Dan Niles is the founder of Niles Investment Management, a strategist for the tech sector. It's a pleasure after all these months to have you back.
Yeah, well, we talk all the time, but yes, it's nice to be back on your pod.
And boy, have I been thinking of you, just looking at the capex investments that have been announced of late. What are your first thoughts when you see capex investment going from $150 billion in 2023 to almost $700 billion this year?
Well, I mean, it's a question of who's getting return on investment on that and what is that doing to their cash flows, and it's very different for each of those companies, which we can get into. But yeah, I mean, before Oracle reported results in mid-December, my thoughts for capex for this year was around up 25%, you know, maybe to 30% is where it would end up for calendar 26 for the big five guys, the three hyperscalers plus Oracle and Meta. And then after all of those capex increases, which just finished up, it looks like the numbers are closer to up 60% capex growth versus up 25 to 30. So you go, 'Wow, those are big numbers.' And so then the next thought is, well, who's going to have still positive free cash flow? And then who had their estimates go up versus go down after that? And really that interplay determined how the stocks kind of reacted to their news. And so I think this year is going to be a very interesting year because I think, you know, the spread between the winners and the losers is just going to continue to grow.
How sustainable is that level of investment? And you already mentioned the hyperscalers, the stocks not necessarily always reacting positively. Many of these shares are now underperformers, Microsoft for example. Is the market starting to look at this kind of level as a capital risk to those companies?
Yes, and it should be because not everybody's going to win. I think this really started to change in late October of last year, I think. And you can go back to Oracle, right? Oracle in the middle of September put up those huge numbers and the stock was up 36% the next day. It was the biggest one-day percentage increase of any stock in history with over half a trillion in market cap. Then fast forward, the stock's down over 50%. Because people really started to question, well, how is OpenAI, which is doing run rate revenues of $20 billion exiting 2025, going to pay for 1.4 trillion in capital commitments, especially when their free cash flow is getting worse for the next several years? And you've seen that kind of concern spread to some of these other names where people are going, well, who really is in a good position in AI? Well, clearly Google is because they have a leading model in Gemini. But you look at Amazon, they don't really have their own frontier model, right? And even if you look at Microsoft, they rely on OpenAI, which we just talked about is a risk point, which is why Microsoft's underperforming. And then you look at the earnings that came out, and it's very different in terms of growth rates for all of these different companies. Microsoft kind of disappointed with their growth rate around Azure. It actually decelerated a bit. Meanwhile, Google's growth rate in Google Cloud grew 14% faster in the current quarter than the one before. And you know, and then you have Amazon kind of in between where their growth rate accelerated by a few percentage points. So, and that really kind of informed, between that and the capex numbers, how the stocks reacted coming out of earnings, because I think ultimately Google will be the winner in consumer AI, not OpenAI, because they can, they have a leading model, but they also have tremendous cash flow to fund their capex. They have end devices through the Android ecosystem and their own Google Pixel phones, and then obviously they have search, which is a great monetization engine. And so they really have the full stack from top to bottom where you don't have that with some of these other players.
You know, it's kind of funny to me because there used to be a time not that long ago, maybe a year, year and a half ago, where people were saying that, 'Oh my god, Google's lunch is going to be eaten by AI. Search is going to be a bloodbath. That's their source of income.' And now suddenly Google is emerging as the huge winner in the AI race. When did that switch happen? And looking at the stock now, great performer the last 12 months. When did that sentiment switch happen?
The sentiment switch happened when the Google antitrust remedies came out.
You mean the Google and Apple remedies basically, right?
