Amazon and Google's Mounting Cash Burn Doesn't Break the Bull Case

Welcome back to The Rundown, interview edition. Today, I am talking to Mark Mahaney, the Senior Managing Director and Head of Internet Research at Evercore ISI. It's been a rough week for tech stocks, so this was a great time to have Mark on. In today's conversation, we got into Amazon's earnings and why Mark is still bullish on the company. We also dig into what caused Google stock to have its biggest sell-off in history, and when we might see the hyperscalers start becoming free cash flow positive again. This was a really fun conversation. Mark has some very interesting takes. And I think you guys are going to really enjoy this. So let's get into it. All right, Mark, I appreciate you coming on the rundown today. It's been a hell of a week. There's a lot to talk about. So let's just dive right into it. It has been a busy week, no question about it. Like most earnings weeks, maybe there was a little bit greater than normal volatility. Well, yes, there was. You had a record correction in Google's stock off of print, at least dollar-wise. Somebody said it was percentage-wise, too. I'm a little surprised by that. I know it's been very early on. I remember seeing some big fall-offs in Google stock or gap-ups, too. On Prince, but maybe it was also the biggest percentage move off of Prince. So yes, unusual volatility for the sector with that one name. Not just that, we also got the Anthropic S1 finally to lead off the week. So none of us got any sleep this week. We're going to talk about all of it. I want to start with Amazon, though, because they reported earnings last night. And, you know, the numbers were pretty solid. I mean, AWS growth came in at 39%. That really stood out. It was the sixth straight quarter of acceleration. Thank you so much for your participation. So investors initially didn't really like it at all. I mean, the stock was down as much as 7% this morning. I think it's recovered some of that now. Amazon's one of your top picks. I'm really curious, like what was your take on their earnings and why do you think the reaction, especially initially, was so negative? Okay, Zaid, I refer to this as an expectations correction, not a fundamentals correction. And I look at it today, it's essentially trading in line with the market. So yeah, you know, like sometimes the aftermarket Things can get exaggerated in the aftermarket, both to the positive and the negative, because you've got a little bit of thin float and it's usually shoot first, ask questions later, or buy first and ask questions later. It works both ways. So there were three things here that the market didn't like. First, that AWS growth did accelerate from 37% to 39%. It was the sixth quarter in a row of acceleration. The backlog jumped 80 billion. It's the second biggest The biggest increase in AWS backlog. Ever. And the margins for AWS are really good, too. Like, they were similar to last quarter. There wasn't any profit problem. But the problem was the bogey in the market. And there were 3P alternative data providers that were telling investors. We're telling hedge funds that AWS is going to print 44%. So the market was sort of thinking, well, we're going to be in the low 40s. It was really kind of a lot of noise. I mean, sometimes I find these alternative data providers are very accurate, and sometimes they just aren't. Sometimes it's just a crapshoot. Anyway, this was a case of... People, you know, like there's, acceleration is acceleration. And my take on it is we still have an inflection point going on with this business in terms of accelerating growth. Heck, they even told it, they even, I mean, rare of rarities for Amazon said that they're going to be able to grow 40% for the next two years. I mean, Amazon never makes statements like that. And then they think they can get, make AWS a trillion dollar business in five to 10 years. Jassy was in his bag in the earnings call. He was in his bag. And this guy's super sharp. I've known him for a long time. I've tracked him for a long time. But anyway, that was the... So the first reason it traded off was the market wanted 44% or 43% or 42% and it got 39% sell. Second one is... The advertising revenue growth was only 12% year over year. Like it really sharply decelerated. They had to spend some time on the earnings call. And I think they kind of helped people figure out, well, you have Prime Day comps in here. Like it really helped Q2, it hurt Q3. And then there were a couple of other things. Q2 had some NBA, Q3 didn't. And, you know, that's an impact for Prime Video. So when they spent the time going through it, they were able to explain to people that actually the ad revenue growth was pretty consistent. Third thing is the cash flow, cash flow, they had record cash, free cash flow loss. of I forget now 30 or 40 billion in the street that thought me too it would have been about 10 billion but and when you dug into it it was all due to the timing of capex like they didn't change their capital expenditure plans for the year they didn't change those it's still going to be 220 billion so really what happened is everybody thought there'd be 50 billion in capital expenditures in Q3 well they were 75 billion which means you're going to have 25 billion less than you had thought in Q4 and don't tell anybody this, but That means they actually could be free cash flow positive in Q4. They're still going to go into a hole next year in terms of free cash flow, but it was a one-time timing issue with CapEx. So I looked at all that and I said, you want to sell this thing off 7%? Fine. You have a right to do that. I'm buying it. This is my number one pick in the space, and I think