You probably still have, I think, another two or three years in this sort of AI bull market cycle before things get really insane. I know some people think things are already really insane. I think if you actually look at the numbers though, for the most part, a lot of these companies, they are supported by fundamentals. We still believe we are earlier than most investors would assess at this point that the probability that there is upside to some of these smaller companies because the numbers are smaller, the law of large numbers working in their favor when it comes to growth is higher. I think the downside to disbelieving in AI and having it play out if you're a mega cap is that the ones who believe and the ones who invest then win and you become relegated to probably not being a mega cap anymore. It won't be a race to the bottom that they'll actually hold up some of their pricing more than what I think people expect because at the end of the day if there's value that's being created we think there is a lot more value in thinking than there is in networking it won't be as steep of a price decline as what many are anticipating.
Any discussion about Deepwater or Intelligent Alpha products and strategies is not a solicitation to buy or sell any securities and past performance is no guarantee of future returns. Additional information on both firms can be found at their websites deepwatermgt.com and intelligentalpha.co.
Gene and Doug, welcome back to Access Returns.
Today we're going to work through the Deepwater 2026 predictions piece that you put out a little while ago. I believe Gene, this marks the 10th annual prediction piece from you. So, it's been a decade of kind of looking at the technology landscape and making these calls each year and sort of seeing how things play out. So, I always love these types of pieces and discussions. One, they're really interesting to talk through because there's so much change happening, but two, you know, you can always go back 12 months from now and we can see how much we got right, how much we got wrong.
Exactly. Yeah. And kind of realizing it kind of puts some humility into this business, I think, a lot of times.
Yeah. So the conversation today, I mean, we'll kind of, it'll give our audience a chance to sort of scour the tech landscape with you guys, two experts in the field. So you know a lot about AI, sort of the big capex buildout, Mag 7, and a bunch of other stuff. For those that haven't or for those that want to actually see the full piece you can go to the Gene site. It's genemonster.com. And then also, I'm a big fan of the research that Deepwater is consistently putting out and that's deepwatermanagement or MGMT abbreviated.com. And then lastly also want to support the Intelligent Alpha business and the ETF and strategies that you know Doug and his team are running there. So you know please go to these sites and support these guys. And the world's greatest ticker GPT.
Yeah, nice. That is an awesome ticker.
Generated with ChatGPT, right Doug?
That's right. Fully on brand.
So yeah. So before we get into Gene your prediction list and hear from Doug too, I wanted to go back and this is very interesting. So I wanted to go back to Doug a tweet that you put out at the end of 2023 and we'll put it up on the screen here, but let me just read it. So now keep in mind this is you know basically more than two years ago you tweeted or wrote my highest three to five year conviction idea is that AI will culminate in a bubble bigger than the dot boom. It's the nature of major tech innovations to create bubbles. AI isn't close to a peak. We're in 1995. So I wanted to hear from you kind of has anything changed in your view and where do you think we are now in this evolution with AI?
Not much has changed I think is the short answer in terms of just the structure of that thesis. So I still think we probably have, you know, we're two years in. We probably still have I think another two or three years in this sort of AI bull market cycle before things get really insane. I know some people think things are already really insane. I think if you actually look at the numbers though for the most part a lot of these companies they are supported by fundamentals very rapid growth. Obviously the data center buildouts have supported the chip side. I'm still kind of waiting in many cases on the software side for some of these companies to really apply AI. But yeah I still think we have two to three years of a pretty strong AI powered bull run left in markets.
Yeah. So, it's pretty amazing the amount of money that is kind of being plowed into this from every aspect of it. And so, yeah, that'll be, this is an interesting opportunity, I think, for investors. So it'll be interesting to see how this all plays out. All right. So, getting to the prediction piece. Let's kind of dive into this. And I'll just kind of throw these out there and then we'll riff on it a little bit. So, the first one is, and you know, I was thinking about this. So, the first one is the NASDAQ will end 2026 up more than 10% driven by the AI trade. And what I was thinking about is I think if the NASDAQ generated a 10% return, a lot of investors might be actually disappointed just given how strong that index has been over the last, you know, 5 to 10 years. The annualized returns are crazy. But just talk us through like how you think those conditions, the path that's needed for the NASDAQ to generate that return, how that might play out this year.
