Oracle Slides by Most Since January on Mounting AI Spending

0:01Bloomberg Audio Studios Podcast Radio News.

0:11Bloomberg Tech is live from coast to coast with Caroline Hyde in New York and Ed Ludlow in San Francisco. This is Bloomberg Tech. Coming up, Oracle falls after reporting a jump in spending on AI data centers and other equipment. We're going to discuss the impact, and there is impact, on the broader tech sector. Plus, Disney agrees to invest$1 billion in open AI and licensed characters for use on the Sora generative video platform. And more tech earnings with chip design software maker Synopsis, our conversation with the CEO later this hour. Meanwhile, we turn to some breaking news, and it's geopolitical in nature at the moment.

0:53So, Ed, we are looking more broadly at the Ukrainian president, Vladimir Selinsky, floating the prospect of putting the issue of territorial control in a peace deal to a referendum. Now, this comes as Kiev comes under mounting pressure to agree to terms to end Russia's war. Moscow has insisted that Ukraine withdraw its forces from the eastern Donbass region, which the Russian military has failed to capture in its nearly four-year invasion. And I'm afraid, step aside, Jay Powell, the focus is all on Larry Ellison and what is happening at Oracle and how much money is being deployed in the build-out of infrastructure, Ed.

1:29The anxiety is coming from Oracle. This is a two-day chart to make a visual representation of the point. It's a big drop. Right now, on track, if it closes at this level of decline, its biggest drop in 21 years, 24 years, sorry, going back to 2001. Very simply, capital expenditures, several billion dollars beyond what the street thought they would be in the quarter. Cloud growth, we're talking cloud growth way up there, particularly in the infrastructure unit, but just shy of very high expectations. Interesting right now, if you look at the chip space, NVIDIA, also like Broadcom, big declines.

2:10And at the index level that you were just showing us, Caroline, those names are part of the drag. So we need to dig into what's going on. Yeah, and we've got Broadcom numbers up to the bell as well. Let's focus in on Oracle. Bloomberg's Brody Ford has been covering this story. And we knew that they were spending an awful lot on capital expenditure. but$50 billion and not enough in terms of the revenue to show for it. Yeah, AI anxiety is right. I mean, the big question for Oracle and all of its peers right now is, is building AI infrastructure a good business to be in? Wall Street understands it takes a lot of upfront costs, but it seems like every single quarter, it's more than we thought it would be, and the revenue recognition is further out than we thought it would be.

2:52And so Oracle kind of gave the exact message that Wall Street wasn't hoping for, which is we're spending more and it's taking a little longer than you expected to see the revenue uplift. But it's a tradeoff with growth. So I'm just going to go through some of the numbers, right? The street thought that capital expenditures in the quarter would be about$8 billion. They ended up being$12 billion. But overall cloud growth, get into that, and then infrastructure growth, that is growth. it's just that it's basically in line with consensus what more is it that the street wants yes it's big growth but the question is is it empty calories right i mean what is it i mean the first question on the earnings call last night was please tell us how much it's going to cost to put up these data centers give us a total number it was a level of you know i'd say aggression that you don't typically hear on an earnings call because it's just still not clear i mean no one's put up gigawatt scale data centers of four, five of them at once, there's a lot of questions about when Oracle is putting up these big data centers for OpenAI or Meta or NVIDIA, what is the true cost profile?

4:00And the jury is just still out. The jury is out. And also the questions are abound of their overall exposure to one key player. $300 billion is how much OpenAI is saying they will invest via Oracle. But when are we going to see that? When is Oracle actually going to be able to book that as real revenue? Right. A lot of us remember in September when Oracle stock went gangbusters and they were the favorite child of the market for a couple of days. But since then, what's happened is OpenAI has made a bunch of commitments with really big billions numbers with a lot of other vendors. And it's made the market say, wait a second, can OAI really pay all of these bills?

4:41And so that concentration risk on OpenAI is a key one for Oracle and one the company has tried to push away by saying, look at the big deals we've had with Meta or NVIDIA or others. Bloomberg's Brady Ford with the reporting. Thank you very much. Let's get more with Alex Zuckin, Managing Director, Senior Analyst, covering enterprise software at Wolf Research. The price target,$275 on this stock, trading at$192.82. Okay, we accept there's some anxiety. They are spending more than you lot thought they would. They have some growth. Where in the equation are they sitting for you? Yeah, I think right now it's a timing mismatch.

