Closing Bell - Closing Bell Overtime 10/19/26

Episode Date: October 19, 2026

From the open to the close, “Closing Bell” and “Closing Bell: Overtime” have you covered. From what’s driving market moves to how investors are reacting, Scott Wapner, Melissa Lee and Michael Santoli guide listeners through each trading session and bring to you some of the biggest names in business.  Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

The bell's bringing an end to the Trady Day at the NOA C, LXP Industrial, ringing the bell and at the NASAC Obsidian Therapeutics, doing the honors. Welcome to the closing bell overtime. We're live from Studio B at the NASAC market site. I'm Melissa Lee, alone with Mike Santoli. Stocks mostly higher today as markets reevaluate. Last week's tech sell off Dow, slightly lower, S&P 500 higher, right around 7,700.

NASA composite with the strongest rebound, although well off the highs of the session. For sure, the AI trade need, mostly bouncing back. Sandisk and Micron both were higher. Sandisk. snapping a seven session losing street. Twilio rebounding after a 10% loss on Friday. Still, core weave the stock at the center of last week's weakness lower once again today. Much more on the software trade coming up.

Bond yields, meanwhile, on the rise as the recent safety bid eases somewhat adding a few basis points higher across the curb. Earning season picking up tomorrow, 3M and GM reporting in the morning, Netflix and SpaceX reporting in the afternoon. We'll get you set up for those reports. But first, the first take on the close, it was a very interesting session. There was an effort at the beginning of the day, but as you went into the close, you definitely saw a leg lower, especially across the board when it comes to the chip names, NASAC 100. Yes, I don't think there's really a lot of strong conviction behind a lot of these moves.

I think as a positive, and there was a lot of talk over the weekend that we were gearing up for something uglier. It seemed like forced liquidation potentially last Thursday and Friday in some of the semis. So they got a little bit of relief. I think it makes some kind of sense. Credit spreads did not blow out again. But even under the surface, all you really show is the most oversold stuff in the market get a modest bounce. Banks were down 1%. Market breadth overall was negative. So I do still think we're kind of fighting this out. You have the fundamentalists out there saying earnings are going to bail us out.

Everybody else saying, you know, financial conditions are tightening. Credit conditions are different than they were just three months ago. Yeah, I think it's also interesting to see if you pull up all the activity on, for instance, a core weave, right? You had the Goldman's Act's upgrade of core weave. You had almost every firm out on the street defending. core weave today and it still didn't do the trick. You had Intel upgraded by JPMorgan still closing at the lows of the session. It's falling on deaf ears at this point. It's sort of like, we still don't know if the situation is resolved at all, if anything is different from last week.

No, it's tough to know exactly what might change that equation. Obviously, we have a Fed meeting next week. I think that there's some hesitation in there in terms of whether, in fact, this is exactly what the Fed was trying to do, is essentially create a little bit more of a higher hurdle and tighter conditions. in the market. So we'll have to, you know, see if anything kind of breaks the fever. It has been a rough go, as we've been talking about, for the tech trade. NASDAQ tumbling 4.5% last week on concerns about AI debt and spending unnerving investors. Corweave became the posted trap for that sell-off, plummeting 25% last week.

It was his biggest one week drop in nearly a year. The company's credit default swaps reached distress levels as well. And while the NASDAQ did post gains today, Corweave, as we said, unable to hold the gains. it was earlier up about 7%. So now, are concerns actually over? Are they easing? Joining us now with Tyler Ratke. He is city's co-head of U.S. software research. Tyler, good to see you. Thanks for having me. Certainly we'll get you on the whole setup for software and IT spending trends and everything. But CoreWeave, I mean, you have the market questioning the capital structure, really the business model of the neocl clouds. How do you assess it? Look, I think this is simply a financing issue, right? I mean, it's not helpful that the 10-year

rates have gone up, you know, 30% off the lows in March, right? That has ripple effects to anything that's not investment grade. So I think the bond market is really in control of this narrative right now. It's very tough to fight that, especially ahead of a Fed meeting, ahead of some big midterm elections. I think if you look everything from a demand perspective, demand could not be stronger. You look at the spot rate for H-100s, these GPU rental prices. They're basically at record highs. The AI demand, you know, as we think about the labs out there that are talking about going public, is very, very robust, right? So these companies are locking in contracts at the highest prices