For starters, put the 10% into perspective. Understand that obviously we've had much more than that. If you look at the NASDAQ over the past year up 23%, year-to-date we're up a half a percent but as a general rule for a growth investor better 7%, 6 to 7% is probably your benchmark. So like in the world of investing, I understand we've been spoiled like a 100 basis point 1% growth is considered like a standard deviation in terms of performance. And so like that statement even though it may come across as feeling a little bit diluted, it's still a bold statement just especially given where we've been. And the substance of it is pretty straightforward. It's just what Doug talked about. From my perspective, Doug says two to three more years left before things get, what was the word, insane. I still think we're, I'm probably even a little bit more bullish, but it'd be splitting hairs. So whether it's two to three or three to four years left. And I think that that's just hard for grounded investors to get their heads around and present company included. And so if in fact that does happen, we can talk about what's going on on the fundamentals of these companies. But if in fact we still have another two to three to four years left, it's highly likely that the NASDAQ continues to have some strong double digit plus performance. If you think about in the simplest construct, I think right now the street's looking for the Q's to grow earnings somewhere mid-teens and the Q's trades around 26, 27 times forward earnings. And so if you just hold that multiple and that multiple, let's call it, it's a little bit elevated from historical sort of averages and norms. It's not excessively elevated. It's not even at levels that we saw back in the 2020 era. And so if you just kind of hold that multiple and you get mid-teens growth in earnings that sort of supports your plus 10% hurdle right there. I think and it's always important, I think, right now in the environment we're in to kind of address like what is the bear case or what is the skeptic's counterargument to that. I think a skeptic would probably point to it and say, okay, well, yeah, earnings are sort of inflated because they're not fully showing the effects of the capex buildout that depreciation is scaled there. And so, if you look at on a cash flow basis, things look a little bit tougher, but even on that basis, you're still going to have cash flows growing kind of probably high single digits if I had to take a guess. And again, you get maybe just a tiny bit of multiple expansion and you still get your 10% hurdle. So I think no matter how you slice it, the numbers are there fundamentally to support it.
One of the narratives that has kind of come more to the surface here is and to your point Doug that you know there's so much spending on this capex that some of these asset-like companies you know have moved maybe to they're more moving into like asset heavy and investors have kind of started to look at this a little bit more critically I think to some extent. But where I want to go there is, you know, if that skepticism is sort of there in the market, is that sort of like a healthy thing sort of at this stage of where we are?
I think it is. I think it keeps that sort of downward pressure, that realism in the market where not everybody is playing by the same playbook. And at Deepwater, we kind of referenced this before in public comments we've made, but we've looked at the market through this lens of kind of a five-part framework in terms of how far are we in a bubble, like how bubbly are we? And the bottom line is in that framework, we're not super bubbly. There's a couple of yellow lights, but there's no flashing red lights right now. If we look at some of the market mechanics in terms of leverage in the system is a big one, right? How many people are trading on margin? How much margin excitement is there? How correlated is everybody's thinking in terms of optimism in the market. Those are the things I think that are sort of held down by what you just described, Justin, people saying, you know, look, this is a wholesale shift in terms of some of these businesses. You need to start to think about them differently because they're not just these really high margin IP businesses anymore. So I think that does help us keep sanity. But it probably won't last forever. Like I said, I think we get a bubble eventually. So I think those skeptics will largely fade away in time and that's the nature of these bubbles is they just get beat up for so long when markets go up 10, 20, 30% a year eventually they just get drowned out and nobody hears them anymore.
So I think the second one here will be pleasure to many active managers' ears in the sense and the number two prediction is small cap tech will outperform the broad, you know, larger sort of tech trade in 2026. So what are your thoughts on this field?
