Power Lunch - Power Lunch 12/23/25

Episode Date: December 23, 2025

CNBC’s Tyler Mathisen and Kelly Evans take you through the heart of the business day bringing you the latest developments and instant analysis on the stocks and stories driving the day’s agenda. “Power Lunch” delves into the economy, markets, politics, real estate, media, technology and more. The show sits at the intersection of power and money. “Power Lunch” gives viewers a full plate of CNBC’s award-winning business news coverage, plus a healthy dose of personality from the show’s anchors and the network’s top-notch roster of reporters and digital journalists. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Strongest economic growth in two years, bonds sell off stocks on pace for a record close. Welcome to Power Lunch alongside Kelly Evans. I'm Steve Leasman. In for Brian Sullivan. We're starting with a key read on the U.S. economy, and it's stronger than expected data shows the economy at its fastest pace in two years driven by a resilient consumer and business spending. That strength is changing the conversation on interest rates. What this growth means? for your money from stocks to rates to the Fed Outlook. We'll tackle all of that with David Zervos, chief market strategist at Jeffries.

He's sitting right over here with a great outfit on. Plus, a new era in the obesity drug revolution. US regulators have approved the first GLP1 weight loss pill, a tablet version of Novo's Blockbuster Wegovi franchise, marking a major shift in how these drugs will reach patients and will look at the names to watch in the space. Plus, the Trump administration is holding off on new Chinese semi-terrorism,

underscoring a shift for, from escalation to stabilization. The pause creates near-term clarity for chipmakers like InVIDIA, which is hitting its highest level, Steve, in more than a month, up nearly 3%. But let us begin with the markets and the economy. Today is stronger than expected GDP data, dimming the odds of a rate cut in January. According to CME's Fed Watch tool traders now only pricing in a 13% chance of a quarter point cut at the Fed's next meeting. On Monday, those odds stood at around 20% prediction market platform.

Cali showing similar numbers. Remember, jobs growth has been sluggish since April. Our next guest warned stronger growth coupled with little job creation could become a double-edged sword. Let's discuss the decoupling of jobs and growth and the impact you could have on the markets with our next guest, joining us for more. David Zervos, Chief Market Strategist at Jeffries. How'd I do reading the prompt? You did great. What do you think? I was really impressed. The reading thing with the words one after another, I'm good at that. You just put it together perfect. But here's the question. How good are you? you at productivity? I think we're all getting better at productivity.

Tell me about that. That's what the data is telling us. Tell me about that, David. It's telling us that America is more productive. That was today's data. By the way, that's really been the case, Steve and Kelly, since for the last almost three years. Go back and look at the strength of growth in 2023. It was 2.8%.

The unemployment rate rose that whole year by 310s from 3.5 to 3.8. Go to 24. Growth was 2.9%. The unemployment rate rose from. 3.8 to 4.1. So we've been watching this movie unfold of higher unemployment rate, very strong growth, near three, or possibly now, we're going into the fours. The president likes to talk about fives. We'll see whether we can get there. I think it's a story that is really not just AI. Well, it's not AI at all. The first part you talked about, right? I mean,

23, 24, that's before. I've got theories about why. 22 is when it launched, but it goes back to 2020 even. I think it's COVID. I really think it's COVID. That's where I'm at. The rolling five-year average is two and a half percent. Five-year average is now a big change from the late 20s. We just got really good at doing stuff with less people because we were forced to do it. Three things. Work from home.

Zoom. And the other thing is getting the right people into the right jobs that was allowed from work from home. I love this story, but let me just say here's a part I don't understand back because I hope this is true. We've got a lot of time, Kelly. I know. We're going to talk. We're going all the way. This is true.

You know, I love the rise of the room. But how can it be. that the era of the great hiring, what did we call it, when everyone was getting hired and getting paid crazy amounts of money, right? We're unwinding that now. How could that era have been more productive? That's what I don't understand. I think at the end of the day, well, first of all, we revised away a lot of those jobs that were created, we thought were created. So that's one. Well, people's raises were not revised away. Like some of the job hopping going on back then, you're like, this is crazy. Right. And at the bottom end of the income distribution was where you had

some of the biggest rises in 2021 and 2022, look, I think it's a confluence of many things. And one thing I will say is that looking ahead, and that's what we should all be doing, is looking ahead, not necessarily spending so much time on the past. The exciting part of this is that we are supercharging this productivity story. So if the productivity story was initially the kindling of COVID, the fire is now, I think, going to be deregulation, the tax changes that are coming in the one big, beautiful bill. all of the sort of productivity-enhancing storylines that are coming from this new administration.