5:26I think they're getting overly penalized for a number of elements. I think there's kind of a vote of no confidence from this market around the ability to grow and serve OpenAI, both on the OpenAI front, quite frankly, and the Oracle front. And I think that that's a mistake. We truly believe the compute demand at the moment is insatiable. It's not just insatiable for open AI. It's insatiable for other foundation model companies. It's also insatiable for other hyperscalers. I think even on the Microsoft call, they noted how much more demand that they had in the quarter for Azure that they couldn't supply.

6:03So we're pretty confident that the demand is real, that the demand is there. and we continue to see accelerating growth with the company stating pretty clearly they, A, expect to maintain investment grade credit. B, they don't believe they will need to spend more than and quite possibly significantly less than$100 billion in debt. And on OpenAI, A, they believe they're very much good for it, but that the demand is truly fungible. And if OpenAI can't make those commitments, there are a lot of other. No, no, I appreciate that. Negative free cash flow is$10 billion, right? And the debt pile you point out has hit$106 billion in debt.

6:44This chart is Oracle five-year CDS. Oracle CDS at the highest level since 2009. Credit default swaps, the cost of protecting debt against default. It's a chart, but it's a reaction. And the three points you just made, clearly, there's a large portion of the market that doesn't share that viewpoint with you. yeah look i think the cds chart uh i think or it was important for oracle to state that they they expect to maintain investment grade credit um worthiness i think that there's also a number of flexible options that they presented on call yesterday like the bring your own chips in addition to the ability for customers to lease so i think that they are uh very uh vigilant about the level of debt that they're willing to take they have another a number of other options including issuing equity.

7:34But at the end of the day, the overall CDS level is still relatively low for a company of Oracle's size and stature. And we don't believe that that is indicative of any kind of risk for them to fail to meet their commitments. Alex, I loved your note. And I love the idea that you're saying this is a tough crowd, basically. But you did reiterate outperform, but you vastly decreased your price target from$400 to$275. Is that more because the market is just not going to push up the stock in the way that you actually think could be vindicated? Do you still think$400 could be vindicated? It's just the market's not going to go there?

8:16Yeah, I think it's two points. I think it's important to point out that since our last price target, the comps have compressed, the relative multiples have come down. And number two, look, we are vigilant of current performance as well. I think for the stock to work back to some of those more lofty angles, I think you need clear evidence of meaningful outperformance on cloud revenue and the IS side, not in line or slightly below. And as the confidence kind of returns to seeing those investments converting to revenue at higher and higher eclipse to where at this point, you know, it looks like Oracle is going to be the fastest growing mega cap software company.

8:58And it is the cheapest mega cap software company. So as those two things start to converge, I think you could get back to those lofty highs. Alex, what's so interesting about your coverage range is that you're a software guy. You're also looking at Adobe and we've just had Adobe numbers. But of course, Oracle has made this huge pivot into almost now infrastructure eclipsing its software revenue that it brings in now. But just taking a step back, the whole AI bubble debate and whether or not we're going to be good for productivity gains in the amount that we're currently spending. Do you think that there is a bubble or that we're in a fine kind of space?

9:35I think it's always hard to call a bubble when you're in the middle of it. We do see a tremendous amount of productivity enhancement that is happening in the industry. And I think depending on where you are and what part you're looking at, you can have a completely different picture. I think at this point, if you're an engineer or if you're a software developer, you can't imagine life without agentic coding. You can't go back to a time when you weren't using one of those tools. And the availability, the abundance of compute is increasingly leading to a renaissance of software development and code creation.

10:10I think if you're kind of a traditional knowledge worker, you're still kind of waiting to see that 10x, 100x improvement in your capabilities. And I think that's natural because those things are a little bit more nuanced depending on where you work, what your role is, how much data, how much risk your firm is willing to take. So I think it's a natural progression. And folks have to remember, this is a journey. We are still way away from that destination. But it's an extremely exciting time to be in this space. And I do think that we are on an accelerating growth path for many of the companies that we cover.