There's three general pieces into it and predictions based more on numbers two and three not number one. I'll say number one is like a mean reversion that typically when you've had outperformance that we've had of the Mag 7 over the past couple years it's just natural that investors start to look for other pockets. I think that's not the justification that we have. Our belief is that, continue maybe starting just with that growth piece. If you just look at the numbers and the growth rate of the companies that are smaller cap because their revenue is typically smaller since we still believe we are earlier than most investors would assess at this point that the probability that there is upside to some of these smaller companies because the numbers are smaller, the law of large numbers working in their favor when it comes to growth is higher. So said another way is that our first prediction which is basically rooted in a belief that we still have two to four years left of this AI trade should, because of the law of large numbers, should have an outsized benefit to what's going on with some of these small cap companies. So, and we put small cap at sub 500 billion. I don't know in what kind of strange alien universe like that is a small cap, but that's kind of how we think about it. Not the Mag 7. There's another piece to it that we thought about last December and it seems to be playing itself more out here, but is this idea of rate cuts and it's becoming more clear that we are going to get rate cuts this year independent of maybe what's going on with inflation. The last reading just recently got is inflation is remaining relatively steady, stable in the high twos think around 2.7 but still higher than where the Fed has talked about. So under normal circumstances you'd say that rates would remain but of course there's a political dynamic to it that is likely going to cause rates to decline. And that comes back to this, this is like good old-fashioned mechanics of how stocks work, that declining risk-free rate benefits both large and small companies. I mean, both get a lift. I mean, inflationary periods typically good for equities, but the smaller ones tend to get more of a lift because they're historically more risky and because they're smaller businesses and so they tend to have a more, they tend to trade more on what goes on with inflation. So, I realize that second piece isn't really like a hardcore tech view, but I think it is something that is important and likely that we're going to get these rate cuts. And from our perspective, not a political statement, but I believe that we are going to see more inflation depending on the size of the rate cuts, but if we see that, that's usually good for equities overall.
So, you're basically thinking it's going to be non-Mag 7 tech will outperform Mag 7 tech when you make this. Is that basically the idea?
And it's interesting too because like last year I believe two of the Mag 7 outperformed the S&P 500. So you're starting to see this a little bit already. I think five of them underperformed last year.
Yeah. We have at Deepwater we have an ETF's been around since 2018. LUP last year up 42%. We're in 500 billion and less in that ETF. So we saw some of that last year. I mean there is and that's all tech and I think Mag 7 was probably up something an order of like 25% or I guess maybe have to pull Tesla out of that but it definitely we saw outperformance relative to the Mag 7 the small cap so kind of a continuation of that.
I know at Deepwater you guys are, you know, investing in both public and private companies, but just and this might be hopefully it's not too broad of a question, like when you're assessing a smaller cap tech company that clearly has different qualities than the mega cap tech companies do from a fundamental standpoint. You know, could you just spend a minute or two sort of describing how you go about analyzing the business from, you know, an opportunity risk standpoint? Do you look for moats? I mean, what are the things that you guys are looking for in small caps that might not necessarily be, well may not be one of them, but be there from large caps or mega caps.
I think the biggest thing if you think about the mega caps, Gene and I have spent our careers you know covering those stocks investing in those stocks the common trait sort of across the board with those is that they're all platforms right they enable a lot of the sub 500 billion dollar companies Gene just talked about small caps and tech as we like to call them they enable many of those companies I mean Uber is built on top of Apple and Google you think of that Airbnb in many ways you could argue the same hundreds and hundreds and thousands of different vendors are built on top of Amazon. And so when we think about these small cap companies, the question we like to ask is what would have to happen like what would have to be true for this company to become a mega cap company? Can they be a platform is usually what the question boils down to. And so as we think about those companies, we do invest obviously in public markets, but we also do a lot of investing in private markets. I think that that platform company is very important right now as we think about the AI era and some of the model builders because those are the ones that most obviously I think could become platform type companies. There's going to be a lot of businesses built on top of ChatGPT on top of Claude on top of Gemini and so you know that's, you need those companies that ultimately do prove to be platform companies because they can just compound year after year after year.
So number three, it gets into something people have been talking about a lot, which is this idea of all this AI capex spend. And I think people outside of the tech community are just saying regularly now, this has to slow down. This has to slow down. It can't keep going. But you're taking the opposite side with your prediction. You're saying hyperscaler AI capex growth will be over 50%. So can you talk about that?