That, to me, takes us to the next level. And that is what's sort of showing up in the latest data, and I think is really the story for 26. The question, Steve, and this is the most important question for the economists out there, is how quickly does that sort of excitement on the productivity side cause people not to hire?

And could you really have a three and a half to four percent growth rate with the unemployment rate at five, by the end of 2026. And I've been saying this for a number of months now. I don't know, it's not my base case, but increasingly, it's looking like that story has some possibility. I think the jobs are coming.

I think it's gonna come with a lag. I hope you're right. I worry about the creative destruction. We want the jobs to come. That's coming from the AI. We want the jobs to come, but the problem if they come is, it's going to slow, like,

it reminds you of Instagram when it first launched had like 12 employees or something. I mean, they must have how many thousands now? Doesn't AI feel like that to you? We're at the stage where you can do a lot with very little. But, like, it's coming. The amount of just content moderation that's going to be involved with some of these platforms,

it feels like there's going to be so much new job required to support all of this. Anyway, I'm just curious what you're doing. Maybe. I don't know, Steve probably has a stronger view on it than I do. I just, I see what I'm hearing from C-Suite's and corporate CEO and leaders. And they're saying the same thing over and over again, which is you can do this without hiring more people. that you have these tools, go use them.

And I think that's what's getting said in our industry, in the finance industry. I will settle the debate and tell you exactly what happens. Good. Let's go. What happens is what you said initially and what Kelly said later. Yeah. It's the J-curve. You come down in jobs and then you come. What happens?

Let's stop hiring, okay? Let me look at what's going on in my company. Wait, this guy's doing zero growth, but this guy's doing 10% growth. I need to hire more of this guy and fewer of that guy. And I'm not pointing at Kelly for any particular reason. You're very productive and a wonderful person here. So I don't mean that specifically. But that's the way it works.

You wait to see who is having the added productivity. Who can do it? How they can do it? I'm also a follower of SAES law. Yeah. Which tells me that the amount of demand for labor is equal to the supply for labor. So we will get there, but there could be some disruptions. But David, I want to turn to you on the Fed on a couple different.

fronts here. Yeah. Let's start off with how all of this factors in, looking ahead, as you suggest, to Fed policy. There's a big debate. Do you take a flyer on the common productivity? Do you take a flyer on the supply side? President Trump today saying, we should be celebrating good news, not mourning it because of the inflation. So that says he wants everybody to buy stocks at all times. He wants to do 20% growth in America, but we'll put that to the side for the moment, that notwithstanding, the idea that this productivity boom, which is inherent in both the data we've seen in the past, inherent today's data, suggests that the Fed can run the economy a little bit hotter, and then when hot growth numbers come out, I should not fear for inflation.

That is correct. That is the correct answer. It is the Allen Greenspan 1990s debate that he was having with the staff and the street, and he basically said, stop telling me the rates, Stop telling me to push this economy down because of risks of strong growth leading to inflation. And he won that debate. And he was absolutely willing. I wonder if the guys in the back and get my picture that I had. It ended with a bubble. Okay.

It was a pretty fun bubble. I mean, it was a heck of a one. And it was great without acknowledging how that. Bubbles are always fun until they're not. We might, you know, it's, you know. We're clearly going to repeat that. Let's put it that way. Guys, listen, no, no, no.

I'll take 96 to 2000 all day long for the next four years. I will settle the debate again. Okay. by telling you that you were right short-term and Kelly of right long-term. What happened was he let it ride, he let rates stay the same. By the way, at a higher level, both the higher real and nominal level for rates. For sure. Inflation plunge, he was right about that. However, in 1999, as Rich Claret had just wrote in a blog that I wrote,

read about and wrote about, he had to start raising rates. And you had a return of inflation, which was interesting in that it was the backside of the productivity boom that saw that. So I guess the idea is, look, the Fed's job, in my opinion, is to allow the punch to be served, but be there to take it away before stuff gets out of hand, before people start dancing on the tables. I wonder if they're even going to do that, though, because they might look and say that six-month, relatively shallow 2001 recession, was it that big of a deal if the trade-off was that you got six years? I think it was even revised away. What was revised away?

The two consecutive quarters of negative growth in 2001. It was shallow if it, yeah. Here's a fun fact about the 2001 recession. If you go back and read the NBR press release, which I'm probably one of only seven people in the whole world who know this, there was not a recession until 9-11. Yeah.

It says that we were not going to say it was a recession until 9-11 came along. So it wasn't necessarily the busting of the bubble. That created the recession. We actually came through that. I think it ended in October. But why? The spending boom.

The important thing is the why we came through it because it was an equity recession, not debt and leverage. Next question on the Fed, David. Who's your favorite guy? Who's going to get it? What is everybody the same thing? What does it mean?

