Stocks Rally As China Trade Fears Diminish… And Bank Results Ready To Go 10/13/25

0:02Live from the Nasdaq market site in the heart of New York City's Times Square this is fast money. Here's what's on tap tonight. Easy go, easy come. Markets rebounding sharply after Friday's sell-off. Did you already miss your chance to buy the dip or is there more volatility to come? And another AI deal shares a Broadcom surging after inking a new partnership with OpenAI. All the details and what it means for the red-hot semi trade. Plus, we're counting down its bank earnings kickoff. Jamie Dimon goes mining for rare earth investments and blooming opportunities. The deal's powering the fuel cell provider to brand new highs.

0:34I'm Melissa Lee coming to you live from CTOB at the Nasdaq. On the desk tonight, Tim Seymour, Dan Nathan, Guy Adami, and Julie Beal. And we start off with a big market rebound on Wall Street today. The S &P recouping about half of what it lost on Friday, jumping more than a percent and a half for its best day since May. The small cap Russell 2000, the tech-heavy Nasdaq seeing even bigger gains, each up more than two percent, while the Dow added nearly 600 points. The action coming after President Trump appeared to walk back threats of a new 100 percent tariff on China, posting on Truth Social yesterday, quote, don't worry about China.

1:07It will all be fine, exclamation point. Tech and discretionary leading the way today with the recently beaten down semis soaring more than 4 percent. So was Friday's pullback just a blip in the now three-year-old bull market, or will today's comeback start to lose steam? Hello, Melissa. Well, listen, I mean, you mentioned 50 percent, 60 percent, depending on what you're looking at, we recoup. But I'll say this. I think the technical damage done last week and in sort of the days leading up to last week are enough where I don't think Friday was necessarily a blip. And I think the market was looking for an excuse to sell off.

1:41It found it in the form of the comments from President Trump. But I don't think it's the end of it is what I'm saying. I do think there's some legs to the downside yet to be seen. And, you know, I think, again, the die was cast with some of the technical moves that we saw. Alia Yachta-est, as they would say in Latin. Whoa, whoa, whoa. Wow. And I took four or five years of Latin and that went right over my head. Well, this happens a lot. So this is just a rest, a rest in the selling, but it will continue? Interesting. Guys, folks, this isn't just a blip in terms of, hey, just kidding. There's no question we saw from the market's reaction two things.

2:16One is that this is a nervous market. We had seen volatility creeping up towards the end of the week. Even without that move, it was a choppy week at best. We've kind of run out of gas, although we're going to talk about OpenAI's new headlines today. But the dynamics of getting a lot of macro, a lot of headlines without a lot of bottom-up substance. That's the great thing about earnings season, which is coming down the pike here. But I don't think there's any question. The reaction we had on Friday tells you that this is a market that is nervous. It also tells you, and I think we said this on Friday, is I'm not sure there aren't a lot of people that want to buy dips out there.

2:50Right now, that mentality is, I think, well in place. But the news flow needs to support it. Well, we did. I mean, we saw the OpenAI deal, which we'll get to. But we saw real buying, excuse me, in semis. The places that were hardest hit on Friday got the most buying in today's session. China stocks had a really nice rebound. I wonder without that OpenAI Broadcom announcement this morning, if you would have had that sort of outperformance by a lot of these names. Obviously, you know, Broadcom was up 10%, right? It wasn't down 10 % on Friday. And I think when, you know, you kind of just take stock of all of these deals that have been announced over the last month, whether it's OpenAI and NVIDIA or the two of them together or Broadcom.

3:28I mean, the list goes on and on and on. AMD, right? You have to say to yourself, they were coming in kind of white hot into earnings, right? So when you see all of these deals and the scale in which they're announced, you have to say to yourself, what are they going to leave? What else is there, right? And so we're seeing some analyst days. Oracle has one this week. They're not going to report again for a couple months. That's going to be really interesting. What that company has to say, especially when you think about some of the stuff that's leaked out of the company, about the margins in the group that is expected to get hundreds of billions of dollars in revenue over the next few years or so.

4:00And it just seems to be a lot of these companies are trading off of a level of enthusiasm that we know that we've seen before, but the scale in which we're doing it. You know, OpenAI has done a trillion dollars worth of deals, a trillion dollars, and they're going to take years and years to play themselves out. So that is one of the things that is fueling this tech market. That's why I think earnings season is coming in the nick of time, because I think we're going to hear numbers out of banks. We're going to hear their business. We're going to get the barometer on the economy at a time. We have a void of macroeconomic with the government shutdown.