the most important thing is AWS growth is, you know, 40% at this base is incredible. That means they're going to be growing AWS revenue by $100 freaking billion in 2028. Highly profitable. This free cash flow hole is not going to be as deep as the market fears. Well, wow. I mean, where do I begin? Well, first of all, I think the market did listen to your note that you put out after the earnings because the dip obviously got bought up. I want to dive into the cash flow number a little bit. Like you said, it came in a lot more in the red than everybody thought. So what you're saying is Amazon spent a lot more money in Q3 building out their infrastructure than they initially thought, but they don't expect to spend even Even more in Q4, and therefore the... Cash flow might come in positive in Q4. Did they mention that at all in the earnings call? I might have missed that. No, they didn't. You didn't miss it. They didn't. Okay. No, but I'm sorry. They had told us before earlier. They raised their CapEx earlier in the year, the capital expenditure guidance, but they kept it. They told us the June quarter would be 220 billion. And that meant that they were going to have to spend something like that implied 125 billion in the second half of the year. Right. I mean... CapEx is always very lumpy, the spend, and so the market has said, oh, it's going to be $50 billion in Q3 and $75 billion in Q4. Well, it turned out to be the opposite for a variety of different reasons, and it doesn't really matter. What matters is that the full-year CapEx number is unchanged, and so the only reason they missed cash flow in the cash flow forecast in Q3 is that the CapEx spending got pulled in to Q3. So that means, you know, it's a sneaky call. call I'm making here, but I actually think they could be. It'll be a positive free cash flow quarter in Q4, but on a trailing 12-month basis. Look, this is a company that is committed to investing aggressively. They're willing to go negative in terms of free cash flow. They've done this historically, too, and I think they're doing it for the right reasons. I mean, they're investing in a business that's going to be doing 40% revenue growth at scale, and it can reach a trillion-dollar business. Yeah, you've got to invest to get those kind of returns, and it's And it's a very profitable business too. Yeah, I think I saw your note saying that you were modeling, what, over $330 billion in CapEx next year? Yeah. So I guess at some point, I mean, at some point, they have to show some return on all that spending. Thank And you think there's going to be a deceleration of CapEx, what, after '27? you so Uh, yeah, I want to be careful how we phrase this. So I think there'll be a deceleration, you know, much like in percentage terms in capex spend, probably in 27, the dollar amount is going to be big. Like, yeah. for If I get these numbers right, and this is rough, but, you know, having they spent about $110 billion on CapEx, capital expenditures, in 2025. me. It's going to be $220 billion in 2026. And my guess is that it's $320 billion in 2027. All they've said is that it's going to be meaningfully higher. I don't know what that means, but it's going to be meaningfully higher. But it's going to be something like that. Like, okay, if it was $400 billion, you know, I'd be surprised if it was that high. But it's possible. It's also possible it's $300 billion, but I think the market is bracing for like, yeah, that $330 billion. I mean, could you possibly invest that much? And the answer is, yes, you can, especially if you think that this is a business that's going to generate a trillion in revenue a year in five to 10 years. If you really believe that, and a profitable trillion in revenue, like they're running at 40% operating margins for this business. And so, yeah, you got to spend. They're going to get that kind of return, and I think they're really well positioned for AI, for cloud, and then this number, this backlog number, and I've forgotten the number now, but I think their backlog is now at, shoot, I think it's like $580 billion. Yeah, that's right. So they already got visibility for the next four years into almost, they have visibility into over a trillion in revenue already, already locked in. Not quite locked in, but mostly locked in for the next four years. That's more than anybody else has, and that's a really good position to be in. Yeah, I think I saw Jassy say something like, I don't know if he said on the earnings, not on the earnings call, but somewhere else where like AWS is sold out for the next two years or something like that. It's crazy. I want to move on from Amazon and talk about the big story from this week was the Google earnings. I mean, the stock got crushed on Thursday. And, you know, I mean, they also reported 100 plus percent growth. On Google Cloud, their free cash flow was also negative $9 billion. Is the market just freaking out about the free cash flow? Or was it because their search business came in under the expectations? Where do you see, what do you see was the reason for just the wild reaction of the market to Google? Zaid, I think it's the search business. Now, I'll also I'll also say that when. You know, when a company is going through this kind of investment cycle, mistakes get magnified. I understand that. And there's a mistake going on here. Like, search revenue growth decelerated much more. This wasn't an expectations correction. Like, you know, Amazon didn't accelerate enough. I mean, this was a case where the search revenue growth decelerated and decelerated by more than, you know, was in most models. And then they had to talk through the reasons why. So they mentioned currency. Okay. Tough comps, we