Author: Dong Jing
Source: Wall Street News
Original Title: Wall Street Conducts "Market Sounding" for OpenAI IPO, Investment Institutions Unimpressed?
OpenAI may still be at least half a year away from going public, but Wall Street's预热work has quietly begun. Several investment banks are actively approaching public market investors to gauge their views on the上市prospects of the parent company of ChatGPT—and the responses received are far colder than expected.
On March 9, according to a report by tech media The Information, sources revealed that multiple investment banks competing for OpenAI's上市underwriting business have started conducting "market soundings" with public market investors. The Information interviewed 11 public market investors for this purpose, most of whom do not yet hold OpenAI equity.
The respondents generally adopted a cautious stance towards this IPO, with core concerns集中在two points:一是unclear profit prospects—OpenAI itself predicts that it will continue to burn cash until at least 2030;二是excessive valuation—the company is currently completing a new round of financing at an $850 billion valuation, equivalent to 28 times the expected revenue for 2026, far exceeding Nvidia's price-to-sales ratio of approximately 12 times.
The report stated that the "cold" market sentiment reflects the deep-seated矛盾faced by this potential largest IPO in history: investors generally recognize OpenAI's leading position in the AI competitive landscape but remain reserved about whether it can achieve reasonable pricing in the public market. At the same time, the strong rise of competitor Anthropic is further分散investors' attention and enthusiasm.
Valuation Controversy: 28x Price-to-Sales Ratio, Where's the Premium?
OpenAI is currently completing a new round of financing at an $850 billion valuation, with participants including Nvidia, Amazon, and SoftBank. This figure has already made many public market investors hesitant, and its IPO pricing may be even higher届时.
Based on the expected revenue for 2026, $850 billion corresponds to a price-to-sales ratio of approximately 28 times. In comparison, Nvidia, regarded as a benchmark for AI investment, currently has a price-to-sales ratio of about 12 times.
The report stated that Bob Lang, founder of trading firm Explosive Options,直言:
"I do think OpenAI is an excellent company with a strong moat, but I don't think any valuation on the first day of trading is a good deal for investors."
He said he would likely not participate in public market investment in OpenAI, especially given its higher valuation multiple than Nvidia.
Lang also pointed out that the real beneficiaries of this IPO will be those early investors and hyperscale cloud computing companies that already hold shares—they will use this as an opportunity to cash out.
Well-known short-seller Jim Chanos used Nvidia as a reference point to question OpenAI's valuation logic:
"Nvidia basically has a monopoly on the market, with rapid growth, extremely high profit margins, and ample cash flow. So why would you give OpenAI a higher valuation?"
Path to Profitability: Burning Cash Until 2030, Will the Public Market Accept It?
According to the report, OpenAI itself predicts that the company will continue to incur losses until at least 2030. This timeline makes public market investors, accustomed to scrutinizing profitability, quite uneasy.
Some investors worry whether the funds raised from the IPO can support OpenAI until it reaches profitability, or whether it will still need to raise funds again届时, thereby diluting existing shareholder equity.
Mark Malek, Chief Investment Officer of Siebert Financial, stated that even if OpenAI finds it difficult to achieve significant profitability in the short term, he would still consider building a position after the IPO, but would strictly control the position size—this is the same strategy he used when investing in Palantir.
Palantir currently has a high price-to-sales ratio of 49 times, with growth far exceeding peers, but Malek believes Palantir's risk is still lower than OpenAI's because its cost structure is more flexible.
"If Palantir loses a government contract, that's bad, but they can lay people off. If you spend five years building a data center, you can't say 'forget it, don't want it anymore.' Palantir is driving a Formula One car, while OpenAI is driving a fully loaded freighter."
Analysts at J.P. Morgan noted in a report this January that OpenAI's move to introduce ads in ChatGPT helps retain users, but also observed that客户sentiment towards OpenAI was "mixed" after the company announced large-scale chip and data center spending plans.
Not everyone is观望—some investors have explicitly stated that once OpenAI goes public, they will consider shorting its stock, betting that the public market's tolerance for its long path to profitability is limited.
Chanos holds a similar position. The core logic he conveys to clients is: "You should be long the output of chips, short the places where chips are stored." The implication is that operating data centers itself is not a high-return business, and OpenAI's business model is highly dependent on large-scale computing infrastructure investment.
Chanos also pointed out that there is a severe lack of financial information about OpenAI on the market currently, making in-depth analysis difficult. But he expects that once OpenAI formally submits its IPO application, the public market will engage in fierce debate over its competitive landscape:
"Is this a winner-take-all, or is the market fragmented like cloud computing? Or like search engines, where one company becomes the standard and maintains it for a long time? For now, the various models are still constantly surpassing each other."
Anthropic's Disruption: Competitor Diverts Funds and Attention
OpenAI's path to an IPO also faces potential pressure from competitor Anthropic.
At this week's Morgan Stanley Annual Technology Conference, Anthropic CEO Dario Amodei disclosed that the company's annualized revenue run rate has doubled to $20 billion. Anthropic recently completed a new round of financing with a valuation of $380 billion, and its enterprise products like the AI programming tool Claude Code are seeing strong sales momentum.
The Information previously reported that Anthropic expects its cost expenditures on AI model training and operations in the coming years to be significantly lower than OpenAI's. Some investors are beginning to believe that,凭借success in the enterprise customer market—customers who are willing to pay a premium for AI services—Anthropic's long-term profitability may be better than OpenAI's.
As Anthropic also prepares for an上市, the IPOs of the two companies could form a competition, further分散investors' funds and enthusiasm. Investors like Chanos have explicitly stated a preference for Anthropic's relatively restrained computing power investment strategy, viewing it as a more prudent and sustainable business path.
Twitter:https://twitter.com/BitpushNewsCN
Bitpush TG Discussion Group:https://t.me/BitPushCommunity
Bitpush TG Subscription: https://t.me/bitpush
