Alphabet to Blow Past Investor Expectations for AI Spending

0:00They told us to expect change. They warned us about the transition. But honestly, they forgot the best part. This is the chapter where we finally focus on us. LifeMD delivers expert menopause and midlife care right from your home. From hormone health to holistic wellness, LifeMD helps you feel your best for the best years of your life. LifeMD, it's just getting good. Visit LifeMD.com slash goodlife. The thing about AI for business, it may not automatically fit the way your business works. At IBM, we've seen this firsthand. But by embedding AI across HR, IT, and procurement processes, we've reduced costs by millions, slash repetitive tasks, and freed thousands of hours for strategic work.

0:50Now we're helping companies get smarter by putting AI where it actually pays off, deep in the work that moves the business. Let's create smarter business, IBM. Being a small business owner isn't just a career, it's a calling. Chase for Business knows how much heart and effort go into building something of your own. Manage all your business finances, from banking to payments to credit cards, all in one place with Chase's digital tools. Plus, access online resources designed to help your business thrive. Learn more at chase.com slash business. Chase for Business. Make more of what's yours. The Chase mobile app is available for select mobile devices.

1:27Message and data rates may apply. JPMorgan Chase Bank N.A. Member FDIC. Copyright 2026. JPMorgan Chase and Company. Bloomberg Audio Studios. Podcasts. Radio. News. This is Bloomberg Business Week Daily. Reporting from the magazine that helps global leaders stay ahead. With insight on the people, companies and trends shaping today's complex economy. plus global business, finance, and tech news as it happens. The Bloomberg Businessweek Daily Podcast with Carol Masser and Tim Stenebeck on Bloomberg Radio. Alphabet shares, which were down 1-2%, now up about 1.6%. No, down as much as 7.5%. All right, right, at its low.

2:15Sorry, that was the last trade. But now they're up actually more than 2 % here. So quite a swing back to the upside here. Okay, the company said it will spend$175 billion,$285 billion this year. That's the headline compared with the$119 billion that analysts expected. Fourth quarter sales for Alphabet, excluding those traffic acquisition costs, those partner payouts,$97.23 billion. It surpassed$95.2 billion expected on average by analysts. I want to bring in Bloomberg Intelligence, Senior Global Head of Technology. He's going through the numbers as we speak. I saw him running into the studio. He's here in our Bloomberg Interactive Brokers studio.

2:53We have so many questions for you. First, I just want to start with the CapEx. This number is so much bigger than analysts thought they would see. We're talking$175 billion to$185 billion in 2026. $119.5 billion was expected. Why? And when you think about the year-over-year growth rate for 2026, we're talking about almost a doubling of the CapEx. They were around$95 billion for 2025. So at the high end, we are talking about doubling. Look. That's huge. We thought 2025 was a big year. No. And after the meta print, it became obvious that these companies are going to step up in terms of their CapEx spend.

3:34But as you've reminded, this is different than meta because Google can use this CapEx to then sell to other companies for Google Cloud. Which actually saw a pretty nice sequential acceleration. Meta Platforms doesn't do that. Meta Platforms doesn't do that, but these companies also consume a lot of the GPU and the AI infrastructure capacity in-house. And look, they gave a token count number again, which I always look for. The 10 billion tokens per minute? 10 billion for their APIs. So that just goes to show that trained LLMs are being used in the real world. And in the case of Meta, yes, there is a higher uncertainty around how their GPUs are being used.

4:19But in the case of Alphabet, you see it in the print in terms of Gemini MAUs going up to 750 million, the API token count going up. So cloud revenues accelerating, search growing 17%. That's all AI overviews. Only thing that disappointed was YouTube. And I just don't know why. So that wasn't a disappointment. There was. And look, Meta just came up with a quarter where they grew 25%. And they raised the guy to 30 % next quarter. So from that perspective, YouTube was light. And I don't know the reason why. Okay. So that's certainly a question that's got to be pushed on the call. Going back to search, I mean, they talked about exceeding$400 billion for the first time, annual revenues in total.