know that. Then they mentioned some smaller issues, but that have been out there, like some of the Asian retailers have had regulatory challenges in Europe, so they've cut back from marketing in that region. It's actually a sizable number. And then there's some World Cup comps. The truth of the matter is YouTube got a lot of ad revenue off of World Cup, and so it just created a tough comp from Q2 to Q3. But then they had to talk about this issue of the monetization of AI searches. And so they go on their earnings call and they say they had the fastest query growth, search query growth in five years, but the revenue growth decelerated. So, you know, how does that happen? Well, what happens is AI, AI overviews, AI mode, kind of the AI search element of Google, that's becoming a bigger part of the pie. But they don't have as many ads in there. Or there's questions about, can you monetize that as well as traditional search? And I'm going to strongly go out on a limb here and say, I think they'll be able to monetize it just as well as traditional search, if not better. But it's going to take a while to prove that. And in the meantime, you've got decelerating search revenue growth in Q3. And then they warned about it again in Q4. And then the comp gets tougher again in Q1. So you probably have a couple of quarters of deceleration. And the bears on... The bear argument, if you got deceleration, if you're a bear and you have deceleration, you win. So the bears are going to win this argument for a while. So I think that Google is likely to be, barring some major new product news, is likely to be kind of range bound for a while. If you're willing to look more than a year out on Google, I think buying Google at 18 times earnings is a great long. Oh, yeah. The valuation is definitely attractive. But I mean, I want to go back to something that you just snuck in there. You think that Google will be able to monetize their AI answers at the same, I don't know what the right word is, efficiency or- Same rate or better. Same rate or better. Or better than 10 blue links, which is the greatest business model ever created in human history. Yes, I think so. So here's my pitch. You know, and I talk with a lot of ad agencies and ad buyers about this. And then I also just track the metrics at Google. And I've tried to talk with as many people in the industry as I can about this. The better the lead that Google generates for the marketer, I.e., the more likely that it converts, you know, marketers pay on a cost per click basis to Google. If that lead, you know, that when the click comes over and it goes from, you know, these conversion rates are always low, but, you know, at scale, they matter. If that conversion rate goes from five percent to six percent. That's an enormous, that's a much more valuable lead. And Google, the search is such a liquid market, like marketers immediately bid more for that. And so what I think what happens is, and just, you know, this is a logic test for you. And I may be wrong on this, but like when you do that 10 blue link search and you do, you know, vacations in, you know, vacation packages in Hawaii, you know, that's not that qualified of a lead. If I go through Google AI overviews and I say, yes, April, the second week of April, role in 2027, family of five, including a toddler, you know, we want a family-friendly resort, like, you know, you go through that AI search mode and you're getting more and more qualified so that when I finally, when Google finally sends me over. to Hilton or to Booking.com or to American Airlines, I mean, that lead is much more likely to convert and the marketers will pay more for it. So that's my simple explanation of what I call agentic intent. And I think that you give Google more intent, they'll give marketers more qualified leads that convert better. Marketers bid more for them. It's a real virtuous cycle. And I think it's going to start showing up and accelerating search revenue growth again in the middle of next year. That's my pitch. That's a, I mean, that's a great pitch. Um, I hope you're right. Uh, but I, but it is interesting though, like how everyone was talking, like Google was dead two years ago, right? Oh, chat GPT is going to crush them. Search volumes, like you said, have just never been higher, but now we're finally starting to see some of the impact of kind of moving over from 10 blue links to AI search. And I guess now they're kind of going through that rough patch right now, and it's going to take them a few quarters to get out of that rough patch and get to the other side. And it's happening at the same time where they're dropping. You know, 50 plus billion dollars on CapEx every quarter. And I guess that's what it's like a perfect storm and markets freaking out. Yeah, I'm sorry. You've got you got revenue growth deceleration. You got free cash flow going from positive to negative. What could go wrong? That's wrong. And I don't mean to, you know, like I take this stuff very seriously. It's just that that was the setup. And man, I tell you, these the free cash flow dynamics have changed so much. I went back and just check these numbers at the beginning of this year. We in the street were assuming, or the street was assuming, that Meta, Amazon, and Google were going to generate $200 billion in positive free cash flow in 2027. Now, we are estimating $100 billion in negative free cash flow. Like, what a shift. And, you know, the market's question is… Well, you know, you better have, when can I, when do I know whether I'm going to get an ROI, you know, ROI, you know, ROIC, ROI, return on that AI spend. I think there's some good data points that you're getting it, especially if you look at their