Bonus Episode: Tesla, Meta and Microsoft Earnings
52:24 Share

89% of business leaders say AI is a top priority, according to research by Boston Consulting Group. The right choice is crucial, which is why teams at Fortune 500 companies use Grammarly. With top-tier security credentials and 15 years of experience in responsible AI, Grammarly is how companies like yours increase productivity while keeping data protected and private.

See why 70,000 teams trust Grammarly at grammarly.com slash enterprise. Meta's open source AI, available to all, not just a few. Here's Steve McCloskey, CEO of Nanom. With Meta's open source AI model, Lama, we built a tool to help scientists to discover treatments for diseases. Learn more at ai.meta.com slash open. This is Bloomberg Business Week with Carol Masser and Tim Stenevek on Bloomberg Radio.

It is Bloomberg Businessweek. I'm Tim Stenevek. That is Carol Masser. And we are following what's going on with all the companies that are reporting right now. So far, in terms of the big companies that we're following this afternoon, Carol, we're looking at shares of Tesla, which are lower in the after hours. In addition to that, we're also looking at shares of Microsoft, which are following after their report. Yeah, absolutely.

We've got Microsoft down almost 4% here in the aftermarket. As we said, some disappointment over maybe some of the numbers in the second quarter revenue. 69.63 billion estimate was 68.92. That's actually better. Second quarter Azure revenue was up 31%. The estimate was for a gain of 31.8%. But the second quarter cloud revenue, that missed estimate. And so that is

Second quarter, intelligent cloud revenue, $25.54 billion. The estimate was just a tad above that at $25.89 billion. So again, maybe some concerns over that. And expectations are high right now. Okay, he doesn't ignore it, but he doesn't cover it like some of his colleagues. Maddy Singh is Bloomberg Intelligence Senior Technology Analyst.

As we await meta earnings to cross, we're going to make him go a little out of his element and speak to Microsoft. Because you can't ignore what's going on with Microsoft. Mandeep, give us an idea of what sticks out to you from this company. I mean, look, the intelligent cloud number, Miz, probably that has to do with, you know, lower adoption of 365 co-pilots. They've been touting that and

the expectations had gone up and look it's one of those products where

The ROI so far hasn't been there. The promise is there. And that's why they are emphasizing the Azure number, the $13 billion run rate, the 175% growth. Look, it's stellar by any comparisons. And from that perspective, the infrastructure side looks solid right now. It's the co-pilots and the application side where you're seeing some hesitation. And look, I think that's

The native integration with OpenAI, as long as it's there, it should help all their products, but there's more competition. I mean, Salesforce has been launching their agent. So there is more competition in the application layer now than there was before. And the point is you will build it, but will they come? And that's what it's kind of like what they're waiting for, correct?

if you're asking enterprises to pay $30 extra per user for that co-pilot feature, it better be good. I mean, that's the key takeaway. It's like, if you want to upsell such an expensive product, it has to have some level of productivity. So remind everybody where Microsoft stands in terms of the LLMs,

Because you mentioned a company like Salesforce. Is a company like Salesforce building its own LLM or are they building an agent on top of an existing LLM? Maybe from Microsoft, maybe Lama from Meta, maybe ChatGPT from OpenAI or Claude from Anthrop. Exactly. So they are not building their own LLM at the

same time they have gone big in their agent force concepts and they are saying AI agents is where they are doubling down on and they don't need to own an LLM so that's where I think with the open sourcing you know of deep seek and the fact that pricing is coming down all

All that is positive for every application software provider, not just Salesforce, even Workday or SAP. They can integrate LLM functionality now. And we saw software names certainly rally this week. Having said that, Microsoft has especially unique exposure to OpenAI and ChatGPT in terms of the, what, $13, $14 billion or so that they have invested. So does that put

them on a certain level of vulnerability, perhaps? No, because they are benefiting from the infrastructure revenue, that $13 billion run rate. It's all Azure cloud consumption, nothing to do with the co-pilots, just companies that are leveraging their infrastructure to run their cloud operations or deploy their applications. So that meter is still running.

And it's running very fast. I mean, 175% growth just on infrastructure is very impressive. It's just the application side, the lift from Copilot, I think that's where they'll probably see more competition and probably expectations were too high. - Well, speaking of expectations, we're expecting Meta Platform's earnings to cross very soon. We have no idea what those shares are gonna do in the after hours.

Revenue estimate coming in at close to $47 billion. What's the most important line item that you're looking at here? So Meta raised its capex right before we had this deep seek. $65 billion. Okay, was that before? Because that had to be on Mark Zuckerberg's radar. It was, but still they chose to do it. Now, the fact that this is really a positive development for open source, everyone is thinking Meta won't have to spend 60 to 65 and they will take it back.

So that CapEx number is still the key for me because Meta has been known for spending, spending big on reality labs and now building data centers. - Are you saying that there's a chance we could hear a pullback in that CapEx number and investors might cheer that? Investors cheered the investment number on Friday. I was surprised to see that. - Not so soon. I think with the deep seek, the fact that everyone is very excited about open source

Meta coming out and saying we can monetize llama our model which really no one was sure How are they going to monetize it suddenly? I think that's that becomes a very valuable proposition and I think just his plans around how do they plan to Monetize AI and really not going above 65 billion That's the other risk because you know, you could revise it further upward at least I think those fears have gone away that there is 65 is a cap and

We talk so much about AI. I mean, for Meta, it's still though advertising, like something as simple as that is still such an important indicator. That's their business. I mean, 45.6 of the $47 billion, Carol, that's expected to be reported is advertising. And I understand it all plays together, but still, that's the metric. And the best way to compare is compare Google search revenue growth with Meta's social media ad revenue growth.

So Google Search is a $200 billion business growing at high single digit at best. Meta is close to $160 billion, growing 18%. How do Reality Labs products help meta platforms sell advertising better?

