← The Rundown
Loading summary
Transcript1 lines
- [00:00]
A
Welcome back to the rundown for another weekend deep dive. Today we are talking about Oracle and their giant gamble on AI. Oracle has spent the last 40 plus years as a boring database company. But now they're going all in to become a critical AI infrastructure company and they're borrowing a lot of money to do it. These moves by Oracle have gotten the attention of the markets and there is growing concern that their big bet might not pay off. So in today's episode we'll break down Oracle's business model and how they went from a database company to one of the hottest AI companies in the world. World. We'll also dig into their partnerships with OpenAI and Nvidia. The bull in bear case moving forward. And if Oracle will be the poster child for the 2025 AI bubble, we got a great one for you today. Let's dive in. Now, before we sink our teeth into Oracle's business model, let's start with a quick history lesson and look at how Oracle has evolved over the years. Because the AI transition isn't the first time they've made a huge pivot. Oracle was founded back in 1977 by Larry Ellison and two other co founders. Bob Miner oats these guys read an IBM research paper about a revolutionary new way to organize data on computers called a relational database. And they basically turned that idea into a company. And what's crazy here is one of their first big customers was the CIA. During the height of the Cold War, the CIA needed a database to store all their super secret information. So Larry Ellison helped the CIA set up this internal project which they called Oracle. The CIA deal gave the company money and legitimacy. And Larry liked the name so much that the CIA used that he eventually called his entire company that. And here's the thing about Larry Elson. He's known to be one of the most competitive, aggressive CEOs in tech history. Throughout the 80s and 90s and 2000, Oracle absolutely dominated the database market. If you were a Fortune 500 company, you pretty much used Oracle as your database. That made Oracle a critical part of every big company's IT department. And Oracle was printing money. But Starting in the 2000s, there was the rise of cloud computing. And that was a problem for Oracle. The Oracle's databases typically lived in a company's server room. But with the rise of cloud computing, businesses started ditching on site servers and moving workloads to the cloud to companies like Amazon's AWS and Microsoft's Azure. You know, the cloud was a more flexible option than just buying more Oracle boxes for the server closet. So Oracle eventually started building out their own cloud platform to compete. It's called Oracle Cloud Infrastructure or oci. But Oracle was late to the game compared to Amazon, Microsoft and even Google. So Oracle is now betting the entire company on turning their cloud business into critical infrastructure for for an AI future in hopes of catching the other cloud giants. So let's dig into Oracle's cloud business and their huge bet on AI. For a long time Oracle was like the landlord of the corporate world. Corporations paid them a massive check every year to license their database tech and Oracle also got paid again to provide support. When things broke down it was a beautiful high margin business model. But today things are different. Oracle has spent the past decade pivoting to be a full on cloud company and and today the cloud business makes up half of their revenue. Now digging into Oracle's most recent quarterly report, their total revenues were $16.1 billion. With cloud accounting for 50% of that total revenue. A year ago, cloud only made up 42%. Today their old school software database business only accounts for 36% of their total revenue, down from 44% a year ago. Cloud is now both the biggest and fastest growing segment for Oracle. Last quarter cloud revenue grew 34% while the old database software business shrank 3%. But Oracle's cloud business is actually split two buckets. There's the cloud application and the cloud infrastructure. And it's Oracle's cloud infrastructure business, or OCI that is the main story right now. It's what investors care about the most. This business includes Oracle's data centers where AI workloads live, powered by massive GPU clusters. And the demand for AI is what's really driving OCI's growth right now. Now Oracle's AI backlog has absolutely exploded. Last quarter cloud infrastructure revenue grew 68% to $4.1 billion. The cloud application business only grew 11% to 3 point billion. I mean, the demand for AI is so much that Oracle has to build more data centers and buy more GPUs. So now let's dig into the bull and barricades for Oracle moving forward. So now that we have a cleaner picture of what Oracle actually does, let's break down the bull case for Oracle Emerging as a AI winner. Oracle has a massive deal pipeline thanks in large part to OpenAI. Back in September, the two companies struck a $300 billion deal for Oracle to supply the ChatGPT maker with computing power over the next five years beginning in 2027. These are sales generated by Oracle's cloud infrastructure segment that we just talked about. Oracle expects their OCI revenues to grow 77% to $18 billion this fiscal year and then increase to $166 billion over the next subsequent four years. And most of that revenue is already booked in Oracle's remaining performance obligations, AKA rpo. Basically, this number shows what future revenues are locked in with existing contracts. Now, that revenue hasn't been recognized yet, but it's a clear indicator of growth and and provides a level of forward visibility. It was this metric that really got investors attention earlier this year. It exploded to $523 billion last quarter, which is a 438% increase from a year ago. And that number hasn't slowed down. According to their latest earnings report, it was up $68 billion from the previous quarter. Management says this increase was driven by new commitments from Metta, Nvidia and others. Right now, Oracle only commands just 3% of the global market share in cloud, trailing