Oracle Corp (ORCL) — Q1 2025 Earnings Call Transcript

Thank you, Sarah, and good afternoon, everyone, and welcome to Oracle's First Quarter Fiscal Year 2025 Earnings Conference Call. A copy of the press release and financial tables, which includes a GAAP to non-GAAP reconciliation and other supplemental financial information can be viewed and downloaded from our Investor Relations website. Additionally, a list of many customers who purchased Oracle Cloud Services or went live on Oracle Cloud recently will be available from the Investor Relations website. On the call today are Chairman and Chief Technology Officer, Larry Ellison; and Chief Executive Officer, Safra Catz. As a reminder, today's discussion will include forward-looking statements, including predictions, expectations, estimates or other information that might be considered forward-looking. Throughout today's discussion, we will present some important factors relating to our business, which may potentially affect these forward-looking statements. These forward-looking statements are also subject to risks and uncertainties that may cause actual results to differ materially from statements being made today. As a result, we caution you against placing undue reliance on these forward-looking statements, and we encourage you to review our most recent reports, including our 10-K and our 10-Q and any applicable amendments for a complete discussion of these factors and other risks that may affect our future results or the marketplace of our stock. And finally, we are not obligating ourselves to revise our results or these forward-looking statements in light of new information or future events. Before taking questions, we'll begin with a few prepared remarks. And with that, I'd like to turn the call over to Safra.