Well, so their bet is on that meta AI, which for which they have 600 million users now. And the form factor, their own form factor is the glasses that you can distill a model. And everyone is talking about distillation right now because that's what DeepSeek showed us, that you can distill a larger model into smaller model. Maybe that model can run on your glasses. Okay. How does that help them sell more ads? Well,

they will learn so much more about you as a person that they can show you a more personalized ad when you go to Instagram or Facebook. And that's where, you know, Meta's strength has always been the quality of their ads. So still at the end of the day, that's what it's all about. Yeah. At least for now. I just want to remind everybody, I'm looking at our live blog, which is waiting for Meta's fourth quarter earnings. Our

Kurt Wagner says they usually drop a few minutes after the market close at 4:00 PM New York time. Still waiting on them. He says it's a bit of a rare delay. We usually have them, like we said, right after the market close. So we're awaiting them. As soon as they come across, we're going to have Mandeep break them down. Also in studio is our own Caroline Hyde, who's co-host of Bloomberg Technology and Bloomberg Television. And Caroline, we've also had Tesla earnings. We've been talking with Mandeep about Microsoft as we await Meta.

Go where you want because it just feels like both Microsoft and Tesla have seen some pressure here in the aftermarket. And also, I don't know if you've broken down ServiceNow, but suddenly software became quite the thing of this year, of this week, because we all suddenly freaked out about AI infrastructure and hardware. But was software going to be gaining from these deep-seek revelations? ServiceNow actually looking pretty lackluster. Outlook on an

slower than had been anticipated. We were thinking AI was going to give it this massive bump. Overall fiscal year sales outlook fell short of those expectations. Subscription revenue going to be $12.7 billion 2025. Market want to see more. So their shares are dropping hard. And I think this is where we're going to have to hear more from Microsoft as well is

Is the OpenAI integrated offering not selling at the pace that they previously expected? I've heard from experts out there saying it's priced too high, ultimately, at the moment. Will DeepSeek force OpenAI's own offering to get cheaper, Microsoft's offering to get cheaper, and therefore it becomes more ubiquitous, we will start using it more, but ultimately they get more bang for their buck in the long term? Does it mean that these companies could potentially pull back on some of the CapEx they've already announced this year?

I think that was interesting in the, and we talked about it when the numbers dropped, Amy heard the CFO talking about discipline.

Now, last year we were not talking about discipline. We were talking like, let's throw $80 billion at this. We just heard another half a trillion being thrown at it from Oracle, SoftBank, and OpenAI. I think you'll start to hear this. This is wise spending. This is return on AI spending. So not quite year of efficiency like we saw in 2023, but now disciplined spending when it comes to capital. We are spending wisely on your data. Just because you say it doesn't mean you're going to spend it ultimately, too. I do want to point out ServiceNow reports an additional $3 billion for share buyback. Oh, nice.

Yeah, but still down about 9% here in the aftermarket. But, Mandeep, come back in because we were talking about the news we got this week from DeepSeek that the beneficiaries seem to be the software providers. Not every software provider though is the same, correct? Well, I mean, a company like ServiceNow in theory should benefit from lower LLM costs. I mean, all the software companies could arbitrage against the LLM providers and say, "If I have a cheaper open source option, I'm going

I'm just going to go with that because the quality of responses are very similar. So from that perspective, I mean, the news just came out. So obviously they didn't embrace open source before. And now a lot will change in the next three to six months. But for now, I think, to Caroline's point, expectations are high this quarter. And a lot of these companies have touted AI agents in a big way. So the street wants to see the revenue. And probably it's not showing up right now.

I had a really interesting conversation with an expert who's helping integrate generative AI into enterprise offerings. And here's the gap, right? We've been talking deeply exuberantly about how it's going to make our productivity so much better. Ultimately, enterprises need to know that their models work.

And they need to stress test them. They need to have the right companies in there ensuring that, well, when you're saying it's going to get you a better mortgage outcome, when you're looking at insurance inquiries, you're actually getting the right output. Companies are trying to integrate it, but they've got to do this with some rules of the road. So it actually takes a bit of time. OK, speaking of rules to the road, perhaps a good segue to get to Tesla because shares are down about 2.7% right now. It seems, Caroline, like it was the adjusted EPS.

that came in shy of estimates, 73 cents per share versus estimates of 75 cents per share. Is it about a lack of efficiency on the supply chain, on the product line? What is it? - They're saying they're not getting the manufacturing savings that they wanted to see.

- Have they tried the blockchain? - Yeah, have they tried blockchain? Have they tried just making everyone have to come back into the office? But more, I mean, Elon's been on that for a long time, but more affordable models is where it's at. This is a company that is having to compete from a price perspective in China in a big way. And look, we saw deliveries did not live up to expectations. So already they're not selling as many, they're having to reduce the costs. It's interesting they're seeing the automotive gross margin ex-regulatory credit still at 13.6%.

This is a company that basically makes an awful lot of money because other car manufacturers can't get the climate initiatives out there quickly enough. I wonder if that'll dial back. But ultimately, they did say they achieved record deliveries in China in fourth quarter. I wonder how they managed to be able to just

keep on competing there. And no mention of robotaxis. I mean, the big bet was this administration should help expedite the rollout of Tesla robotaxis. That's what investors are betting on. In the letter, they did say this is going to be the year of supervised fully self-driving. But we're still waiting to see what the guidance is from the federal government. The cyber cab is on track for 2026. So ultimately, they're talking about still being wanting to get that

cyber cab out there. In the meanwhile, we're having to see whether that cheaper price point new vehicle will happen. He's saying new vehicles remain on track for the output start in the first half.

When, where, how? I just want to update. Meta shares are just down about five-tenths in the after hours, five-tenths of 1%. Kurt Wagner on our live blog says, the latest Meta or Facebook has ever posted earnings is 4.16 p.m. We have now a new record because it's 4.20 p.m. here on this Wednesday. He says, in all seriousness, this is definitely unusual. I've asked the company what is going on and will report back if I get any news.

