Closing Bell - Closing Bell: (Deep)seeking Answers 1/28/25

Episode Date: January 28, 2025

What does the future of investing in AI look like and has the game truly changed? We discuss with Big Technology’s Alex Kantrowitz and Alger’s Dan Chung. Plus, some are calling the Deepseek development a boon to startups looking for cash. Rich Heitzmann from FirstMark gives his take. And, we discuss some big moves in the Bitcoin miners – and what’s at stake for that group. 

All right. Thanks so much. Welcome to Closing Bell. I'm Scott Wapner live from Post 9 here at the New York Stock Exchange. And this make or break hour begins with today's tech bounce following that brutal deep seek sell off. NASDAQ is up sharply today and it is led by NVIDIA, the stock rebounding after suffering its largest market cap loss ever one day ago. And other market caps, mega caps are on the move as well. Just as many will report earnings this week. We'll ask our experts over this final stretch what's now at stake for those companies and investors, too. Apple, Meta, Microsoft, Amazon, Alphabet, they're also higher today. As for, well, there's Chipotle. It's negative.

As for the broader markets, a fairly muted day ahead of tomorrow's Fed decision. No move expected, but the commentary will be closely watched for clues on what lies ahead. It does take us to our talk of the tape, Seeking Answers, what the future of investing in AI will look like and whether the game has truly changed. Let's first bring in our dear Jabosa with the very latest on the ground out in Silicon Valley. What are you hearing today? I can only imagine, Dee. So, Scott, I think consensus is that, yes, the game has changed. But does that change the investment thesis? Maybe, maybe not.

Wall Street and Silicon Valley alike are still trying to digest with this new phase and building AI foundational models, AI software companies, applications, what this means going forward. Consensus seems to be that more AI equals more AI. This is going to be a good thing. It's just like, you know, when we got electricity, we didn't use less light bulbs. So more is going to beget more. And that will be positive. For the hyperscalers, for the foundational models, the public ones at least, the big tech, they're diversified. They have ways of distributing these use cases, which now the focus is on. So they're likely going to be OK.

Are they going to be spending the same amount, tens of billions of dollars in CapEx? That's an open question. And that's certainly going to be top in mind as we go through earnings season, certainly as it kicks off this week. Now, the political side, Scott, that's a little bit more complicated, right? Because although DeepSeek is an open source model and that is seen as generally good for developers and for the development of the industry of artificial intelligence, there's also the consumer side of this. And I just checked a moment ago, DeepSeek is still number one on the App Store, you know, still unseating chat GPT. That has very different implications.

That's not sort of an open source model that can be tweaked by users. That is users getting and giving information to a Chinese AI lab. And that's sort of the rhetoric you hear from the White House, from Washington, which is a little bit divided, you know, President Trump willing to say that this is competitive and could lead to positive outcomes, with others saying that this is a real threat. So certainly being debated at a number of stages and by different groups. It's certainly shaken things up, you know, to say the least. Has it put many, obviously the incumbents, on the defensive a bit out there?

It's a good question. I mean, you have a company like OpenAI, right, who led the way with a proprietary closed model. There's so many questions about what its business looks like going forward, how it's able to make money. I mean, I was talking to Ali Godzi, he's the CEO of Databricks yesterday, and he said that every company now, every model is going to be an open source model, whether they like it or not. There's obviously huge implications for an open AI whose business is getting people to pay for that model or for its APIs. But like I said, with some of the bigger tech companies, for an NVIDIA, which is not just a hardware company, but really an ecosystem that's going to have a place in the inference stage, which is where we are now, they're going to be OK. But for some of these companies and the VCs as well, by the way, that have poured billions

and billions of dollars into foundational model companies, much more of an open question. I appreciate the reporting, Dee. Thanks, dear Jabosa. Let's bring in now big technologies, Alex Kantrowitz and Aldridge Dan Chung. Alex is a CNBC contributor. It's good to have you both with us. Alex, you first. I mean, how do you think this

changes the conversation about AI and investing in it? Well, look, if you believed in the AI story before last week, you believe in it even more today. The mere fact of the matter is that building AI has just become far cheaper than it was previously. And if you're, let's say you're a box of the world or you're a service now of the world and your strategy was around implementing AI into your product, this is the best thing that's ever happened to you. Now, look, if your business is entirely built