4:31So I think it's important. Really interesting, though. You know, here's gold having the kind of a day that would indicate. And I realize we've maybe we've carved through a new asset class, although we've been talking about this for a while. But gold's move today tells you something. And it tells you that even where you had a rally back in risk assets, gold hasn't necessarily been trading. well, like a risk asset, but that's not today's move. Yeah. Julie Beal, I mean, Dan, I was worried about you, Dan, because I thought the glass half full kind of silver lining, Dan, was gone. But the silver lining to all of this, Julie, is that we didn't recoup all of Friday's losses and that it is a better setup going into earnings season.

5:05Yeah, I agree. It's kind of an unusual setup going into third quarter earnings where the revisions have really been mostly higher. Oftentimes, you see revisions kind of coming in and management's team just kind of meeting guidance where they are. I'm a little bit worried that if we've been raising expectations for the last three months, that if we don't meet those raised expectations and we're at valuations that are pretty rich wherever you look, that you could see a more material correction. So I think it makes sense to be kind of thoughtful about the companies that you're buying. I don't think you want to be overly levered to too many of the AI themes going into this quarter, because to me, expectations look pretty high.

5:44You know, Guy, we were on the desk on Friday. Would you take the day off? I was not here on Friday. Long weekend or something like that? I mean, something like that. But Tim, Mel, and I had a really good conversation, didn't we? And one of the things with the market, you say so ourselves. Yeah, well, no, but we were not sitting there on a Friday afternoon with the stock market down 3 % closing on its low, saying that we're going to get some massive follow-through because we've been through this before, especially when it's sort of tariff-related, especially when it's by tweet, that sort of thing.

6:12And so, you know, yeah, it took about a week and a half back in April to kind of sort this stuff out. I just didn't think this was going to be the sort of thing where we're going to kind of be piling on for a week or so. But it does suggest that if we can't go and take out the prior highs in the S &P 500 as we get closer to earnings, then expectations are still high despite being a couple percent off those all-time highs. I'm glad Tim brought up precious metals because the gold move should be surprising a lot of people, but should also be a bit of a warning sign as well. There's something going on clearly.

6:42And now it's not just gold. Now everybody all of a sudden is a silver expert. But you have silver levels we haven't seen ever. And ever is a pretty long time here, Mel. So the fact that gold can move the way it has in the wake of the dollar going up, dollar going down, all environments where historically it might not necessarily do well should be a bit of a warning sign. Yeah, I think the precious metals trade is rooted in a little bit of the unrest that has been this is why you want to own gold. I do think there is an element of platinum and palladium and silver just as underperformers here that that are really we've seen this cyclicality.

7:17I've been investing in gold for 20 years, but platinum, palladium and the other PGMs tend to trade as a group. And even though you don't have some of the scarcity effects, I do think it's interesting. I also think it's interesting that that copper continues to also have a bid today. Today, copper had a very big bid. And you've got a dynamic where we talk about the linkage to the buildout of data center, utilities, an electric grid, power consumption in this country. Copper is a big, big input. I think that's a trade you stay to. Yep. Well, let's talk more about the AI trade. Broadcom, the latest chipmaker, to announce a major deal with OpenAI, the two companies partnering to build and deploy custom AI accelerators, sending Broadcom shares higher by almost 10 percent.

7:57Mackenzie Cigales has got all the details on this one. Mackenzie. Hey Mel, so this is OpenAI locking in its hardware advantage, striking a blockbuster deal with Broadcom to co-design and deploy 10 gigawatts of custom AI accelerators purpose-built for its own models and set to roll out in the back half of 2026. It is a direct challenge to NVIDIA and AMD, moving beyond those off-the-shelf GPUs and into homegrown silicone. Now Wall Street had speculated that OpenAI might be Broadcom's mystery. $10 billion buyer disclosed in its latest quarterly earnings. But Broadcom's president told CNBC this morning that it's not.