I was trying to see if there was any interesting activity here in the aftermarket ahead of the actual results. But again, it's just down a hair, down about four-tenths of a percent, but still waiting on Meta's results. I don't know. Mandip, you follow them, right? They're usually out by now. I mean, it reminds me on when NVIDIA, I think it was a few quarters back, they were delayed and everyone was like, what's going on? And sometimes I think it's one of those situations where...

Yeah, there isn't much to it. It's just there is a delay and you just wait it out. But look, I think in the case of Meta,

I mean, there are so many things that I think investors are looking forward to, especially there was a number around reality labs growing 40% for 2024. And when you look at the consensus expectations for this quarter, it's like 3% growth. So I don't know how they grew 40% for the full year. - Caroline, come on in here. Is that because the, I don't know, the glasses exceeded expectations? Again, we're still waiting for these numbers to cross.

Reality Lab's a tiny, tiny portion of the company's full revenue. I mean, a tiny fraction. But an area where they're investing a lot and where they think there's a real future. And people got very excited about the product. Did you see how often Ed Ludlow's been wearing his metacross? I've heard they're great. They really did blow people's minds as to how exciting the future could be of just the integration within your everyday work. Suddenly, you know, smart glasses became cool. And thanks, Ray-Ban, on a large part of that.

Yeah, Google Glass is calling. They're saying, hey, we had this 10 years ago, but it didn't stick. And what you want is the zeitgeist, right? Sure, people might not buy that iteration. What about the next iteration? What about the fact that this is going to, in some way, be infused within your work-life environment as well, and suddenly you'll be able to have four screens wherever you are, not just two? Yeah.

I think it suddenly just showed the power more than anything of Mark Zuckerberg to interlace generative AI everywhere. For me, the big standout has been the fact that I do use generative AI within my Instagram. So what do you do? Just out of curiosity, how are you doing? I've got my meta AI sidekick and I ask it questions. I get it to like design what one should wear for an event. I give it pictures of my fridge and get it to recommend what I should be cooking. Do you follow it? Do you use it? I mean, do you like take these things and actually do something? Yeah, yeah,

Yeah, I've definitely said what they should cook. Sometimes the fashion sense I like to take a little. But to be clear, you could also use ChatGPT from OpenAI to do that or Cloud from Anthropic to do that. But I didn't have to download an app.

I already use WhatsApp every day of my life to talk with my family. So I don't subscribe to the expensive versions of, I take the free version of ChatGPT. You don't have $200 a month just to walk around on ChatGPT? Not for my personal agent to get me a flight to wherever, which I think is a really cool idea. And I think that shows the promise of agents. And we hear Salesforce talking about the promise of agents and stuff like that. But are you able to do...

for free with Llama, what you could pay ChatGPT to do? Is that an issue? I have to say, personally, I'm not running the two against each other. And I use ChatGPT for other things. I'm using ChatGPT to articulate what my au pair, because we will have to rely while I'm at work on wonderful people looking after your children, how her week should look. And I just speak into it, dictate it. So I'm using different things for different points. But for me, I actually end up going towards the meta offerings a bit more just because they are here. We're seeing interesting...

an agreement from the Wall Street Journal talking about Trump signing an agreement calling for Meta to pay some element to settle a suit. So maybe there's some breaking news there. But I am, yeah, I am using Meta AI just because it's already an app that I'm using when it's WhatsApp. I just want to mention Microsoft is not down as much as it was earlier. It's now down about...

three quarters of one percentage point. Tesla is now down only about nine tenths of a percentage point. I want to mention IBM because we are seeing this one rally in the aftermarket following its earnings. This one's up about 8% here. IBM coming out with results, kind of older tech, if you will, but jumping after fiscal year revenue growth. Outlook tops estimates. We mentioned ServiceNow, and we are seeing that one move in a big way. We've seen that move to the downside. Some concerns there. We do have, as you said, that

headline crossing from the Wall Street Journal. Trump signs a settlement for Meta to pay $25 million to settle a suit. So perhaps that is why we might be waiting ultimately for Meta's results. Meta in the aftermarket right now is just down about one quarter of 1%.

And I've been also looking at the NASDAQ 100 E-mini futures pretty flat in the aftermarket. And S&P 500 E-mini futures, they're down about 0.4%. So just kind of rolling everything together to see kind of market reaction. Tesla, though, higher now by about 1%. That's what I'm saying. It's like kind of fascinating to see the bump up. You're right, now up about 1%. And I think what's interesting is remember NVIDIA, you were talking to him earlier about the hit NVIDIA took.

The shares are climbing now after hours because of Microsoft's long-term CapEx commitment. So maybe even though we don't get NVIDIA numbers until February the 26th, Microsoft's already coming out standing by its capital expenditure commitment. If we get Meta committing to that $65 billion once again, maybe we do start to see NVIDIA clawing back some of the loss that we've seen. And a good reminder, too, that we still have a lot of details that we're waiting on about DeepSeek and what exactly they were able to do with what.

So some investigations going on, some investigations, Microsoft got their earnings out on time and we're doing an investigation. The war room of engineers reportedly. So I want to bring you back in there. Could it possibly be that there's a delay because of this settlement and they want to kind of like roll it all together? That mountain looks too small to matter. It doesn't matter. You know, 25 million for meta it's like, and that was under the, over the Trump suspension to the use of Facebook and its family of apps.