on infrastructure, that's an issue because we don't know what's gonna happen on the infrastructure front. But ultimately, I think this places the emphasis where it needs to be, which is on the application. What are we building with the AI itself? We've talked so much about data centers, about cloud services. Have we really talked enough about the products that are being built with this technology? Because if it is as revolutionary

as everybody's saying, where's the revolution? And I think that's going to be the question. If this AI trade is going to continue, we're going to need to see applications. We're going to need to see concrete uses of the technology because the infrastructure side of it isn't going to be as exciting as it was now that the models are starting to be commoditized, starting with the deep sea Gar one model. This, Dan, a game changing moment for you as an investor. You look at the rundown of stocks that Alger owns in its 35 fund portfolio, Amazon, Meta, NVIDIA, Broadcom, Microsoft, Marvell, Vertiv, and on and on and on and on. You're highly exposed to this space. Are you less optimistic today or questioning it in any way?

We're always assessing where the risks and rewards are, but overall we're actually more optimistic. What DeepSeek is showing is that costs can be brought down dramatically and that dramatically supports if not enhances the opportunity in the real world to implement AI driven solutions right if the cost is lower it really tends to drive much more usage and consumption and so we're quite excited about still infrastructure providers key component providers providers, but also, yes, it's actually quite a bit earlier, I think, than perhaps we would have thought,

where we might see really interesting real-world applications. And of course, those real-world applications are enabled by the lower cost. I mean, when you saw the sell-off yesterday, what was going through your mind? That this was another overblown sort of announcement and a hyper reaction, probably by quantitative trading models.

But how can we be so sure? I mean, are you skeptical that DeepSeek was able to do what it did for the cost that they suggest they did it for? So I just convened the entire Alger tech team and e-commerce team, which is pretty significant. This is like an all-hands-on-deck. Yeah, for everybody in tech, AI, and senior PMs. And so one thing we'll note is, look, the DeepSeek model uses a technique called distillation.

That means it actually leaned on or learned from Lama and learned from some of the other chat GTP4. So the big models. So it's riding on the backs, if you will, of the masters. So yeah, we'd still like to see sort of more validation of exactly what the cost would be to implement and, you know, I think the people at DeepSeq have been very open. They do acknowledge that and, you know, I think the people at DeepSeek have been very open. They acknowledge that. And, you know, but they're still demonstrating that their architecture, and that's the key insight here, that their architecture could be one that rapidly brings down cost. Seems, Alex, if anything, it has changed the conversation as it relates to open source. It was no surprise yesterday, for example, in the mess that were many of these names,

Meta was green for obvious reasons, right? Because of open source. That's right. Look, we're about to move from a moment where the proprietary models were the leaders, the OpenAI GPT model, maybe the Anthropic Cloud model, to a moment where open source leads. And basically, someone who works in the industry put it to me this way. You would have OpenAI working on their model, Claude working on their model, I mean, Anthropic working on their model.

Now you have the entire open source community coming together and pushing forward, right? The DeepSeek model builds on Lama. Lama's going to build on top of that. You're going to have meta building on top of what the DeepSeek team is doing and vice versa. And so you're going to have this collective effort in the entire open source community to take these innovations and move it forward.

And what that's going to do, like I said, is going to commoditize the models. And so they're all going to keep working together. Proprietary is not going to be able to make as much money as they would otherwise because the open source is doing a good enough job. And that will definitely shuffle the deck of cards that we've been playing with up until now in AI. I mean, if you have any skepticism about DeepSeek, where does

it lie? And do you feel like part of the story is not being told correctly? There's been some guests on the network today who have called you know what on a lot of this story. So, Scott, I think this is the most important part here. You can't really have skepticism on the results. The model weights are being downloaded and deployed by people in Silicon Valley today. It's live in perplexity right now, you know, slightly adjusted from what it is. So the fact that you can run this model as cheaply as you can is the core innovation here. And that's gonna really change what it takes to build this technology.

Now the questions you might have is, did it really take them as little money as they're saying to build and develop this model, just a couple million dollars? Probably not. But ultimately you can't argue with the output, and the output is actually what changes the game here, not the process.

You feel like this is, Dan, a 99-ish moment for some of these names, only in the sense of the ones that got a big halo effect. And there were many that we had deemed the next great AI derivative plays. You know, the NVIDIAs and the hyperscalers are sort of maxed out. All the money had gone to them from investors. So where does the money then need to go? It needs to go to some of the software plays. Well, OK, we're tired of doing that. Well, what about the power plays? Right. That's the next great area. The utilities and other things that just were up, you know, 500 percent last year. Is that where the big issue lies today? I don't think so. I think it's still too early for that. So, for example, Microsoft is one of the hyperscalers, right?