Everybody put at the kids' table, a little shorter, and you ask them. We're serving you, Franks, and chicken tenders. What is this little table? Just shut up and eat your chicken tenders and tell me who's going to be the Fed chair. What's really at stake, and I like Steve's point from yesterday, which is, could the president be persuaded that a pick like Waller, which potentially might be less comfortable to him, would have more gravitas at the Fed to actually implement the policies the president wants? I think Steve made a good point yesterday or the day before when he said that. I think all of these candidates come from the supply side. They're all going to be thinking about this productivity story.

They're all going to be thinking about accommodating the initial phases of job destruction that might come from the AI productivity gains. And I think the market will applaud that. And that is a positive. All of them. All of them are going to be thinking like that. So there's no sort of traditional Keynesian storyline that's going to make its way through this Fed.

post-May with a chair of either. I know we have to pivot, but you're positing a Fed chair who doesn't follow the Phillips curve and does not see the trade-off between unemployment and inflation. Not in the middle of a positive supply shock, Steve. Not in the middle of a supply shock, but otherwise, yes. I know getting you screamed that. We've got to run it. We've got to move on. I'm sorry. I think that's right. And I think all of them have that, have that capability, that belief, so I think we're in a good place. I think we're in a good place with all of them. Quickly then, lastly, before we go, the tax refunds. are coming. Chris Kruger just said, watch out, Obamacare could take, you know, some wind out

of the sales, if those premiums go up. Do you have a Q1 expectation for how that all shakes out? I think on the demand side of the equation, you've got a lot of these countervailing forces, and they're not strong enough to be the dominant force. The dog, geez, sorry about that. The dominant force is the president, maybe. Time for an interview. The dominant force is the supply side, Kelly. I mean, that is what we're seeing. This productivity, story just dwarfs, I think, the tax-free funds, the sort of Obamacare ACA... The affordability crisis, it's all going to...

I think so. I worry more about the sentiment in the job market. That's what we're easy. What about the inflation in today's report? No. The last inflation report was incredible, and people have tried to dissuade us from thinking that that was incredible. It was incredible, and I do believe the rent story is for real.

It is coming down. It's coming down dramatically. the immigration behind that, that's that immigration storyline behind that is real, as the Secretary of the Treasury has been saying. I think we are not going to be debating some surge in inflation in 2026. That's not the story. We might be in the Greenspan world of debating that inflation is only too quickly. Well, if I said anything that makes any sense, I must have misspoke. So go ahead. And we appreciate you leaving it on that clear point as well. David, thanks, as always. Great to see you. Great to see you guys. David's your most. Let's get the

bond report. Rick Santelli has more. on how the markets are reacting to all of these cross currents. Rick, what can you tell us? Well, you know, I like the way the charts tell a picture. If you look at twos and tens when the number was released, a couple things should jump out at you. The twos definitely seem to be more aggressive in holding the upside. I don't think that's for any big reasons to explain the Fed

or acknowledge percentages and probabilities. I think it's purely the next chart, which is 10 years bucking up against that 419 to 420. resistance. That chart goes back nearly two weeks and stopping it there, it makes perfect sense. And if you open the chart up, should we settle above 419 and it doesn't look like we will, that'd be a three and a half month high yield close. To me, a bit of the surprise today was the dollar index. When the data was released at 8.30, it definitely turned up. It had been moving down, but it really doesn't alter the general picture. Any close today below 9814 is a two and a half

month low. And to answer your question more directly, Kelly, I personally still don't see any path to get to 2 percent so many could disinflation, especially those that were potential picks for Fed chairs. But ultimately, just like anything else in our economy, when the economy gets good, competition for capital gets hot. And that's why usually, historically, a good economy may have higher rates, specifically on the long end. It isn't necessarily a bad thing. I think the big negative here is the fact that we have one report with hot inflation. We have other reports with cool inflation. I need to get a better GPS. And in terms of consumer confidence, I would take a very big shovel and shovel them all right into the garbage

dump because Main Street media has polluted the well and this administration doesn't get a lot of credit that the markets give it. And I think that's what the issue is driving negative confidence back to you how much concern was there down in the pitch there when it came to the issue of the higher inflation numbers versus the stronger growth numbers well i think there definitely was a little nervousness and you know but it didn't really change much because what we see down here is it's the june meeting that seems to be the only meeting many of my sources think may be in play for an actual quarter point cut rick we really appreciate it thank you so much sir 417 on the tenure this afternoon.

