Oracle’s Halo Effect… And CME Group CEO On Retail Trading 9/10/25

0:00Hi, it's Melissa. Before we jump into today's show, I've got something exciting to share. On December 11th, we are hosting a special edition of Fast Money Live, trading the holidays right here at the NASDAQ Market Site. You get to watch a live taping of Fast Money, meet and interact with the traders, and, of course, celebrate the holiday season with us. It's stocks and cheers in the heart of the city, Times Square in December. You will not want to miss this. Tickets are available now at CNBCEvents.com slash Fast Money. Live from the NASDAQ Market Site in the heart of New York City's Times Square, this is Fast Money.

0:29Here's what's on tap tonight. Outstanding Oracle. Shares of the software giant surging 36 % and adding nearly$250 billion to its market cap. What it says about the state of the AI trade and all the companies that benefit from the gains. And a rapid rate drop. The yield on the 10-year treasury closing in on 4 % ahead of tomorrow's CPI print. Will the report solidify hopes for a Fed cut next week? And what could it mean for the markets? Plus, fintech firm Klarna makes its market debut. The trade desk tumbles on a Morgan Stanley downgrade. And Robinhood unveils a slate of new products, how they will affect the stock as it trades near all-time highs.

1:04I'm Melissa Lee. Coming to you live from Studio B at the NASDAQ. On the desk tonight, Fanoa Nison, Karen Feinerman, Courtney Garcia. And from the Robinhood Summit in Las Vegas, the one and only Dan Nathan. We start off with Oracle's earth-shattering surge, the stock hitting an all-time high, up as much as 43 % today, closing out its best session since 1992. The move boosted the software company's market cap by$244 billion, putting it within striking distance of becoming the next trillion-dollar tech stock. The strength coming after, Oracle said, remaining performance obligations in its latest quarter more than quadrupled.

1:36That's a measure of bookings, by the way. More than quadrupled from a year ago. One big customer, likely OpenAI, the Wall Street Journal reporting today that the two companies signed a$300 billion deal for computing power. That would be one of the largest cloud deals ever. Oracle's move having a halo effect over hyperscalers, Broadcom, NVIDIA, Taiwan Semi, Micron, AMD, they all jumped. Not to mention data center stocks like Corwee, Vertiv, Dell, and Supermicro. And power suppliers like Bistra, Constellation Energy, GE Vernova, Eaton, and Talon. The biggest winner of all might just be Oracle founder and chairman Larry Ellison, who saw his fortune increase by more than$100 billion with a B today.

2:15Helping him pass Elon must have become the world's richest man. But the rising tide didn't lift all votes. Cloud giants, Amazon and Salesforce down more than 3 % each today. So did Oracle's results just draw a line in the sand between the haves and the have-nots in the AI trade? We were all sitting here last night digesting the quarter, watching the stock go stratospheric, hitting record highs in the after-hours session, Karen. And here we are. It actually carried through into this session. It did. It carried through and it traded as high as, I think, 345. I actually did a put calendar because I just thought, wow, this is just this is just nothing.

2:50It's not deserved. It was a giant, stunning, stunning release. But we still we talked about this last night about the margin on the additional revenue and how margins will be compressed. The overall dollars of margins, of course, will be much bigger. But it just seemed to me like, wow, this is just a frenzy now to own this stock. and I wouldn't be surprised if, I mean, we know that AI is in the early-ish innings, but we also know this is not a linear thing, right? And earlier in the week, there was a story about AI adoption is slowing, which if we step back and, you know, squint a little, it'll be the tiniest little, you know, nothing little tiny dip.

3:31And so I know the overall trajectory, of course, is up and to the right for a while, But this kind of move on a company this size is just, it's enormous. And so would I be short it? No. Right. But to buy some puts, I think the risk reward got, they got crushed, obviously. I thought that was compelling. Right. But this also comes as we got that great data point from Broadcom with its stock pop. It also moved sharply higher in today's session, like 10 % or so. Yeah. Yeah, so we talk about broadening and how that kind of builds confidence for the market. We wanted to see something outside of NVIDIA that really had some sustainable pathway.

4:11We called in question whether or not the CapEx spin would continue. And we really felt that it was NVIDIA and everyone else. And perhaps at one point it was AMD that was the second player. Their quarter was a little bit less than Stellar. But then, as you mentioned, when you bring in Broadcom and you think about inference and you think about kind of extending out just the data center play, I think there is a lot of positive there because at the end of the day, since April, this really has been the AI, AI adjacent. I mean, you mentioned the utilities that are kind of tertiary placed to that. And I think this does show some sustainability.