They're about to report this week. Azure is growing roughly 35% year over year. Microsoft has said that about 10% of that growth is coming from AI applications. So they're still going to benefit from essentially all of this computing needs to run somewhere. It's running on the cloud, and the hyperscalers benefit from that.

I think what will be interesting in this moment is important for at least a beginning of a signpost is, okay, who has proprietary data versus, you know, easily available data, for example, is all of the works of Shakespeare, right? It's going to be hard to build a business model on top of that. But very proprietary data, for example, Amazon, Netflix, Microsoft, Google, what do they have? They have user data, a meta. You know, what are real people doing real time? What are the personalized data that isn't out there in the open but is extremely important,

for example, of course, for e-commerce and advertising? And then, of course, the software industry. What are they going to build on top of this? So we're looking now with a lot of excitement actually because of DeepSeek's result. Agentech AI, so personal agents, customer service agents, those are probably going to be some of the first examples of AI really proving the business case out. And we think we may see them earlier rather than later as a result of this. Do you have as much confidence today in NVIDIA as you did last week in their ability to beat and then continue to, if not raise their guidance to the degree they had in the past?

Because most think that that's not possible anyway. But enough to justify the excitement and the move in those shares. Well, yes, I do. I mean, we're doing a lot of research on this. And one of the things to understand is that training consumes a lot of computing power. But inference, especially at the edge, i.e. when you do something on your iPhone that requires any kind of agent AI, any kind of AI function, that inferencing function could be multiples of compute power needed.

And the interesting result, DeepSeq is run on NVIDIA chips. It's just run on last generation NVIDIA chips versus, say, upcoming generations or highest end. Blackwell is coming from NVIDIA. We think there'll be plenty of people pursuing both ends of the AI spectrum, right? There'll be some still pursuing the absolute high-end cutting edge. Think about internet search. There was Yahoo.

Then there was Google. But in between, there were Ask Jeeves, AltaVista, Bing has been around. That battle to be the ultimate dominant AI model is going to continue for a while still. And at the same time, with the lower cost, a lot of more pragmatic, practical applications can be possible, particularly through software and hardware. Alex, who do you think the biggest losers are in all of this, at least for what we know today, which is not a lot.

Yeah, it's OpenAI and it's Anthropic. It's the companies that have been building the proprietary models. I think you see Sam Altman over the past 24 hours really scrambling, making promises about pushed up releases. He just posted a selfie with him and Satya Nadella, like we're supposed to take something out of there. If you have an open source model that's effectively equaling the performance of some of your state of the art models, you have some questions to answer in terms of what the value proposition that you're bringing is. I mean, they're going to have to continue. They're in the hits business now.

They're going to have to continue bringing new innovation forward and advancing beyond what we just saw from DeepSeek. Same with Anthropic. There's this thing called Chatbot Arena that ranks the LLMs top to bottom. And DeepSeek right now is tied for three, and you have Anthropics, Claude, and the teams right there. It's a company that got $4 billion in investment from Amazon last year and might be raising another $2 billion. So it's really a rough moment for the proprietary models. I think the NVIDIA sell-off, by the way, was a little bit overblown yesterday.

But I do see a reason to worry over time, which is basically that if this is all commoditized and they can, you know, they might have an issue because if we don't have the applications to run on this stuff, then over time the demand for the infrastructure will be down. Getting 7% back today. Points well made. I appreciate the conversation, guys. Thanks to you both, Dan. Thanks for being here.

Alex, we'll see you soon. Alex Kantu is joining us. You did give us a good segue, too, because one of the key questions is how all of this impacts the open AIs of the world. That company and its co-founder responding today. Our Kate Rooney is following that side of the story for us. What are we learning here today, Kate? Scott, well, Alex teed us up well. OpenAI, as you mentioned, and other AI leaders,

think of Anthropic, for example, they were able to avoid that public repricing we saw in markets this week, but this absolutely has an effect on those companies. I'm talking to investors on those cap tables out here. There's a little bit of anxiety, I would say, about what deep seek means for these companies. It could really change the economics here,

the subscription models. They are closed source. And then the price as well. It also begs the question, are these companies overcapitalized? So if this tech, as you guys have been talking about, is so much cheaper to build, think of the top private names.