5:02And you talked about some of the other usage points. So the payoff, the ROI, we're clearly seeing when it comes to Alphabet. that? I mean, for a business that's a$225 billion run rate to grow 17 % at an incremental margins of 70 % or 80%. I mean, search margins are probably the envy of the world in terms of, you know, how much money they print from search ads. How come my Google Home is so slow though? Because since they added AI, it takes forever to get an answer. Well, I mean, look at - Maybe you need a new Google Home. Yeah. Or you need to update that connection. our connection's pretty good uh mandeep fourth quarter gemini app monthly active users 750 million the estimate was for 650 million monthly active users uh what does that exactly measure is that people using the standalone gemini app or is that me who's never used that app and instead just using mobile google search but gemini within google search and no so they do bundle gemini with their Google One subscriptions.

6:01So if you are a paid user, then you could buy their bundle where they will sell you cloud storage and other things as well. But these are standalone Gemini users. So we should look at this as compared to monthly active users for ChatGPT. Yeah, 900 million versus 750. What do we know about people who use Google's products the way that I use them? Which is like, recently I started using ChatGPT and Cloudless because I've realized that the AI function within Google search has gotten pretty, pretty good. Yeah. When you look at it from like an LLM perspective. And they just released this personalized intelligence, which basically connects their family of apps, you know, Gmail, Google search, Chrome usage to Gemini.

6:47So they can personalize a lot of things for you than just, you know, you searching based on a prompt, which is what you're doing on ChatGPT. And that's the advantage that Google has, is their family of apps. And that's why Meta is really building their AI data center capacity, because if you want a personalized agent, you want to give them the context around what you're doing and how you're spending your time. And Google knows that because of Gmail, because of YouTube, and all the family of apps they have. Let's just remind everybody where we are. Alphabet blowing past investor expectations when it comes to AI spending reported fourth quarter revenue to that beat expectations.

7:29Not typically a good thing. No. Blow past those expectations. The Google parents saying it will spend$175 to$185 billion this year compared with the $119.5 billion that analysts had expected. Fourth quarter sales, excluding partner payouts, were$97.23 billion, surpassing the$95.2 billion expected on average by analysts. And I'm looking at the stock in the aftermarket. As Tim reminded me, down as much as 7 % initially. It was also the rallying, and now it's just down about seven-tenths of a percent. We're talking with our own Mandeep Singh. We want to bring also into the conversation Dan Ives, Global Head of Tech Research over at Wedbush Securities.

8:11He also has the Dan Ives Wedbush AI Revolution ETF. Alphabet, by the way, is the fund's third biggest holding dan is with us from miami good to have you here uh dan with tim and mandeep uh your reaction to alphabet look i mean i think it's just further i think it solidifies what's happened in this ai revolution i mean look at the spending look at the monetization and look it goes back to look at this tech earning season palantir look at hyperscalers last week as mandeep talked about i think to some extent in terms like this white knuckle this nervousness that we're seeing it It just further solidifies the spending, the monetization, the cloud, the moves that enterprise are making.

8:53Why did we see shares move all over the place, Dan? Seven and a half percent was the initial reaction to the downside. And then you saw shares turn positive, unchanged. What's going on with the instant moves? I think these knee-jerk reactions, especially when you're looking right after the print sometimes, I think a lot of times they're misguided. You got to listen to the conference call. You got to see ultimately what, you know, when you kind of peel away the onion in terms of guidance as well as CapEx. But look, I think it's one. It's very easy to get caught up in days like today. Right. In terms of just even so far we've seen in terms of the last few weeks in tech.

9:33I think it's very important. Mandeep obviously talks about it a ton. What's the narrative? Is spending there? Yes. Is CapEx accelerating? Yes. Are enterprises moving full steam ahead? Yes. Look, and I think that's keys to the puzzle. Is this, how great, in your view, is this increased CapEx versus expectations for NVIDIA and Broadcom? Look, it speaks to the narrative that we saw with Microsoft, that we've seen with Meta, we've seen across the board. It just further speaks to this arms race that's playing out across AI. got. And then when it comes to cat-backs, obviously AMD stock sold off, but again, this is like room wasn't built in a debt.

10:20You got to think about this is multiple years, re four trillion that's going to be spent. And I think that's, look, that's the one thing where, you know, and I've seen it being in CS, being in Davos, Mandeep and I were there.

10:38the spending is unlike anything I've ever seen, you know, going back to late 90s in terms of spending. Hey, I just want to point out, shares of Alphabet, because they have been bouncing around here in the aftermarket, they're now down about 1.3%. NVIDIA, by the way, is up 1.7 % in the aftermarket. You've been listening to Dan. Listen, everybody says these guys have to spend, these hyperscalers, right, in order to compete. Is there some truth to that, some reality to that? Because I feel like as these guys have reported, we get different market reactions. I mean, in the case of Alphabet, not only have they trained their own model, but the Anthropics model is trained on Google TPUs.