Yeah, I just think about the stories we've been putting out. They're paying billion-dollar fines in the EU, so this is like a drop. This is nothing. I mean, it's kind of fascinating to see that we're just kind of waiting. But Tesla's on time. I know. Tesla's on time and Meta is late. What a turnaround for everybody. And Tesla's now up about 2.6% in the aftermarket, so quite a turnaround here. Wow. Yeah, it is notable, Meta's delay, and we have got...

requests out for comment on what might be holding them up but this

This is, I feel there's going to be a lot of redrafting of messaging going on since Monday. And look, what a very difficult week to have your earnings on when you're trying to understand the competitive threat of DeepSeek to a Lama model, which are both open source, but at the same time, the way in which it could enhance their products. Everyone read DeepSeek as a good thing for LLM, for Lama and open source LLMs. But maybe there is a question of ultimately, why should you be backing AI homegrown open source versus deepseek?

competing versus China. I mean, Mandeep, do you think that the narrative like substantially should be different because of what we got on Monday from Deep Seek? Like,

Like the conversation that we are having around AI in terms of the spend, was it smart for us to kind of all of a sudden shift it felt like the narrative or time will tell whether, again, there are so many questions still about the reality of what they did. Yeah, I mean, look, last six months you've been talking about scaling laws nonstop. It's like everything is how big are the data centers going to get? How long will the scaling laws hold? So from that perspective, I think this does put

a question mark around the scaling laws. How big does the LLM need to get to improve its reasoning? Well, a smaller LLM can give you a similar reasoning as an OpenAI latest model.

Look, I think with LLMs, because it's so hard to understand everything about how these LLMs are trained, the data sources, the algorithmic improvements, it's still a black box. I mean, things are getting better in terms of the transparency, but in the end,

Common people cannot interpret what's going on in terms of building these LLMs. And it's really hard to explain it to someone why an LLM's quality is better than the competing LLM. It's just almost impossible because it's all vectorized math at the end of the day. Do they become commodities, Caroline?

I think that has been the question for a while that will ultimately just move, toggle between Chachi PT, your meta offering, what Anthropic has been up to. That doesn't mean that you can't

still see an awful lot of money having to be invested in these latest and greatest underlying foundational models. Whether eventually the money will accrue to the application layers, to the software layers. Well, I do think there's a compelling case for Google to make that behind the scenes, if you're using an AI agent and you ask it to do something complicated, Google and actually Microsoft too with Bing has that search engine built in

that can actually do that stuff and perhaps they make money the old fashioned way. - More of a vertically integrated offering as well. They have to have the foundational layer to then sell you the software applications as well. But would you rather be a ServiceNow that just has the software margins rather than having to do the underlying investment? What's been so interesting has been that shift as well from investment coming ultimately from private sources. What we've seen with the announcement of AI infrastructure money coming to the US is, hi SoftBank,

Oh, thanks, Auroch. The chipset, like nowhere to be seen for the last few weeks of late. Now, what happens to Project Stargate? I mean, the whole timing of it, Meta raising their CapEx, Project Stargate announcement, and then DeepSea coming out. Just think of the sequence of events. And they're not in sync with each other. So that's where, you know, if you're trying to follow the big trend,

it's not very clear what the big trend is right now in terms of scaling versus you know how do you go about data center investments uh medium to long term and do you buy the inference argument though where we that was the argument of nvidia is you're still going to need all of our very expensive gpus for inference and we're still if we're all ubiquitously using generative ai more and more and more yeah

Well, that's when the Jevons paradox comes in that Satya wanted to lead to. You need the data centers for cloud. Yeah, more than that, I think compute inherently is fungible. So if I'm using certain chips for training now and I don't need to build a million chip cluster, then I can use some of that compute for inferencing because it's the same GPU.

So from that perspective, I do buy that argument that if AI is more pervasive and accessible, then I'll use more of that chip capacity for inferencing because everyone realizes there is ROI in deploying generative AI. - I mean, this is the technological cycle at work, right? I mean, stuff comes out of the gate initially. It catches all of our attention. We've been obsessed with really kind of a very few companies over the last couple of years. Here we are more than two years in on this, right? It makes sense, you guys.

Caroline Mandeep that we're going to start seeing other companies come to the forefront. There's going to be some competition. The model might be tweaked a little bit here. Yeah. I mean, and look, the commoditization, I do think, you know, the moats around, oh, you've used this LLM and you can never get out of that. It's so sticky.

I think that is going away because now you can swap API from one LLM to another and that should be doable. And companies will arbitrage against which is the cheapest model provider because it impacts your margins. I mean, look at the scale of some of these LLM deployments. We are talking about billion users. So billion users are using a generative AI query.

That's all API revenue that these companies are getting. And if I'm paying the highest price per API, that's going to hurt my margins. So I'm going to go with the cheapest option that's out there. I mean, Caroline, when you think about the conversations you have over and over, what changed so dramatically this week? Is it just kind of a rethink of these long-held beliefs about AI? And I think for a lot of people, it was felt that people acted before they really knew. So it's just...

sell, I've seen this headline, I'm worried about the valuations of these companies anyway. NVIDIA was trading at 41 times. We see generally everyone elevated versus the NASDAQ. Get out, then ask our questions. People started to come back to software after they asked those questions. I thought, ah, maybe this could actually change the game and make things even more margin rich for these types of companies. But I think the question is still out of ultimately what access did they have to NVIDIA GPUs? Ultimately, they say that they got them totally legitimately

Nvidia would say the same. How much it is ultimately, Spark inferences the use. And what does it mean in terms of ultimate infrastructure investment? I think people for the short term, most analysts I speak to say we're sticking by it. Like we're not going to see a CapEx pulled back anytime soon.

Interesting stuff. We're going to have to wait also to hear what David Sachs, the AI czar for the Trump administration, has to say. All right. I think we're going to let you go. I'm really fearful to do that to both of you because I know the minute you get out of your chair and walk out of our studio.

All right, but Mandeep, we're going to let you go. And hopefully we'll maybe even bring you back a little bit later on. Mandeep Singh, of course, a key member of our technology team at our Bloomberg Intelligence Unit. 89% of business leaders say AI is a top priority, according to research by Boston Consulting Group. But with AI tools popping up everywhere, how do you separate the helpful from the hype?

The right choice is crucial, which is why teams at Fortune 500 companies use Grammarly. With over 15 years of experience building responsible, secure AI, Grammarly isn't just another AI communication assistant. It's how companies like yours increase productivity while keeping data protected and private.