Yeah. And this kind of comes back to just culturally Doug and I approach investing is to be grounded on what's going on and not take the bait when people are trying to like one up themselves and there's a group around AI and investing that's like this one-upsmanship. I just want to be clear like that's a pretty bold prediction that we're going to see greater than 50% capex growth. The reason that comes from is we believe that essentially within the mega caps at least they have to all kind of run capex at a similar pace as the one that's running the fastest and there is a piece around that which is kind of you know if you're not spending appropriately you may fall behind and that could have a very bad outcome for you and if you do end up cutting if you end up being right and end up not spending as much. The benefit isn't that great. So, you kind of need to run it whoever is running it the fastest. And right now, that happens to be Meta. Now, most of these companies haven't given guidance for calendar 26, but Meta said more about 26. And based on their expectations, it's going to be up just over 50% at least where it stands as of a few months ago is that 2026 the overall. So, if we're going to take the four that give capex guidance, that would be Amazon, Meta, Microsoft, and Google. Those it anticipates right now the street's looking for around just over 30%.
So, that's the delta. We're sitting ahead of them reporting December earnings at plus 30. We think that by the end of the year it will be plus 50 in part because of what's going on with Meta but also if we just look at what's going on with the broader buildout and look at what Jensen Huang and their CFO's comments were at the JP Morgan fireside at CES last week is that they had this 500 billion Blackwell and Rubin cumulative revenue target through the end of calendar 26 and they said at that event and that would basically imply around 50% capex growth that they said that they're going to exceed that number. Now some of that comes from China but in general they're saying they're going to exceed that target. So I think when you put all this together is like the two big indicators that investors look at is you know what's Nvidia saying and then separately what are the hyperscalers saying about capex and we think the commentary is going to continue to point to a measurably higher number than where current expectations are.
Do you think these companies still see this as a winner take all type situation? You're talking about they're all going to match whoever the biggest spender is? Like do they see this as a situation where they have to spend with everybody else or they risk losing this whole thing?
I think they do. We've kind of had this view where it's kind of a Pascal's wager, Blaise Pascal, the do you believe in God or not believe in God? And the example here is four box quadrant, right? Do you believe that AI is going to be huge or do you not believe that AI is going to be huge and then AI is huge or AI is not huge? I think the downside to disbelieving in AI and having it play out if you're a mega cap is that the ones who believe and the ones who invest then win and you become relegated to probably not being a mega cap anymore. You're probably one of those sub $500 billion small cap tech companies in that case. And so I think all these players realize that this is a moment where it is a potentially fundamentally earth-changing technology. And so if that does turn out to be true, the investment dollars we're talking about, they're substantial, but they will look minuscule. They will look irrelevant in 5, 10 years if the bull case does play out and AI is as big as some of us think it could be.
How do you guys think about this idea that the people doing the building will be the beneficiaries? You know, it's easy for people to point right now to something like the fiber buildout and say, well, all the companies that built that didn't do well, but the companies beneath them that took advantage of did really well. And I wonder if that's overly simplistic looking at this, like because these are obviously way better companies. They're less debt-ridden. They have huge cash flow. Like, how do you think about the beneficiaries here relative to the people building it and then the people that ultimately sit downstream from them?
So that's a pretty well traveled or that's the kind of the well-traveled view about why this isn't sustainable that this is profitless prosperity that when Sam Altman talks about token pricing going down a 100x per year I mean it's like just defines remarkable in terms of the amount of pricing pressure I think it all leads to this idea that these models will become commoditized and you know how can these companies ultimately have some form of return.
There is a dynamic where and this is, is it the Jevons paradox? A UK economist had a theory that when the price of coal in the late 1800s was dropping dramatically that the usage would go up more than enough to offset the decline in price and it would be ultimately good for coal producers. In fact, that happened and that's something that Sundar from Google has talked about a year ago when DeepSeek came out that this reality that the usage will increase. So I think from one perspective we're going to see more use case and Doug and I have talked a lot about this. I know Doug you've been a believer that it won't necessarily for like the most advanced thinking of these models. It's not that it won't be a race to the bottom that they'll actually hold up some of their pricing more than what I think people expect because at the end of the day, if there's value that's being created, we think there is a lot more value in thinking than there is in networking. Going back to your fiber example, we think there's more value in thinking that the pricing even though some tokens could go down 100x that in aggregate it won't be as steep of a price decline as what many are anticipating.