11:15And now you have got Apple saying they will use Google's chips. So three out of the big players are, I mean, we talk about applications and how good Alphabet is in search. This is about their chip prowess. And, you know, the fact that three frontier labs or companies are using Google's chips, that's a testament to their mode and how well they have integrated vertically. And the Apple thing is huge, I think. That's why they are really ramping up the CapEx here. Right. Because they can see the demand just from one company. I mean, imagine if Apple wants to roll out Siri powered by Google. I mean, that's what they're relying on, right?

11:54I do think there is a lot of embedded business that's coming from Apple in that CapEx guide. Broadcom up by 7 % after hours as a result of this print. Mandeep, you were doing some multitasking. You were looking at your terminal. What else sticks out to you from this print? What are we missing? I think after the Microsoft print, I thought maybe, you know, there were concerns around the cloud growth. And Microsoft talked about allocating some of their capacity to first party. This is a supply constrained environment. and everyone is trying to be very careful how they allocate that capacity. Every hyperscaler has that challenge.

12:34And that's where, you know, an Alphabet really is going big because they have$150 billion in free cash flow. They can afford to go up to$180 billion in terms of CapEx. Companies that don't have that operational cash flow, they are the ones who have to think about, okay, how much do I raise my CapEx? But Alphabet has that$150 billion operating cash flow. Do you have a calculation, Mandeep, for how much of the infrastructure spend a company like Alphabet uses for its own services versus selling to third parties? I mean, you can model it based on the cloud revenues that they are generating. And that cloud business is close to a$70 billion run rate business.

13:14Fourth quarter cloud revenue,$17.66 billion, beating estimates of$16.2 billion. Yeah, so$70 billion annual run rate. Compare that to an Azure, which is$100 billion plus run rate. Wow. And you can see all these companies were spending CapEx close to each other. Microsoft obviously was spending more because they had that open AI partnership. But, I mean, about, I would say, 30 % to 40 % capacity in the case of Alphabet is getting allocated to cloud. And they are using a lot for AI overviews and their internal family of apps. And then the training component is huge when it comes to these LLMs as well.

13:51All right. Yeah, go ahead, Carol. Continue to watch Alphabet here in the aftermarket, now down about 2.3%. Dan, as you continue to look at this quarter, listening to Mandeep, you know, what is it that you want to know more about when it comes to Alphabet's business right now? Look, it's all about cloud. I mean, if you think this arms rate is going on between Microsoft, between Google, between Amazon. It's really trying to get a sense to what customers are going after. And strategically, what are they seeing? The monetization. That's really front and center. Because as you said, as Tim talked about, it's not just about Alphabet.

14:26It's about the ripple effects across the sector to what we're seeing. You're trying to put pieces together in the puzzle. Looking so far, you look at the last few weeks, it's basically telling you spending is accelerating. Dan, I'm looking at the Ives U.S. Equity. It's the Dan Ives Wedbush AI Revolution ETF. The third biggest holding is Alphabet. Are you, you wear a lot of hats and you have a lot of positions. So I don't know if this is something that you do day to day, but would you add to Alphabet after this print? Yeah, look, I mean, it's in the Ives AI 30, you know, in terms of our winner list.

15:04And look, and I think it obviously, you know, it speaks to our view. You can just look at these things over 24, 48 hours. It's our view. We look at the winners and that's one of the core winners that we've identified. And it's our view, like tech stocks are going to continue to move a lot higher this year, but we're hitting a white knuckle period. But you got to look at the spending and the spending, you look at every single aspect of it. Alphabet's one, go back a year ago, year ago, no one would touch us, right? DOJ, breakup, search is going to get ruined from AI. I look at it today. Yeah, it's just kind of interesting.

15:44I keep just thinking over the last week or so and how the different AI trade, we've had some ups and downs here. You know, the race is on, right, in terms of we've talked about that there could be and there will be multiple big players in this, Mandeep. Is it safe to assume that the companies that we talk about, the Mag7 and these big hyperscalers, they are all going to be a big part of this going forward? or could, I don't know, could someone like an Alphabet be one of the leaders here? I think right now they are positioned to be one of the leaders simply because of the vertical integration and that chip aspect is huge.