As I mentioned, the big ones there, OpenAI, Anthropic, XAI, they've raised a combined $50 billion raising cash in the long term, or too much cash can often dampen return on equity, among other things. And then regardless of how cheap DeepSeek was to build here, the efficiency gains, they're real. So investors in those US companies argue, well, enterprises aren't going to want to use that Chinese-built AI. But then you also have the domestic startups here who are looking to basically recreate that and undercut on price as well. Bottom line, it turns up the heat on these

companies to justify those valuations and really wow the world with their next models. The pressure is on there. The bulls would argue price was coming down anyway, and OpenAI and others can supercharge the usage of these to sort of offset that. Sam Altman, though, speaking out, calling DeepSeek's model impressive on social media. He said he promised to, quote, obviously deliver much better models and pull up some releases, Scott. So a lot to watch. Yeah, expected that response, I suppose. And we'll see if there are more in the days ahead. Kate, thank you for your reporting. That's Kay Rooney. Some are calling the DeepSeek development a boon to startups looking for cash. Rick Heitzman is the founder and partner at Firstmark Capital.

He joins us from down in Miami today. It's good to have you on on a day like this. How are you thinking about it from the VC's perspective? Yeah, thanks, Scott Scott for having me on. What we're seeing now is it's still really early in the AI race. We saw with valuations and what was happening was people were expecting that the game was already won, they've picked the winners, they valued them like winners.

But what in reality of it is we're still in the second or third inning and we're still seeing a lot of both the infrastructure and the application layer playing out. I'm wondering what you do think about the issue of startups who are maybe feeling new life today in their own ability to raise cash. And that field is now more wide open than it was perhaps a few days ago, literally. I want you to listen to what the billionaire investor Mark Cuban told me on Halftime Report today and get your response to that.

I think you're going to see a lot more entries into the AI space. I think you'll see a lot more private investment into small up-and-coming companies. I can't tell you how many conversations with private companies I've had, and now it turns out I was wrong where it was like, well, shoot, you're going to have to raise billions and billions of dollars just to compete with these guys. And the question was, how big a model and how much of an investment do we need to make?

That all just changed. And so I think you'll see a lot more companies get funding that otherwise might not have gotten it because it won't take as much money to compete. How about that, Rick? What do you think about what Mark Cuban told me? I 100% agree. I think you're going to see that DeepSeq shows that with a very little amount of capital, if you have a superior product, you can compete with the end user.

It shows you could build on top of the existing models and even the open source. And also shows that we're so early in this game that there's still a lot of venture opportunities, especially at the application layer. And 2025, Scott, as we've talked about before, might be the year of AI agents, both on the consumer as well as the enterprise side.

Wondering, too, the implications for investors in our public markets. And I also thought a point that Mark made that I wanted to run by you was there aren't enough public companies in this space to invest in quite yet outside of the hyperscalers and the NVIDIAs of the world. And that's one of the biggest issues that now, if these companies maybe are going to have a better chance of being funded, if there ultimately

is going to be a day where more of these companies are going public, listen to what he said about that and get your take on the other side of it as well. So I think one of the fundamental issues with the market right now is no companies are going public. You know, there isn't. Back then, you know, Amazon was a new company. Dell was relatively new. You know, there were incumbents that we all looked at as being, you know, the default powerhouses. But there were new companies coming public that we can look at and say, okay, maybe these guys have something unique. I mean, if it was today, Google probably would

not be public. Amazon might not be public. You can go down the list. And there are companies right now that are up and coming in the AI space that are private. And because they're private, nobody gets to invest in them. Well, you know, typical investors don't. And you don't really get much conversation about them. And I think that impacts the conversation and the market movements that we saw yesterday and today. What do you think about that? Yeah, I agree with some of what Mark said.