Yeah, and basically the judge said no, no problem. And so there was no breakup that needed to happen with the company, etc. And if you look at the stock, it just started to rip right after that because all of a sudden Google could go on the offense instead of being on the defense. The other thing is the new Gemini models that came out are better than what ChatGPT had. And you can really go back a few years before that, which is when ChatGPT first came out at the end of 2022, Google was really like nowhere. They were still struggling with, you know, getting out models that had accurate information. They were more interested in the models being politically correct than returning correct answers. The two founders then came back to the company, Sergey and Larry, and started really working on, hey, we want to produce a model that's good and the best models. And that focus by Google again really started to get Gemini to catch up where now if you look at the leaderboards across all the different metrics people use to judge AI, Google in general is at the top of the leaderboard now, not ChatGPT. And so it's really that progression from the founders coming back, no antitrust real remedies, and then the model getting to the top of the leaderboard, and you put it all together and that's where we are today.
Now, I just want to be clear on this. OpenAI is going to introduce ads. By definition, Google's going to lose some share because OpenAI is going to introduce ads into ChatGPT, right? Because they're going from 0%, right, no ads, to some sort of ads. And so, by definition, you're going to pull something out of the Google ecosystem and the Meta ecosystem, right, the two big leaders in online advertising. So, it's going to be very interesting to see how both Google and Meta stocks respond when that does come out. And think of it kind of like when TikTok came out and the damage that caused to both Google and Meta, you could see a similar short-term reaction to those stocks. So, I just want to balance out, you know, those views right now because as we've seen recently, right, the market's moving very aggressively on news releases before we even get actual results. You would think it shouldn't matter because we all know the ads are coming out, but as we've seen recently, right, news releases can really move some of these stocks around. And Google's had a great run and the capex numbers have gone up a lot. You're, you know, they can still fund it, but your free cash flow is obviously dropping quite a bit based on the capex numbers moving a lot higher and the estimates really not changing very much.
Yet Google is so much more than Gemini. We have the Waymo story now, valuation at around $126 billion. You have potentially the IPO of SpaceX. Google is invested in SpaceX. If you look at Google outside of AI, what do you see looking at this company? What makes it attractive in your opinion?
You know, I think the biggest advantage they have, quite honestly, is the fact that they've got nine different products with over a billion users each, and you can use that to train your AI models, and all of this is their data, right? So, they can train on it for free because as consumers, we sort of give up our rights to have them train on it when we use these products for free. And so, I think that also helps a lot of different parts of their ecosystem. And they've done a really good job, as you said, of getting into EVs, having an exposure to the space build out as well through some of these other investments that they have, and pretty much the only big tech company that also owns their own power provider. So, they're also set up very well on that angle.
Very smart. Let's dive into this whole cash flow debate. Amazon is going to be pretty much free cash flow negative the next couple of years. And you know, looking back into the '90s, I remember before I go into this chart, I remember the times where people were saying, 'Oh my god, I mean, Amazon is so highly valued, they're not even profitable yet.' Amazon understood very early that it's not about profits, it's actually about cash flow. Now it seems like everybody is chiming in on the same idea that, well, profit is not necessarily what matters, long-term growth opportunities matter. They all seem to be following in this path. Well, going back to this chart, we see that Amazon is outspending their operating cash flow on capex, Meta and Google most of it, and only Microsoft still clears capex. At what point, Dan, do you think the market stops rewarding revenue growth and starts punishing tanking or even negative cash flow?