Thanks, Ken, and good afternoon, everyone. Before I go to our Q1 numbers, I thought I'd take a moment to review some of the things that you'll be hearing about over the next couple of days from Oracle. We are at Cloud World in Las Vegas, and Cloud World is where we come together with our customers and partners to share experiences and showcase our latest products and services. Our customers are our best people when they share how our technologies transform their enterprises. The innovations from our labs and research centers, in combination with feedback from our customers, have helped us build superior products and services. You'll hear about new cutting-edge features within OCI, database, analytics, Fusion, NetSuite, and our industry applications. We will also be showing new capabilities that we've been working on for a while, including embedded AI agents in Fusion and those that drive productivity and efficiencies for our customers when they're rolled out. As an Oracle Fusion customer myself, I take great pride once again in my team's ability to have us announce earnings and give guidance nine days after the quarter ended. Many of our customers ask us how to replicate our results. At Cloud World, we will be having a lot of Oracle playbook conversations this week. You've already seen today's announcement of our partnership with Amazon Web Services, which has now joined Microsoft Azure and Google Cloud in making OCI and Oracle available in their respective clouds. Needless to say, we think our multi-cloud strategy will expand the ubiquity and popularity of our differentiated technologies, especially the Oracle Database. Larry will share more details in just a moment. But now to Q1, which was clearly another outstanding quarter, with total revenue at the high end of my guidance and earnings per share of $0.04 above the high end of guidance. Currency was essentially in line with my guidance. And as usual, I'll be discussing our financials using constant currency growth rates as this is how we manage the business. Total cloud revenue, that’s SaaS and IaaS, was up 22% at $5.6 billion, with SaaS revenue of $3.5 billion, up 10%, and IaaS revenue of $2.2 billion, up 46%, on top of the 64% growth reported last year. As a reminder, we exited the advertising business last quarter, which had the effect of lowering the total cloud applications revenue by 2% this quarter. Total cloud services and license support for the quarter was $10.5 billion, up 11%, driven again by our strategic cloud applications, autonomous database, and OCI. Application subscription revenues, which includes product support, were $4.8 billion and up 7%. Our strategic back-office SaaS applications now have annualized revenues of $8.2 billion and were up 18%. Infrastructure subscription revenues, which includes license support, were $5.8 billion and up 14%. Infrastructure cloud services revenue was up 46% and up 49% when you exclude our legacy hosting services. Our infrastructure cloud services now have an annualized revenue of $8.6 billion; OCI consumption revenue was up 56%, and demand continued to outstrip supply. Cloud database services, which were up 23%, now have annualized revenues of $2.1 billion. Very importantly, as on-premise databases migrate to the cloud, on OCI directly or through our database cloud services with Azure, Google, and AWS, we expect those cloud database revenues collectively will be the third leg of revenue growth alongside OCI and strategic SaaS. Database subscription revenues, which includes database license support, were up 4%. Software license revenues were up 8% to $870 million, including Java, which saw excellent growth. So all in, total revenue for the quarter was $13.3 billion, up 8% from last year. Shifting to gross profit and operating income, the gross profit dollars of cloud services and license support grew 9% in Q1 as our cloud businesses continue to scale. The gross margins of both cloud applications and cloud infrastructure have been climbing higher. We continue to display operating expense discipline with Q1 operating income growing 14%, and the operating margin was 43%. The non-GAAP tax rate for the quarter was 18.9%, with non-GAAP EPS at $1.39, up 17% in USD, up 18% in constant currency. The GAAP EPS was $1.03, up 20% in USD and up 22% in constant currency. Included in my guidance at the beginning of the quarter was the expected completion of an assessment of the useful lives of our server networking equipment, including an increase of the estimated lives from five years to six years, effective at the beginning of this fiscal year. This change in accounting estimate reduced Q1 operating expenses by about $197 million. At quarter end, we had nearly $11 billion of cash and marketable securities, a short-term deferred revenue balance of $11.5 billion, up 2%. Operating cash flow for Q1 was $7.4 billion, while free cash flow was $5.1 billion. On a trailing 12-month basis, operating cash flow was $19.1 billion and free cash flow was $11.3 billion. Our remaining performance obligations, or RPO, is now $99 billion, up 52% in constant currency. Now while we typically see a seasonal decline of RPO in Q1, we signed several large deals this past quarter, resulting in a sequential increase in RPO compared to the decline that we typically see based on our experience over the previous five years. Further, our cloud RPO grew more than 80% and now represents nearly three-fourths of total RPO. Approximately 38% of total RPO is expected to be recognized as revenue over the next 12 months, which reflects the growing trend of customers wanting larger and longer contracts as they see firsthand how Oracle Cloud services are benefiting their businesses. We spent $2.3 billion on CapEx this quarter. Given the demand that we see in our RPO growth and the additional demand we have in our pipeline, I expect fiscal year 2025 CapEx will be double what it was in fiscal 2024. As always, we remain careful to pace our investments appropriately and in line with booking trends. We now have 85 cloud regions live, with another 77 planned, and more to follow. We have public cloud regions, dedicated cloud customer regions, national security regions, sovereign regions, Oracle alloy regions with our partners, and multi-cloud regions with Azure and Google Cloud, and now shortly with AWS as well. This sizing flexibility and deployment optionality of our cloud regions continue to be significant advantages for us in the marketplace. As we've said before, we're committed to returning value to our shareholders through technical innovation, strategic acquisitions, stock repurchases, prudent use of debt, and dividends. This quarter, we repurchased 1.1 million shares for a total of $150 million. Additionally, we paid out dividends of $4.4 billion over the last 12 months. The Board of Directors again declared a quarterly dividend of $0.40 per share. Before I dive into specific Q2 guidance, I'd like to share some overarching thoughts and the benefits that I expect will bring over the coming years. First, the Oracle database is thriving, and the multi-cloud agreements we now have with Microsoft, Google, and AWS make it easier for our customers to run their Oracle databases in the cloud. Second, we are rapidly expanding our OCI capacity to meet the demand that you see in our 52% RPO cloud growth. Third, while much attention is focused on our GPU-related businesses, our non-GPU infrastructure business continues to grow much faster than our competitors. Finally, our strategic SaaS apps continue to grow, while we are starting to see more and more of our industry-based cloud apps come online. All these trends point to revenue growth going higher. We will discuss the implications of these positive trends at our financial analyst meeting on Thursday. For fiscal 2025, we remain very confident and committed to full year total revenue growth growing in double digits and full year total cloud infrastructure revenue growing faster than last year. Let me now turn to my guidance for Q2, which I'll review on a non-GAAP basis. If currency exchange rates remain the same as they are now, currency should have about a 1% positive effect on total revenue and as much as a $0.03 positive effect on EPS. However, the actual currency impact may be different. Total revenues are expected to grow from 7% to 9% in constant currency and are expected to grow between 8% and 10% in USD at today's exchange rate. Total cloud revenue is expected to grow from 23% to 25% in constant currency and 24% to 26% in USD. Non-GAAP EPS is expected to grow between 6% to 10% and be between $1.42 and $1.46 in constant currency, and non-GAAP EPS is expected to grow $0.08 to $0.12 and be between $1.45 and $1.49 in USD. My EPS guidance for Q2 assumes a base tax rate of 19%. However, onetime tax events could cause actual tax rates to vary. And with that, sorry it was so long. And with that, I'll turn it over to Larry for his comments.