4:44Karen makes a great point in terms of the slowdown of adoption. But I think the change in the rate of adoption does not necessarily mean that the target addressable market is significantly lower. So as long as I think we see continued spend on data center and to Oracle's point, I think a reason why you see a company this large have this size of outsized move is because, keep in mind, the CapEx spend and the FreeCastle that they were foregoing in order to get into this market. And so I think what you're seeing is some somewhat of a follow through, some confidence in that spend and some broadening out from the one name that really has been at the center of this whole thing.

5:22Do you remember the quarter when NVIDIA said revenues were going to jump by? Yes. It was like a$9 billion surplus. Yes, from 7 to, yeah, 11. Yeah. Right. It was a crazy guide hire. Crazy, stunning. And it's sort of like what we saw in Oracle yesterday, where the guidance was just so strong. You rethink how you view this company, and you rethink this company's trajectory and place in the AI trade. Yeah, it almost seems like it's an error, right? I mean, it's like that big of a move. You have to, like, double check and say, did I read that correctly? But I mean, this company, if this is in fact accurate, and it is, I mean, this is what they're reporting, this demand is huge.

5:57I mean, I think the question is, we actually had a guest here yesterday who brought up a good point, that can they meet the demand? Like at some point, is there like chip constraints or energy constraints to this kind of demand that they're seeing? But the demand is absolutely there. When you're looking at where investors are pricing this right now, it's now trading about 45, 46 times forward earnings, which is much more than Microsoft and Amazon, which are both in like the low 30s right now. So, I mean, this just completely popped. People are really trying to jump into this trade here. CapEx is something to take a look at.

6:26I mean, they were spending under$2 billion in 2020 on CapEx. It's now going to be$35 billion in fiscal year 2026. So, yeah, I think the question is what sort of profitability and is there going to be any sort of constraints to this? But the AI tread clearly is not over. And I think that's what you're seeing with this sort of demand. Yeah. So, Dan, I'll ask you, I feel like you're the right person to ask in terms of what the caveats could be in the guidance that we got last night. Well, the caveat is that we just don't know. I mean, these RPOs, this, you know, this backlog that we're supposed to get here.

6:58You know, this is something that needs to be achieved over the next couple of years. The spend to do that, I think Courtney just mentioned that CapEx. I mean, they were, you know, single digit, you know, billions as a percent of their total revenue. Their competitors are spending 25 percent of their revenue. So this is a catch up trade. We also don't know if this is a market share thing. And if it's a market share thing, then that means, listen, we already knew that this OpenAI contract and some of these others were coming this way. Why? Because they're probably competing on price. We also know that this is a company that's desperately trying to cut costs because they know that this spend is really going to weigh on their margins.

7:35So at the end of the day, if we get closer to that digestion phase of this data center build out, then these are going to be the ones. Oracle is going to be the ones to feel it first. CoreWeave, these neoclouds are going to be the ones to feel it first. And when you think about the balance sheets that the other hyperscalers have and their ability to spend the way that they have, it just is not the same category, in my opinion. And I'll just say this on the Oracle. You're talking about 40%. This stock had a$680 billion market cap yesterday. If you're finally joining the party here, if you're finally saying we need to get more long Oracle up here near a trillion dollar market cap, I think you're kind of doing it wrong.

8:18And I think there's probably other ways to express this view in a manner that actually makes more sense in a less crowded field. So when I think about their market share, yeah, can they grow market share off a really low base? And is that going to be the sort of eye-popping sort of guidance that they can give? But I'll just mention this, and maybe they can pull up Broadcom here from December. Remember they gave that huge guidance rate looking out to 2027, and stock went up 40 % in a couple of days? Well, go back and look how much it gave it all back. Now, I know that this company is executing.

8:49I know that they've gotten a bunch of big cloud orders. But at the end of the day, we don't know that this backlog is going to materialize. And so to me, I just think this is a bit frothy right here. All right. Well, our next guest says we haven't entered the AI bubble just yet. He just raised his price target on Oracle to 370 from 270, reiterating his buy rating. Ben Reitzis of Melius Research joins us now. He's the head of tech research. Ben, great to have you with us. You've had that buy rating on Oracle for about a year. So that's been the right call. Got more right last night. But in terms of the big stock pop and where you see the stock going at this point, you know, how do you answer some of those those questions?