16:19I mean, there will be a time when the focus will be on CapEx efficiency. How much revenue are you generating from the CapEx? And for a company like Alphabet that is not paying the NVIDIA premium in terms of buying their chips, you know, at such a high price, it will be an advantage. You know, they will have a much bigger infrastructure with a much higher ROI just because they are doing their own chips. Are they not using any NVIDIA chips? They are getting NVIDIA chips, but more for their cloud business, the external facing, for all their training. That's why the point about Anthropic training their models on Google, Google themselves training their model, and probably now Apple doing some custom version.

17:00It's all trained on Google's chips. That's a huge advantage. When is CapEx enough? When are they going to say, because the efficiency question comes into play here, when are they going to say we have enough capacity because we've become more efficient? Does that happen? That's a demand supply question. And look, everyone has so far who has reported has said they are supply constrained. Microsoft would have grown faster had they gotten more supply. So from that perspective, until we reach some sort of an equilibrium, I think these guys will continue to step up. And it comes down to your operating cash flow.

17:41Like, yes, there is a cap. I don't think they will go beyond, you know, let's say$200 billion because of that operating cash flow that they have. And that's where a meta, yes, they have a phenomenal business, but they can only go so far. So the luxury that these companies have is the core business generates so much cash. Is it still questionable? I mean, I just look at what this stock has done down 7%, then it rallied a few percentage points still down. I mean, what does that trade tell you, though? Does it tell you anything? Well, I think right now people are just trying to make up their mind on that CapEx number.

18:15To your point, how big is too big? Are they really stretching themselves? Because now they are going beyond their operating cash flow. And they will have to raise some money. despite the cash cow search business that they have, they may have to raise an extra money from somewhere. I mean, increasingly, we've seen these big tech guys, right, going to the debt market. Yeah. We've seen that trade. And we want to broaden out and talk a little bit about Qualcomm as well. Dan, we're looking at Qualcomm shares. You cover this as well. They are tanking in the aftermarket. We've seen this stock under pressure, down about 9 % here.

18:49The Qualcomm story, they gave a tepid forecast, sign of a shaky phone market. They see second quarter revenue,$10.2 billion to$11 billion. The estimate on the street is$11.18 billion. So kind of pulling that forecast in. Qualcomm, what's the narrative here? What's the theme here? What's the story? Look, I think, I mean, you continue to focus more on data center. Yeah. Revolution. When you look at smartphone market and you look what we see look around the edges there's weakness i mean we've talked about in terms of asia checks and you know obviously you know this is something that's going to be a focus in terms of you know can they navigate it how long the sort of softness will be there look i think it just speaks to overall what we're seeing in this market like you're gonna you're gonna have winners and losers it's a stock pickers market and it's one where i think you hit on it before when you told Mandeep, you're now starting to see separation in different pockets.

19:53And I think that's what's happened during earnings, Houston. I mean, Qualcomm is also, like a lot of companies, looking to increase their AI exposure. They're trying to figure that out, right? Have they figured that out? Well, that's it. I mean, they've been offering up some products. They announced some products late last year to go up against NVIDIA. I mean, is that real? in your view look there's one chip in the world fueling the ai revolution and that's invid right and and it's led by godfather and and the reality is like look jensen they're years ahead of anyone else i mean if you look look a third rate nvidia chip h200 restricted in china still above huawei now okay like demand supply we've talked about continues to be 10 to 1 12 to 1 but look it's just like AMD Intel or Google's do.

20:48I mean, this is an arms race playing out, but obviously NVIDIA is playing another game. What about when it comes to inference? Because there've been concerns that you don't need the same capacity or same type of capacity for inference and other chips could be just as sufficient. Google's own, for example, what's the story there? Look on the edges. Yeah. I mean, like, look, the reality is like when you look at metal, like they need more chips like they can't get enough from nvidia so again they're i mean google is sort of coop petition on that area but the reality is like others are gonna play when it comes to inference they'll get that market but it also just shows like everyone right now is going 100 miles an hour trying to make sure they're in a position to gain on the chip side on the software side and on the build out because we're still early days in the build out.