Coming up, the weight loss drug wars are heating up. Novo shares are surging after the FDA approved their pill. Where does that leave Eli Lilly? We'll discuss next. Welcome back. Shares of the Uygovi maker, Novo Nordus, are getting a huge boost as the FDA approves their weight loss pill. It's the first drug maker to win approval for a GLP1 oral truce. treatment for obesity. Evan Siegerman is head of health care research at BMO. He's only got a

market perform on Novo. He says this is a big win, but sees a challenge road ahead. He's here on set with us. First of all, welcome. Thank you for having me. It's great to be here despite the weather. Yes. And as you say, despite this having been a hoddle, hold on for dear life year for health care broadly, but a nice finish, no? I mean, go back to the summer when you had the health insurers, you know, collapsing and all this stuff going on. Just kind of give us a landscape coming into this Novo announcement of what you think is happening with the space? For sure. So just a bit overview on the year, kind of weird optimism into the Trump presidency. Then we had Liberation Day. Everything fell off of a cliff. Turmoyal at FDA.

And then in the past six months, like the back half of the year has been one of the best segments for BioFarm, I think, in a decade. Are you serious in terms of returns? Yeah, I think it's the best. We looked at it. I think it's the best in a decade. So strength there, people are kind of getting back to the sector. Because we were told that vaccine uncertainty was going to mean no drug, you know, was going to hurt the development pipeline or just approval uncertainty writ large. But you look at these names and a lot of them are up. People feel optimistic. But going back to Novo. So obviously we're in the obesity war with Lily and Novo. I have an outperformer on Lily. I think I called it the goat of obesity last year on this program or CNBC.

And, you know, I downwind to Novo shares back in April because they just had lost their way. So they are making. good progress. Does this change your mind? Because the CEO told our reporter, Anika, earlier today, that, and I think the market has said this as well, that this one performs better, it's superior, it's a better product. It is important to the story, but if you talk to Nova, they're worried about kind of revenue growth next year. I think when you compare the Wagovi pill, as it's now called, to Lillies-orferglypron, you have more weight loss with the Wagovi pill. It needs a lot of API. It's like 25 minutes. What's API? Great question. It's the active drug in the pill. I see. So it's like,

25 milligrams, and the high dose of the shot is 2.4 milligrams. So it's 70 times more. When it comes to Lily, I like Orphaglipron. You don't have what's called the food effect. That's a big deal. So when you take the Wagovi pill, you have to take it first thing in the morning and wait a half hour before you can do that. You do that every day. The Orphaglipron, Lily's pill, doesn't have that issue. Regardless, yes. The fact that this new pill is so highly concentrated, as you described, Do you think that presents side effect risks? I mean, there's always side effect risks with GLP-1s, but I think that it's pretty well known if you tie trade up.

And I think when it comes to, you know, Novo, they are making progress, right? I'm not hating on the stock. I see some issues. There is a clear valuation disconnect, and folks constantly ask me, do I get in now? I want to see them make more fundamental progress when it comes to kind of revamping their strategy. Before you get to the 0.24 versus the 2.4,

versus the 25, there's a bigger difference. The shot versus the pill. Sure. How big a deal do you think that will be for average Americans? How many people don't want to do the drug because it means a shot versus are happy and are happy taking a pill? So if you asked me two years ago, I would say no one wants a shot. Now the needle's like this big. I don't think most people care.

It's amazing. And they like the convenience. It's once a week. They do it on a Sunday. Then they forget about it. So there is definitely a subset of patients who don't want a shot, needlephobic, get sick of it. But you talk to a lot of folks who are on Wagovi or Zepound, and they don't mind it. Where are we on the curve of the adoption of this drug?

How much has happened and how much more is there to go? And how much of what there is to go is already priced in. Oh, that's a lot of questions there. So the last one is the most important part. Yeah, exactly. I think when it comes to Lilly, there's a lot priced in this huge $100 billion plus mark. In terms of where we are, I mean, there's probably five, six million patients who are on a GLP1 for obesity in the United States. We look at script data.

And how many? Am I 140 million are obese or something like that, right? So we're not that deeply penetrated. So there is still room, and that's why, you know, there's still room for Novo to win, for Lilly to win. And I think what's going to be very interesting is what happens when you have longer actings. Other players like Pfizer coming into the market. So there's still penetration. So if I could just understand what you're saying, the market will not price in the expected

or even reasonable growth because it's not sure which medication will capture the bulk of the growth. Well, that's what, to a point, I think on the Lilly side, I think you have Orphaglipron, Trezepitide, Retachritide, a lot of different products that. Names that you didn't know three years ago. I knew a few of them. You know, no, it's fine. It's that, you know, they're targeting all segments of patients from lower efficacy to higher efficacy.