Amazon, Microsoft, Google and even Alibaba. The top account for 62 to 63% of the market. But some on Wall street think that it's early innings for Oracle. Wells Fargo recently initiated coverage on the stock with an outperform rating and a 280 price target. And they said that they think that Oracle can grow its share of the infrastructure cloud market to 16% by 2029. Oracle is banking on two main things to differentiate themselves from the other cloud companies. For one, it's the relationship with Nvidia. Oracle's data centers are stacked with Nvidia GPUs and Oracle has a strong relationship with Nvidia which so they're able to get access to the latest chips before the other companies. The other thing that Oracle has going for them is that they're a neutral player. Tech analyst and author of the strategic newsletter Ben Thompson calls Oracle the TSMC of AI infrastructure, similar to how TSMC is a neutral player when it comes to manufacturing chips. You know, they make chips for Nvidia, Apple, Broadcom, pretty much everyone. TSMC doesn't try to compete with their customers by designing and making their own chips. That's one way that TSMC was able to take market share from intel over the past couple decades. Oracle could also play that same neutral role when it comes to AI cloud companies like Meta and OpenAI might prefer working with Oracle because they don't have to compete in other areas like they would if they were working with Amazon's AWB or Google's cloud. So that's the bull case for Oracle. That's how they differentiate themselves from other hyperscalers. But as Oracle builds out their AI infrastructure footprint and their data centers, they're taking on a lot of debt to do that. I think it's a good time to talk about the bear case for Oracle moving forward. All right, so let's now move to the bear case when it comes to Oracle because while there's plenty of upside, there are some real risks to evaluate. The one that we have to start with is that unlike the hyperscalers like Meta, Amazon and Google, Oracle is funding their AI build out with debt. Oracle is borrowing a lot of money and their cash burn accelerated in the last quarter with free cash flow dropping to negative $10 billion. The company now has over $106 billion in debt, according to Bloomberg. And Morgan Stanley estimates their debt will balloon to $290 billion by 2028. S&P Global says that Oracle's capex spending is projected to climb 60 in fiscal 2026 to $36 billion after more than tripling already in 2025. The problem for Oracle is they don't have the same rock solid balance sheet as the other top tech companies. Companies like Google, Meta, Microsoft and Amazon all have existing businesses that print a lot of cash, so they're able to use that cash and fund it towards their AI buildout. Oracle doesn't have that. In fact, they're the only hyperscaler to have negative free cash flow right now. And I think markets are now recognizing that risk. The price of Oracle's credit default swaps are going up. A credit default swap is insurance against a company defaulting on its debt. So the market is sending a clear signal that they're worried that Oracle might not be able to pay back its debt. If Oracle fails to show significant revenue materializing from all this AI investment, I think their credit default risk is only going to get higher. The other concern when it comes to Oracle is their customer concentration. Most of their money is expected to come from OpenAI. Now Oracle has signed major cloud deals with OpenAI. But what happens if OpenAI can't make these payments? That's a question that Oracle executives don't have a good answer for. And OpenAI is facing a lot of competition now from companies like Google. So when you combine the debt burden along with the customer concentration risk, investors are now starting to price in the real risk that Oracle faces. At the time of this recording, Oracle stock is down nearly 50% from its peak back in September. And this is wild. As I'm recording this, we Just got breaking news that Oracle has pushed back the completion dates for some of their data centers that is developing for OpenAI from 2027 to 2028. These delays are due to labor and material shortages. According to Bloomberg, this news is sending Oracle stock even lower today. And it's just another example of how ambitious Oracle's original timelines were when it came to their data center build out. And that they could face further delays which could delay the revenues they make from these data centers. The interest payments on all that debt they took out, that doesn't get delayed, they still have to pay that. So what's my take here? Well, I think that Oracle could end up being the poster child of the 2025 AI bubble. They are quite literally betting the entire company on AI by taking on a lot of debt. And when high levels of debt enter the picture, that's when bad things can happen. Looking back at previous bubbles, the reason companies went bankrupt is because they borrowed too much money. Now Wall street is starting to get nervous. You know, on one hand, I do respect Oracle for going big here. Larry Ellison is now 81 years old and he could just be chilling on his Hawaiian island right now, but instead he's betting his company on AI to dethrone Amazon and Microsoft. And look, this bet could still pay off, but the risks are really high at this point. If you're investing in Oracle, it's like buying a highly leveraged call option on the future of AI. If you're right, it could pay off massively. If you're wrong, it could end in disaster and we might look back at Oracle's huge bet as a warning sign that the AI bubble was about to burst. Well, all right guys, hope you enjoyed this weekend's deep dive episode about Oracle. Let me know in the comments what you thought about today's deep dive. As you could probably tell, I'm pretty bearish on Oracle, but let me know in the comments if you think that it might be an overreaction. Thank you guys again for listening, watching and commenting. Shout out to Mike and Connor for all the work behind the scenes and we'll see you guys back here tomorrow.