cloud businesses. But, you know, I understand it's the biggest investment cycle we've ever seen. Do we really know it's going to work? At some point, we're going to have that moment where the light switch is going to go off and you're going to get this re-rating opportunity. The stocks will gap up. And my guess is you're going to get that in the next 12 months for all three of those names. Well, I do find it also very funny how like every quarter. All the executives, they're not shy about it. They're like, yeah, we're going to spend more money. We're going to spend more money. We're going to spend more money. They've been saying it for the two plus, three plus years now. And for the most part, everyone was okay with it until now. And I think it's because free cash flow has just gone negative and debt has gone up. And now everyone's like, okay, you said you were going to spend more money and you did, but now you're hurting stock buybacks. Now you're hurting free cash flow. Now it's breaking my models. I don't like this. Yeah, I mean, yeah, those three hyperscalers, you can throw Microsoft in there, too, were wonderful cash flow generating machines, high margin cash flow generating machines that were generally known as being light on CapEx. I mean, they didn't have any big distributions except for Amazon. They didn't have, you know, massive physical footprints. Well, Well, that all changed. Some people refer to it as the industrialization of the Internet. You know, maybe. But I mean, it's a it's a it's a reasonable metaphor. I mean, the question you got to ask Ask yourself, is there really going to be a dramatic economic impact from AI? Is there going to be a lot of revenue and profit potential coming out from this AI technology shift? That's the action question you have to have a comfortable answer to. I'm in that camp. I think this is the biggest technology shift. This is bigger than cloud. I mean, it's growing. It's based on cloud. It's bigger than mobile, but it's based on mobile. Like it all, you know, these accumulate up to where we are in AI. I mean, this is to me, you know, like the telecom AI boom at the end of the 90s. And boy, that led to a lot of dislocation. But one big difference, and I'm going to insist on this, one big difference is I was there back then, and there were companies building out infrastructure without demand. It was called dark fiber, laying dark fiber. There's no dark fiber here. Like every single bit of compute that's being built now is being, there's a buyer for it. So that wasn't the case back then. I mean, I think there really is a there there now. And, you know, like I think the average, you know, retailer can, the average investor. Retail investor can see this. You use Google. You tell me that Google hasn't improved, that search product hasn't improved dramatically in the last three years. Oh yeah, AI search is great now. Yeah, and by the way, thank you, ChatGPT. ChatGPT got Google out of prison. I'm exaggerating, but that's how Google essentially won the DOJ case because the judge said there's obviously great competition here. And ChatGPT sort of kicked Google and got them to... dramatically improve their search product than they did and that's why you improve the search product people use it more and it became better for marketers too there's a real another virtuous cycle there. Do you think there's like one more chapter left when it comes to like the funding story of all right so it started off with with cash flow yeah then it started off with uh well then then I guess it's debt and now it's stock issuance right so like Google did a massive Over the summer, Meta is doing it now. What, uh, you have, I mean, like you have IPOs, like what's, what's next? Like, Oh, is, is everything. Is there a limit to where things get tapped out, where then all of a sudden the bond market's like, hey, if you want to borrow $50 billion, this is the spread, and it's much higher than it's ever been. I mean, I'm just really curious to see if it ever gets there. That's also part of this massive ----. I mean, the free cash flow story's changed so much, Zade. You captured it right. I'm sorry, these were companies, and I'm just going to focus on these three, Amazon, Google, and Meta, that were gushing cash, or at least Google and Meta were, And Amazon, you know, It was always a little thinner on the cash flow, just different business model. But they were generally, and now they're going deep into free cash flow holes and they're going to the markets for liquidity. So that's a dramatic change. And of these three, I mean, Meta, Amazon, Google all traded off on their earnings, but for dramatically different reasons. And for Meta, it wasn't the results, it was that they announced that they were going to, we found out they were going to go to market and raise, whatever, $95 billion. $98. $98, $98 billion. And so, you know, that dilutive event, selling the stock. But, you know, they needed to do that. I think Meta, I'm going to give you a slightly contrarian view on this. I think that story is now in the past because... Google got ahead of this and raised $160 billion over the last 12 months. Amazon got ahead of this and raised $100 billion. They don't do equity. Amazon hasn't raised any equity since 1995, i.e. their IPO. They just do debt. That's just a cultural thing at the company, and I respect them for that. So I'd be shocked if they did an equity offering. They haven't done it in the last 30 years, so I don't know why they would do it now. And they can tap the debt markets. Well, they already did. 100 billion. So I don't think either of those companies, Google or Amazon, needs to come back to the markets. And I think that I assume that