What the heck is going on with Oracle?
It doesn't matter whether you care about this stock or not. The news coming out of it are a must-know for anyone interested in financial markets or the economy in general.
TLDR Summary
Oracle is on a mission to rival the Big Three cloud providers and its cloud ambitions are part of a theme that will determine GDP growth in the US and globally as a whole. They are guiding for a 10x revenue increase in cloud computing in just four years which would transform Oracle Cloud from a nobody with 3% market share to a leading player with a business larger than AWS is today.
Most importantly, this guidance is based on contracted revenue, i.e. revenue that their customers (most importantly OpenAI) basically only can get out by declaring bankruptcy. It lends credibility to Oracle’s technological capabilities and to the AI compute infrastructure boom as a whole.
In my opinion, ERP providers were originally in the pole position to rule the cloud. They had the customer relationships, the technological expertise and the data advantage. But they got played Big Tech. They were too hesitant and too comfortable. This may be the moment where the table turns. The moment where Oracle starts playing Amazon like Amazon started playing Oracle 15 years ago.
It’s frankly hard not to be completely blown away by Oracle’s latest investor communication. I need to rethink my skepticism of my latest AI related posts. Reality is obviously not black and white, it’s grey. My perspective has changed to a lighter shade of grey this week.
Company background
Oracle grew up as a relational database provider as many companies did in the enterprise software space. As such, they were actually well positioned to lead in cloud computing when it took off. They had the customer relationships, the know-how and the data needed to succeed.
They are the record keepers for large enterprises. Every business process runs through them. This gave and gives them great insight into corporate processes and it makes their services very sticky once they are in.
It would have been a great selling point to tell their customers to remove their costly and complex on-prem systems and simply subscribe to a cloud service. As a German, I have often argued that SAP should have built AWS. Oracle was late to the game as well. They didn’t push aggressively into the cloud space until the late 2010s.
As a result, others are leading cloud computing. The global cloud computing industry generated $99bn in revenues in 2Q25. The Big Three, AWS, Azure and Google Cloud dominate the market with a combined market share of 63%. So far, Oracle has been on fourth place with just 3% market share.
Latest earnings
Oracle reported their 1Q26 earnings on September 9, 2025. Total revenues grew by 12% YoY to $14.9bn, mostly driven by the cloud segment which grew at 28%. Operating income advanced by 7% YoY, net income was flat.
So far so unimpressive. What was impressive though was the 359% (!) increase of remaining performance obligations, in short RPO, to $455bn.
RPOs are contracted future revenues. Some of this is billed and as such recorded in the balance sheet as deferred revenue. But most of it is not and therefore won’t show up in the balance sheet. It merely gets disclosed as an off-balance sheet item.
A day after earnings, the Wall Street Journal reported that $300bn of the $317bn RPO increase relates to a five year computing deal struck with OpenAI. It’s the next step for OpenAI to diversify from Microsoft. Previously, they contracted with Coreweave and also with Google. $300bn is $60bn per year, about 4x what they likely will spend on compute this entire year. OpenAI’s pace of securing computing supply can only be described as astonishing, especially given the financing of all of that is still very much in question.
Much of the Oracle/OpenAI partnership gravitates around the infamous Stargate project, the essence of which is Oracle buying and operating NVIDIA GPUs for OpenAI. For example, in May 2025, Reuters reported that Oracle plans to buy $40bn worth of NVIDIA GPUs to power OpenAI’s new US data center situated in Abilene, Texas.
Management also provided mid-term cloud revenue guidance in conjunction with the sharp RPO increase:
The enormity of this RPO growth enables us to make a large upward revision to the cloud infrastructure portion of our financial plan. We now expect Oracle Cloud Infrastructure will grow 77% to $18 billion this fiscal year and then increase to $32 billion, $73 billion, $114 billion and $144 billion over the following 4 years. Much of this revenue is already booked in our $455 billion RPO number, and we are off to a fantastic start this year.”
Safra Catz, CEO, 1Q26 earnings call
This means that Oracle’s FY30 (which ends in May 2030) is guided to come with $144bn in cloud revenue. It would then be larger than AWS is today! Wow.
Even more importantly, much of this will likely be AI cloud computing, i.e. deep learning for data/content generation. In the article below, I estimated that only about $40bn of the Big Three’s $260bn cloud revenue this year will be ‘AI’.
It’s absolutely possible that Oracle will emerge from all of this as THE AI cloud operator, playing Amazon like Amazon played them in the original cloud revolution a decade ago.
Oracle’s share price exploded based on this possibility. It finished up 36% on the day after earnings. Some consolidation has happened subsequently, but the stock is still up more than 20%. A monster move for such a megacap. I wouldn’t be surprised if this runs much more in the near-term once the consolidation is complete and the implications start getting fully digested. Most people online are making fun of the ridiculousness of the numbers and the capex requirements this comes with. Purely sentiment speaking, there is more upside than downside.
All of this is seemingly coming out of nowhere. It begs the question:
Why is Oracle suddenly so successful in Cloud?