Thank you, Safra. Today, Oracle has 162 cloud data centers, live and under construction throughout the world. The largest of these data centers is 800 megawatts, and it will contain acres of NVIDIA GPU clusters able to train the world's largest AI models. That's what's required to stay competitive in the race to build one of the most powerful artificial neural networks in the world. The stakes are high and the race goes on. Soon Oracle will begin construction of data centers that are more than a gigawatt. Building giant data centers with ultra-high-performance RDMA networks and huge 32,000 node NVIDIA GPU clusters is something that Oracle has proven to be very good at. It's the reason we're doing so well in the AI training business. It's important to remember that we first developed those high-performance RDMA networks to interconnect our Exadata CPU cluster hardware that powers our Exadata database cloud service. The Oracle Database Cloud service running on Exadata and Exascale RDMA clusters provides an order of magnitude better performance, scalability, reliability, and security than other databases. And it's still the world's only autonomous, fully self-driving database. Our large and loyal customer base understands and appreciates the many technical advantages of using the Oracle database. And those customers wanted us to find a way to make the very latest and best Oracle technology available on other clouds, in addition to OCI. We found a way. With today's AWS announcement, our customers will be able to use Oracle's latest Exadata and Exascale RDMA clusters with the latest versions of our database software from within Microsoft Azure, Google Cloud, and AWS. This will enable customers to use the Oracle database anywhere and everywhere. That has always worked well for our customers and for our database business. We believe our cloud partnerships with AWS, Microsoft, and Google will turbocharge the growth of our database business for years to come. Back to you.

Thank you, Larry. Sarah, if you could please poll the audience for questions.

Thank you, Larry and Safra. There are many positive aspects here, but I would like to ask about margins. You continue to report strong cloud numbers, particularly in OCI, but the guidance you provided seems quite challenging to achieve. We also expect upside in RPO, and Safra, you mentioned the sequential increase. The last time we saw a sequential increase was due to the Cerner acquisition, which added RPO. I assume a significant portion of that increase is also related to OCI, which you've indicated is primarily cloud-based. This suggests there is more growth ahead. Given the ongoing shift toward the lower margin OCI, and while I understand this is evolving, how should we consider overall margins in relation to profit for the company in the future?

Let me begin by highlighting that our database and autonomous database, which is part of OCI, is starting to grow significantly. Our multi-cloud agreements will further enhance OCI's gross margins. Even in this quarter, our gross margins increased as a percentage, despite the larger OCI footprint. Our business is just beginning to reach substantial scale. We have designed OCI to be highly automated, which aids in its management, leading to increased profitability as we expand. While percentage improvements are encouraging, our operating margins continue to rise as well. OCI encompasses not just basic storage and compute, but also many additional capabilities, including the database, which has strong margins. Additionally, our SaaS business, which is also performing well and scaling, benefits from our growth in OCI, contributing to improved high margins in the past quarter. Larry, do you want to add anything?