they've seen. They're telling us that the margins are higher. Now, the cost of capital, that's super important, right? Like, that is part of this model. But our view is if this is simply a financing issue, those higher financing costs will get built into those contracts when they go out and negotiate with customers. So we think this gets figured out, but again, in the very near term, as we have all this uncertainty on the financing market, I think that's why you're seeing a lot of the volatility today. We're getting a lot of news out of N-scale, which is also in the neocloud space, and there was a report from Bloomberg that Microsoft is walking away from a lease. And that may signal that perhaps hyperscalers are in the driver's seat when it comes to walking

away because of bottlenecks. In this case, it was because the UK didn't have enough power. But certainly we face the same issue here in the United States, potentially with power grid bottlenecks, but also zoning bottlenecks, et cetera. Can you speak to how solid these contracts are and if they can, in fact, be canceled? Well, you have to deliver on a contract, right? That's kind of what it comes down to. We've seen, I think, Google walk away with the contract with Crusoe, right? So we've seen this thing play out in the past.

Interestingly, with these bottlenecks, whether it's power or regulatory that you talked about, This actually increases the scarcity value of the things that the neoclouds are selling, right? If it is harder to stand up capacity to build these sites, the supply that's out there today and very limited supply is only going to be worth more. But look, I think you've got to do the right things, right? You have to put these in the right places. You can't route things through residential areas and expect there to be no issues. You've got to go through the right permitting process. I think that will improve, sort of we get past the midterms, especially it's very politically.

sensitive right now. But yeah, I mean, I think the hyperscalers, they need this capacity. They're willing to do it in-house. Microsoft builds a ton of their own data centers, right? And so, you know, I think that that's a good sign that Microsoft has confidence in their own ability. I do think there's, you know, this ongoing debate about just once things are built out to a greater scale, where the profitability is going to settle at, what layer of the whole food chain? And the market seems to be saying neoclads are at the end of the line. Right? I mean, in other words, you have the ability of either these big companies to find other solutions beyond just paying up for these massive data centers. They're trying to do things more locally. And then you have the

software layer where it's kind of implementing them on the enterprise software side. Yeah, absolutely. And look, most of my coverage is kind of more of these software companies, the Microsofts, the Palantiers. I mean, these are some of the best margin profiles on the planet, some of the highest cash flow. So I really do think that's what. where the value is headed. And in particular, enterprises are kind of grappling with, do we use frontier models? Do we use open source models?

Do we use small models, large models, right? But the software ecosystem is really what's kind of selling that vision to enterprises. You think about the things that Microsoft and Pallentier and some of the others are doing. So ultimately, we think those are the best positions, especially Microsoft. They're not relying on the debt market, right? They're still free cash flow positive, yet they're still. spending, you know, tons of money on KAPX every year. So you really think the value is going to be accrued at the software level, not necessarily

the neocloud? It sounds like there are a lot of questions about how much is going to be needed later on. And so, therefore, isn't it sort of a question mark as to whether or not you should be in the neocloud layer? Well, look, I mean, I don't think they're mutually exclusive, right? I think, you know, as we know about the tech chain, right, like the semiconductors and the hardware layers where we're seeing, you know, those stocks have massively outperformed software since the AI boom. Now you're starting to see some software stocks really, you know,

respond favorably here. But look, I think there's fewer question marks about Microsoft's long-term profitability. Their long-term value, you don't have to make a call on GPU depreciation, financing, et cetera. So I think that's sort of where our highest conviction ideas are, kind of more of those pure software ideas. Tyler, good to see you. Thank you. Thank you. Thank you. Casey City. Well, Netflix is set to report earnings tomorrow after the bell. The sock has been under pressure so far this year, down more than 25%. The report comes during a time of change at the company. Today, it was reported that the head of its film unit is out after three years. Last week, the company's chief content officer was shown the door. And the company also cut 6% of its total jobs. Will tomorrow's results get the street back on the streamer's side? Join us now as Needham Senior Internet and Media Analyst Laura Martin. Laura, great to have you with us. Nice to see you. It sounds like you think.

the stocks sell off or, you know, the sagging nature of its stock is overdone at this point. You want to take a look at ads, which you're actually optimistic on? Yeah, they've said they're going to do $3 billion in ad revenue this year, which would be sort of a doubling from last year. And I'd say it's long overdue and about time. But that form of revenue has an 80% margin, which gives them upward pressure on their margin and free cash flow return on assets. So that's great.