Designed to fit the needs of business, Grammarly is backed by a user-first privacy policy and industry-leading security credentials. This means you won't have to worry about the safety of your company information. Grammarly also emphasizes responsible AI so your company can avoid harmful bias. See why 70,000 teams and 30 million people trust Grammarly at grammarly.com slash enterprise. That's Grammarly at grammarly.com slash enterprise.

This show is sponsored by BetterHelp. BetterHelp has been revolutionary in connecting people to mental health services. Using BetterHelp can be as easy as opening your laptop or your phone and clicking a button, and the session begins.

Clients are able to choose in what way they would like to communicate with me, whether video or on the phone or chat texting. BetterHelp is there when you need it, and that's what makes all the difference. Visit betterhelp.com slash podbusiness to get 10% off your first month. Therapists were compensated.

We are awaiting meta results. Also a key member of our technology team here at Bloomberg News, of course, Caroline High. So lucky for us, she's going to stick around, one of the co-hosts of Bloomberg Technology. We do have another guest, though, that we want to bring into the conversation who really follows the tech sector very closely and invests in it. Yeah, I want to bring in James Chokmok. He's technology analyst at Clockwise Capital. Clockwise Capital is the core equity and innovation ETF. The ticker is CHOCMOC.

time. Biggest holdings are the companies we're talking about, Meta, Amazon, Microsoft and Apple, along with smaller companies, including Palantir, Spotify and Netflix. The ETF up about 35 percent over the last year, 4 percent year to date. So let's go through it and bring in James. He joins us from Miami this afternoon. James, good to have you as always with us. I want to start with your impression of Microsoft shares falling after reporting slowing growth in the cloud unit. What sticks out to you?

Yeah, the big thing there with Microsoft is that the bogey was a couple of points higher than where they reported on Azure. The street was looking for around 33, 34% growth to keep the momentum in the stock going. They posted around 31, 32%.

So it was a couple of points light, but the main thing is that the undercurrent, the trends that you're seeing in productivity and the shift to the cloud and AI offerings continues. You're still talking about a three handle on that growth rate. So we'll see what they say on the call as far as CapEx is concerned. We do think that's going to start to come down here as we look into 2025.

So I think it's going to be all right. You know, we hold a position in it. We are underweight relative to the index for this very reason, because of the run-up, but

Overall, I think everything's okay with Microsoft. I'm going to just jump in because we are getting Amy Hood, the CFO, giving an interview with our own Bloomberg reporters saying that they remain constrained in cloud capacity. Remember, we got that sort of shock move that OpenAI was going to be building out AI infrastructure with Oracle. Of course, they signed a deal with Oracle in the summer because...

Microsoft just can't build them up quickly enough. Talking about the bookings rising 67%, partly due to the open AI commitments and commercial bookings were better than they forecast. They're still trying to talk about how much more percentage point add AI is giving them to Azure more broadly. All right. So, James, so remains constrained in cloud capacity from the Microsoft CEO. What's your read on that?

Yeah, I mean, they're going to have to keep building. I'm not suggesting that CapEx needs to stop. It's more just what is the growth rate of that spent? You know, you saw massive investment years in 2023, 2024. And we think that that spread will be growing at a slower clip. Do you change in turn? Sorry, James, do you change your your holdings? Do you sell Microsoft?

Well, we actually trimmed Microsoft into the quarter today. You know, we're now at 3% weight. It's around 8% in the NASDAQ. So we are underweight there. But I would say, you know, we still want to hold it. You still want to be there. The company is incredibly important. But at the same time, we think that there's better opportunity elsewhere with companies that offer and stocks that offer better convexity. Okay, so where does that money go? You trimmed your holding in Microsoft. Where are you spending it? What are you buying?

So we're looking at a lot of software names right now, you know, with the disruption in the market that was caused by DeepSeek. I don't think you can take that lightly. I think that you have to re-examine all of your assumptions top down.

You know, from that, we did cut Nvidia before the market opened on Monday and cut that exposure. And we've been rotating to software names. We picked up Datadog, for example. You know, we think that rates will continue to trend lower.

You know, this year, obviously, two estimates baked in, but I think that that trend will help a company like an upstart. You know, we continue to like Spotify. We actually grossed up our meta position heading into the quarter.

So we'll see how that plays out. I don't know if the numbers are out yet. No, we're still awaiting a meta. I want to ask you something, though, from our BI team, our Anurag Rana writing about Microsoft and says, Microsoft's close relationship with OpenAI makes it better positioned than most of its software rivals to capitalize on increased generative AI spending. You agree with that? I mean, they know a little bit about software, certainly Microsoft. Yeah.

I'm not sure about the health of that relationship, to be honest with you. Why? So, well, I think that DeepSeek certainly changes the game a little bit. I think that they'll increasingly have less reliance on open AI and...

And as their costs come down, they'll continue to be able to drive the inference and all the AI tools for their customers. So I'm not sure exactly how healthy that relationship is or how it will play out. But at the end of the day, Microsoft continues to be well positioned. Actually, I would say following DeepSeek, we came away incrementally positive.

on Meta, on Microsoft, on Amazon. So those are the three that we like following that news. And then we became incrementally negative on NVIDIA. All right, hang on for a second. We're talking with James Chokmok, partner and tech analyst over at Clockwise Capital. Caroline Hyde, co-host of Bloomberg Technology, also in studio with us. And just walking in, our own Anurag Rana. He is with our technology team with our Bloomberg Intelligence team.

He's a senior industry analyst writing. I am reading from the research he just put out on Microsoft. Microsoft now, Anurag, just down about eight-tenths of a percent. So it's definitely come off the lows that we saw earlier in the aftermarket so far. Tell us what we need to know.