I still, again, being it's so early that most of the companies are not public and investable by most people. And that's fine because we're still trying to figure out what the rules of the game are. You're going to start to see people that are working on the AI infrastructure, companies like CoreWeave that will probably come out in the near term. OpenAI might go public this year. So there's going to be the beginnings of that coming out. But on the agent side and the application layer, you're going to start seeing those companies get ready to go public in 25 and 26. And that's still being sorted out, who the leaders are, who has the first mover advantage,

and who are gonna be the fast followers and create tremendous value, and really be all that's promised in AI. You have a take on Nvidia and how much it sold off yesterday, what the broader implications truly are for that company, if in fact DeepSeq was able to do what it did for such lower cost? Yeah I think that everyone believed there was a conventional wisdom

that the winners were the winners and it was a wildly capital intensive game that you had to play and therefore you either had to have the backing of a hyperscaler like Microsoft, if you're open AI, or you had to have such a huge infrastructure like Nvidia to be successful. And that's been proven wrong. In addition, the closed models that cost a tremendous amount of capital to build and maintain were the only way to go. And now you're showing that open source work.

So because of those two reasons and the conventional wisdom that companies like Nvidia were a winner, they were priced to perfection. And so some of the air coming out of that is probably healthy and what it's showing is there's going to be more of an open field in every level of the AI activity chain for startups to compete as well as existing software companies and infrastructures to compete going forward. Does that mean that perceived moats are going to evaporate or frankly might have overnight? They have definitely evaporated, especially the capital moat. So OpenAI, even things like

Anthropic were saying, hey, if you can't raise $5, $10, $20 billion, you're not going to be able to compete in this arms race because it's a game of capital intensity. And therefore, they were kind of blowing out the bottom end of the startup community and even public companies that couldn't keep up with that CapEx race. Yesterday proved that that's not true. Yesterday proved that you could be a capital efficient

startup and find an area to compete. And you can compete on a broad basis, or even compete on a vertical application where you have proprietary data and are able to tweak an open source model to make it work better. And that really opened up the playing fields with lesser funded companies and companies that want to focus on a particular part of the whole ecosystem. And so I think that's great for the consumer.

It's great for competition. Great to get your insights as well. Rick, thanks for that. Enjoy Miami. We'll see you soon. That's Rick Heitzman, First Smart Capital. Thank you. With his take on what we are witnessing in these markets.

We are getting some news out of Washington at the moment. Eamon Javers has that for us right now. Eamon? Scott, we've got a new post on social media from the White House press secretary, Caroline Leavitt. In the new post, the press secretary is saying that the White House is aware of the Medicaid website portal outage. She says, we have confirmed that no payments have been affected. They're still being processed and sent, and we expect the portal will be back online shortly. Now, the shutdown of this Medicaid portal affects obviously millions of people, billions and billions of dollars

in spending to doctors and hospitals inside the country that benefit people who are on the Medicaid program. Scott, all of this resulting from confusion around today's executive order, last night's executive order from the president directing federal agencies to disrupt their spending, put a pause on federal spending. And the White House has spent much of today clarifying exactly what they meant by this. And Scott, it may be that a lot of this confusion is stemming from one comma in a sentence in the original OMB memo, which said that federal agencies must temporarily pause all activities related to obligation or disbursement of all federal financial assistance, comma, and other relevant agency activities that may be implicated by executive orders.

So it may be that what the White House intended here was not to stop down all federal financial assistance, but only that federal financial assistance, which was implicated by executive orders from President Trump relating to things that are important to the MAGA agenda, such as DEI, gender terminology, and the like. And so what we're getting now is some clarification from the White House that this is not a freeze of all federal financial assistance. The details in this matter, Scott, that comma may be responsible for $100 billion worth of confusion today and a lot of anxiety throughout the federal government and the entire payroll system of who gets paid by the federal government, including states, nonprofit organizations, universities, and a lot of publicly traded companies as well.

Scott, back over to you. The dreaded punctuation mark. That's right. Gets you every time. Eamon, thank you very much for that clarification. That's Eamon Jaffers. We're just getting started. Up next, looks like deep sea concerns could actually benefit the cyber software stocks. We'll tell you why. We'll run through the names that could see some serious gains just after the break. All right, welcome back. We're tracking some big movers today in the energy space. Pippa Stevens has that for us. And this is related to this story as well, Pippa.

That's right, Scott. And the nuclear power stocks that got hammered yesterday on the Deep Seek News, fighting to make back some of their gains with Vistra and Talon both at more than 4 percent, while Constellation is hovering around break-even. Now, a number of Wall Street analysts saying the declines were not entirely justified. Morgan Stanley calling yesterday's plummet an overreaction, saying U.S. AI infrastructure will continue to grow rapidly, while Evercore ISI said it was shooting first and asking questions later. And Jeffrey said the drop was a buying opportunity for Talon as the dust settled.