10:44Alex, I'm going to recap just really quickly. Capital expenditure is slightly ahead, a few billion dollars of where the street thought it would be. But impressive growth in cloud. The stock, a decline of 14 to 16 percent would be if we close this level down the most in 23, 24 years. Is Oracle good at technology, Alex, very quickly? Yeah, I think Oracle is great at technology. I think Oracle for multiple decades has been able to make the pivot to the new exciting secular domain. It takes time like it did for Oracle in the cloud. I would argue that they've made the pivot to the AI compute cycle much faster.

11:23And I think over time that will serve in their favor as their ability to take their existing customers and convert those customers to both an AI cloud first and AI native world with a full stack experience. Not that dissimilar from what Microsoft has done will put them at the forefront of a lot of these secular waves. Alex Zirkin of Wolf Research. Fantastic talking to you. Thank you very much indeed on all things Oracle and beyond. Meanwhile, coming up, Disney. It agrees to invest$1 billion in OpenAI and license characters for use on Sora on ChatGPT Image. We'll dig into it next. This is Bloomberg Tech.

12:14Disney will invest$1 billion in open AI and license more than 200 characters for use on the Sora generative video platform and in chat GPT images. It's all to generate short, fan-prompted videos and stills. Bloomberg News Entertainment reporter Hannah Miller joins us now along with Dave Lee from Bloomberg Opinion. And Hannah, I go to you for the news first, because this is a big step in IP licensing in the world of generative AI. Yeah, there's been a lot of anxiety about how AI will affect the entertainment industry. And Disney's support, its investment in open AI is a huge development. I mean, this is something that, you know, we didn't really expect.

12:53There's been a lot of tension and controversy with AI in film. Dave Lee of Bloomberg Opinion I get what OpenAI gets out of this a library of creature characters 200 of them what does Disney actually get out of it? Well look I was watching an interview with Bob Iger earlier and it struck me that he's been in this game a long time and I think he's actually got a very good deal I think he's gone to OpenAI and he said your Sora app which people are using is only fun when you use essentially without deals stolen material IP from other companies whether it's Disney or Nintendo or even the Premier League.

13:29And I think Bob Iber says, right, you need this. You need consumer attention. We know there's a lot of pressure from Google on OpenAI. And so what Disney gets, he gets a slice of this big, exciting AI company. They get that billion dollar investment with the chance to do more. They have a deal that is only exclusive for a year. So if we find that consumers are really into making these short videos, which I'm not too sure they will be long-term, then he can shop it elsewhere and make more deals like it. And he also gets to go to Google and say, look, you know, OpenAI have made a deal, so now you should do as well, because that's how this is going to run from now on.

14:05So I think, honestly, I think Bob Iger has shown his experience here and made a very smart deal indeed. Hannah, the company's also disclosed that basically Disney becomes a major customer of OpenAI. What other technology work are they going to do together? Yeah, I mean, it's great that they get to use these open AI tools. You know, they will be able to give access to chat GPT to their employees. But I think what we can see from this is this is just the beginning, that there could be more down the line. What's really interesting, Dave, is we've heard in the same breath, Bob Igo really confirming that he's sent a C-synthesis to Google.

14:43We know that they've been litigious about their IP when it comes to mid-journey, when it comes to Chinese generative AI opportunities. Are we then, after this year of exclusion, Just going to see a whole myriad of deals being done. And what will Hollywood and creatives think about it? I mean, there's two sides to it, isn't there? I think, yes, there'll be deals done, or at least there'll be attempts and maybe more lawsuits instead if the deals aren't made. You will see this increased uneasiness. And we're already seeing out of Hollywood of people saying, well, what's Disney's endgame here if they're going to start using OpenAI's technology?

15:18I think this is if anything quite troubling development for the AI companies because you know if OpenAI is going to have to give away a billion dollars of its firm to all these people who want intellectual property well then there's not going to be enough of the company to go around and you know I have to think to myself you know if you're someone who's backing OpenAI you're thinking okay we've got Mickey Mouse he's not allowed to talk probably and they're going to be short clips is this really getting us to super intelligence like is this an important part of that i i i i don't see it personally i feel quite cynical about this entire deal i have to say but there's a lot there that i think is troubling for ai companies looking to do some of the same things next year dave dave just real quick that's what i'm trying to understand who has the balance of power here which side needs the other more the the studios in hollywood or the the ai labs i do you know i think the question is for the studios in hollywood they can live without without working with these AI companies.