Yeah, it's all the guidance on the conference call. We want to hear that Azure growth is going to pick up over the next 12 months. And I think that's really the biggest thing that we see. Now, that's assuming all the investments they have made in the data center side is picking up because I just saw the news go by where the CFO is saying that they are still supply constrained. Now, the supply constraint doesn't mean it had an impact

on the current quarter growth, which was below the 31, 32 that they gave, or will it impact in the next quarter? I think that's the single biggest question right now. So is it fair to say this is totally supply side and not demand side? Can I just jump in? Because we do have meta finally. Please. And the fourth quarter revenue is beating expectations. $48.39 billion is coming for fourth quarter. It was expected to be $46.98 billion. And they're pushing us forward to a guide of the first quarter to be $39.5 to $41.8 billion.

That's slightly shy of the midpoint where the expectations were for 41.6 billion. So forecast a little bit light. Fourth quarter, it beats on family of app revenue doing well. They're seeing reality labs operating loss of almost 5 billion, but the estimate was for even more than that.

So I think you are seeing perhaps a slightly shy forecast, but they're seeing first quarter revenue of $39.5 to $41.8 billion, and the market wanted $41.7 billion. Stock bouncing around. It was down 5%. It's now down about 3.8%. Now it's down about 5%. It is bouncing around. Also some commentary in terms of the CapEx. We anticipate our full year 2025 capital expenditures will be in the range of $60 to $65 billion. So a little bit more about that, which we just got news last week.

The company also says we're not providing a full year 2025 revenue outlook. Is that a surprise to you, Caroline? Yeah, usually they are able to give clarity. And it's interesting that they're giving so much clarity on their capital expenditure, perhaps not clarity on forward revenue at the moment. Is that because they think it's just going to go through the roof? Is that because they do feel that they can't give as much?

forward-looking guidance at the moment. But whether it's an FX perspective, costs and expenses, they're still dictating. But I'm just scrolling through the CFO and they expect first quarter 25 total revenue to be in that range, which reflects 8% to 15% year-over-year growth or 11% to 18% on a constant currency basis. But what's

We are not providing fully a 2025 revenue outlook. We expect the investments we are making in our core business this year will give us an opportunity to continue delivering strong revenue growth through 2025. Also saying the regulatory landscape in the European Union, the United States could impact business. We do see the stock now just down about 1%. Again, the key that everybody seems to be focusing on is that first quarter sales forecast trailing at the midpoint. Also saying fiscal year depreciation expense about $2.9 billion less than

than before. Again, bouncing around big time in the aftermath. Yeah, I want to draw your attention to what's going on with Reality Labs. The company saying that revenue coming in at $1.08 billion, shy of estimates of $1.11 billion. I don't want to say there was a whisper number ahead of this, Caroline, but

Folks were bullish on Reality Labs going into this. There was that report from Insider that came out earlier this week that talked about growth in that unit. How are you, again, a very small portion, but a very important part of the company. How are you looking into that number? Yeah, I do think that overall the losses are...

Big. So you think the losses should be more of a focus than the revenue? Well, I think revenue 1.08 is slightly less than anticipated, but at least their operating loss is slightly less than expected. But I think there's going to be a lot of digestion of restructuring that's currently going on in that business. And the fact that they're shifting over to the overall unit COO and they're putting up some operational changes at play. But how much is Mark Zuckerberg going to speak to his commitment to this part of the business? How much he wants to invest in Reality Labs?

how much of that $65 billion up to CapEx and AI is ultimately going towards those sorts of products. All right, we're going to keep watching and reporting on Meta. As we mentioned, stock is bopping around here in the aftermarket, really a focus on the outlook. But right now we're looking at just a slight move to the upside of about one quarter of 1%.

I want to go back to Microsoft for just a moment because we have Anurag Rana in here, another big one after the close, which has been bouncing around too. It's now only down about eight-tenths of 1%. The Bloomberg audience thinking about these earnings, the crazy week that's been in the tech community around AI. What should be top of mind? What's top of mind for you?

I think the biggest question is, they have a commercial booking number that just was absolutely spectacular. It was driven by Azure commitments and the OpenAI work that they are getting. But the question is, can you fulfill that demand or you're going to go to third-party providers to take care of it?

And that's really what it is. I mean, frankly, at this point, more color in that, you know, the big commercial booking number would be helpful. And also about what does Microsoft CEO think about all the things that are happening with DeepSeek? Because he did tweet something very interesting that, you know, if the cost curve goes down, the adoption rate improves, which is good. But at the same time, you know, he has a very strong partnership with OpenAI and others. So there's a lot, I think,

we will learn about the tech landscape in general from him, not just about Microsoft's cloud numbers. Now, frankly speaking, let's say for the sake of argument that they do miss guidance for 3Q. I think people will still give them a pass over the next several days, may not give them a pass tomorrow because the bookings number is so strong. Another headline on Microsoft to add jobs and infrastructure, Gen AI and Reality Labs. That's meta, right? Oh, did I say? Oh, Microsoft. Microsoft.

They must have made a mistake. Okay, so... There is a headline. We will... We are going to keep covering... This is coming fast and furious. I do want to do a big thank you to Anurag Rana for joining us. He is going to go off and listen to more of The Call.

and as well as have more notes out for Bloomberg Intelligence. I'm putting pressure on you, unfortunately. I'm sorry. Thank you for stopping by. Usually in Chicago, joining us here. I do want to get back to James Chokmok. He's technology analyst. And it is a correction meta, which makes sense because of Reality Labs. Technology analyst at Clockwise Capital. We gave you a little break, James, to jump in and look at those meta numbers as they were breaking just past 435. What sticks out to you?

Yeah, I mean, the big thing is the first quarter guide coming in a touch light. But the fourth quarter itself was solid pretty much across the board. So we'll have to decipher what exactly is going on in the first quarter guide on the call. But as it stands, if you look at 4Q in a vacuum, you know, it seems like it's all systems go. But just got to drill down a little bit more.

All right. We want to... That's why I think the stock is bouncing around so much, down five to up three. Everything's bouncing around. Meta is now down about two-tenths of a percent. Mandeep Singh, who follows Meta for us here at our Bloomberg Intelligence team. We're giving him a workout today. Get your steps in, buddy. Get your steps in. What's your reaction to the results? Well, so one is obviously the guide for 1Q is lower than consensus. The big number that sticks out to me is the total expense guide of $114 to $119 billion.

