Shares of Oracle (NYSE:ORCL) rocketed higher earlier this year when it announced a deal worth about $300 billion to provide computing capacity to OpenAI.
Since then, however, Oracle has retreated just as drastically and is now down almost 35 percent in the last three months.
Is Oracle stock undervalued today, or did the stock’s selloff accurately reflect the risks that surround the business?
Where Is Oracle Trading Now?
After coming off of its highs earlier this year, ORCL is trading at about 35.7 times trailing 12-month earnings, 9.0 times sales and 86.7 times operating cash flow.
Though this is quite a premium price tag, analysts still appear very bullish on Oracle’s potential returns. The range of price targets for the stock runs from $175.14 to $400, with an average of $292.85.
The most bearish target would represent a downside of just 7.8 percent from the current price of $189.97, while the average target implies an upside of over 54 percent.
Remaining Performance Obligations at Oracle
Looking into the revenue projections offered by Oracle, it isn’t difficult to understand why its shares surged in the wake of the OpenAI agreement. Oracle expects annual cloud revenue to grow to as much as $144 billion by the beginning of the next decade, well over double the total revenue it has achieved in the last 12 months. Total remaining performance obligations stand at about $455 billion, of which the majority is accounted for by OpenAI.
It didn’t take long, however, for the market to start rethinking its bullishness on Oracle. Chief among the worries raised by investors was the concern that OpenAI may not actually be able to hold up its end of the agreement. OpenAI reportedly produced about $12 billion in ARR around the time the deal was made public and had only raised around $60 billion worth of lifetime funding. With massive losses still ongoing at the AI startup, it’s unclear whether even its strong growth will enable it to meet its obligations to Oracle.
Underlying this concern are related worries about the capital Oracle will have to deploy to ramp up its capacity to meet OpenAI’s needs. CapEx is expected to come in at about $35 billion for the current fiscal year and rise to around $80 billion in 2029.
Though necessary to fulfill the terms of its agreement with OpenAI, this level of spending introduces additional risks for ORCL shareholders. The business already has nearly $130 billion in long-term debt, and the high level of spending Oracle will have to maintain over the next five years is expected to produce half a decade of consistently negative cash flows.
Due to a convergence of these factors, Oracle has also become something of a canary in the coalmine for those worried about a growing bubble in AI stocks. Oracle will have to deploy significant amounts of capital now in exchange for future revenues that are largely tied to a single high-risk AI customer. If OpenAI stumbles or artificial intelligence itself proves to be less valuable than currently expected, Oracle and many other businesses could see a large selloff.
A final point of concern about Oracle is the increasing worry that data center infrastructure built today could have a rather short productive lifespan. Between the rapid pace of improvements that could make GPUS purchased today obsolete within a few years and often high maintenance costs, the infrastructure that Oracle and other hyperscalers are building today could struggle to produce positive profits. Though this problem is far from unique to Oracle, it does add one more layer to the increasingly complex web of valuation issues that the business appears to be facing.
So, Is Oracle Undervalued?
Right now, much of Oracle’s valuation seems to hinge on OpenAI’s prospects. It’s worth noting that Oracle has maintained a fairly optimistic tone. Management recently pushed back on reports that shortages of labor and materials would delay the opening of new data centers for OpenAI into 2028. So far, management insists that the OpenAI deal is on schedule and that Oracle is confident in OpenAI’s ability to pay.
Even so, the risks around OpenAI’s association with Oracle are hard to ignore. Oracle, already carrying a significant debt load, will have to leverage itself even more highly to build capacity for a business that could be ordering well beyond its means. As noted above, the infrastructure it’s building may well also have a very limited lifespan, calling the economics of the new data centers into question.
This brings up the question of whether Oracle could be a good value if OpenAI fails to come through. After all, AI computing capacity is in extremely high demand, and Oracle could find other customers among hyperscalers as it continues its push to become a leading compute provider. Unfortunately, both the high levels of debt Oracle is taking on and the expected restrictions on its cash flows could be deeply negative without customers lined up for the new infrastructure being built.
At the end of the day, Oracle’s valuation seems to come down to whether or not OpenAI can pay for $300 billion of data center capacity. One of the main tools it could have in attempting to do so would be raising capital through a large IPO, something that could happen within the next two years.
As of right now, though, OpenAI hasn’t made concrete plans to go public. Even so, OpenAI’s status as the central hub of the AI world could give it considerable capacity to raise funds, especially if the AI industry can get through the current worries about a bubble forming.
If Oracle’s revenues really skyrocket as expected, the stock could prove to be drastically undervalued. A more realistic case, however, could be moderate undervaluation caused by a sudden deterioration of investor sentiment. Oracle’s sudden selloff likely has left it a bit undervalued, though only OpenAI’s own future performance can reveal exactly how much. Even with the challenges it’s facing, Oracle appears well-positioned to become a key provider of infrastructure to the AI world. This would seem to support long-term growth at Oracle, even if the immediate outlook does appear a bit rocky.
The author has no position in any of the stocks mentioned. Financhill has a disclosure policy. This post may contain affiliate links or links from our sponsors.