It's already been doing it for the last couple of years. I mean, you can go back to mid-2024, and if you bring up a chart of all those companies, you'll see from the middle of 2024 to kind of the end of 2024, the stocks were all struggling, including Nvidia. Because if you remember the June quarter of 2024, Amazon, Google, Microsoft all reported, they all took up their capex, and the estimates for all three went down a little bit, and the stocks all struggled, and they struggled for basically the back half of that year. And then inference came out in 2025 and really started to ramp, and that got these companies going again. And now you're seeing a return to what happened in kind of mid-2024, which is another big increase in capex from all these three players, but the estimates in general didn't go up, right? And for Amazon, they went down. So, and you saw the stocks react negatively off of that. So this has been going on really since the middle of 2024. But as you transition from training in 2024 to inference in 2025, that drove, you know, better performance. But now as you're moving into 2026 and you have this huge step function move up in capex spending from what we were thinking, which is 25 to 30% growth, now it's closer to 60% growth. And now you've got your free cash flow numbers getting revised lower. People rightfully are going, well, can all of these guys win? And especially when you see the difference in growth rates in their cloud businesses, that's a very good question to ask because you go, wait a minute, like Google's really accelerating their growth, and meanwhile a company like Microsoft, which has exposure to OpenAI, the growth's not really changing. And Amazon, you know, is somewhere in between. And so I think you are seeing that being questioned, and I think you're going to see it being increasingly questioned as we go through this year. That's why coming into this year, I thought it'd be a very volatile year. I honestly thought we'd have more of these problems near the end of the year because in the beginning of the year, my thought was, well, OpenAI is going to raise that hundred billion, Anthropic is going to raise their tens of billions, and people will get optimistic again. That clearly has not been the case to start off. And so it tells you that investors rightfully are getting more concerned in terms of who's going to ultimately win because not all of these companies can. I mean, that's just not the way it works in tech, right? If you think back to the tech bubble, how many winners did you have in e-commerce? Really one, Amazon. What about in social? Really one, Meta, you know. What about in streaming? Really one, Netflix. What about in audio? That's really Spotify. And so you've only had really one winner per vertical. And right now the market's obviously saying, 'Oh, you know, we've got five to ten companies that are all going to win,' that unless something is very different about this cycle versus the last, that's probably not the way it's going to work out.
So winner takes it all. That's also one of the reasons Uber is having issues in regards to Waymo eating market share in the robo taxi area, including Tesla, which is a whole debate to itself. But I want to stick to the AI story. Looking at OpenAI, Oracle, and Microsoft to a certain extent, Microsoft is, they're both a bit of a proxy for OpenAI. There's been a certain sentiment bounce. OpenAI is claiming that ChatGPT growth is now back to a monthly rate of over 10%, and Wall Street is hoping that with $100 billion they're raising in additional funds, that the financing of all those deals, including Oracle, are a bit safer. Plus you have $20 billion in stocks being placed, plus debt is being issued. What is your take on OpenAI, also taking the rise of Anthropic and Gemini into account?
Well, I think OpenAI's got a huge problem because for one simple reason: Google has trained five billion users on the planet that you can enter in a question and get an answer for free. OpenAI gets over 50% of their revenues from consumer. It's going to be very hard for them to convince people to pay, especially when you're getting answers from Google for free, and Google doesn't need to worry about that, right? So that's the first issue with OpenAI. The other issue is Anthropic, because Anthropic gets 75% of their revenues from corporate. Well, corporations are used to paying for things, right? So they're looking at Anthropic and saying, 'Of course, we've got to pay.' And Anthropic right now has the best coding models or assistants out there. And so corporations really like it. That's why Anthropic is growing so fast. And meanwhile, OpenAI is playing catch-up in that market. So OpenAI is really wedged between Google on the consumer side and Anthropic on the enterprise side. Meanwhile, their cash burn is getting higher each year for the next several years. That's why I think OpenAI has got a big issue. And Anthropic also, which I found interesting, and you may have seen the Super Bowl ad that they had on this, Anthropic has said, 'Hey, we're not going to put ads on our platform.' OpenAI has to put ads on their platform because they have to figure out a way to monetize the 95% of consumers that right now are not paying to use ChatGPT. And so it shows you the different strengths of these companies by what they're having to do, right, as well. And so that's why, yes, short term, do I think everybody will get all excited about, you know, OpenAI raising that money? That's what I thought coming into the year. Obviously, that's not the way it's played out the last month and change. And so that tells you something. Is there going to have to be a bounce because things got very oversold? Obviously, but it's never a straight line down or up. But that's how I'm kind of thinking about this. If I look out a couple of years from now in terms of OpenAI relative to their competitors out there...