16:14So they might as well just get as good a deal as they can get. And if they're not happy with it, they can walk away. And that's perhaps why Google was in a stronger position. It'd be interesting to see how the talks now have to carry on with Google and Disney. Because OpenAI realized that they don't have any other business with Disney. So there's no incentive for Disney to sort of bend more to their will. I think in Google's case, you know, Disney just signed a new multi-year deal with YouTube TV, for example. So there's a much more sort of close meshing there between those companies. I wouldn't be surprised if some agreement came out of, you know, around Soar and Google quite soon.

16:50But yeah, I think that's where we're going to see these sort of interesting power plays. But if you're an owner of intellectual property, really good stuff, particularly, like I say, like Nintendo or football companies, I think you're in a very, very strong position. It really is notable, though, I think, that some of these large language model developers would say and chatbot owners are saying, we are amazed at how much people are using them for image generation. Even if it doesn't lead us to super intelligence, it's an addictive element to the use and maybe outperforms, therefore, and who wants to be using it and keeping 800 million people wanting to use it every single week.

17:22But, Hannah, when the power play comes back to the worker, you reported long and hard about the strikes that happened in Hollywood. What do you think the narrative is Bob Iger needs to give to his workforce? Yeah, he needs to give reassurance. I mean, people are worried about their jobs. Hollywood has been through a lot lately, you know, from the wildfires to layoffs at major entertainment companies. So I think there is a lot of concern here about replacement. If you're going to be able to use AI tools to animate, you know, why do you need humans behind it? Bloomberg News reporter, Hannah Miller, Bloomberg Opinion contributor, Dave Lee.

17:56Thank you to you both. We are sticking with tech earnings and we're going to deep dive on chip design software maker Synopsys now. Just out with its own fourth quarter results, giving an upbeat first quarter forecast. This following a significant equity investment from NVIDIA earlier in the month that sent shares higher. And maybe we're just pairing some of the previous gains. We're just down by 1.2 percent on the day. Synopsys CEO Sassim Ghazi joins us now. And even though your shares are down a little bit, the analysts seem to feel positively about these numbers. But JP Morgan asks that maybe you're being slightly conservative on your full year 2026 guidance, are you?

18:31We're looking at it as being balanced and pragmatic regarding FY26, given the global and the overall environment. We guided at 9.6 billion, and that's after finishing up FY25 at 7 billion. So we're very excited about the opportunity ahead. chip design right now there are some bigger picture discussions to be had a great debate about companies big and small looking at custom a6 through to doing more in-house versus not bothering because why nvidia is doing such a full stack offering what are the trends that you're seeing particularly when you think about your backlog sasim because the analysts are also talking about your backlog, Ex-Ansys.

19:15Yeah, we have a significant backlog entering FY26 at 11.5 billion. And what's driving that backlog is the strength across the portfolio. As for your question, most hyperscalers, they're looking for multiple alternatives to how to build and optimize along the stack. The silicon is an essential component, as you know. And almost every one of these hyperscalers, they have a strategy along three vectors. One, they buy merchant chips from NVIDIA, AMD, Intel, et cetera, and they build their software stack. Then there's the custom vector or ASIC, and they have their own customer own tooling, what it's called, where design the chip all themselves.

19:59For Synopsys, all three vectors are a fantastic opportunity because you need the essentialness of what we provide in order to achieve that silicon design all the way up to the system. It's the first time we've spoken to you since that$2 billion endorsement investment in your stock coming from NVIDIA. And at the same time, NVIDIA is trying to gain inroads once again back into China. You've got headwinds in the China part of the business. How do you see that part of the world evolving? So the NVIDIA investment was really driven and it's a significant endorsement to our strategy to provide engineering solutions from silicon to systems, where NVIDIA has been talking about the whole physical AI, et cetera.

20:44That cannot happen without having physics simulation inside the product development. And that's what Synopsys provides. As far as the China market and opportunity, FY25, we faced a significant headwind in China. Our business declined by about 20 % due to these headwinds. What we talked about in our FY26 guide is we'll continue on assuming that the environment in China will remain status quo, the same stress that we saw in 25, and we have de-risked it out of our forecast. Sasim, very, very quickly, 30 seconds. Why does NVIDIA need to have an equity stake in your company? It's exactly what Jensen said.