Let's begin with our SaaS offerings. Transitioning to the autonomous database has led to significant efficiencies. We are currently in the process of migrating Fusion and NetSuite to the autonomous database. Our decision to move everything to autonomy is driven by two main factors. First, with a fully autonomous database, there is no need for human intervention in its management. This not only results in cost savings but, more importantly, eliminates the possibility of human error, providing a substantial security advantage over our competitors. With complete automation, there are no mistakes, and everything operates seamlessly. Furthermore, the autonomous database is genuinely elastic. For instance, if a task requires 500 microprocessors for three minutes, you can access that capacity temporarily and then return it to the resource pool. This operational model is distinct from other databases, which may have some elastic features but are generally not fully elastic. Our approach significantly reduces hardware usage while enhancing speed and efficiency. The autonomous database operates without human labor, making it more secure and yielding much higher margins compared to traditional Oracle business areas. In fact, the margins for our autonomous database business are comparable to those of our SaaS operations, which are also notably high due to their reliance on the autonomous database. We maximize hardware efficiency and minimize labor usage since manual operations present security risks and hinder our expansion capabilities. Each Oracle data center, regardless of its size, maintains identical features and functionalities, differing only in scale. This uniformity enables us to utilize one suite of automation software across the board, which no other company can match. This level of automation and autonomy grants us superior margins in both our database and SaaS sectors as well as in our overall cloud services. Our automated cloud systems lead to very low labor costs and highly efficient networks. With our networks operating at significantly higher speeds, we can reduce costs by half, and our networks are faster than those of competitors, enhancing our growth potential and promising better margins moving forward.

So is it safe for me to assume that then the upside you keep putting up and you hopefully will continue to do will add profit and that profit will actually increase as a percentage of the margins will also increase over time?

I believe there are varying perspectives from different engineers as we transition Fusion to Autonomous Database. I anticipate that our cloud cost savings will be about 50%. It could be 40% or even 35%, but we will see significant cost reductions from our current state, impacting all Fusion customers. This illustrates how we are leveraging faster networks, quicker databases, and increased automation to enhance the security of our products. I want to stress that security is our main focus, but as a secondary benefit, we will also significantly reduce our data center operating expenses.

Great. Thank you, Larry. Thank you, Safra.

Thank you very much and congrats on the great performance. Larry, how do you envision the market transitioning from the AI training phase to the AI inferencing phase? There's some debate out there on whether we have an imbalance or a bubble on the front end of the curve because training is compute intensive and then perhaps it recalibrates differently somehow for the inferencing stage, which might be less intensive? Or do you see the potential for high growth all the way through both of these phases?

A lot of people believe that once they send a child to college, the education is complete and the child can immediately start working effectively. However, that's not the case. The development of neural networks is an ongoing process that requires continuous improvement, and the costs associated with this training can be enormous. When I refer to constructing gigawatt or multi-gigawatt data centers, I mean that for anyone looking to compete with advanced AI models, the initial investment needed is roughly $100 billion. Let me emphasize that, it’s about $100 billion over the next four to five years for those who want to participate in this field. It’s a significant amount of money, and the situation is only becoming more challenging. Therefore, there will be very few companies capable of developing these advanced models. Additionally, there will be numerous highly specialized models. For instance, I am involved in projects using computers to analyze biopsy slides or CAT scans for cancer detection, along with specialized blood tests for the same purpose. These models are often focused and do not usually rely on general models like Groks, ChatGPTs, Llamas, or Geminis. Instead, they are tailored to specific data sets, such as millions of biopsy slides, where other training data may not be relevant. This trend will continue, and we can expect to see more applications like this. If you’re looking at a timeline of five years or even ten years, I wouldn’t be concerned about having trained all the necessary models and merely shifting to inference. There is an ongoing competition for technological leadership that will involve only a few companies and possibly a single nation over the next five years, or more likely over ten years. This industry is expanding, and there’s no indication of a slowdown or shift on the horizon.

Thank you very much.

We're currently in the process of designing a data center that will exceed one gigawatt in capacity, and we've secured a suitable location with access to power. The site already has building permits for three small modular nuclear reactors that will supply energy to the data center. This is an indication of how rapidly things are evolving in our industry.

Just a question more on the database side on the agreements that you just announced today or that you have in place now with AWS. So now that we have all the hyperscaler agreements in place, how do you think about that migration movements from database workloads that are at the moment running on-premise or on cloud customer to the public cloud? I mean how should we think about that momentum? Thank you.

We believe it will speed up. Please continue, Larry.

