Instant Reaction: Microsoft, AMD Earnings

Published Jul 30, 2024, 9:10 PM

Bloomberg Businessweek breaks down earnings from Microsoft and AMD with Senior Portfolio Manager at Synovus Trust Dan Morgan and Bloomberg Technology Co-Host Ed Ludlow.

Hosts: Carol Massar and Tim Stenovec 

Bloomberg Audio Studios, Podcasts, radio News. This is Bloomberg BusinessWeek inside from the reporters and editors who bring you America's most trusted business magazine, plus global business, finance and tech news as it happens. Bloomberg Business Week with Carol Messer and Tim Steneveek on Bloomberg Radio.

Okay, the call has not started yet, Carol. Investors are still taking note of the numbers that we got from Microsoft. Shares down six point eight percent in the after hours. The company reported Microsoft Cloud revenue for the fourth quarter that met the average analyst estimate capex though coming into at thirteen point eighty seven billion, that was higher than analysts wanted to see at thirteen point twenty seven billion dollars.

Yeah, kind of interesting, So let's get to it. Dan Morgan is with a senior portfolio manager at Sonvs Trust Company about twenty billion in acids under management, joining us once again from Atlanta. Hey, Dan, so good to have you back with us. Having said that, yeah, Hey, it feels like all of the numbers are kind of right where the street was expecting. We were talking ahead of it that we needed to see some outperformance in order for investors to be impressed. Are you impressed with these results from Microsoft?

Well, you're right, Carol. You know, the two numbers that really stay out stand out to me. You've already mentioned these as terms of ASER growth only up twenty nine percent. They had guided on the previous third quarter that they would hit growth between thirty and thirty two. The street was actually looking for something north of that. And then Tim, you mentioned that Campex number coming in a little bit higher than expected. Of course, there's some rollover going on right now from the Alphabet report that we came out earlier, when it really raised concerns about, you know, how much gain are you getting in regards to your AI initiatives versus the amount of CAPEX that you're spending. And I was kind of hoping, Carol and Tim that this Microsoft report might kind of dampen those worries and just kind of blow numbers out, you know, ASER up thirty four percent and so forth, But unfortunately it was a little bit shy. So I think that conversation is going to have to come back in in these concerns that we saw after the Alphabet report are going to reignite again.

Yeah, Shares down six point six percent. Right now, help us help us understand exactly, Dan, where we should be seeing the results of Microsoft's investments in AI and in open AI. Where should we be looking in this earnings report to see benefits from AI.

Well, foremost is in the data center, right, Tim and Carroll, That's been the area where we've seen the biggest amount of spending. That's going to be out of their ASER group, which is encompassed in the umbrella of Commercial Cloud and Intelligent Cloud. Those are the two segments that report. Of course, we don't get the hard numbers. We just get a growth rate from Microsoft. That's what they disclose to us. But if you look at intelligent Cloud and Commercial Cloud, they both looks like they both kind of keen in line. I noticed the cloud segment was a little bit shy three point eight to zero versus eight four, which I think was the estimate. The other area that we'd like to get a little more clarity on, Tim M and Carroll is what's going on with copilot. That's kind of buried in a lot of different areas, but it's hard to distinguish it. But we're always hoping that Microsoft, maybe on the conference call, will come out and basically release how many seats they've been able to license on that product, because that's another one of their core AI products. We all know that they ink that deal back a year and a half ago with chat GPT, But that's kind of hard to really figure where that comes into the earnings reports. But you know, the key areas is ASER growth, commercial cloud, intelligent cloud, and then any additional information that we can get on copilot in terms of additional disclosure. So again, guys, Tim and Carroll, it's very difficult, right to really distinguish, you know, we look at these segment data points that came in on this, you know, headlines coming through. It is very hard to distinguish exactly where all this capex is impacting in regards to AI initiatives.

So what do you Okay, So wait, you just threw a lot at us. So what is the question to Microsoft in the aftermarket or on the analyst call Dan to kind of get more clarity to try and figure out if the AI spend is maybe not living up to expectations or it is slowing down.

Right, So if we look back, for example, Carol and Tim and we look at the last quarter they came out and specifically said that ASER was positively benefited about seven hundred basis points compared to six center basis points in the second quarter in terms of increased growth coming out of the data centered unit, which is huge for them. That's huge recipe for their overall growth. So obviously, Carroll Tim, that's a number that everyone's going to be looking at when they publicly disclose of the twenty nine percent growth they did in ASER. What percentage of that or what boost did you get from AI and the past they've disclosed that, So that's a really key number that everyone's going to be queuing in on the conference call.