21:29He sees a great opportunity to make money and to make sure that he's participating and aligning both companies' roadmap in order to accelerate the investment and deliver to this opportunity. Synopsys CEO, Sassim Gazi, it's great to have you on Bloomberg Tech. Thank you.

21:53Welcome back to Bloomberg Tech. This is a story about rising AI spending and that spending rising at a level that is not satisfactory to investors who want to see how it will convert into growing revenues. The story is Oracle, the stock and its shares on track for their biggest drop in about 24 years. That's the story. Capital expenditures in particular, billions of dollars ahead of what the market was expecting. Let's get out with the Bloomberg Intelligence reaction. Anna Ragrana joins us now. I just wanted it to linger. It's severe. And we spent all of yesterday's show saying this might happen.

22:40That's your reaction in your research. What's your thesis when you woke up this morning? I mean, I was surprised that they didn't talk a lot about their supply constraints because, you know, we heard it from Corby, we heard it from Microsoft and so forth. But they did spend a little bit of time trying to explain the funding issues. But it doesn't look like, you know, the market's happy with their explanation. But, you know, one of the big things is when their cloud infrastructure estimates were 69 % growth in constant currency and they came at 66. and anybody who's followed, you know, what happens with Microsoft Azure or AWS numbers is when you miss on a cloud growth rate by even a little bit, I mean, that has an impact on your stock and that's what we are seeing here.

23:22On top of that, they talked about, you know, increasing their CapEx by$15 billion. That was another thing. So there are multiple factors that are affecting the stock. But having said that, you know, you also go back and see there was one day when they announced that big OpenAI deal, the stock had gone up like, what, 25, 30 % at that point. So there's a bit of digestion in here, and it'll take a few years to figure out how much or how long will it take to take that 500 billion of backlog and convert it into sales. Talking of that backlog, many would say that the moon music is still in their favor, maybe why the shares remain higher for the year.

23:56Anurag, I'm looking at this really in-depth survey that you've done of the C-suite over at Bloomberg Intelligence. And one of the key findings really was that all industries cite a shortage of AI infrastructure as a roadblock. So when will we know that Oracle is good for the half a trillion dollars in backlog that they have? Yeah, it's probably the most important questions for all of these companies right now. But you have to separate the OpenAI Oracle relationship with the rest of the hyperscalers because the rest of the hyperscalers, if let's say, God forbid, things go bad next year and we don't need that much AI infrastructure, it's OK for the three hyperscalers because they have a lot of other businesses that they can run.

24:38They have their entire cloud businesses, search businesses, etc. for the three of them combined together. But for Oracle OpenAI, the big question is, OpenAI by the end of the year is going to do 20 billion of annualized revenue somewhere in that range. And it's made commitments of hundreds of billions of dollars. So this is one pocket that needs to be examined a lot more carefully than the rest of the equation. But having said that, we are very confident from what we learned from these executives that AI infrastructure is the biggest bottleneck in deploying all sorts of AI functionalities in any part of whether it's hospitals, financial services, or pharmaceuticals.

25:19Blueberg Intelligence Analyst, Anurag Rana, with really important research. We now want to talk about how that affects your investments. Beth Kindig is here with us, Lead Tech Analyst at the IO Fund. And Beth, we come to you because of your semiconductor expertise, But here is Larry Ellison and others over at Oracle saying, look, we will put in whatever chip you want. We are not beholden on NVIDIA. But what they do seem to be beholden on is a supply issue, not a demand issue. Is that coming down to semiconductors or is it more the build out of data centers? Yeah, I think it's a natural statement for Larry Ellison to say that because we know custom silicon will help drive down those CapEx costs.

25:56From my perspective, we really are at the crux of this issue to where, for example, Oracle's estimated 2026 CapEx was supposed to be$9 billion. We're now over$20 billion, low$20 billion, assuming no more raises. We are significantly higher, 2.5x higher than what was originally estimated. This is what one analyst has called drunken sailor type spending. However, what I would encourage investors to keep an eye on is the moment that there's an inflection. We haven't gotten that inflection yet. And that is really what these very smart tech CEOs across the border banking on. As an investor, I need to pay attention to what these CEOs are saying.