take-all market, or will it be fragmented like cloud computing? Or like search, where one company becomes the standard and stays there? Right now, the models are still leapfrogging each other."
Investors are also questioning whether the capital raised from the IPO will be sufficient to fund OpenAI until it becomes profitable, or if the company will require further financing, potentially diluting existing shareholders.
In a January report on OpenAI's advertising business, J.P. Morgan analysts suggested that OpenAI's strong brand could allow it to introduce ads in ChatGPT without significant user attrition. However, they also noted that client sentiment toward OpenAI is "mixed" following recent announcements of major expenditures on chips and data centers.
**Unique Risks**
Mark Malik, Chief Investment Officer at Siebert Financial, stated that while he does not expect significant near-term profitability from OpenAI due to its rapid expansion, he would still consider adding it to a portfolio post-IPO. He added that he would limit position size if the stock price is too high, similar to his approach with Palantir Technologies Inc., which trades at a forward price-to-sales multiple of 49 despite its high growth rate.
Nevertheless, Malik views Palantir Technologies Inc. as safer than OpenAI because it can control costs. Unlike OpenAI, Palantir Technologies Inc. does not need to build data centers to support its growth. "If Palantir loses a government contract, it's bad, but what can they do? They can lay people off. But if you spend five years building a data center, you can't just say 'never mind'," Malik said. "Palantir is driving an F1 car, while OpenAI is steering a fully loaded cargo ship."
Another significant risk stems from competition, particularly from Anthropic. With AI tools like Claude Code gaining traction in the enterprise sector, Anthropic's business growth is reportedly on par with OpenAI's. At a Morgan Stanley technology conference this week, Anthropic CEO Dario Amodei stated the company's annualized revenue has doubled to $20 billion, with a recent funding round valuing it at $38 billion.
A report from The Information in November suggested Anthropic holds an advantage, as its costs for training and running AI models are projected to be significantly lower than OpenAI's over the coming years. Some investors are beginning to believe that, given Anthropic's success with enterprise clients who have strong payment capabilities, its long-term profitability prospects may be superior to OpenAI's.
As Anthropic also plans an eventual public listing, it could divert investor attention and capital that might otherwise flow to OpenAI. While Anthropic's more restrained approach to procuring or building expensive computing power might slow its growth, investors like Chanos view this cautious strategy positively. "Our core advice to clients is: be long on what chips make, and short on where chips are housed," Chanos said. "Simply put, we don't believe running data centers will be a high-return business."
Comments