the people in finance at Meta are smart enough, I know they are, that they don't raise 98 billion with the intention, oh, we'll have to come back for 50 billion next week. You do this and you plan it out and you match it against what your CapEx plans are so that you get through this cycle and not have to come back to the markets because everybody knows the cost of debt. Everybody thinks that the cost of debt is going And it may well. So anyway, I that's the setup on the stock. So and that's what creates these interesting dislocated opportunities, because I think these companies, you know, was that Musk expression, you know, fully funded or whatever. You know, things could always there could always be bumps in the night and some surprises. But I don't think those three companies need to tap the markets after these rounds that they've just finished. And so now what we're going to wait is, does the revenue growth keep accelerating? And then when when can we say not that we're going to immediately turn to positive free cash flow, but when does when does it get less worse? When is the free cash flow hole or the number start to slightly improve? And then we get the path to back to positive free cash flow. And in three years, four years, we're going to be getting looking at these very large free cash flow generating businesses again. And it's that. There's going to be that narrative shift is coming on the stocks. It's just timing it is a very hard thing to do. Absolutely. That's very well said. I want to end with the Anthropic IPO. The S1 came out on Monday. Like I said, none of us got any sleep this week. So S1 came out on Monday. I think a lot of people were surprised with how decent the numbers were. I think people were expecting a disaster. But Anthropic is sitting on a ton of cash. They showed that serving their existing models, was that 85% margins? I mean, they had some interesting things to say in the S1. What was your take? And how do you think the market's going to react to the IPO next month? So I'm going to punt on some of that. I don't have any formal opinion on the – On the Anthropic as an investment, and I don't know, I have no idea what the, and I'm not going to opine on how the market's going to react to it, but do a note on some of the things that were relevant to the other companies that we cover, because... You've got customer concentration risk potentially for Amazon and Google because they're major suppliers to Anthropic, and they also own equity stakes in Anthropic. So I look at it that way. And the one thing, like here are the facts. The company was free cash flow positive in the first half of this last year. And we also only got numbers through the... June quarter, and it's already, you know, it's reasonably public what their September quarter numbers were, which was a deceleration. And then the question is going to be, well, you know, we should probably have, my guess is that we'll have some October numbers before the month of October. This is how finite this is before the actual offering. I may be wrong on that. But anyway, I think people are going to look at the... Yeah. The revenue growth is, I don't know what it was over a thousand percent in the first half of the year. You just don't see that too often. That's one of the fastest revenue growth rates I've ever seen. So that's factually true statement. And the fact that they are, you know, those gross margins, that's factually true. Oh, I thought one of the most, and it's factually, it's in the, in the document, there's no revenue concentration risk, at least for Anthropic. There's no one customer that's more than 10%. Like that's a good thing. The more diverse. Diverse the revenue bases, the more, well, revenue diversification is a good thing. So they gave you revenue diversification, positive free cash flow, improving unit economics. There's a picture in that S-1 where they show revenue per- It's a great picture, great diagram. And cost per compute. And you just saw how the cost versus the revenue has materially improved over the last six quarters. Now investors have to answer for themselves the question of whether that's going to continue or not. But if it does, and they follow the laws of scale, Here we go. IE margins go up as, as revenue scale. If they follow those, those laws, then one, one, if they do that, then you know, one can have decent confidence in future economics. And that's about it. Yeah. Well, look, it's going to be, it's going to be a fun time. The next, the end of the year is going to be fun. We're not getting a break here. So Mark, I appreciate your time, man. This was great. I really appreciate the insights or contrarian takes and everything. So. Thanks for coming on and we'd love to have you back on soon. Yeah, great talking with you, Sadie. Take care. Appreciate it, Mark. Well, all right, guys, hope you enjoyed that conversation with Mark Mahaney. I really enjoyed talking to Mark. It was just fun having this back and forth conversation and kind of digging into some of the details from the earnings report. My favorite nugget might have been that Mark thinks that Amazon might report positive free cash flow in Q4 of this year. Let me know what you guys thought about the conversation. Are you bullish on the hyperscalers as Mark is? Has the sell-off this week changed your mind at all? Drop your thoughts on Spotify Spotify and YouTube. And while you're at it, consider giving us a five-star rating as well. You know, all that engagement really does help us out and it helps other people find the show. Thank you guys so much for listening, watching, and commenting. Shout out to Mike for all the work behind the scenes. And we'll see you guys back here tomorrow.