Now, Oklo, the small modular reactor maker backed by Sam Altman, up 6 percent and making up some of yesterday's losses with uranium stocks also in the green. Traditional energy stocks, though, not seeing the same type of bounce, although gas-focused driller EQT is an exception up here, about 3%. Of course, it is important to remember that while AI is one theme driving power stocks, it's not the only one. We also have reshoring and electrification, which are also, Scott, raising power demand. Pippa, thank you. That's Pippa Stevens. Software and cybersecurity names are also seeing some big moves today. Seema Modi documenting those for us. Seema?

Well, Scott, the market seems to have decided that the DeepSeek findings are a positive for software given the fierce debate around whether big tech will cut spending on chips. Goldman Sachs says that AI models can in fact run more efficiently at a lower cost. How that will spur more competition among software players building AI AI applications, and increased demand from enterprise customers. Analysts there, they see the benefits of lower compute benefiting names like Salesforce, Adobe, and ServiceNow, which does report earnings tomorrow. Cloud providers, as tracked by the WisdomTree Cloud ETF, take a look at this on pace for 10 straight days of gains, outperforming semiconductor stocks. And then there's cybersecurity. That has also been seen as a good place to hide in the last two days with CrowdStrike trading at a new high, up about

eight, nine percent on a report that DeepSeek had been hacked yesterday. Again, that stock building on its gains going into the close, Scott. All right, Seema, appreciate that. Thank you very much. Seema Modi up next. iConnection COO Caitlin Mallon on the growth in alternatives. We are back on the bell right after this. Welcome back. A volatile week already in the markets as tech earnings and the Fed meeting loom large in the days ahead. One of the biggest trends in the market over the past couple of years has been the growth in alternatives. They are the focus of the iConnections conference in Miami. Caitlin Mallon is COO and joins me now. It's nice to see you. Looks nice down there. Yeah, it's nice down

here. A little bit of a breeze. Good to see you, Scott. Can you give me an idea of who exactly is gathered there and what you hope to accomplish? Yeah, this really is. The iConnections Global Alts Conference has become the place to be if you're in alternative investments. We are an index for the alts industry. We've got about 60% on the hedge fund side,

another 40% of that is a liquid strategies, and there's 1,000 funds. Overall, we've got 6,000 people and then another 1,200 investors. And that's half institutional and then half family offices, fund to fund. So you're seeing every little bit of the industry here. Where are we, do you think, in the evolution of alts as individual investors should be thinking about them? They certainly are all the rage.

On this program, we talk about it all the time. Some would suggest, are we at peak alts? How would you answer that? Yeah, I mean, if you're down here in Miami, it certainly does not feel like we're at a peak. The room down here, 6,000 people, is absolutely buzzing. We have an industry who's excited. This event in particular, if you're down here, it's not just the fund marketers who are looking to raise capital. It's the investment decision makers, the portfolio managers. We obviously have a bumpy week in terms of market commentary,

and everyone's here talking about that. So I think that we're just at the beginning with alts. It's continuing to grow, and everyone's excited. You know, if there's one, I guess, risk factor that we've been thinking about and individual investors need to think about, too, alts are not like regular equities, right? They're, for the most part, illiquid, not easy to trade in and out of private equity and hedge funds and the like. But you're seeing growth in another part that, in fact, is liquid, liquid alts. Can you tell me

about that? Yeah, definitely. You know, down here we have across, like I said, we've got 60 percent of the funds that are here are on that illiquid side of the strategy. Collectively across everyone who's here, it's representing 55 trillion in assets. In this week alone, we'll host 20,000 meetings. And after this conference, 10 trillion of that will likely exchange hands because of the event. So liquid alts is a huge part of that driver. I think that we're going to continue to see interest there because as the market opens up, as people start to see a more flexible regulatory environment, liquid alts are

just a big part of what they want in their portfolio. It's good to catch up with you. Enjoy the conference. We'll see you soon. Sounds good. Thanks, Scott. All right. That's Caitlin Mallon, iConnections. Up next, we track the biggest movers as we head into the close. Seema Modi is standing by with that. Hi, Seema. Got 21 minutes left in trade. A big day in travel. One stock hitting a new high while the others are falling. We're going to get you the full story