And Laura, you would think that at a time like this, when everybody is obsessing over, you know, the AI trade and companies that have too much leverage and where's the next, you know, kind of earnings boost going to come from, Netflix might have defensive properties, you would think, as a business model and where the valuation is right now. It doesn't seem as if, though, the street is willing to kind of underwrite that idea. I think that's fair. You know, for sure, leisure time and media feels defensive. However, their biggest competitor is YouTube. And YouTube aggregates 10 million creators, and the cost for those creators to make better content is falling because of generative AI. Like people can make users, like around the world,

can make better videos cheaper by using generative AI, and that then competes in the short version, three minutes, five minutes, seven minutes, with TV shows and films for time. So I think there's this competitive aspect of, you know, sort of user-generated content on TikTok and reels and YouTube sort of eating into the market share of Netflix and Disney and Peace Guy. In terms of Netflix's strategy to combat that, Laura, we heard earlier this year that they are going to start these micro dramas, for instance, are going more towards that YouTube style short-form content.

And I'm wondering from an advertiser's perspective, do you get concerned that their ability to monetize the advertising is less if they offer similar sort of short, if they went the other way and did less of the sort of the big picture sort of films, et cetera, and more of the short, you know, content, is that better or worse from an advertising profile perspective? Right. So much worse. But the key point is right now Netflix is doing like 20% sellouts. So they have another like three or four years before they're totally sold out in long form, at which point then it matters that short form doesn't monitor. as well, but they have been very slow at growing their ad revenue. I think this is our fourth or fifth year. We're in advertising, and we thought they should be $10 billion after two years, and we're

now like four or five years in, and we're at $3 billion. So they just have a lot of unsold ad units in their long-form stuff, and I think it's smart. Look, the job to be done here is to aggregate leisure time for consumers and monetize it. So you're right to ask, does short-form video or vertical video monetize less, yes, but they're not sold out on the other stuff. So eventually they'll sell out on that and then they have to worry about how to monetize short form video, which is problematic, but not in the next three years. What does aggregating leisure time look like in addition to what Netflix has already been doing with things like podcasts and the occasional live event?

My opinion is they must be in more live. I think they've got to be in news. I think they've got to be in more sports. I think they've got to be in the film business, meaning more windows of monetization. Remember, Netflix started as Pure Play TV, which means you only monetize in one window. And that window, called television, is mature. So now they're growing engagement 2%. And everybody on this call has probably already watched everything on Netflix.

So now their growth is slowing. So they need to add more types of genres, more types of content. and they need to then figure out how to monetize those. So they compete with the Walt Disney's of the world. Or, you know, we have this new sky dance competitor, which is the combination of Warner and P. Sky Paramount. And I think that's really a material new competitor for them. So they need to sort of up their game, in my opinion.

When you say news, Laura, what do you mean by news? Ask an anchor on a cable TV network. Okay, like a CNBC. Yeah, they should buy you. No, I'm just kidding. But they need to do news. They need to have some kind of. news, even if it's targeted at Gen Z. They must have all genres of content in order to amass

the maximum number of audience. Yeah, I mean, who knows? Maybe that blurs with the podcast stuff they're doing. The theatrical window, Narnia, two months next year. IMAX. We'll see if this all morphs the business model along the way. Laura, good to talk to you. Thank you, Laura Martin. Russian President Vladimir Putin reportedly wants a summit to negotiate an end to the Ukraine war, but doesn't want to include Ukraine's Vladimir Zelensky. Amon Javers joining us now with more details. Amen. Mike, that's right. President Trump responding this hour to that Wall Street Journal report that you mentioned today that said that Russian leader Vladimir Putin is pushing for that

three-way summit between himself, Trump and Chinese leader Xi Jinping to end the war in a deal that would not include input from Zelensky in Ukraine. The president said he would take that meeting. We would schedule a meeting based on the deed of and that's a very positive thing for the world. So I would do it if they want to have a, I just heard about it, if they'd like to have it, I think it would be held in China, he just left here. If president she wants to have it, I'd have it. I'm sure we'd talk about other things like trade too, but yeah, I would do that.

Meanwhile, Ukrainian President Vladimir Zelensky lamented the U.S. government's pivot towards supporting Russia over the weekend, calling the U.S. government's reverse. so unbelievable. And the Ukrainian government said that it hasn't been able to get a response from the Pentagon or from SpaceX about its questions regarding the throttling of the company's Starlink and Star Shield Services, guys. Back over to you. All right. Amen, thank you. Amen Jabbers in Washington. And take a look there. We want to show you Global X-ringing the closing bell at Cebo in Chicago, ending the regular trading day for options there. Well, the regular session for options may have just ended, but today marks the beginning of extended trading, a first step on the path to 23-5 trading.

We'll have much more on that next. And we will also discuss plans by Anthropic and Elon Musk to spread the wealth from AI. Will it sway public opinion? Closing bell overtime, be right back. Today, the NYSC extending options trading on a handful of stocks and funds beyond the regular hours, but that's only the beginning as 23 hours a day, five days a week trading begins soon. Oliver Renick joins us now with more. Oliver.