No, I think you're correct. There are two key points. The public cloud is both fascinating and crucial. Oracle found significant success in the database realm many years ago because we emphasized portability. We operated on various platforms like IBM mainframes, Microsoft PCs, Hewlett Packard machines, and many others, which allowed our customers to run the Oracle database in any environment. It was essential for us to ensure that the best versions of our database, including Exadata and Exascale, were available in other cloud environments. We managed to downsize OCI enough to embed an OCI data center within platforms like Microsoft Azure or Google or AWS, allowing it to be fully autonomous and utilize Exadata and Exascale clusters. Accomplishing this was technically challenging, but we succeeded. As I previously mentioned, all our data centers share identical functions, differing only in scale. Currently, the largest operates close to 800 megawatts, while the smallest are around 150 kilowatts, and we aim to reduce that to 50 kilowatts. This flexibility means that numerous medium and large companies may opt for an Oracle private cloud. There's no real difference between our private and public clouds; they are completely identical. For example, Vodafone has six Oracle private clouds to manage their workloads. Costs have decreased to the point where any company can transition to the cloud, reaping its advantages while ensuring exclusivity. This is appealing for those concerned about security or bound by government regulations. We believe that providing Oracle database services on platforms like AWS, Microsoft, and Google is extremely significant. However, we anticipate that private clouds will greatly exceed the number of public clouds as companies choose to host Oracle Cloud behind their own firewalls without external neighbors. Our data centers are highly automated, scalable, and functionally identical, leading us to expect the establishment of 1,000 to 2,000 or more Oracle data centers globally, many of which will be dedicated to individual banks, telecommunications firms, tech companies, and government entities. It’s difficult to predict whether private or public clouds will dominate, but the positive aspect is that we benefit in either scenario.

Okay. Interesting. Very interesting.

Next question, Shara.

Thank you so much and congratulations on the quarter. Very impressive both the quarter and the guide. We've seen a lot of focus on the model training side, but less on applications and inferencing in the rest. You guys have a lot of expertise in the market and in the industry. You already have traditional AI sprinkled throughout all the Oracle products and capabilities. But where do you see the monetizable value of GenAI on the app side? How long do you think it will take for GenAI to be a meaningful revenue, not just for Oracle, but software in general, on the app side, not on the training side? Thank you.

I apologize for dominating the conversation. Let me begin with health care. We assist doctors in diagnosing various diseases. For instance, during a sonogram, I've witnessed nurses and doctors manually measuring a baby's skull and spinal cord, which is quite inefficient. This process should be automated by computers. If the umbilical cord is wrapped around the fetus, computers should be able to identify that and document it. The doctor could receive computer-aided support for these tasks, including analyzing plaque and coronary arteries. We already provide a summary for doctors before they see patients, generated through AI analysis of electronic health records and recent lab results. It informs them of any disease progression or stability prior to the patient consultation. This summary is easily understandable. AI also listens to conversations between the doctor and patient during consultations. If a doctor prescribes medication, the AI ensures the prescription is correct, enters it into the record, and updates the electronic health records based on the discussion. At the end of the appointment, the physician receives a draft to review and approve quickly. Prescriptions are then filled, orders executed, and records updated. We are already implementing these systems. In health care, there are numerous needs, such as reading X-rays and improving user interfaces. Our interface is much more advanced than Epic's. For example, while at Stanford, it took three staff members to locate my son’s X-rays. In our system, you simply ask for the latest X-ray using a voice command. The computer recognizes you by your face and voice, confirming your identity and access rights through AI. As for monetization, Cerner represents that monetization. Our ability to significantly grow our health business is rooted in AI. AI is integral, combining various systems like diagnostics, electronic health records, pharmacy, prescriptions, and user authentication. Many believe AI is a separate entity, but it's not. Our applications are primarily AI-driven, and I have difficulty understanding how others propose charging separately for each component. It's a puzzling notion to me.

Larry, I think you are the first person to explain it that way. Thank you. It makes a lot of sense.

Great. Thanks. And I will echo my Congratulations. Safra, Larry, you guys have had this big inflection in RPO growth over the last few quarters. Could you update us on how you're feeling about supply availability and your ability to stand up data center infrastructure in a time-efficient manner in order to move from contract signing to consumption and convert backlog into revenue? And I guess, are you doing anything different today than, say, a year ago to try and help accelerate these timelines?