That is almost $8 billion higher than the consensus. So what that tells you is the gross margins are going down.

and plus the depreciation costs, the cost that they are spending on data centers, they have to depreciate those servers, right? That's going to eat into the margin. So we are talking about at least a 4% to 5% margin impact for 2025 just from the data center cost, the depreciation cost. And so clearly there is a gross margin degradation.

happening here and the revenue growth is decelerating, which is what we are seeing in OneQ. The CapEx guide is what they gave, you know, on Friday, so no change in that. But clearly the data center expenses are eating into the margins. Which is something Gina Martin-Adams has said, watch with all of the companies reporting, but the tech guys in

as well in terms of margins and seeing some pressure on margins, which is what we're seeing. That didn't seem to be the case with Microsoft, though. So that's the surprise here, that Microsoft didn't call out that kind of a margin impact from depreciation expenses in their quarter, at least from the print that I saw. But clearly it is having an impact on Meta. When you dig into, well,

The CFO has been saying as well. And ultimately, she's talking about how the single largest driver of their expense growth that you speak to is going to be infrastructure costs. They say higher operating expenses, depreciation as well of those assets, of those data centers more broadly, of the chips they have to keep buying again. But then they say employee compensation, the second largest factor, as we add capital.

technical talent. - It's a good time to be that technical talent, right? - It is, it is. You can go and charge a pretty penny if you've got a nice computing degree. - So what is that, how do you read into that? Does that mean that the folks who are the engineers doing the AI stuff are in full demand? Who is that?

Areas of infrastructure, monetization. So interesting they are obviously monetizing their overall business. Reality labs, generative artificial intelligence, as well as regulation and compliance. They are the HR go-gets for this particular year. And they are actually overall saying that headcount has ticked higher, 10% year over year, despite this efficiency mode they've been in. There's also this note, the majority of 2025 CapEx to be directed to core business.

What's the definition of Meta's core business? Is it the family of apps, the Instagrams and the WhatsApp? And to say, okay, we're going to invest in that. So don't worry, investors. We're spending money on the parts of our business that make money. Is that the message there? I mean, Mandeep is your man to ask what's a core business, but I'd say Reality Labs isn't it, even though their company name is Meta?

Yeah, look, I think overall, they will be prudent when it comes to headcount growth. They may have to pay more for the AI talent. That's what they're alluding to. But at the same time, they are continuing to lose money on reality labs. I would have expected, you know, that would be your offset. If you see margin degradation, you offset it in terms of reality labs losses. You're not seeing that. So James Chokmok, this is still a company that...

You know, the message to investors is this is a company that makes money because we're opening Instagram and we're using Reels and we're using the family of apps, the core business, right? That's their message. They don't want people to get worried that this is another, you know, meta, a foray into the metaverse.

No, I mean, I think the metaverse continues to be a long-term vision for the company. But in this kind of market, you can't be looking at things on a moonshot. You have to be able to just focus on the quarter and focus on the year. And in some cases, even just focus on the week. So the duration that you evaluate these companies is becoming a lot shorter and shorter.

as the volatility increases given the increase in uncertainty.

So that being said, you know, for a company like Meta, yeah, that's exactly how I describe it and think about it. And those moonshots, you know, as a TBD, but not part of the calculus as it stands today. Hey, for those who care, NVIDIA is up about 2.1% here in the aftermarket. So we're seeing some movement there. Meta not giving a full year 2025 revenue outlook, Mandeep, is that significant?

No, I think based on the 1Q guide, you can see currency is a headwind. In fact, it's a 3% headwind, so quite significant. And the other thing I can point out in the print is 6% impressions growth. That's the weakest they have had. So 14% pricing growth.

That's a result of AI, so it's paying off in ad pricing. But 6% impressions growth. There are two reasons why it's decelerating. One is they're showing you less ads to make a better experience. Other is people are spending less time. I mean, it's still the, you know, they have 3 billion family users across their family of apps. Incrementally, that engagement growth

is slowing down because they're not showing those ads on what's over election. No, I'm just kidding. It could very well be, you know, a slew of factors, but engagement growth is how impressions grow. And so 6% is the weakest they've had for a while. Well, Caroline, our own Kurt Wagner pointing out that Meta's average price per ad increased by 14% year over year. It increased 10% year over year for the full year of 2024. Is it because

the AI investments they've made are making those ads so much more targeted

better targeted. Or whether people just know that if you're going to be committing any capital anywhere in terms of marketing, it's a very good bet to be doing it on the family of apps that Meta offers. It's interesting, we've had Micmac on the show, which is a company that tracks digital spend in particular and efficacy. And she came on and said, actually, Rachel came on and said the beginning of this year,

We have seen a big dive in the amount of money and marketing being committed to the likes of Meta's family of apps was her take. And the data which she was seeing in her theory was on the back of some of the politicization of Mark Zuckerberg of late. And indeed, therefore, without the

of some of the content and some of the fact checkers that are there anymore. People have started to shy away from committing their brands onto that company. Now, that is Micmac's perspective, and we'll have to see whether it's vindicated through what is said on this particular part of the business. But as you say, if there's a slight pulling back in impressions, but the average prices are managing to go up, they're managing to remain incredibly important to the brands that do remain committed and charge more for it.

Going back to Meta and the business, and I think it's interesting coming off of a Fed meeting today where certainly the Fed chair was asked about policies coming out of the administration, regulatory or others. A little warning from our

our live blog when it comes to Meta, Meta's CFO in the release saying, in addition, we continue to monitor an active regulatory landscape, including legal and regulatory headwinds in the EU and the U.S. that could significantly impact our business and our financial results. James Chokmok, come on back in. I mean, let's not forget there are things that are overhanging some of these big tech companies like a Meta.