8:35This deal is separate and now makes OpenAI Broadcom's fifth major hyperscaler customer. Chip stocks broadly rallying on that news. OpenAI's push into custom chips puts it in the same league as Google, which also co-designs chips with Broadcom. And remember, OpenAI is also getting into the business of building the infrastructure themselves, all of it, a play at becoming a hyperscaler in its own right. Broadcom CEO Hawk Tan put it simply this morning, if you do your own chips, you control your destiny. He's on Mad Money tonight with our Jim Cramer. Mel? Mackenzie, this is very different from the other deals that OpenAI has signed with an AMD and an NVIDIA.

9:13There's no stock element to this. There's no investment element to this. What is the importance of that? Well, that was the first question I asked, whether any equity was changing hands. And the significance to me is that we keep having this conversation about a circular economy where the companies that are selling their chips to OpenAI are in some ways giving them a handout, whether it's$100 billion,$100 billion investment in the case of NVIDIA taking an equity stake in OpenAI or AMD giving equity in its company to OpenAI virtually for free in order to pay for its chips. But this is a notable departure, which seems like a promising sign because so much of the narrative around this AI bubble comes from the fact that the value creation is not necessarily real.

9:56And so that was the big concern. And that's something that's not a factor here. We don't know the exact numeric terms of this deal. That's something I'm trying to find out now, Mel. But then that begs the question of how OpenAI will pay for this sort of deal. Well, it goes back to what Sarah Fryer, the company's CFO, told me a few weeks ago, that equity is very expensive and not something that they want to hand out. So it's all about debt financing. And we started to see more creative structures just last week with XAI and SPV, where essentially XAI would rent the GPUs from the special purpose vehicle using the GPUs themselves as leverage.

10:33And Fryer was teasing at the fact that there are a lot of financial instruments that they haven't dealt with before that they're looking into in order to fund this kind of grow. Because at this point, my back of the envelope math, we're talking about 33 gigawatt build out in commitments made over the last three weeks at roughly$50 billion per gigawatts. You're at$1.65 trillion. They don't have that in the bank. In fact, they haven't even wrapped their$40 billion round with SoftBank that needs to happen by the end of the year. Yep. Mackenzie, thanks. Mackenzie Sigalos. And so there's all this optimism about Broadcom and the impact of Broadcom,$40 billion in AI revenues per year.

11:10That's all fine and good. But then it is the question of will OpenAI, how will OpenAI finance this, how will pay for this? And is that going to be good down the road? January 23rd, Stargate members. So we had Masa, we had Sam Altman, we had Larry Ellison. And right after that announcement, Elon Musk tweeted out something like they don't have the money. OK, so this is a theme that's now going on for like nine months or so. And we just said that OpenAI has done a trillion dollars worth of deals. Well, they don't have the money. And now they're going to private equity and they're doing this is all the hyperscalers.

11:42And, you know, when you think about this, McKenzie just said 33 gigawatts. Now, those numbers, you know, 33 doesn't sound like a big number. That would take like 35 million homes of energy just to power that. OK, just think about that. So we don't have the money. We probably don't have the energy, but we're taking the debt and we're diluting equity to build out this infrastructure. that's going to take three to five years that we don't even know if there's going to be demand for. The other thing is you could say, well, nuclear. Well, nuclear takes 15 years. Big day for nuclear, by the way. We forgot to say that nuclear was limit up today.

12:15Yeah, they take 15 years to build a reactor, right? Solar was up too. And then if you want to build 10 gigawatts, okay, if you do a little math guy, I don't know, you were kind of doing this on the back of the envelope. That could cost you$500 billion to build out 10 gigawatts, a data center. I listened to the valuation, and also we got a new valuation benchmark for OpenAI a couple weeks ago. And she just talked about it. Mac just talked about how the equity is so expensive, they don't want to, you know, they don't. I mean, I hear equity is so expensive, and I say, there's your currency. I would issue more equity here.

12:51I wouldn't raise debt. I would issue equity at a price level when there's no financial detail around these deals. And that's the one thing that we've known. But Dan's right. 26 gigs equals a trillion dollars in debt that needs to be raised. Like so many of these deals, though, what's interesting here is this deal validates Broadcom's custom silicon approach and that they have five different massive clients and they're able to make adjustments to them. We also learned that we don't think that OpenAI was that unnamed$10 billion client. It seems to me that that's anthropic. But again, I think OpenAI should be issuing as much equity here as they can.