Hey Melissa, the world demands 24-7 trading, and today just concluded a big step in that direction for options. We just finished the aftermarket extended trading period for single stocks and ETFs from Sibo and Nisi. As you guys know, taking this closing bell at 415 every day, this has been going on for index options for years, but today was the first time for two dozen stocks and five ETFs. Looking at the tape over the last few minutes, there were some notable activity in Spy, QQQQQQ, But keep in mind, a lot of brokerages also offered aftermarket trading for the deepest and most liquid ETFs. This morning, Q's did the most volume with 7,000 options traded. This afternoon, there was some GLD trading going off as well.

This morning, there was a lot of Tesla. My gut tells me the afternoon or curb session, as CBO calls it, will be a little bit more familiar for traders, given the brokers that have been doing this already. And the fact that in the morning, it's just a little bit of a different demographic. you're going to be dealing with European and Asian traders who kind of pioneered this whole 24-7 trading to begin with in the crypto and perp world. So in the morning, might be a little different in the afternoon, probably a bit more familiar. Yeah, Oliver, it's a good point about it. Being familiar in the afternoon, I've been trying to puzzle around, you know, let's say

earnings season, right? I mean, a lot of companies, either they do or they don't want people to be able to transact easily, you know, as they release that news. So I just wonder how that might, you know, get reflected in this extension. session on these handful of names. 100% Mike. In fact, I think most of the feedback that I've gotten from individual traders or viewers has been about how this is going to affect their trading around earnings. No word yet.

In fact, it seems like the message also was from brokers is that they didn't really communicate a lot with their clients about what this was going to be today in terms of this actually getting off the ground. What's going to be available? I expect more coming over the next couple weeks and months as we get going with this that we're going to find out more about how companies. respond to this as well because to your point those kind of black out brownout periods they might have to change and it might shift some of the timing for that and certainly the way you want to trade

also the implied volatility the way we talk about market maker moves in theory that could change as well it's only been one day all over so far at sievo at least but what what have the volumes been like i'd imagine they'd be pretty thin very thin that's the big thing here i made the analogy earlier and not to hammer at home but i think it's the best one think about this as basically a poker room. The guys who are waiting for a new table to open up, they're usually the ones that either are looking for higher stakes to play because they've got bigger bucks, or it's their preferred games. So in the morning, traders probably need to be very careful. Now, this morning, we saw people going out and buying far out of the money, 14-cent flyers on

Tesla. The price of that type of option is not really going to change no matter when you trade it. But if we're used to trading more liquid deep, SPX or S&P or individual stock options, you want to be careful about the bid-ask spreads you're getting. You want to be very attuned because when there's low liquidity and low volume like there was this morning, we're talking companies like Nvidia, they did 150 contracts. I mean, this is like a fraction of a fraction of what they do throughout the day. So you definitely got to be on your toes in the early morning session. We'll see what the numbers come in for this afternoon. But as more people pour in, in theory, more liquidity, more hours for trading should bring price discovery over the long term.

Yeah, that would be what history would suggest ultimately. Oliver, thanks so much. As Anthropic marches on toward an eventual IPO, the company says it will put 5% of the equity of the company in a public benefit fund, but not in a nonprofit charity. At an estimated $1.5 trillion market cap, that stake would be worth $75 billion. On X, Elon Musk saying Anthropics should give the shares to the public and says that that's what he'll do, pledging to give one share of SpaceX from his personal holdings to every kid in America. Anthropics still needs approval for its plan and statements made on X are not legal in binding. So neither of these plans may ever come to fruition. And immediately everyone is sort of saying, well, maybe that sure makes sense. It's a little bit of a down payment on a societal insurance policy about AI. Or cynically, they want to become too big to fail in everybody's account and essentially become endorsed by the country.

I mean, my first question was, first of all, what are the mechanics of this? How does this actually get done? But also, you know, if you're putting 5% of shares into some sort of public trust, who owns the voting rights to that? Right. So who will control that? That's also an interesting sort of, you know, thing to think about. Yeah, is it founder shares? Right.

Are they super votes? Is there a separate board for it or something like that? I mean, obviously people do pledge these big chunks to foundations and you have their own board, no doubt. You know, and the Musk piece of it, I mean, from his personal holdings, it would be just a sliver of his ownership. $10 billion in total. Not bad, but I don't know if it's life-changing. Mackenzie Bezos, by the way, is given away $26 billion. Actually giving it away.