9:26Will RPOs actually materialize into revenue? How will Oracle pay for the CapEx? What sort of margins will it have on this new source of revenue? Well, it's definitely a negative mix shift and it's good to be here, Melissa. Thank you. And I appreciate your shout out on Oracle. The it's definitely a negative mix shift towards lower margin business, but the growth is really crazy. And what was kind of neat about their guidance was that it's not that much upside for this year or next, but it really inflects in 2028. Their FY28, which overlaps with calendar 27 and then 29 and then 30. And we're talking about growth there, like where we took we were saying that we're going to grow 19 percent in FY28.

10:09And now it's like over 50 percent. So this kind of growth is in a shortage in the market, especially in software, where SaaS companies, software as a service is so, so threatened by AI. The infrastructure companies that made the right investments in cloud and their growth, especially in the out years, are so much better than these guys that they're getting. You got to get paid for it. So with respect to Dan, if they can pull off this kind of growth, they're going to get paid for it. And it's frankly only trading at under 30 times my earnings for FY28, where the revenue growth and flexed to 50 percent.

10:45Dan, that's cheap. So, you know, if they can get the RPO converted to revenue, it's going to work. You know, I would say the big risk is whether OpenAI can pull off getting funding and keep growing because their cash flow is negative. So it all depends on OpenAI being a stable and thriving enterprise. Hey, Ben, I do appreciate the respect, but we're talking about two years from now. We're talking about a cycle that is already in its third year right now. We're talking about nearly a trillion dollars by its main competitors, right, who have basically built out this infrastructure. So, you know, I don't know as far as valuing this company right now on calendar 2027 in one fell swoop.

11:31That seems kind of odd to me. Well, I think you have to discount some of the risk that, you know, perhaps the RPO doesn't come through because a lot depends on SoftBank giving money to OpenAI and OpenAI getting financing and OpenAI being able to execute. So that's the big risk. I mean, we all know that most of the RPO is coming from this OpenAI Stargate slash OpenAI, whatever they're doing with others project here. And this is huge. What is happening, though, that is much better than expected in terms for the chip guys and for the cloud guys is that inferencing has turned into just this huge, huge razor and blade, like the best razor and blade model that perhaps of all time, where it's not that you just get a razor and then like get a few blades for the year, like four.

12:17You get a razor, you train a model, and then you're like getting a thousand blades. And it's really something. And I think it's catching everyone off guard. And, you know, I think you're you're setting off bubble alarms before the bubble even starts. And I think when you look at the industry day in and day out and talk about what's happening with inferencing, you know, that's the big surprise here. These companies are seeing something. There's a reason Jensen Wang said$3 to$4 trillion in data center capex by the end of the decade. There's a reason Hock Tan was so bullish on his call. These are the two greatest CEOs.

12:55You know, I don't want to even put down anybody. These are really good CEOs. So you're kind of saying that what they're seeing is a clown show. And I think it's the exact opposite. I think that people are looking at it and the smartest guys are going, oh, my God. And part of this will go down in history, what you heard last night. So maybe the bubble starts now, but this is what you're seeing. And I would just be careful because we've been building up to this moment. And now AI is starting to have that inferencing halo that could be pretty powerful. And that's our view here. So Jensen said three to four, but you're saying two in terms of the TAM for AI by 2030.

13:40And I'm just wondering, you know, what if Jensen is right? What does that do to your view on any of the stocks that you cover that are in this space? Well, actually, Jensen and I are saying the same thing. What Jensen is doing is he's including the dirt in the ground, the walls, the cooling, you know, all sorts of things. Like he may not do the air conditioners. And I'm including just the compute and networking. So when you take Jensen's three to four trillion, which I think is going to be at the upper bound of that, and then you take a little more than half, 55 to 60 percent, which could be compute slash networking slash other things NVIDIA does, then we're saying the same thing.

14:16And I actually was able to back into that number. We put out a big report Monday and it's just a huge number. And and I think that with regard to that, you know, what we really need to do is make sure that obviously people are getting a rate of return. I actually think the rate of return reports are very suspect. I actually think a lot of our companies are, unfortunately, what's going to happen here is there's going to be a lot of headcount cuts and stuff like that for AI. And we're seeing it at all the companies who have stopped hiring and taking a lot of the leverage. So I think that that shows that AI is showing a lot of return.