I'm going to begin, and Larry will finish. We have significant demand, and it's true that demand is exceeding supply. However, I'm comfortable with that because we are ramping up our supply. You can see this reflected in our revenues and guidance for the upcoming year. We will share more on this during Financial Analyst Day on Thursday. We've implemented several changes, including the automation of our data center setup and deployment, as Larry mentioned earlier. Given the high demand, we need to expand our reach. A couple of quarters ago, we decided to focus on larger locations instead of smaller ones, and that strategy is proving effective for us. We are expanding quickly, not just with public cloud offerings, but also for private clouds and national security regions, which are in high demand. We are accelerating our efforts as much as possible, and automation is a key factor in this. The uniformity of our operations means there’s nothing unique about what we offer, which makes it difficult for our competitors to match us and supports our rollout.

I want to emphasize what Safra mentioned. Our private clouds are essentially the same as our public clouds, except they may have only one tenant and be located in a facility you own. Other than that, they're identical. We own and manage the hardware for you. It's just in a building you own where only you have access. This sets us apart from our competitors, and it's all fully automated. We're equipped to manage thousands of data centers. To illustrate, think of Elon Musk's Starlink, which has nearly 7,000 satellites in orbit, 6,800 to be exact. Managing these satellites is a challenge since they're constantly moving and not in geosynchronous orbit. It requires full automation to effectively manage thousands of such spacecraft. Similarly, running many data centers also requires complete automation and standardization; you can't automate a wide variety of systems. Additionally, what's notable about Oracle is that many senior members of our management team are experts in constructing facilities, power plants, and energy transmission systems. Building data centers involves more than just the structures; it also requires considering energy generation and transmission to the facility. The most efficient approach is to position the power plant close to the data center for minimal energy transfer distance. We have experienced professionals from the utilities sector who assist us in executing these large-scale projects. As a point of reference, consider Elon Musk again. A significant challenge he faced was building the Austin plant, which is the largest building ever constructed by humans. The largest building isn't the Pentagon or the NASA facility; it's the Tesla plant. To create such a plant, you need the proper contracting capabilities to construct it, power it, and implement automation systems effectively—this is the most challenging aspect of creating cloud infrastructure and automation systems. We have a team with diverse experiences that is much stronger now than it was five years ago.

Thank you very much.

Great. Thanks so much for fitting me in. And I'll just start by saying I can't remember a Q1 ever being this exciting. Larry, we've talked about a lot of the reasons why you win in cloud infrastructure, especially by addressing areas of the market where your competitors can't even reach. But in light of many high-profile cyber incidents lately, can you talk about how being more secure, not just in being so highly automated as you've already discussed, but how being more secure is helping to win some of these very large OCI deals especially in the U.S. and other governments around the world? Thank you.

Thank you for the question. I will be discussing two main topics at Cloud World on Tuesday: multi-cloud and security. First, I'm declaring that we are eliminating passwords entirely. They are easily hacked, and attempts to improve them have only made the situation worse. Instead, I will simply enter my email, and the computer will recognize me and log me in without the need for a password. If my family can recognize me, there's no reason a computer can't do the same. Additionally, we are introducing Zipper, our Zero Trust Packet Routing technology, which enhances the security of our network communications. Biometric authentication will make it much harder for fraudulent users to access our systems. Users can opt for various methods of biometric authentication, like fingerprint or facial recognition, making it a convenient option. Furthermore, Zero Trust packet routing will authenticate users at every level, keeping data secure by simplifying network security and separating it from configuration. Automation is also crucial, as it reduces human error, which often leads to cyberattacks. The autonomous database eliminates the risk of configuration mistakes made by human database administrators. Moreover, code generation—where programs are automatically created instead of manually written—reduces the chances of security vulnerabilities and ensures applications remain operational despite disasters. Overall, incorporating Zero Trust packet routing, biometric authentication, automation, and code generation are vital strategies in winning the cyber war. We need better defensive technology to match the growing threat from cybercriminals and state actors. We must strengthen our computer systems, and thankfully, advancements in AI-based security systems are providing us with the tools needed to counter these threats effectively.

Amazing, Larry. Look forward to learning more this week in Vegas.

Thank you, Brad. Thank you, Larry. A telephonic replay of this conference call will be available for 24 hours on our Investor Relations website. Thank you for joining us today. And with that, I'll turn the call back to Sarah for closing.