So, you know, it's doable in another way. If you believe the Tam is like all of the universe in outer space and beyond that, then maybe one share is... 150 bucks a share goes to millions, yeah. Well, AI is drastically changing the way we work, but we've been through changes before. Up next to take a look at women, how women in the workforce have adapted before and could again. And we're also watching a new theme emerging in AI, agent swarms, teams of AI agents that team up to complete your tasks. But will they do more harm than good? Well, as Fox CEO, Aaron Levy.

Every major technological revolution has reshaped women's role in the workforce from the washing machine to the typewriter to the PC. Today, AI may become the next great turning point. Steve Leesman's got the story. Before there was Siri or Chat GPT and Claude, there was... Go ahead, please. For much of the 20th century, connecting a phone call required a human being, and that human was almost always a woman. Telephone operation became the single largest occupation for white U.S.-born women under the age of 20.

By the 1940s, nearly 95% of America's telephone operators were female. Today, the nation's far-flung telephone network has become the nerve system of this democracy. A decade later, automation caught up, allowing America's, to place long distance calls themselves. The need for switchboard operators slowly disappeared. Research shows the women working the switchboards, they mostly dropped out of the workforce entirely or wound up in lower-paying jobs.

Most people are not changing careers very often. They make their career decisions early, and they stick with those careers. If the value of that work, not just the specific job, but the type of work they do, suddenly goes out of fashion, becomes technologically obsolete, they're going to, in general, do something not as good. Just as switchboard jobs faded, millions of office and clerical jobs emerged.

And it was just in time. Not for the switchboard operators, but for the next generation of women and those who could now enter the workforce because of another technological development, the washing machine. This piece is part of a CNBC series, The AI Economy. You can read about it and watch the documentary on CNBC.com. And there you'll learn how the washing machine. seen and the PC had profound effects on women in the workforce. My personal takeaway from the research,

there's no perfect historical analog for AI, but history shows there will likely be job losses on the one hand and new opportunities on the other. What's different this time, Melissa, more women are more highly educated. They hold more professional and managerial jobs. That offers some protection compared to past innovations. But then again, the whole point and the whole promise and threat from AI is that it moves up the human value chain. Sure. And, and see, I'm just wondering because the flip side to it are jobs that are seen as AI proof. Are women represented well in those professions compared to others? Better than they were, but not as much as you would think.

There has been gains for women in executive roles and gains in managerial and professional, but still a lot of these jobs. The latest study that we've seen shows women are perhaps three times more likely to be vulnerable to AI developments than men. Yeah, and I guess you have to make sense. some call on exactly how quickly this change is going to happen, too, to know how potentially disruptive it might be to a bunch of folks. Steve, thank you very much. Steve Leesman. Pleasure. For C&BC News Update with Julia Borses. Julia.

Hey, Mike, a federal judge ordered the Saudi National accused of bugging an open AI conference room to be detained without bond. A court filing says he planted a recording device at the company's headquarters. It also claims he was in frequent communications via WhatsApp with a member of the Saudi royal family who tried to recruit him for a kingdom-backed AI project. Attorneys for the BBC subpoenaed former special counsel, Jack Smith, to testify in President Trump's $10 billion suit against the news organization. The president is accusing the British network of defamation over claims a BBC documentary about January 6th falsely implied he incited violence. Smith, who led a now-drop probe into the president's alleged attempts to overturn the 2020 election, will be asked about Trump's

state of mind on January 6th, 2021. And John Stewart will continue to host the Daily Show one day a week until the 2028 election. Comedy Central announced today he signed a new deal extending his contract for another two years. It was set to expire at the end of this year. Back over to you. Julia, thank you. Coming up, much more on today's market rebound. The NASDAQ got more than 1% today, but that means it's not gone anywhere in the last month. So which is more likely to come next? or a breakdown. Stay with it. Welcome back to closing bell overtime live from the NASDAQ market site. Tech stocks bouncing back from last week's selloff, both the NASDAQ composite and the NASDAQ

100, gaining 1%. The rebound spread throughout the AI trade, memory and software as well. The core weave, which has been at the center of the latest concerns, closed lower once again. Well, last week's moves revealed how exposed the market still is to the AI trade. Excuse me, but we did see a market rotation into so-called safety sectors. Is that a sign the markets can hang in, even if AI concerns linger. Joining us now is Adam Farster, Pia Schroeder's, head of multi-asset Americas. Adam, great to have you with us. Great to be here.