13:29I agree with that. And you think about, so this deal, so the market cap gain in Broadcom today was about rough back of the envelope math, about$150 billion. Think about how many companies have a market cap of$150 billion in the first place. And this headline creates that kind of value. And again, the circular nature of all this and the fact that all these companies seemingly rally on the same type of news, I think it should be concerning. And I'm with Tim on this. I mean, this is the time to use your equity as currency, not necessarily the debt market. Expensive equity means you sell equity. Yeah.

14:01So the 10%, though, gain in Broadcom today, because it's not circular, is it more valid? No, I mean, it's not. So think about this, okay? NVIDIA makes all these GPUs. And NVIDIA, they actually have to use this compute themselves, right, to build out all this software and all this stuff that goes on. They're actually invested in CoreWeave, which is a NeoCloud. They sell the GPUs to them, and then they rent them back. So think about it. They don't want to own these GPUs that they make, that they're trying to sell to every hyperscaler and anybody who wants them. And that's obviously global, too, because they don't want the depreciation of their own chips.

14:36So when you're talking about this build-out of these data centers that could take three to five years, well, Jensen Wang at NVIDIA has already told us that we're done with Hopper. We moved on to Blackwell, right, and we're going to be on Rubin. So what do you think? There's an acceleration, a depreciation that is not being recognized. So when you think about custom silicon, all of these companies now are going to build the chips that they want for their specific tasks, right, that they're going to be doing, which hopefully for them has a longer depreciation schedule. So I just think there's a lot of stuff going on here that I think is just we haven't really scratched the surface yet.

15:10For more on the AI trade and where markets go from here, fast money friend Jeff Mills, General Mills, joins us now. He's chief investment officer at Bessemer Trust. Jeff, it is great to see you. It's been a while. I know. Sorry I'm stuck uptown here. Next time I'll make a visit to the studio. Please do. Please do. How are you feeling about valuations here and now? I mean, I agree with what Julie said in that I think you have to pick your spots. You know, this isn't about piling into the AI trade. I think that there are parts of the market that are still interesting from a valuation perspective.

15:43I think the move in health care might be early on here. So playing the resurgence there, we've actually leaned into a stock like UNH, for example, down a ton, bounced off support at that 288-ish level. They have new management, probably still an important player in the space. So value there. Another company like Medtronic, for example, the valuation is reasonable at 16 times. They've sort of jettisoned some of the less attractive growth businesses, focused more on things like robotic surgery. So I think overall, certainly you can point to areas of the market that are expensive. But if you do pick your spots and you are careful, I think you can be active and still find value.

16:22But in terms of getting exposure to the A trade through big cap tech like MAG7 plus, you know, maybe a dozen others, Jeff, you think that that's just a fool's trade at this point? I mean, is it too fully valued, these names? I think that as long as these models continue to improve, there's a good chance that money is going to continue to chase this theme for a while, maybe beyond the point that we think is logical. So I agree with what you're saying in terms of cautious, but you have to think about certain stocks that are going to be winners and losers and mispricings along the way. I think of a stock like Google in the beginning of the summer, it was trading at a three to four turn P discount to the S &P 500.

17:07So this market gives you opportunities to lean into certain names when the narrative shifts. Taiwan Semi at 21 times, you know, certainly not a ridiculous valuation. Earnings are expected to grow 17 percent next year. And with some of the demand just around AI that we're all talking about, I think there's a reasonable chance that they beat those numbers. So I think you can lean into some of the names. Again, you just have to be selective. I was walking to the studio on Thursday and I saw Jeff Mills from afar. Did you really? Ask him. Ask him. He almost gave me a heart attack. He grabbed me on Sixth Avenue.

17:40So, you know, I'm staying away today. I didn't know you. You're lucky you didn't get pepper sprayed. Fair point. But, Jeff, it's great to have you back. We haven't talked to you in a while. So thoughts on the bond market here? Because it's been pretty tame since the last time you've probably been on. I think I'm one of the few people that think rates are going higher. It doesn't appear that way now. What are your thoughts? Yeah, I think we talked about this the last time I was on a number of months ago. And my mantra has been kind of sideways to down. And I think that's probably the case. You know, I think that, you know, somewhat slower growth, at least over the next quarter or so.

18:12I'm more of a growth bull longer term. But I think when you think about government shutdown, trade uncertainty, at least the appearance of a weakening labor market, you still have Fed rate cuts. That serves as some gravity on yields. The question really is, at what point is the bond market signaling something more sinister with the economy and lower rates actually aren't useful? The short answer is I don't think anybody really knows. But I look to those lows in the 10-year yield in the April time frame around 385. So I think as long as we trade between, you know, 406 where we are now and 385, we're probably OK.