14:52And we're seeing digital labor kind of come in faster than expected, not slower. So that's where we are right now. And we feel like our TAM number of two trillion for compute networking can really make Broadcom, Arista, NVIDIA, as well as a few others go up a lot more even from here. Ben, it's Karen. Thanks for being on and excellent call on Oracle. So given the news from Oracle last night and where all the stocks have sort of repriced, I know you cover the space, AMD, NVIDIA, Broadcom and Oracle, obviously, if you rank them in order on this same news at these new prices. well i mean i think right now we we literally we're going to remember this call a long time i think we we might be entering you know and i don't know what's going to happen geopolitically or whatever but we might be entering like one of those crazy periods now and it's really hard to figure that question out um we also you know at our firm cover a lot of energy companies that uh are great pa stocks for a guy like me but like you know the the uh the energy sector is going to be RedHot, too.

16:02But in terms of what we're thinking, when you see a number like that, like the reason, you know, Broadcom could double, maybe even triple if their AI revenue. We're seeing Broadcom having 400 to 500 billion in revenue by the end of this decade, you know, and getting a pretty healthy revenue multiple on that. So that would mean it at least doubles. Believe it or not, NVIDIA can lose share and still get revenue probably six, if I'm right on the TAM, 600 to 800 billion in revenue. So that one we like. And then if the spending is as big as it's looking and Larry is right, and obviously Dan made a great point that a lot still needs to happen, then AMD is going to get some love.

16:44AMD, you know, there's a view out there that AMD is going to get squeezed and you can't really have Broadcom doing so well and NVIDIA doing so well. But if there's this much spending going around and opening, I can raise money. AMD is going to get a lot of love. And then there's Arista Networks. Arista Networks is great. That company is going to wander a place. So maybe I put them ahead of AMD even. Ben, great to speak with you. Thank you. Thank you so much, everybody. Ben Reitz of Milius. How would you re-rank? Would you re-rank your AI sort of stocks, your favorites, based on Oracle and Broadcom earlier?

17:22I mean, it's tough to have Oracle leapfrog NVIDIA here. I think you probably want to perhaps rethink Broadcom, as you mentioned, particularly around inference. What really calls into question, and this is a would you rather not, is Amazon. Because that cloud growth vis-a-vis peers on the last earnings call raised flags. and now that you're seeing this spin and you have OCI, which is within Google Cloud and Amazon Cloud, and you're still not seeing the growth that we've seen from Azure and some of the others, that makes me scratch my head and want to kind of go back to the table and look as to why they're not seeing the same type of growth that some of the other peers are.

18:06All right. Let's turn now to rates dropping ahead of tomorrow's CPI print the yield on the benchmark 10-year Treasury. Closing back in on 4%, economists expect consumer prices rose slightly more than they did in July with an annualized rate just under 3%. We did see a surprise drop in wholesale prices, PPI, in August this morning. The PPI fell by one-tenth of a percent, while consensus estimates expected an increase. So that was a nice surprise. So companies may not be passing on increased costs here. Courtney, how do you feel about what we're going to get in the 10-year going to 4 %? That's basically an April low.

18:40Yeah, I mean, I think we really need to see how CPI comes out tomorrow because this was like a big surprise. I mean, theoretically, if this is a trend, you would actually see deflation. But I think it's way too soon to actually say that. So you need to see what this other data is going to come out to show what that trend is going to be. But I do think at this point it is indicative that the rates probably are going to continue to come down here if you continue to get these kind of reports. And that is going to be a positive for the markets. But again, we want to see what the report comes out tomorrow.

19:05I was surprised actually the market wasn't up more on this. I know there were some big Dow movers like Apple that took a lot of points. But even the whole thing, spiders, all of it, I'm surprised because the magnitude of that data was big. Now we'll have maybe too big and maybe we'll see something very different tomorrow. That could be a reason. But those auctions went very well. All of that, you know, you have to think if you were on the fence about the Fed cutting. Right. This data is not completely flipped tomorrow. You have to think they must cut. And, you know, two or three weeks ago, that call for 50 basis points.

19:45I thought, all right, that's a little dramatic. That's, you know, labor isn't isn't weakening enough. And we still have inflation. And now both of those same things seem to be turning. So I'm on that. I'm surprised the market wasn't actually at more. Coming up, another IPO pop. How Klarna fared in its market debut and where it stands among the fintech players. Plus, another tough day for the trade desk. Shares hitting a five-month low after a bearish call on Wall Street. What's got analysts so skittish? We'll get some answers. Don't go anywhere. Fast Money is back in two. Welcome back to Fast Money.