after this break. We'll be right back. All right, about 15 away from the bell. Back to Seema Modi for a look at the stocks that she's watching. Hi, Seema. Scott, let's start with Boeing. Shares are climbing after CEO Kelly Ortberg said the company is making progress on its recovery and is stabilizing production. The aerospace giant posted a nearly $4 billion loss for the fourth quarter and had its largest annual loss since 2020, but shares up about 1%. Take a look at Royal Caribbean hitting a new high on a big fourth quarter beat and a growing number of younger customers getting on board, really fueling sales. The cruise operator also announcing it's entering into the premium river cruise market, which is dominated by Viking. Royal up about 11 percent at this hour. JetBlue shares, different story, plunging. The airline

issued a disappointing outlook, saying its unit cost, excluding fuel, will rise as much as 7 percent this year. Here's what CEO Joanna Garotti told CNBC earlier. Quarter one is a trough quarter for JetBlue. We're a leisure airline, a lot of trough weeks in the quarter, and that's what we're seeing. When you adjust for Easter, our RASM improvement from Q4 to Q1 is on pace, if not slightly better than what we've seen in years prior. JetBlue shares currently down about 24% on pace for its worst day ever, Scott. All right, Seema, thank you. Seema Modi still ahead. We'll break down some big moves in

the Bitcoin miners today. What's at stake for those names as well? That's coming up. Closing bell is coming right back. Quick programming note. Don't miss our exclusive interview with Double Line's Jeffrey Gundlach. It is right after Fed Chair Powell's news conference, as usual. Remember, tomorrow's Fed Day. Closing bell tomorrow afternoon, right when the Fed chair is finished with his own remarks. We will speak to Jeffrey Gundlach up next. We get you set up for Starbucks results at the top of the hour.

That and more inside the Market Zone next. We're now in the closing bell market zone. CNBC Senior Markets Commentator Mike Santoli is here to break down these crucial moments of the trading day. Plus, Taneya McKeel on the moves into Bitcoin miners. Courtney Reagan looking ahead to Starbucks. The earnings out in OT. I'll turn to you first. I mean, does this qualify as a pretty decent bounce back?

It is. Certainly on the index level, it's an almost perfect inversion of yesterday. Yesterday looked like a washout based on the NASDAQ and the S&P 500, when in fact the majority of stocks were higher today. You have just about 1% bounce in the S&P 500, even though breadth has been negative and it's been a little bit less inclusive. Tells me a couple of things. One, just before around noon, the NVIDIA fought its way back above its 200-day average. And I think it gave emboldened traders to say maybe this can be a real rebound. And the entire S&P kind of released higher. And so it's Apple, NVIDIA,

Microsoft powering things today. Bigger picture, it tells you the rotational mechanics of this market remain intact. The market sort of repairs and rebalances itself when it can, when yields are tame and when there's not a lot of macro stress. So all to the good, although I would say more of a kind of low volume lift than it is a real stampede back into stocks. Some interesting commentary from down in Miami at iConnections from Point72's Steve Cohen. I want to read you a couple of the headlines, get your reaction to them. Says, I would expect markets to top over the next couple of months if they haven't already. Says, President Trump's immigration and tariff policies

will slow growth if implemented. And says, what happened yesterday with DeepSeek is bullish for AI. Yeah, interesting, the idea of a market top. Who knows if he means like a huge, profound, ultimate market top or just, look, we're due for some kind of downside move. I do think that the focus on policy to be a growth restraint or just kind of dust in the eyes of the economy because of so many things being kicked up

and tariffs not so much being a direct inflationary pressure, but being something that just gives businesses hesitation. Yeah, I think that's one of those things you have to have on the screens at all times. It's a potential risk. So interesting. I mean, he didn't exactly make his living making huge market-wide calls. He's a trader of the tape. But it's interesting how he sees the risk-reward here in the intermediate term.

Says of NVIDIA, great companies are going to be expensive. Yeah. Got to pay up for the great ones. And by the way, it's less expensive than it was a couple of days ago. It really is. It's like a mid or high 20s forward PE right now. If you believe those earnings, it doesn't look terrible relative to the alternatives. Yeah.

So Fed meeting tomorrow, you know, we'll start to get the mega caps coming as well. And the commentary is going to be dominated by this topic now. It is. And, you know, everything from potential macro influences of cheaper AI or whether there's some costs involved in all of this CapEx that we've been doing. I mean, none of this stuff is going to be settled soon. In my mind, the market really was in a hurry yesterday, I think, to say, OK, here are the new winners and relative losers. So I think it's what it's more a matter of the next phase of the game gets just a little bit more complicated. It's a mature AI theme or maturing AI theme.