21:43I mean, think about where we are right now, like monetization just starting to happen, enterprise consumer now starting to come through. They have physical AI and every chip player, everyone in the supply chain, the memory players, everyone trying to make sure that they're in the right spot. So, Dan, you know, watching the trade, Alphabet's been bouncing around a lot. But NVIDIA is up about 1.5 % post-market following Alphabet's earnings. Broadcom's up almost 6%. Is there a trade, a playoff of here? Are you buying potentially shares of Alphabet on any kind of weakness? Like, walk us through. Is there an investable tradeoff of all of this?

22:25Oh, yeah. Look, I mean, to me, use continuing the hyperscalers from Alphabet to Microsoft. All of them. All of them. Because, look, I don't view it as like, it's not a winner take all. It's one that they're all going to benefit. You play the memories we've talked about. You play the chips as everyone, you know, that we've talked from Micron, you know, obviously to NVIDIA, to AMD and others. And look, and then you're trying to focus on who the second, third, fourth derivatives are. But the one thing is like, you can't get too caught up. And look, we've talked, you know, me and you, we've talked about it on the show for years.

23:00You cannot get caught up in some of these weekly or daily gyrations. You've got to understand where the story's going, and the story's solidified. Well, I'm wondering if this is a daily gyration or if this is a generational shift, Dan, and that has to do with the story around software and the decline in software companies that we've seen. just in recent days after Claude, Anthropics Claude, was said to have this ability to do to legal docs analysis that would affect companies like LegalZoom, for example, and it opened this whole can of worms that, wait a second, we're not going to need to pay Salesforce per seat because we can get some of these AI companies to build this old model for us and we won't have to pay a subscription fee.

23:44Is that overblown? Look, I think it's overblown. look look is ai had one for software in some areas yeah but is this overblown yeah i mean like no one you're not replacing salesforce with a model like this and i think that's something where you know you see that narrative really creating a black cloud over the sector but then to some extent now it's in the hands of benioff salesforce mcdermott service now to prove it right you got to prove the monetization. I think that's the opportunity and challenge for software. But look, right now, going back over the decades, we're seeing something really unprecedented in terms of a sell-off.

24:29Have you been buying or selling any of those software games? You have a buy on Salesforce, by the way. Okay. Yeah. And Salesforce is like, look, it's an example. Mandeep and I have talked about it. It's like, is AI a headwind for Benioff and Salesforce? Yeah. But is that changing? Would you tell them when the biggest install base in the world and agent force and the monetization that they're going to be able to do? Are they going through a patch or something? Yeah. But so I think you have to separate out reality from fiction, but it also comes down to like for investors. It's not just about talk.

25:08Show it. Stay with us. More from Bloomberg Businessweek Daily coming up after this.

25:47Hello, hello. I'm Malcolm Gladwell, host of the podcast Smart Talks with IBM. I recently sat down with IBM's chairman and CEO, Arvind Krishna, and I asked him, how can companies use AI to its fullest potential to create smarter business? My one advice to them, pick areas you can scale. Don't pick the shiny little toys on the side. For example? If anybody has more than 10 % of what they had for customer service 10 years ago, they're already five years behind. If anybody is not using AI to make their developers who write software 30 % more productive today, with the goal of being 70 % more productive, so we are not asking our clients to be the first experiment on it we say you can leverage what we did we are happy to bring out all our learnings including what needs to change in the process because the biggest change is not technology it's getting people to accept that there's a different way to do things to listen to the full conversation visit ibm.com slash smart talks

27:05Sonesta Travel Pass makes traveling more rewarding, designed to help you get more out of every stay. Sign up at Sonesta.com to enjoy instant savings, bonus points, and valuable perks like early check-in, late checkout, room upgrades, and free stays over time. With Sonesta Travel Pass, every stay brings you closer to your next reward. Choose from more than 1 ,100 hotels across 13 distinctive brands and unlock the best available rates when you book direct with Sonesta Travel Pass. Here today, Rome tomorrow. Join now at Sonesta.com. Terms and conditions apply. You're listening to the Bloomberg Business Week daily podcast.

27:44Catch us live weekday afternoons from 2 to 5 Eastern. Listen on Apple CarPlay and Android Auto with the Bloomberg Business app, or watch us live on YouTube. We're going to start global and kind of drill down on this Wednesday. First, President Trump out on social at around 1040 a.m. this morning saying he, quote, quote, just completed an excellent telephone conversation with President Xi of China. Long and thorough, he said the call was. He said they talked about a lot of different things. They discussed trade. They discussed military. They discussed the upcoming April trip that the president will be making to China.