What's the question that you have for Satinadella?

Well, I think you know, again, you know, just trying to weigh out CAPEX expense. You know, where is it going? I mean, let's be let's be honest, Carroll Tim. I mean, you're looking at a company it's going to probably spend about fifty billion dollars in campex, So everybody wants to know what kind of ROI are you getting from that? And at this point we really haven't seen a lot of ROI from anybody, So again it's just them laying out that roadmap in and again going into what they're doing and their hopes in terms of how this will come to fruition and pay off for them. Yes, I think that's what everyone's going to be really focusing on Tim and Carroll after that Alphabet report.

I mean, how, how does this response from investors explain how investors are pricing companies right now? Because Microsoft reported a slight four tenths of one percent beat on EPs and then a small one tenth of one percent miss on cloud revenue, and Microsoft shares are just getting punished after hours.

Yeah, you know, Tim, the stock really traded up anticipation of the reporter. Is actually trading at forty four times earnings, about thirty eight thirty nine times Physical year twenty twenty four estimates the average pe on Microsoft over the last five years is about thirty two times earning. So these stocks, including Alphabet, Meta is going to report here in the next day or two. We've got Apple and Amazon all are trading at the very high bandwidth of their multiples because of excitement surrounding AI. So it doesn't surprise me Tim and Carroll that unless this report is perfect, it's going to sell off. And there was enough in here. As you mentioned Tim, beyond just the top line and bottom line beat, you mentioned the commercial cloud unit. Kind of a miss on aser that's going to lead to a selloff on the stock.

Hey, I do want to mention too, AMD just crossing the Bloomberg terminal. Second quarter adjusted EPs sixty nine cents is share that's a penny better than what the street was expecting. Second quarter revenue five point eight billion. The estimate on the street is five point seventy three billion. Second quarter capex one hundred and fifty four million versus an estimate of one hundred and twenty seven point one million, So that is coming in higher than forecast second quarter just an operating margin twenty two percent, that's better than the street estimate of twenty one point eight percent and giving some guidance in terms of the outlook. Third quarter revenue six point four to seven billion dollars, and that compares with an estimate TIM of six point sixty two billion.

Yeah, we're seeing shares move higher sharply now in the after hours as traders look at what's going on. Up four point six percent carroll for AMD right now in the green ceas third quarter adjusted gross margin about fifty three point five percent. That came in below estimates of fifty three point eight percent. Dan, I know you're just getting these numbers as soon as we get them, but do you have any thoughts on what we're hearing from AMD?

You know, Tim and Carroll, the big number that we're waiting for AMD was their data center segment. The expectations for that group was to grow over one hundred percent. Everybody was worried Tim and Carroll coming in his report that we might get some sort of color from Microsoft that they might have been backing off in terms of purchasing AMD's AI chips, which are the MI I three hundred x I three hundred A. That's what the big concerns were coming into this report. But at least based on what you've told me so far, it appears that there's been no aberration in terms of demand for those products. Yeah, they said extremely welcome record.

They beat so record. Data center up one hundred and fifteen percent year over year, So that number coming in high.

Two point eight three billion versus one point three two billion year over year. So all right, step to watch, Dan Morgan, thank you so much. AMD A different story.

Yeah, I AMD's up by two point six percent right now. The company reported revenue of six point four to seven billion dollars versus estimates of six point six two billion dollars. Adjusted earnings per share came in just above estimates at sixty nine cents, and then data center revenue coming in at two point eight three billion versus estimates of two point seven five billion dollars. With us now to break down these numbers and more, we got Ed Ludlow, co host of Bloomberg Technology on Bloomberg TV. He joins us here in the Bloomberg Interactive Broker studio. We were just talking to Dan Morgan over at Sonovas he said data center revenue, they wanted to see growth over one hundred percent. It came in at one hundred and fifteen percent. Is that what has investors so excited?

Yes, And to put it in Layman's terms, it more than doubled year on year, and it shows sort of the evolution of AMD from a standing start to taking a tiny little bite out of the market share that Nvidia has for AI accelerators. We actually didn't learn that much in that release.

You know.