What do you make of the sell-off last week? And it sounds like you think that yields have this pull higher just because of what's going on around the world. So what are the ripple effects of that then? You know, I think we're seeing this competition for capital in the fixed-income markets now. So it's not just concern about inflation, although I think if we look at the inflation prints, you see there's a lot more pressure than people we're acknowledging just a month ago in the inflation prints. On the technology side, you're getting to the show me state, which is, are you going to actually deliver the earnings? We're hearing all of the talk about how much you actually need to return to justify this capital investment.

So people are getting to that point where they're starting to think about, how do I sort out the winners from the losers? What's off balance sheet? What do I need to worry about? So as the questioning process goes on, we have still, a, you know, S&P 500, that's like 40% the AI trade. And, you know, these ripples in the credit markets, things that are at least suggesting that some of these late cycle disturbances are kicking up. Is that something to be concerned about,

or is it just another gut check for the bull market? I think you're looking at a mid-cycle correction. I mean, famous last words, but we think there's still a lot of earnings power within technology and the broader market. It is interesting outside the MAG7, you're seeing the earnings really coming through, we're going to all be very attuned to what happens in this earnings season. We want to stay long equity markets in our portfolios.

However, you have to watch for these signs of deterioration. You could think it's like 2006. You know, it's the early warning shots that could go on for a very long time. But how do you put resilience into your portfolio when you have such a concentrated market is the point question? What are those signs of deterioration that you're looking for? Have we checked the box for? some of them already?

What we're looking for is, do you see a slowdown in any of the capital expenditures? I mean, I think that's the obvious point. When you look beyond that, though, is what do companies actually do with their spend? And I don't mean the hyperscalers. I mean companies like Schroeder's,

you know, we are spending a lot of money to put these AI tools into place. We see a lot of benefits from them. But will we actually get the returns before you actually get the compression on the inflation side of the future. So what are the ways that you would suggest, you know,

adding resilience to a portfolio if that's what needs to be done? Okay. So the first thing is we're seeing that barbell outside of the U.S. and global portfolios. You look at the performance of value stocks outside the U.S. It's been strong. Now, you have to be careful because the semiconductors appear in those value indexes. So you have some pretty concentrated returns in Korea and in Taiwan.

But we look at the cyclical exposures in your, where the manufacturing sector continues to be strong. We see some concerning signs about maybe a slowdown, but I think that's a mid-cycle slowdown in Europe. And then we look at the opportunities in like financials in Europe. That also looks very strong. You're getting to a point where yields are compensating you in portfolios. We're staying away from credit, but you don't need to be short the government bond market.

And then you think about diversification from the dollar, so things like gold that can give results. that can give resilience into the portfolio. In Europe, are you concerned at all about what's going on in France and the ripple effects that will happen to EU? Yeah, right now we haven't seen the ECB say they want to step in to support what's happening in France. We think ultimately that will happen. They're not going to let France completely fall off the side of the cliff.

But France has got to sort out their own fiscal house. And if you look at what's happened over the weekend, that makes the picture much harder for the politicians to sort out. And of course, we have the elections coming up, and people are concerned about what does a Le Pen presidency in France potentially mean. So if the ECB is waiting out there maybe to be some kind of a backstop, I mean, what does it mean in terms of actual monetary policy, Europe versus the U.S., which is looking to tighten a little more? Yeah. So we think the U.S. is going to continue to tighten. And in Europe, we think that you're very nearly at the end of that tightening cycle.

You don't see the domestically driven kind of inflation that we've seen in the U.S. So if you look at the U.S. inflation print, it's coming from services. That's very domestic. It's not just the energy side of the equation. Whereas in Europe, if you get a pullback in energy prices, it looks like they are in a much more balanced state, and the ECB doesn't need to hike as aggressively. Adam, we're going to leave it there. Good to see you. Thank you.

Thank you. Ahead is a group of AI agents working together are positive for getting work done, or could it pose a real threat to safety. We'll take a look at the next evolution in AI agent swarms next. with Boxio, Aaron Levy. And Amazon won, one of the big winners in the SB 500 today. Morgan Stanley naming it a top pick, raising its price target to 350 from 335, implying upside of roughly 40%.

The analyst there believes Amazon Web Services, growth, and profit runway is underappreciated. Closing bell over time. Be right back. Herce was a chatbot to talk to. Then it was an agent to help you. Now, it's agent swarms, groups of agents working together on large tasks. In a long post on X last night, Box CEO said, in part, that Agent Swarms are going to be a very big deal because they are not limited by some of the constraints we face now. He concluded by saying this changes a lot of fields as well as even the competitive dynamics in these fields.