Yes, that's true. But that's the case always. All of these companies constantly face regulatory scrutiny, both here and abroad. But I think with the change in administration, some of that may alleviate here.

at least the state side. We'll see what happens overseas. But overall, I mean, that's an overhang I think you constantly have to grapple with. So, you know, it's something that we're kind of immune to at this point. Caroline, is the regulatory landscape in the U.S. less worrisome for meta platforms now than it was a year ago, given Mark Zuckerberg's

relationship with Donald Trump, his visit to Mar-a-Lago, his trip to Washington, D.C., what he said on the Joe Rogan podcast. Great point. Dana White on the board. I mean, they've just settled $25 million for kicking Trump off their family of apps previously. So maybe we'll see some culmination of what that previously no love lost will currently end up being. I think people had felt actually the realignment of

of Zuckerberg, whether that's from a fiduciary duty perspective or whether that's just actually where he feels in terms of personal persuasion, have ultimately will bear fruit when it comes to doing business. Look, we've seen from Oracle, from OpenAI, from SoftBank, from you name it, Dubai wealth magnates and real estate experts

How do you do business in America now? Well, you go visit Mar-a-Lago. You pay some money to attend ultimately the inauguration and you see what deals can be done. And I think many would say that what Zuckerberg's been doing is in line with that. I think the regulatory landscape from an M&A perspective becomes different with a new FTC, new ultimate overseeing of what happens with the SEC. But I mean...

I'm sure there's still some... He's got some way to go to make up for some of the lines that were said last year of maybe he'll end up in jail because of Trump's irritation with past misdemeanors. I'm only going to mention some of them because there's a ton of earnings out here after the close. But just to rehash, Microsoft down about 1.6% in the aftermarket, certainly off its lows of the earlier trade right after earnings. Tesla now up about 3.4%. It had been down...

about 4% or 5% here. IBM has rallied about 8% after its earnings. We've been talking about Meta. It is up about 1%. And NVIDIA, which doesn't report until late February, is up almost 2% in today's trade here in the aftermarket. James Chokmok, it's been quite a week. You know, one of the things when I think about big tech in particular is

is the environment for rates. And we had a Fed decision, right? And it seems like Fed rate moves to the downside. Cuts seem to be putting off more and more this year here in 2025. How are you thinking about valuations and the rate environment and what that potentially means for some of these names? Or is that not as important versus maybe hearing about their AI spend or other things?

Evaluation is certainly a factor when it comes to rates. Now, I'd say big picture when it comes to rates, we do think we perhaps have a slightly more dovish stance than some of the market. Well, we do kind of agree with Powell that the rates are likely trending toward zero, not zero, toward 2% inflation rate.

We think that the market's currently expecting two rate cuts this year, June and December.

pretty probable uh scenario um so that's good um that being said the stocks that we're primarily focused on and we leaning into are the ones that have not only upside to estimates but also upside uh to their valuation multiple now across the big tax tech space uh there's not that many of those uh i'd say with the exception of meta really and to a certain extent google um

alphabet. The other ones are pretty much in, you know, Apple's up there, Microsoft's up there. You know, Tesla's certainly not cheap. So, you know, it's something that we have to be mindful of. We're there because, you know, you have to benchmark yourself to a certain extent on the top 10 names in the NASDAQ, but that's certainly not our oversized. And we've only got about a minute left here. Does TikTok matter to the revenue guide for the future of meta? And just quickly.

Yeah, I mean, I think TikTok matters to all of these companies. We'll see what happens, but it's certainly something to be mindful of, cognizant and follow extremely closely. If it does come into domestic hands, that could change the calculus to a certain extent. So I think meta will be fine regardless. TikTok is obviously already here, but what does that mean for the valuations and whatnot? So

My eye's full of it. Not worried about it right now, but definitely keeping an eye. All right. Well, we are keeping an eye on a lot of stuff, and we ain't done yet because we've got more stuff to come this week. Our thanks to our external guests and also, of course, our incredible in-house team. James Chokmok, thank you first. Partner and technology analyst at Clockwise Capital there in Miami. Caroline Hyde, you rock. Bye.

covering so much about the tech community. And be sure to check her out on Bloomberg Technology on Bloomberg TV at 11 a.m. Wall Street time. She is one of the co-hosts of Bloomberg Technology. Yeah. Just a big thank you to our Bloomberg Intelligence team, too. Both here in the studio today. Mandeep Singh covering all things meta and more. And Anurag Rana on Microsoft. Check out their notes. They're out now on the Bloomberg Terminal. And be sure to check out

all of the write-throughs on the stories and on the names that have been reporting here because they are moving in the aftermarket. That's going to do it for Tim and me. Have a good and safe evening. Ever notice your dog slowing down and having health issues and wonder, what can I do to make them better? Well, my friend, add Rough Greens to your dog's food for 90 days and I guarantee you'll see changes that will amaze you. Greetings, naturopathic doctor Dennis Black, inventor of Rough Greens here, and I invite you to give your pup the Rough Greens 90-Day Challenge.

In the first 30 days, you'll see shinier coats and increased energy. By day 60, your dog will have a stronger immune system, less shedding, improved joint function, all due to the live nutrients that you've added to their diet. And at 90 days, better digestion, reduced inflammation, improved heart health, and you may even have reduced their cancer risk.

Fetch your dog a free Jumpstart trial bag today. Go to roughgreens.com. Use promo code 90DAY. That's R-U-F-F greens dot com. Use promo code 90DAY. You just cover the shipping. You don't have to change your dog's food to improve your dog's health. Just add a scoop of Rough Greens.

Donald Trump has already changed the way we think about the US economy. Now he's back in the White House and Bloomberg's Trumponomics podcast is here to help. I'm Stephanie Flanders, Head of Government and Economics at Bloomberg. Whatever the big question of the week is, we'll have something interesting to tell you about it in a lively conversation with the reporters and analysts closest to the action. Listen to new episodes every Wednesday and follow Trumponomics on Apple Podcasts, Spotify or wherever you listen.