And, you know, in terms of the Fed itself, going to be, you know, interesting scrutiny in the press conference about the basis of their decision making and forward looking guidance and all the rest. Although we don't have a dot plot to chew over this time. And we'll talk to Gunlock right on the other side of it, as we always do. I'll come back to you in a minute. Let's talk crypto. Taneya, tell us about these Bitcoin miners. Yeah, Scott, Bitcoin miners were some of the hardest-hit names in the deep-seek-driven sell-off yesterday, and they extended those losses for much of today while the broader market rebounded.

Take a look at Core Scientific and IREN, both down more than 30 percent yesterday and about 6 percent earlier, but have really fought their way back from those lows now. Bitdeer jumped today and Benchmark's Mark Palmer puts this on revenue diversification. Scott, most of these Bitcoin mining firms have both a Bitcoin mining operation and sell power to AI companies. Bitdeer, additionally, has plans to sell their mining rigs this year, and Wall Street is very, very excited about that. Now, analysts from Piper Sandler and Cantor Fitzgerald suggesting today that the sell-off in hybrid miners was overdone. Mike Colonese from HC Wainwright, however, has a more contrarian view on the AI opportunity, says pure-play miners like Merrill Holdings and CleanSpark should outshine the hybrids later this year.

Both those stocks were up most of the day and are ending here slightly lower. Scott. All right, Taneya, thank you. Taneya McKeel. All right, mega caps start tomorrow. Starbucks is tonight. Courtney Reagan. Hey, Scott.

Yeah, so it's going to be the first full quarter for Starbucks with CEO Brian Nicol at the helm. He started September 9th, so he was there for the last call, but not overseeing the quarter. So investors are really hoping for a jolt to spur the stock price up. It's only up about 7% over the last 12 months. Now, the coffee giant expected to report a drop in same-store sales. It's gonna be the fiscal first quarter they're reporting,

so that would mark the fourth straight quarter of declines for that particular metric. Consensus is for a drop of 5.5%. That is smaller than previous quarter's 7% decline. We'll see what we get there. International comparable sales also expected to see a sharper decline than the broader number. Investors are going to be listening, of course, for any discussion of traffic trends, both in the U.S. and China market specifically. And you know that Starbucks has fallen short of revenue expectations for the last four quarters.

It hasn't posted an earnings beat since November of 2023. So expectations are probably a little low going into the print. And Starbucks suspended its previous guidance as Niggle took over. So investors are going to listen for any new guidance if we're going to get any now that he is there and has been there for a quarter and has a plan in place. Starbucks also hinted at job cuts. We're wondering if, in fact, those are on plan or not. So still a lot of questions here to go in the next, what, 15 minutes or so when we get those results. Scott. All right. We'll see in a little bit when those numbers drop. Court, thank you. I mean, nickel has gotten all but a free pass to this point.

And at some point that's going to run out and we're going to have to see some tangible results of a turnaround. For sure. Fact in store. Yeah. Got a lot of credit up front for what might happen there. And I agree that the general sense on the street is that expectations are in check and they're pretty low in terms of the near-term results and maybe even the guidance. But the stocks had a pretty aggressive move. So market-based implied expectations might be a little bit higher than they are in terms of the official consensus. But very interesting. It's been a real laggard among the big kind of elite brands of consumer discretionary. So see if anything and also see if there's any color, as there sometimes is, on consumer macro. I think there's great comfort right now with the condition of consumers coming

into the year for good reason. You know, the credit situation in households looks fine. The Atlanta Fed's at 3.2 percent annual real GDP closing 2024. A lot of that is the consumer strength. But we'll see if they have any color in terms of more near term how they came in. There was a little bit of noise in today's consumer confidence report that was suggesting that there was a bit of a rethink of the post-election excitement about consumer prospects. Thank you, Mike. I'll see you tomorrow. A big day, of course, the Fed decision, Meta, Microsoft. Again, we have Starbucks coming up in OT. And we have a pretty decent rebound today for some of these mega cap names that sold off so dramatically one day ago.

NVIDIA, of course, among them up at the highs of the day into the close near 9%. Bounce back. Some of the other big names, too, within the mega cap also looking pretty nice and green today. I'll see you tomorrow in overtime.