20:16Buy now, pay later. Company Klarna making its market debut today. Shares opening the day at 52. That's well above their IPO price of$40, but losing steam throughout the session, ending the day up just over$5. The stock now has a market value of over$17 billion at its peak in 2021 funding round. It was valued at$46 billion. Dan, we should also note that, I mean, the range is much lower. So this is, you know, by all accounts, a very successful IPO, even if it fizzled during the session. It's a 20-year-old company, too. It's been around the block. Yeah, I think that's worth noting. I mean, a lot of these IPOs that we're seeing, they're not kind of newbies here.

20:53And so for them to kind of test the market, I think, is interesting. This did not trade well. Let's be clear, right? If you look at Chime in the FinTech spake, that was an IPO in the last couple of months. That has not traded particularly well. When I think of buy now, pay later, they might be doing very well in this environment where you're just starting to see the labor market weakening a little bit. You're starting to see the bite a little bit from tariffs and trade and the like. But there's going to be another phase of this on the other side of this. If we do see unemployment going up towards 5%, we're going to start seeing delinquencies.

21:26And that's when these models really start to have a difficult time. So, again, you talked about those valuations in 2021. They were a bit silly. You look at it right now, and it's a different cycle for a lot of these names. So I just don't find it particularly interesting, this Klarna or Chime, for that matter. Yeah, I mean, a lot has happened since 2021 in Buy Now, Pay Later. Affirm is the closest comparison, I would say, to Klarna. I have Affirm. Yeah, I like Affirm, although it's had a nice run-up. Although it was down today on this, I don't know if people thought, oh, Pardin's going to trade fantastically well on a firm will too, and then it didn't.

22:00Pardin didn't do badly. But I kind of, I do like the space. I do like buy now, pay later. There's a lot more Fast Monday to come. Here's what's coming up next. Rotten Apple? Shares of the iPhone maker dropping the day after its big product reveal. Why the stock continues to struggle and whether it can get back into rally mode. Plus, the state of retail trading. what one exchange CEO sees in store for investors as markets, rates, and trading tools continue to change. You're watching Fast Money live from the NASDAQ market side in Times Square. We're back right after this. Welcome back to Fast Money.

22:44Shares of the trade desk falling today after a downgrade from Morgan Stanley. Analysts pointing to mounting competition in ad tech from the likes of Amazon. Julia Borson's got the details here. Julia. Hey, Melissa. Well, the Trade Desk finished the day down about 12 percent on news that shows growing competition from Amazon's ad platform. Now, this morning, Amazon Ads and Netflix announced they're partnering for Amazon to offer programmatic buying on Netflix's ads platform. This gives advertisers using Amazon's DSP access to Netflix's ad inventory. Netflix is also partnered with The Trade Desk, Yahoo and Microsoft.

23:20But this news feeds growing concern about Amazon eating into The Trade Desk's market share. Now, ahead of this news, The Trade Desk was already down about 4 percent on that downgrade. It's now down about 60 percent year to date. Morgan Stanley cutting its price target, writing, Amazon's DSP is expanding faster than expected. And key questions now are whether restructuring can meaningfully reaccelerate growth in the near term and whether the trade desk can defend its leadership in connected TV against competitors, including Amazon. Melissa. All right. Julia, thank you, Julia Boorstin. Trade desk has been a roller coaster ride for a long time.

24:01Dan, really threatened by some of the other, you know, ad offerings out there, even like a Netflix and on top of Amazon. Yeah, you know, last quarter they had their first revenue miss in, I think, 33 quarters. And, you know, that's the sort of thing that just kind of highlights a decelerating sort of growth story. And that's why you've had that volatility over the last few years. You talk about competition from the likes of Amazon. We know that that ad business has been a great driver of profitability for them, especially when you think about just that kind of North American retail and how low the margins are.

24:36and also picking up some of the slack that you've seen from AWS as it's seen decelerating growth. So when I look at a story like this, I say, well, they also have problems with like the Metas and the Googles and the walled gardens that exist there. So this is a tough place to be. I don't even know if valuation works right here because that out year probably has to come down as far as estimates for earnings and revenues. Yeah, I mean, if you want to be in the space, you might want to look at a Meta or Netflix, which has the offerings that are being used. Exactly. Right, which got them into this somewhat.