1. Oracle is big in enterprise data and that is the most valuable data.
Oracle is the largest ERP vendor in the world and proprietary enterprise data are some of the most valuable data out there. This makes them a natural winner in inferencing, i.e. generating insight and content based on source data.
“It's AI inferencing that will change everything. Oracle is aggressively pursuing the AI. And we're not doing badly in the AI training market, by the way. With inferencing figure, oracle is aggressively pursuing the inferencing market as well as the AI training market. We think we are in a pretty good position to be a winner in the inferencing market because Oracle is by far the world's largest custodian of high-value private enterprise data.
Larry Ellison, 1Q26 earnings call
2. Oracle has made its databases AI-ready to capitalize on its data advantage.
Enterprise data is very valuable, but it has so far proven difficult for enterprises to fully leverage them with AI tools. A controversial MIT paper made the rounds some weeks ago for concluding that 95% of organizations are seeing no business return from their AI spending. Part of the reason may be that the AI models don’t have full access to all data required to unleash their full potential. This may have to do with the way enterprise data is organized.
Like most other ERP vendors, Oracle has historically stored data in rows and columns which is optimized for transaction queries, but not for AI models which require unstructured data such as text, images or audio to capture semantic meaning. Oracle claims that their data vectorization solves that issue.
With the introduction of our new AI database, we added a very important new way for you to store your data in our database. You can vectorize it. And by vectorizing it, by vectorizing all your data, all your data can be understood by AI models. Then we made it very easy for our customers to directly connect all their databases, all their new Oracle AI databases and cloud storage, OCI Cloud storage to the world's most advanced AI reasoning models, ChatGPT, Gemini, Grok, Llama, all of which are uniquely available in the Oracle Cloud.
After you vectorize your data and link it to an LLM, the LLM of your choice, you can then ask any question you can think of. For example, how will the latest tariffs impact next quarter's revenue and profit? You asked that question, the large language model will then apply advanced reasoning to the combination of your private enterprise data plus publicly available data.”
Larry Ellison, 1Q26 earnings call
3. Oracle is a trusted record keeper
Part of the limited success of enterprise AI projects so far may also be related to the hesitation of companies to share the most secret data with AI models, not knowing how they will be used. Oracle promises to connect the companies data safely to the most powerful AI models.
“But unless you have a database that is secure and reliable and linked to all of the popular LLMs, and we've done all of that, unless you have that and you have to tell me who else has that besides Oracle. Unless you have that, it's going to be very hard for you to deliver a ChatGPT-like experience on top of your data as well as publicly available data. That's a unique value proposition for Oracle. And that's because, again, we're the custodian of all of much more data than any of the application companies. They have their application data. They measure their customers in tens of thousands. We measure our customers in millions of databases. So we think we're better positioned than anybody to take advantage of inferencing.”
Larry Ellison, 1Q26 earnings call
4. Oracle claims to be orders of magnitude cheaper than the Big 3.
This claim is hard to verify or rationalize, but it’s at least worth noting that Oracle management claims to have a huge cost advantage in cloud computing:
“We have gotten the entire Oracle Cloud, the whole thing, every feature, every function of the Oracle Cloud down to something we can put into a handful of racks, 3 racks, we call it Butterfly that cost $6 million. So we can give you a private version of the Oracle Cloud with every feature, every security feature, every function, everything we do for $6 million. I think the cost for the other hyperscalers is more than 100x that. So we can actually give our customers clouded customer, the full cloud customer.”
Larry Ellison, 1Q26 earnings call
“We do not own the buildings. What we do own and what we engineer is the equipment. And that's equipment that is optimized for the Oracle Cloud. It has extremely special networking capabilities. It has technical capabilities from Larry and his team that allows us to run these workloads much, much faster. And as a result, it's much cheaper than our competitors.”
Safra Catz, CEO, 1Q26 earnings call
5. Oracle is both application and infrastructure
AI has put SaaS into a very difficult spot. Software moats are drying up because the cost to develop software is falling rapidly. On the other hand, pure infrastructure businesses like Coreweave for example will have to deal with commoditization pressure as well. Oracle combines enterprise software expertise with enterprise infrastructure which may prove to be a powerful combination, providing a seamless experience for customers.
Analyst reactions
In my opinion, this Oracle earnings release is a profound one, similar to NVIDIA’s 1Q24 release in that it may build the foundation for a multiyear narrative.
But don’t take my word for it. Check out the exhilaration expressed by some of the analysts on the call. The vibe on the call was something to remember.
“I have been doing this a really long time, and I tell my old team, pay attention to this, even those that are not working on Oracle because this is a career event happening right now, and it looks -- it's just amazing.”
“Great. And I think we're all kind of in shock in a very, very good way. Larry, there's no better evidence of a seismic shift happening in computing than these results that you just put up.”
“Great. I'll echo my congratulations on this momentous quarter.”
“I originally was going to ask you if the new Oracle AI database really opens up the general enterprise inferencing market. And based on your script, it sounds like the answer to that question is hell yes.”
Sincerely,
Rene