Novo is trying to figure out their footing. I think will also be very interesting is what happens in Medicare. That's very important. There was an announcement back earlier in November opening up access in Medicare, and that's 30 million patients potentially that are going to have unfettered access. But we're still, in other words, in the growth and adoption phase of this, because the next place my mind goes is when we had Gilead and Hep C and all these treatments back in the day where they did so well they became their...

own overhang for the stock. You were never going to be able to comp them. You were never going to be able to sort of come out of those growth rates. So the Gilead Hep C was unique because it was eight weeks of treatment and it was curable. And I don't think that they really forecasted kind of the demand properly. And then Abby came in an undercut pricing. So I guess the other analogy would be to the cholesterol drugs where everybody takes them. And as a result, they cost $5 and, you know, there's not a lot of profit. So that's a great point. And I think when it comes to Lily and Novo, their prowess is manufacturing. So in 2030, when you might have a generic, they're going to be able to make it cheaper and at a higher volume than a generics player and still

able to compete. And they went on volume versus price. We're seeing that now. And we have till 2030, I guess, for that generic, which tells you how early we are still. We are still there. It's 2026 next year, though. Evan, thanks. Thank you guys. Appreciate it. Evan Siegerman of B. Moe. Coming up, Lulu Lemon on pace for its worst year in more than a decade. Our guest sees a comeback story, though, we'll ask him why in Market Navigator. Time now for today's Market Navigator. And let's tackle Lulu Lemon. What a rocky year it's been akin to the great financial crisis, honestly.

The shares are down 45% year to date, which means it's on track for its worst year since 2008. But Lulu is trying for a year-end turnaround. Shares are in the green for the fifth straight week. Can this momentum continue? Let's see what our next guest thinks. Tony Zhang is Chief Strategist at Options Play. Tony, what do you see? What trades are you making here?

Yeah, so I think the best way to play here for Lulu Lemon is to go out to January and sell a 210 at the money put option. Earlier today, you can collect a little over $7 for that at the money put option expiring in around three weeks time. which is about over 3.5% of the stock's value. And if that gets put to you by that January expiration, you're looking at being able to acquire these shares at around 202 and a half or so, which is around, you know, a little over 15 times forward earnings, which I think is a really compelling valuation when you consider the fact that, you know, the stock is starting to see

some growth internationally. You know, U.S. sales are still flat. That's part of why the stock's down 45% for the year, but that international expansion, China's up 46% year over year, and they're very, and they're in the early innings of that international expansion. And that's really what I think is quite compelling about this and, you know, why the, why the management team has continued to authorize a billion dollars of share buybacks, and we're starting to see that stabilize the stock. That would be fascinating if China experiences an athleisure boom, kind of like the one that we went through and that has become old news now. We had on the women's side, then the men's side. And now you say maybe China becomes their next growth lever.

Anything else about the stock? I mean, you mentioned a trade that's pretty short term. Do you have to just kind of wait and see on this one, how they execute in the early part of the year? So we just saw the earnings announcement come out. And that's really why we're starting to see that confidence in China, in Tier 1, Tier 2 cities in China, seeing that sales growth. But, you know, my view here is long term. I think if you look at the chart here, it's been range bound between 160 and 190 for that. the better part of four or five months now. We just broke out above that level here. My initial upside target here is 250. So when I'm looking at selling this put option, I'm looking at this as a long-term investment. The put option is just my vehicle to try to acquire these shares at a lower price. Exactly. Try to get that lower price. And then there you can see the upside as it

heads if you're right towards 225, 250 or beyond. Tony, thanks so much. We appreciate it today. Thank you so much. Over to you. Okay, Lulu isn't the only name looking about. back next. We'll reveal this mystery chart despite tumbling this year. It's one of our next guest's top pick for 2026. Our mystery chart before the break was service now. Shares are trading lower today after the company announced it's acquiring cybersecurity startup Armis in a nearly $8 billion cash. that's expected to close next year. Service now hopes the deal is going to bolster its cybersecurity capabilities in the age of artificial intelligence and more than triple its market opportunity for security and risk

solutions. CEO Bill McDermott, he joins Squawk on the street this one and to lay out the strategy behind this multi-million billion-dollar deal. This is about making a strategic move to accelerate growth and we see the opportunity for our customers because in this AI world, especially with the agents, you're going to need to protect these enterprises. Every intrusion is a multi-million dollar problem, and we're here to solve it for our customers. Service Now investors certainly hoping this deal could turn around.

The stock's recent underperformance. This shares are falling almost 18% in just the past few months. And sticking with Service Now, let's get a power check on the name. Our next guest is bullish on the stock heading into next year and thinks now is the time to buy into the enterprise space. Let's welcome Eric Clark, the chief investment officer at AcuVest Global Advisors. Eric, it's great to see you. And what does Service Now signal to you is going on more broadly here? Well, I mean, 2025 was a difficult time for software in general.