18:47If we start to push lower, I think the bond market is telling us that there's something problematic going on in the economy that I don't necessarily see right now. Jeff, you know, there's always a seat for you here at this table. You're welcome anytime. The general. I always appreciate that, guys. Jeff Mills. By the way, he's got huge shoes to fill. The CEO of Bessemer now just left as a CIO. that's what Jeff just kind of stepped into right there, and she's a rock star. That's Holly McDonald. Wow. Good for Jeff. Good for the general. Rock star. By the way, that was my nickname. Am I getting credit?

19:20No, I mean, come on. Oh, you gave it to Jeff. Sure. I thought you meant rock star. You're kind of a rock star. Anyway, getting down to what Jeff was saying, Julie. Oh, okay. Now is not the time to be bearish. Be wary about the valuation of certain AI trades and be long health care. Yeah, I kind of agree with all of that. Genuinely, I do think that there's a lot still of good news to come probably, and I wouldn't be an aggressive fighter of the AI trade. But I would be really suspicious of anything that is AI adjacent and doesn't have strong earnings in the here and now. Part of the reason why you can be confident about something like a meta is that even if their potential to sell AI goes away, the rest of their business is an unregulated monopoly.

20:04And the AI that they have only makes it stronger. So those types of businesses, I think you can still feel confident about. I think health care is kind of an interesting place. It's been so maligned. And Badger Meter, that's a cane holding for the last 14 years. So I support it, too. What was the name of the company again, Julie? Badger Meter. That's a small company. Badger Meter. BMI. I could use a Badger Meter. Well, you could use a meter for a lot of things, Tim. Meantime, check out shares of Microsoft, lagging both tech and the broader market over the last few months. Microsoft down a percent since August.

20:36While the S &P 500 is up nearly 3%, the overall tech sector has rallied more than 6%. Dan, you flagged this one. What is ailing, Mr. Softee? All right. One of the first beneficiaries, if you think about it, they made that$13 billion investment in OpenAI starting, I think, in 22 and kind of leaned into it in 23. And so they were, again, you know, a lot of that market cap that they accrued over the last couple of years was because of that investment. But they don't have the models, right? And they also have the same situation where if they're buying all these GPUs, they have to deal with this depreciation sort of schedule.

21:09So they've also been unloading a bunch of their compute to some of the neoclouds too. So I guess the picture here is like, what's the step forward? Until they start telling us that their customers are using this compute, which they are, but using the models on them and they're basically able to charge them, I think the stock is probably stuck in the mud for a little bit. But that's a story we're really interested in hearing on their conference call. Yeah. At its peak, I mean, I think Microsoft is probably trading close to 32 times next year's numbers, which is obviously a lot more expensive than a broader market and far more expensive than Microsoft historically trades at.

21:42And a company that is not growing nearly, in my opinion, as robustly as they were at sort of the heyday about a year and a half or two years ago. So it's a valuation thing as well. And I think that's part of what's struggled here. But just look at Microsoft versus the S &P over the last two years and you're flat. And you would be expecting more out of the high growth part of the market. So agree with these guys. Coming up, investing in America. how J.P. Morgan is looking to align its financial plans with U.S. interests, and where CEO Jamie Dimon says the big bank is putting its money to work. Plus, a number of fast movers in today's market rally, the headlines behind the moves in metals, makeup, media, and masks.

22:17Oh, makeup. Don't go anywhere. Fast Money is back in two.

22:26Welcome back to Fast Money. J.P. Morgan pledging to invest$10 billion in areas key to U.S. national security, including critical minerals and so-called frontier technologies like AI and quantum computing. Shares of rare earth stocks like critical metals, MP materials, and USA rare earth all soaring again today. Quantum stocks like D-Wave and Regetti also jumping by double digit percent. Of course, the seeming off ramp, you know, between China and the U.S., that certainly helped the rare earth trade today. Helped the rare earth trade. It should help J.P. Morgan, too. I mean, if you think about the smart money, you have to believe JP Morgan has been in circles where, first of all, Jamie Dimon, I thought he's done a masterful job of trying to position around the challenge for him of embracing blockchain and crypto, but then pointing out where blockchain was critical to where the financial industry is going.