Joining us now exclusively to explain his optimism about Agent Swarms, Box CEO, Aaron Levy. Aaron, great to have you with us. Hey, thanks for having me on. Appreciate it. Agent Swarms at the surface sound terrifying to me. Explain how they can be deployed and what sorts of problems you're thinking about. Yeah. So, first of all, obviously, the industry needed to come up with a different name. So, you know, that's our bad.

We're going to work on that one. But the best way to think about agent swarms is if you just think about, you know, some, you know, group of agents, it could be, you know, 5, 10, 50, 100. There's some examples of Open AI using 10,000 agents to work on some collective task where the agents are actually able to collaborate together. just as people would on whatever tasks they're trying to pursue or goal they're trying to pursue. So the recent examples have been some of these great math breakthroughs where you have lots of agents running around looking at research, you know, pushing the limits of what's possible

based on their level of intelligence and then coordinating in some central system where they can leave messages and coordinate together. Now, we're obviously seeing some of the risks of these agents forms. So this will be, you know, one of the ways where we see maybe cyber attacks begin to happen but on the very positive side, imagine any kind of problem that is constrained by our collective ability to go attack a set of information or be able to build tools or automate software to be able to make any kind of discovery or push some workflow forward. So this could be life sciences breakthroughs. This could be continuing to automate the software development lifecycle. This will obviously show up in areas like financial trading.

So it's going to be a very big deal in the productive use of AI. both in the medium and long run. So, Aaron, what has to happen for the infrastructure to get to a place where you could have just this background hum of all of these agents constantly working? I mean, at what cost, I guess, is some of the questions? Because even with things like Muse, the question is, what does the cost metadata to deliver on some of these relatively simple consumer applications? Yeah, so I think you're going to see it at the market obviously split in two ways. So on the consumer side, the subsidization will likely come from the actual agent vendor themselves. And they're able to subsidize this due to either advertising or transactions happening through their platform and commerce to some degree.

So I think the consumer side will be kind of a very simple equation about, you know, can you flow commercial activity through your platform? On the enterprise side, it'll eventually converge with the ultimate value that enterprises are getting from these agents. And that will be whether life science, companies can accelerate drug discovery. It will be whether financial services firms or trading firms are able to, you know, create better alpha in the market. It'll be, can cyber firms or companies that need to protect themselves from cyber attacks, can they use agents to better protect their organizations? And so that's ultimately where the ROI will manifest. And to your point about infrastructure, you do need a lot of build out of infrastructure to make this all possible.

Some of that comes down to the raw materials like the GPUs and the CPUs that agents need to power their tokens and the sandboxes they operate in. But there's a lot of software infrastructure that you will also need. So the ability to protect data that these agents are working with. That's obviously the role that Box plays in this ecosystem. So how do you make sure agents can access and work with unstructured data in a secure and well-governed way? This is going to mean that the systems of record that agents have access to will grow in importance because you will want to make. sure that those systems are governing the effectively creating the rules of the road for these agents swarms to operate in. So we think there's going to be a modernization of infrastructure

that enterprises will need to be able to run these agent swarms. And then obviously you're going to need the raw compute as well, all of which will continue to grow in the future. Embedded in this idea, Aaron, of the agent swarm, do these agents have a level of judgment? because if they're collaborating, they're deciding what, effectively, what to do next, is that judgment of some sort? Well, you know, all of this, it's, first of all, great question, and it's going to follow the general capability or intelligence curve of these models. So the reason why you're hearing this now be discussed,

and this wasn't really practical to talk about six or 12 months ago, is the models themselves were just not smart enough or capable enough to let you have long-running agents that were coordinating, with each other with a set of tools and in a shared space. That just was not something that the models were capable of doing. So as the capability level of AI continues to improve, as the coordination ability of AI improves, and to your point at the judgment or intelligence or intuition

or alignment to our values or our safety improves, then you can basically let the agents work for longer, longer periods of time or on wider and wider scoped projects. Now, humans will have to stay in the loop throughout the entire workflow that is being run. And so that might be regular check-ins. That might be looking at the final output of what was produced. That's obviously the work of validating what the agents are actually working on or setting up the prompts and the goals properly. So this does not mean that all of a sudden all of our systems become autonomous.

This is just the ability to now throw more compute at discrete problems inside of an organization, where we did not have that as a variable before in the economy. Three years ago, there were very few areas in a company where you could say, I will put compute in and get productivity out. That was a very sort of finite part of the economy where you do that, maybe in trading, maybe in media, a few other places. But now, for the first time ever, you can say we're going to put compute in to a drug discovery process. We're going to put compute into a manufacturing automation process, and we were going to get business value out. Again, with humans in the loop the entire time, that is not changed from any of this. This is just our ability to now have the agents be able to do more work inside of their tasks.