You know, the narrative was AI is going to disrupt every software company, sell them all. And that's, you know, we've always thought that was a bit of a lazy approach. You know, it's a little bit more nuanced. You have to find the businesses that are going to benefit. And Service Now is right at the top. We absolutely love the name. Certainly we love it as a dislocated kind of misunderstood growth stock down here for sure. Right.

And I know Steve's looking at your guitars there in the background. I got two strats in a L.S. Paul, but I didn't see much more than that right there. I couldn't see more. Service, I'm just wondering, is there anything that you think you'd tell people about this stock or some of the others that might have been left behind and you think can play catch-up now? Well, I think if you take a step back and you say, listen, And every company we know wants to deploy AI throughout the enterprise where they can. And so your choice is you can do it yourself one by one.

You can do it with a Band-Aid strategy or you can find a partner that's going to allow you to do it fully across the enterprise. And there's a lot of data points that show the companies that get the most amount of benefit from AI deployment are the ones that do it on a comprehensive basis. And so in order to do that, a service now gets really competitive. because of this AI control tower, where I call it the one door. You know, you can go in one door with service now and be connected to every agent, every model, every data source, every hyper-scaler across the board. And so they're going to help you integrate AI across the enterprise because it isn't easy to get all of those benefits that takes time and money. So most companies will need a partner and that's service now. So I love the acquisitions. And you have to give Bill the benefit

the doubt. I mean, since he was the CEO in 2019, the stocks outperformed the S&P by about 700 basis points annualized. He's earned the right with his team to bet that they're going to do the right thing for customers and the stock will follow. So it's just now you get it on sale down 26% year to date, which is a very rare thing. Can you back up for me? Maybe Kelly does this every day. I don't really quite know. What are this particular issues that AI brings? to cybersecurity for companies that and who other along with service who are the leaders in addressing that well i mean service now is is pretty unique because they're they're one of only a few companies that allow other companies to connect to so many other data sources you know that the

industry is so fragmented um and new solutions come on board but but you need something that connects to everything and speaks to everything in order to broadly deploy the AI. So obviously cybersecurity is a big part of that, but it's just one part of the enterprise, the tech stack, if you will. And when you have something that speaks to other parts of the tech stack, you know, they talk about east to west in the business and then north-south on the tech stack. That's a very unique solution. And it's pretty complex, which makes service now really. valuable. All right. Now, turning to the e-commerce sector, your pick for next year, Amazon? I mean, that's, you know, kind of boring. What am I missing there? Well, it is boring and it's

underperform. But, you know, Amazon historically has underperformed when they're in their big TAP-X spend cycles. But, you know, the last report was a great, was a great testament. The retail business is firing on all cylinders. Margins are still expanding at scale. That's pretty impressive for a four or $500 billion business segment. There's a lot of operating leverage there when margins can expand still. And AWS started to reaccelerate too. So I'm not sure why the stock hasn't moved too much. It's kind of been flattished year to date. But historically, when the business is doing well and the stock market doesn't care, that's usually the opportunity for some good catch-up opportunities. And I would put Mercado Libre, which is kind of the Amazon of Latin

America in the same boat. It's pulled back and there's a great opportunity there because e-commerce is not going away. It's just being, it's a bigger and bigger part of everybody's lives because it just does so much for us on a very low friction perspective. Okay, Eric, thanks. Just move your head a little bit to the left so I can see the guitar coming out of your head there. I can't quite see what that is. I wish I could tell you that were all mine. I'm in my buddy's office. I don't play anything. It looks like an SG, but I don't think it is. Anyway, thank you very much, Eric Clark. Let's get over to McKenzie Zagalas for a CNBC news update.

Hey, Steve. The U.S. reportedly moved a large number of special operations aircraft and multiple cargo planes filled with troops and equipment into the Caribbean this week. The Wall Street Journal says U.S. officials and flight tracking data confirm the escalation. It comes as the U.S. ramps up military pressure on Venezuelan President Nicholas Maduro and as U.S. has begun seizing Venezuelan oil tankers in the region. The Trump administration says it will start to garnish the wages of student loan borrowers who are in default in the new year. According to the Education Department, beginning on January 7th, roughly 1,000 borrowers will receive notices informing them of their status.

The number of notices will then increase on a monthly basis. And the Justice Department is suing the state of Illinois over new state laws that aim to protect immigrants. The laws bar immigration officials from making arrests at and around courthouses and require hospitals, daycares, and universities to have procedures in place to handle civil immigration operations. The Justice Department argues the laws threaten the safety of federal officers. Steve, sending it back to you. Thanks very much. Thanks, McKenzie.