The expectation was that data center revenue would grow more than one hundred percent. You're in year. You're going to have to wait for the call because the number we want is the full year forecast for their AI accelerator family, the mi I three hundred and with that will know if things are going better than expected in their efforts against in video or just as we thought they were.

Any questions that you've got besides that for AMD.

Yeah, Look, the market for high performance GPUs or AI accelerators that we train large language models on looks as follows in. Nvidia will probably do one hundred billion dollars of sales in that market this year to AMD four billion, maybe a little more. We'll find out Intel five hundred million. So the real question for me is what's different between the three of them. You know, they all talk about how they benchmark against one another in terms of energy performance and their ability to process data. But the real question for investors is, like, if we're going to believe that you can grow in this market against Nvidia, in which cases are you? Why would a big enterprise company or a META, for example, choose your A accelerator over another. Meta's very very solidly behind in video, right, it's buying hundreds of thousands of GPUs from them, And I think that's really interesting that you know, what do all these things get used for? Because right now it seems like in video seems the de facto choice for large language model training.

What do all these things get used for? Because the reason I'm asking this question is because last week we saw alphabet cap x come in above estimates. We saw it come above estimates today with Microsoft. A lot of that has to do with the buildout of infrastructure around AI. What is this getting used for and when our investor is going to start to see it show up in the bottom line.

This is a really worthwhile conversation to have, even in the context of earnings where you kind of focus on the numbers. When we talk about chips used for AI, we're not talking about sort of a bagg of chips you throw over your shoulder and look, I've got some. The reality is they go into many, many racks of servers, and those servers get lined up into massive data centers. And while we don't have a sense yet of how much money has been made through services or software and AI, not really. I mean, look at Microsoft's earnings as an example. We'll get to it. I think people need to understand that the workloads have increased massively. The training of the models requires an enormous amount of compute, and so Microsoft Aws, Amazon, Google Cloud Platform have no choice but to buy these chips because they work for that purpose and build more data centers and the nervousnesses will hold on. Whereas all the money going on the other side of the big moneyment.

Sages, it's kind of like, build it and they will come, Well, will you know, are they coming? Will they continue to come and will payoff? Be there? So I mean, I feel like that's still the big million, tillion dollar crush.

So let's jump into Microsoft, right. They very narrowly missed on their cloud services revenue or growth in cloud services revenue. I'm doing the math, but basically they had fifteen percent overall top line growth from just below thirty percent growth in the Azure unit. And the headline on the Bloomberg terminal eight points of contribution to as your revenue growth came from artificial intelligence. We readheaded it, so it must be important.

Eight hundred basis point. Dan Morgan said, what was seven hundred bass points as your increased growth for the data center group? So he was looking for, I think a number that was comparable or above that. So that seems impressive.

And sequentially it's an improvement from the prior quarter, but it doesn't really tell us anything. And this is what I mean by the money machine. So capex is the money going in. It's the number of dollars Microsoft spends, right, But what we don't have a good clear Layman's explanation is how many dollars are coming out the other side. You'd hope it was several more dollars for every dollar you put in. Cloud growth is important, but AI contribution can mean many things. It can mean running workloads, or storing data or doing all kinds of things. It's not evidence of sales of a generative AI tool.

Can it be better placed advertisements against Instagram reels ed too?

So this is why I think a lot of people are very interested in meta because we've talked a lot about how in the first half of this year the majority of the game on the SMP five hundred is attributable to those mag seven games. The idea that we believe the infrastructure will pay off in top line growth, in the EPs growth, even if they have stretched valuations. A lot of investors I speak to see a point of difference in meta. They're making all the same infrastructure investments. They're a developer of large language models with open source, but they have social platforms where you can believe more near term that you drive growth in your existing advertising based business because you're using smart artificial intelligence technology. There are many investors I speak to that believe that Microsoft doesn't have that. I mean, with respect, you guys probably have a longer standing relationship with Microsoft than I do. You know, in the early nineties when I started using computer, that's what was there. They've always been good at selling software. They still do that. It's just it's not clear to us which the AI part is. What they've really done is just make their existing software offering is a little bit better.

I've heard good things about copilotes in the sense of on Microsoft teams. I mean, I think Carroly, you did this during one of your prep calls for some of the conferences that you've gone to, you know, instead of taking notes during that, yeah, you know, copilot will actually spit out some pretty solid notes. Take the notes with And I've heard from some folks in financial firms who use it ad for getting actionable items return to them after a car.