How do you guys view the Mag 7 in light of this? You know, one of the things that was great about these companies is they were very asset light companies. They had strong cash flows. Do you when you look to like what they become now, does it change the way you view them? I mean, do you view them as completely different companies than you did before? And I know it's on a case-by-case basis, but I'm just wondering like this has been such a quick change in the types of companies these companies are. I'm wondering like for people inside the tech landscape like you guys, how are you viewing them now relative to how you viewed them before?
There's this, it's going to sound like a weird start to my answer, but there's this poet named David Whyte and he's written a handful of books. He's kind of like a modern philosopher, I think you could say. And his books aren't really necessarily poetry. They're really explorations about things he's learned about human nature. And one of his ideas that has stuck with me as I've read some of his books is this idea of like the seasons of life. And one of the most important things about humans is do you know what season of life you're in and sort of matching your actions to the season of life. I think the same concept of seasons is true in investing. It's true in companies. I think we just went through this season really for 20, 30 years probably where asset light building on top of the internet. You just mentioned the fiber providers who created all the infrastructure that the Googles and the Amazons of the world took advantage of to build these for the most part very high margin businesses. I think we're in a different season now where there's a lot more value in the current season to owning the infrastructure because to the extent you can own the infrastructure and Google's kind of proven this out I think to some degree with TPUs. To the degree that you can own the infrastructure and customize it to your particular use case, you might be able to lower your cost of delivery in a way that you couldn't do if you were relying on third party infrastructure. And so that to me is the new season. And so yes, it is a different regime than we're used to over the past 20, 30 years, but I think the companies that get that right and figure out how to extract maximal economic value and I feel pretty comfortable that the Mag 7 are experts at that. I think those will be the companies that continue to do well.
Are you saying, Doug, that there's more value to good old-fashioned infrastructure today than there was 15 years ago?
I think that's absolutely true. Yeah, because I think the application of cost of intelligence and cost of intelligence is not purely commoditized whereas you know bits running through the internet was largely a commodity.
Yep. The networking piece was commoditized but I mean we kind of saw it this week with the Google Apple announcement. Google Apple traded off a little bit kind of more or less in line with the NASDAQ. But you see Google trading up after being up I think 38% over the last three months traded up a percent and a half day one a percent and a half because their cloud business is going to get a boost from providing results to Siri. So that may be I guess exhibit one your honor.
So this next one, I know you wrote this last year and one statement from one company can completely change this, which we were talking about before we recorded, but you had said top private AI companies will stay private through 2026. So how are you thinking about that right now?
Well, I mean, talk about humility here. I'm pretty sure I'm going to be wrong on that at this point. I thought it was going to be SpaceX and then we'd see kind of a flood in 2027, but that one comment was from OpenAI that they're looking for a person who has experience with the other Mega Caps, the other Mag 7 to run their investor relations, which probably puts an IPO somewhere this year, early next year. And Doug, maybe you can talk a little bit about why that in itself changes the whole speed around which other companies want to move.
Well, I would say there's that aspect, right? There's value in some cases of being first, particularly if there's companies coming out that are very similar doing what you do. I mean, Gene and I back in the way back in the day when we were on the sell side, you know, that was something that you would tell potential banking clients to think about, you know, right? If your biggest competitor is thinking about going out, there's value in getting out before them because you get to tell your story first. You get to talk to investors first and that's useful. The other thing though I would say though if you think about what's going on the structure, why are some of these companies staying private for so long and maybe what are we seeing because Gene and I do spend a lot of time investing in these later stage private companies. There's been just a massive amount of private capital available to these big AI winners, you know, companies like OpenAI, Databricks, xAI. And I do think I mean this is just kind of a gut sense in terms of just what we're feeling from the market. Nothing in particular but gut sense is I think we are probably getting to some limits for some of these companies in terms of how much appetite there is for private capital to invest in these companies. You know tens and tens of billions of dollars in many cases have been invested in these companies and so I think some of these companies are getting close to maybe tapping as much as they can in that part of the market and the next logical place for them to go of course is the public markets.
Do you guys ever think tech will fix the IPO? Because I'm just thinking like the IPO is such an inefficient thing. Wall Street makes tons of money on it. Like I know they've tried direct listings and stuff like that, but I would think the world of tech would eventually figure this out and a way to get these companies public without you know paying all those fees to Wall Street.
I thought something, I thought Google was going to be a Dutch auction or...
Yeah, they did. They did Dutch auction.