Up next, a, sorry, excuse me, oh, there it is, A-Dash-I. Just so you didn't think it was A or A-E. AI is entering a new phase models are becoming commodities as competition heats up. So who is pulling ahead that answer next? Welcome back. Welcome back. The AI industry is shifting gears looking to cash in on its models by rolling out more consumer-friendly products. Steve Kovac has the story. So, Steve, what does consumer-friendly look like or feel like? You're a perplexity guy, right? Yes, I am. You talk, we talk about this

in the newsroom all the time. We do. That is the perfect example, because the large language models that perplexity uses aren't made by perplexity. It's made by open AI and anthropic. And this is kind of what we've been seeing others take on, including soon Apple, this idea that, okay, we don't have the resources or the money or the desire to build these LLMs ourselves. So we're just going to build on top of it and kind of choose off the shelf. So why am I paying perplexity then? You're playing perplexity, who in turn is paying the model makers. What are they giving me that I wouldn't do if I went to open?

Well, they built a really great app, a really great search. It's the interface. It's the, exactly. It's the fact. 20 bucks a month? That's what I pay for chat. Which is the same thing for chat chabit. But you've clearly thought that they have a better product.

The app is better experience. I upload spreadsheets in there, and I say calculate regressors and correlations for me, and I couldn't be happier. There you go. And so you're happy with that. And that's exactly what Apple wants to do, by the way, next year with Siri. They tried it themselves. We're going to build our own large language model. We're going to make this happen.

Didn't work. They couldn't get it off the ground. And now, we've gotten so many hints, including my conversations with Tim Cook over the last few months, that they're just going to go out there and partner with someone and make it happen themselves. Real quickly on that, which means. maybe is for or against the commoditization argument, I'm not sure. People ask, you know, is this a good buy for Apple here? Not asking to answer that, Steve, but are they really bringing LLMs to Siri next year? And if so, could that? They say they are. That would be huge.

And so then the question becomes, where is the monetization? Okay, is this a goodbye or whatever? They've said they're giving this away for free. Apple intelligence is a free thing. You buy the phone. It comes with the phone. It works only on the newer devices. So therefore, they're going to hope you have to go ahead. you need a 15 pro or better in order to use this system. So their hope is, okay, I have a 14 or younger. I'm going to go out there and buy one of these new phones in order to get Apple intelligence. Or, and let's take it a step further.

I'm going to talk about this a lot next week. Devices, because this new artificial intelligence part of Siri is going to enable them to do glasses. It's going to enable them to add more to AirPods. It's going to enable them to do smart home stuff, all sorts of other things. But they have to get this right first. Steve, you're talking about commoditization, but I feel like this story could easily be labeled differentiation. That too.

But on the product layer, not the LLM layer. Let me finish that. Let me finish that thought, which is that are we, does perplexity have a chat GPT? Do they have a unique advantage or value add at the front end that others can't come in? Why wouldn't chat GPT the makers of the LLM say, you know what? The hell with those guys, I'm both going to do the front end and the back end. And chat chvety does that. That's why chat chvety is.

that's why chat chabit is chat chabit they make both the user interface right and the large language model they all do in certain ways but the idea here is that other people are going to be able to take it off the shelf like your last guest was talking about with service now same concept they're going to do it for the enterprise service now isn't out there building their own lLMs nor is sales force and so many of these other companies that talk about their AI products a lot of time they're using existing models and and figure out which is the best one i talked to judson outthoff He's the CEO of Commercial over there at Microsoft, and he told me their customers are asking them for more choice when it comes to what models they use. Obviously, because of their relationship with the opening AI, they're all in on opening AI.

Well, he said, we're looking at the quote right now. Our customers are demanding model diversity. They want to be able to do other things on our products. So co-pilot is already putting some Anthropic in there. I expect to see more of that next year. Same thing, again, going back to Siri, using Google and maybe some others. We got to go, but here's my question. Are we the point, what's, what's, what's, when you say commoditization, what's the oil here?

Is the knowledge, the oil, the facts, the oil? Oh, or the weights of the, yeah. Is that what's corn? And what is the Gucci bag, so they're not commodity? The knowledge base is the same, right? The knowledge base is the entire internet. It's every book. It's how that is synthesized, and there's so-called weights on each model that kind of measure how they respond to you.

That's the secret sauce. But again, it's getting harder and harder to distinguish. If you use Gemini or if you use perplex, if you use any of these, it's hard to distinguish what am I using. The LOM is the oil, and that is the commodity. And the front end is really the- And the user experiences is the thing that's you. And that's what we're going to see more of next year, I think.