It's reassuring that they'll try to keep us in the loop. I appreciate that. Well, that part's up to us to some extent. But that's the goal. Let's try to make sure. Aaron Levy, thanks so much. Appreciate the time today. Industrials have been significantly underperforming the broader market over the last three months.

Up next, we'll discuss whether earnings season could turn things around with 3M set to report before the bell tomorrow. Closing bell overtime. the NASDAQ market site. We'll be right back. Industrials have lost some steam into earnings. The XLI industrial's ETF down nearly 7% over the past three months. With GE Aerospace, 3M, and RTX, all reporting tomorrow can earnings restart the industrial trade? With us now to discuss is Stephen Tusa, Wells Fargo managing director and senior analysts covering the electrical equipment and multi-industry sector. And Stephen, it's great to have you on. You say that essentially, you know, AI

is kind of the only horse driving things in terms of the overall industrial story and the growth. Where does that leave you in terms of figuring out potential winners and losers into earning season? Yeah, thanks for having me on, first of all, and it's great to be here at Wells. I think you talked about it going out of the break there, that the setup for this earning season has changed dramatically since, you know, I had my break about five months ago. performance has been pretty bad. There's now a lot more risk priced into these stocks, but there's a lot of trepidation around the concentration of growth, which no doubt is all about this data center, CAPEX. And that's why in the last couple of weeks, the risk reward into the quarter has changed with this data center related sell-off on a couple of those data

points that came out last week. I think the setup is pretty good. And you can definitely start to pick away at some of these things that have pulled back. Stocks like Avertive and Johns Controls, a couple of our picks on the upside, look a lot better today than they looked a couple weeks ago. Group is still relatively expensive, so the earnings are going to have to come through. But I think by and large, they will. And I like the fact that people are much more negative now than they were three months ago. At the same time, Stephen, I mean, when you talk about a name like Avertive, we also show

G. Bernova on that screen, too. I mean, these names are closely tied to this data center trade. So do we need that overall trade to pick up? So if we're looking at, for instance, concerns continuing to linger over the trade, over Corweave, over all of that, should we expect this particular subsector of your coverage to remain beleaguered? Yeah, look, I'm not a tech analyst. I think there are people that are better positioned to call the trading nuances day-to-day

on certain data points. I kind of look at the fundamentals on supply demand. Demand is very strong, and we are in very short supply. And so ultimately, the question is, you know, when is peak? We are mid-cycle, but when is peak? And I think right now the market is debating whether 27 is peak. And if you look at AI data center-related KAPX out there from the hyperscalers, you actually see a pretty slow rate of growth for 28, which is what people are debating.

I think that number is ultimately going to be a lot higher. And I don't think the market is going to discount, you know, a peak three years ahead of time. And so we're very comfortable buying this weakness, you know, despite what may happen, you know, day to day. So that's kind of the core of the call. And if KAPX continues to, you know, outperform to the upside, then the revenues for these companies are going to blow away numbers. And ultimately that matters because, frankly, I mean, Vertev 340 to 200 are done. downside cases now 175, our upside cases north of 300, that's a very different risk reward than when the stock is at 250. So we've chopped a lot of wood on this trade in the last couple of

weeks, which makes it set up a lot more attractive. And we think 28 is definitely going to come in much better than probably the street thinks right now on CAPEX. We did mention we are going to be waiting for 3M numbers. This is not a company particularly geared to the AI CAPX. So how does that set up into the results? Yeah, so 3M is one of the lower growth companies. Management has done a phenomenal job of taking out cost and being more productive. Margins are up very nicely. The debate right now is around growth. And, you know, literally something close to four on organic will be viewed positively and something below three and a half will be viewed negatively. They've got to show some acceleration exiting this year. And with their end markets like automotive, consumer electronics,

a little bit of housing here and there, a lot of consumer exposure. I really struggle to see an upside case on that organic. So we're a little more cautious on that front. And we'd rather buy beaten up stuff with a lot of growth than, you know, try and play that trade. All right. Fair enough. Yeah, a lot of those AI-connected industrials down 25, 30%. We'll see how that leaves them as we get this quarter underway.

Steve, thanks very much. Appreciate it again. and good to have you from your new home at Wells. Great beer. Thanks. Thanks for having me. All right. So we were saying, you know, to start things, pretty tentative recovery, only in the most hard hit parts of the market as we get into a little bit of a heavier earnings flow tomorrow. Yeah, definitely. We can do it for overtime today.

Fast money starts right after this.