Skepticism is mounting about the viability of Oracle’s multiyear deal with OpenAI, a company that has yet to report a profit.
Oracle’s stunning results on Tuesday inspired confidence from the market about the longer term viability of AI demand.
However, the bulk of this tally is tied to just one customer: OpenAI .
OpenAI has already committed more than $100 billion to other AI and neocloud projects while having about $12 billion in annual recurring revenue.
Davidson managing director Gil Luria, OpenAI would need more than $300 billion in revenue to “justify that level of spend.”
Oracle’s multiyear agreement with OpenAI, a business that hasn’t turned a profit yet, is becoming less and less viable. The market became more optimistic about the long-term sustainability of AI demand after Oracle’s impressive results on Tuesday. Oracle’s stock surged 36% on Wednesday, marking its highest day since 1992. However, given its high leverage and customer concentration, questions are beginning to arise about the company’s growth prospects. According to Oracle, its remaining performance obligations, or RPO, have increased by 359 percent to $455 billion from the previous year. The market is placing a wager on OpenAI’s ability to fulfill its commitment, though, as the majority of this total is linked to just one client: OpenAI dot. Beginning in 2027, the AI company agreed to purchase $300 billion worth of computing power from Oracle over the following five years, according to a Wall Street Journal report that CNBC verified. With roughly $12 billion in recurring revenue annually, OpenAI has already committed over $100 billion to other AI and neocloud projects. In line with DdotA. According to Gil Luria, managing director of Davidson, OpenAI would require over $300 billion in revenue to “justify that level of spend.”. Luria is one of the few Wall Street analysts who, following Oracle’s quarterly report, maintain a neutral rating for the company’s stock. According to Luria, “the report that it came almost entirely from OpenAI significantly tempers our enthusiasm for Oracle’s backlog announcements.”. Given Microsoft’s ownership of the company, he anticipates that this problem will be fixed by next year, but he pointed out that OpenAI’s nonprofit status is restricting its capacity to receive the $40 billion it raised in March. Oracle stock, which is up 75% for the year, may increase by less than 3% from Friday’s close, according to Luria’s $300 price target. Friday saw a 5% decline in shares as investors cashed in on gains following this week’s more than 25% surge. Mark Murphy, an analyst at JPMorgan, is likewise rated as neutral on Oracle, according to Murky. He claimed that after this week’s gains, the risk-reward ratio appears balanced. Although that suggests shares could drop more than 7%, he raised his year-end price target by $60 to $270. The fact that over 70% of Oracle’s overall revenue comes from recurring revenue pleases Murphy. Additionally, he anticipates that because the company is a “relative safe haven” in the software industry, it could gain from a wider market rotation into value. The analyst’s primary concern, though, continues to be the hazy profiles of Oracle’s contracts. There are at least two other contracts worth’multi-billions’ that Oracle is closing outside of OpenAI, but we don’t know if those multi-billions total $4 billion of the $455 billion RPO balance, or $10 billion, $50 billion, $100 billion, etc. “There are many tech companies that have net debt, little cash, and/or will be burning cash for the next several years — that list is not limited to Oracle itself,” Murphy wrote in a note to clients on Wednesday. “This is a huge swing factor when trying to assess customer concentration risk and likelihood of customers being able to fund $455B worth of future payments.”. The analyst went on to say that “deeper thinking takes longer to settle in, sometimes requiring quarters or years, while short-term trading responses tend to be headline driven and controlled by the instinctual mind.”. Oracle’s cash holdings give a different picture than its peers, despite the fact that its stock gains moved the company to the tenth-largest position in the S and P 500. At the end of the May 31 quarter, the company had $11.2 billion in cash and short-term investments, a substantial decrease from the “Magnificent Seven” cohort. For example, Alphabet, the parent company of Google, and Microsoft each have a cash moat of roughly $95 billion. Indeed, Oracle’s fiscal year 2025 free cash flow of $5.8 billion represented a significant increase over the $394 million reported the year before. Oracle stock performance for ORCL 1Y mountain during the previous 12 months. According to Gary Marcus, a professor at New York University and the founder of Geometric Intelligence, Oracle’s massive stock move this week may be an indication of a peak AI bubble. Marcus asserted that many people are unaware of OpenAI’s vulnerabilities. To support his argument, the AI researcher pointed to Chat GPT-5’s lack of a technical advantage and the unsatisfactory response to its release. “There are numerous reasons to question whether OpenAI can actually accomplish this. dot. There are numerous ways for things to go wrong for them, including having a customer with whom you have an agreement who might not even be able to pay the bills, and Oracle, of course, having its own financial problems and not really having the chips,” Marcus stated in an interview. “I find it all to be highly speculative. dots. If they did indeed collect all $300 billion, I would be shocked. Some people have warned about the risks of an interconnected AI ecosystem on social media, which depends on strong collaborations between software developers, GPU and infrastructure hardware suppliers, and AI model creators. “How is this going to work exactly? ORCL has to buy the chips and take on more debt, while OpenAI has $10B in revenue but will spend $60B/yr in CapEx for five years,” said Ophir Gottlieb, CEO of the boutique market research firm Capital Market Laboratories, in an X post. Shanu Mathew, portfolio manager at Lazard, also questioned the Oracle-OpenAI deal, questioning whether “so much of AI feels levered to whether OpenAI can eventually generate meaningful revenues and keep financing massive amounts of infrastructure dot.”. AI is still its largest client, even when it comes to token usage (inference)—agents, vibe coding, etc. It has the feel of a small house of cards. —MacKenzie Sigalos of CNBC helped with this story.