Especially Apple's going to be the best example of that. Siri built into everyone's iPhone, being able to talk to it like that. Not powered by Apple technology necessarily. Powered by Google, maybe Anthropic. A little open AI is already integrated there. The future of AI in three minutes with Steve Kovac. That will be the launch to watch. He's better than AI.

March is, spring is going to be huge for Apple. Not just turning 50, but this big AI launch. 50. Or A. Dash I, as we say. Aye. That's the way it looks, A-Dash-I. Thank you.

Thanks, guys. Coming up, the U.S. accusing China of unfair trade practices in the chip industry, but the government is holding off on new tariffs, at least for now. We'll explain why next. The U.S. is planning to raise tariffs on Chinese chip imports, but not right away. Semiconductors from China will face a 0% tariff for the next 18 months, with higher duties set to kick in starting in mid-2020. Let's bring in Christina Partsenevolous to explain. Christina. Well, it sounds like a big olive branch to Beijing, Kelly, but 18 months of zero additional duties while the two countries try to keep

their trade truce intact, but if you dig into the actual numbers, this isn't the concession appears to be. China accounts for less than 3% of U.S. semiconductor imports. You can see to the far right of your screen, about $642 million worth, according to U.S. import data. And Biden's 50% tariff on Chinese chips, which kicked in on January 1st, actually stays in place. So there's still a 50% tariff. So President Trump is essentially delaying additional tariffs on a sliver of the market that's already heavily taxed. Why do I say that? The delay does give President Trump flexibility. He's preserving the option to impose tariffs if the China truce falls apart without having to take any action right now. But what's actually covered here

are legacy chips, the older technology in cars, appliances, telecom equipment, not necessarily all of the GPUs we talk about often. But most semiconductors don't even enter the United States as stand-alone chips. They're already inside your phone, your laptop, your car. So today's announcement doesn't necessarily touch those finished products. We are still waiting for the results of an investigation of the semiconductor industry launched in April, Section 232. If the government finds chip imports impair U.S. national security, that could actually reshape the semiconductor supply chain. Potential tariffs on chips from all countries possibly extending to finished electronics. President Trump, we know months ago, floated the idea of 100% tariffs in August,

but eight months later, we still have no details. Officials privately say those tariffs might not come anytime soon as well, according to Reuters. Meanwhile, U.S. companies still have to plan around major uncertainty in what could be far more disruptive than anything announced today. So, Kelly, we're at the status quo, essentially. Steve? Christina, does this help at all people who are planning and figuring out how much their expenses will be next year in terms of chip importation?

No, because there's been so. many waivers and extensions provided for a lot of these companies thus far. So it's on a case-by-case basis. So it's very difficult. I'm on a rampage, not against tariffs, which I don't like. Against uncertainty. Right, but my rampage is against uncertainty. And I believe the Trump administration should really spend time to figure this out.

People are trying to plan. It's like, tell me what the rules are. And I think Christina's report makes clear that that's, we're not getting any closer. Tax code makes my head hurt these days. I think my dad had a good point. Like, it's, didn't Reagan try to simplify everything? Just like, let's get, and now it's like back to this. We're going to have Scott Lindsaycombe on Squawk Box on Friday, on Friday morning.

And he wrote the definitive piece about tariff complexity hell. If you see these graphics, Kelly, you will not, you go from headache to migraine. And this is what small business is dealing with trying to navigate the tariff tax code right now. I do sympathize. We'll have much more on Power Lunch right after this. Welcome back. Kelly and I sitting here puzzling over the GDP numbers today. Don't want to take anything away from strong growth. But when you dig inside it, you find this health care services factor is on the rise. There's the change in GDP. And we do know that

health care services was a big part. It was in the quarterly services survey. Right. And that jumped up. And that's a part of why the unexpected part of why consumer spending was higher. Well, right. You get a 4% GDP print, you think. Wow. And then you look within that, and it was three and a half percent was the annualized pace of consumer spending. It sounds great. But within that, there was this huge jump in health care spending, which I don't think anyone feels. There's another piece of this I haven't had a chance to talk about, which is incomes were flat, spending was up, savings rate went down. There's two things that could be happening. People who are really confident spend from their savings and people

who are really stretched spent from their savings. Yes. It could be one, the other, or depending upon ink or growth, it could be both. And I like, appreciate that all the economists looking at this say, like, we're not sure which way this is all going to be resolved. We have to wait to find out. It will flatter fourth quarter GDP, but we have to go now. Steve, it's been a pleasure. Thank you so much. I'll see you again tomorrow. And thanks for watching Power Lunch. Closing Bell starts right now.