Yeah, it's a good way of boosting one's efficiency or aggregating information very quickly. But at the same how much does that move a financial thinking yeah, And what we're talking about here, I think the show me the money conversation. It's not a consumer like me at my PC thinking wow, AI is wonderful. These are enterprise customers and you want to see they're committing big contracts to deck out their ten thousand person company with Microsoft Copilot. That's where the money is to be made.

I'm just thinking about the conversation we had was at last week about workers saying, Okay, all the bosses said AI, our generative AI was going to help me be more productive, And in fact, it's stressing people out. It's stressing people out, and they're not necessarily finding those productivity gains. Are we ed? You know, when I think about technological cycles, is it just going to take some time for this to kind of work its way through in terms of jen AI really making an impact on how we work and creating those productivity gains.

I mean in the near term. That's why Microsoft's down seven percent and after hours because investors had hoped it would be now and it's not. It is probably it's.

Still many months or years ago wish, right.

But you also have to say that thirty percent top line growth year on year for the second quarter in a row of azure is pretty good, you know, in any other sector industry you go, Wow, what an impressive set of earnings, right, even if you slightly miss estimates. There's something. I'm trying to tie all of this together like it's one quarter print in a big newsweek, but I think that it's very telling the comments of Mark Zuckerberg and Jensen One yesterday on stage in Denver that right now, AI is for the knowledge worker. It's for a computer scientist who has to do it for their job every day. And the vast majority of businesses, even if they know they have to invest in it, haven't quite worked out what they're supposed to.

Do with it.

But that's not what the promise of AI is going to be.

Now.

Promise of AI is, you know, for lack of a better term, getting rid of customer support agents who you call because it's so expensive for a company to pay them that you know, the idea is they'll create programs that will do their work for them.

Yes.

So like one example is a startup called in the Van. I just pluck it out of my head because I saw a billboard I drove past it the other day. Think about like if you have some travel disruption, or you are providing technology to a like package holiday website, and you go, my flight's cancel. Get me on another. The vision is that the agent has enough intelligence to say there's a flight at four, I'll move you on to it our refund. I'm doing this of my own volition based on the data available to me, and it's getting there.

Yeah.

But also the consuming's willingness to use that is important. Like I'm not using AI to write my scripts for Bloomberg Technology every morning because I still feel that please please keep me in my job as a journalist. I'm better suited.

We've tried it. It's track right, So maybe I'm a bad prompt engineer. Maybe that's the problem, but we have not been able to pull it off yet.

Ed.

Yeah, although we talk about it. If there is a point where it is actually really good in terms of writing introductions, how great it would be to free us up to do things like booking guests and doing things like that or researcher what have you, kind of move it to a higher level. All right, so alphabet, yeah, Microsoft, So now we move on to what Meta. We're going to move on to Amazon. What are you thinking about the rest of the space.

So the thing about alphabet is it goes first and you learn a lot about the week that's to come. But the big takeaway from that earnings. Even though we focused on Capex, which you've been talking your guests about for forty eight hours, and we focused on cloud growth, the most important takeaway was it was their search business that was the biggest contributors to growth. And they didn't expressly say that it was AI that did that. They just said it was our search business. They old fashioned bread and butter search business. And so when you look ahead to a Meta or an Amazon, you're going to get AWS cloud growth from Amazon, You're going to get Meta talking up at Capex and it's AI agents. But it's very likely that they'll also say growth came from one of our many other businesses. That's why they have some differentiation. But that might also be a cause for disappointment.

Do the increasing visibility of AI agents within what's Happened within Instagram? Does that do anything for the bottom line?

They hope?

So.

I mean, I don't know how many of Meta's platforms you guys are on. On most of them, yeah, I like Instagram What's Happened? I use the Meta AI quite a lot in the search function. Right, it's basically another chat GPT for want of a better comparison Lama, Yes, exactly, it's better AI built on Lama. But the probably most interesting example is Facebook Marketplace, where people are buying and selling their own secondhand goods, right or dealing is a small business as an advertiser on the platform. I think that's where Meta really sees AI agents adding value. It just automates and makes more efficient that business transaction.

All right, everybody, that's a wrap. Ed Lo Love, of course, co host of Bloomberg Tech on Bloomberg TV.

Bloomberg Surveillance

The economy and the markets are "under surveillance" as we cover the latest in finance, economics an 
Social links
Follow podcast
Recent clips
Browse 3,632 clip(s)