Welcome to Okay Computer. I am Dan Nathan. That is Gene Munster. He is the managing partner at Deepwater Asset Management. Gene, welcome back to the podcast.
All right. You and I, if you're viewing this, if you're watching it on the YouTube, you and I kind of look like twins here. It's fine. You know, we did not color coordinate, but it looks great. I appreciate you digging my fit, because I dig your fit, too. Okay. How's that?
I love it. Right back at you. I always love your office here because there's something near and dear to my heart right over your right shoulder. It is the what the what was it called again? The Apple speaker is something really simple like that.
It's called iPod Oh my goodness.
Yeah, there was it was it had some odd name.
Yeah, we'll come up with it. Our team will find it while you and I are talking. But I also love that you have the iPod mini in that. You and I have talked in the past. I got all that stuff, too. So someday we'll just nerd out and bring it all up and we'll have like
We can do an episode just on nerding on old tech.
That would be super super fun. You and I got a lot to talk about because it's shocking. We are heading into Q1 earnings season already. And I think that given the environment that we're in, there's so many moving parts, right, with the mega cap tech, how they trade in the public markets, but then the valuations that we see that keep skipping higher in the private markets. We'll talk a little bit about that. I got to hit Microsoft with you a little bit. I know this is one that you have not been that in favor of. It acts so badly. I wonder if the sentiment is too negative. The relative strength in Google. And I think that is a story versus Microsoft. Apple fold. A little debate here. We're recording this in the afternoon of Tuesday. You're listening to this on Wednesday morning. And then just this OpenAI saga. Man, it just sounds like it's gone from bad to worse despite the fact that the valuation keeps skipping higher. And then this is really cool. You just got off the set, I think, of CNBC are doing a hit and you have a piece coming out on CNBC.com detailing your views on SpaceX. We'll get a little preview of that in front of this IPO that they've confidentially filed. I think that was last week. All right, Gene, let's start with the Mag 7. You know, I like to call it the faithful eight. As you and I are talking right now, the only two names in that group that are up are Broadcom and Google for reasons that I'd love for you to go into a little bit. This is a deal with Anthropic for compute. It's Broadcom with working with Google on Google's tensor processing units. Break this down for us in a market where the NASDAQ was down one and a half percent earlier in the day. These two names were two of the biggest bright spots. I have a feeling that if Anthropic was publicly listed, it would have already also been up. Why the positive sentiment about this deal?
Well, it's a validation that Anthropic is in a good place and that they need more capacity. And as part of this too, if you look at Anthropic's, their revenue, it's measured by kind of their monthly revenue run rate, but has gone from nine billion at the end of 2025 to 30 billion now just announced. Now the big step up has, I mean that is just a spectacular step up and the fact that they are doing this with Google and Broadcom really suggests that even though they've had this big step up that there's more to come that they're seeing even a higher growth rate down the road. So from Google's perspective and Broadcom's perspective, it's validation that they're in a good place and that's going to be some business. From Anthropic, as you mentioned, if they were a public company, they would be up today because what they're really signaling is that they just can't, they simply can't keep up and they need more capacity. And so the narrative that this is somehow slowing down or topping out, these announcements kind of would say something different that we're still at an acceleration phase.
Yeah. You know, Ben Thompson of Stratechery was breaking this deal down this morning and you know, he's talking about a, you know, I mean, no one's talked about the implications for Nvidia on this or no one that I've actually read. I don't think in Ben's note this morning that he really kind of addressed it one way or another, but they did spend some time talking about the amount of TPUs that Google has that they're using for their own models that they have in GCP, right? And how this is a huge differentiator. And I think that was a big part of why Google caught a major bid, you know, mid last year. And I know that you were coming up with multiple reasons why Google was a buy. This was going back to the first half of 2025. It really was though, I think, the TPU story that kicked it into high gear late last year. What do you make of, you know, Broadcom despite being up, you know, 5% today is still down about 25% from those all-time highs which was back in December when they were reporting, you know, that fiscal Q4 and the stock, you know, this was not a name, it's got that great backlog, you know, they gave really good revenue guidance. Why do you think the stock has just not performed well especially, you know, I could see them giving back this move, you know, tomorrow in a tough tape or something like that. It just feels like investors are just, you know, have their finger on the trigger here in most of these mega cap names.
Yeah. And also on the infrastructure, the hardware infrastructure. And I think what it comes back to is just some of the math that's going on in investors' minds about what can go on with capex spend. Obviously, the custom silicon piece that Broadcom plays into. They can grab market share from Nvidia, but they're still playing in the same ocean and as the tides go up and down. And you know, maybe I kind of jump into that's always been a topic around earnings is what are these kind of the four kind of primary hyperscalers going to say about capex and just to kind of rewind I think that that is what's playing into what's going on with Broadcom's underperformance. You also look at Nvidia. What they said at GTC a few weeks ago suggested that revenue next year is going to be up 30 plus percent, excuse me, 40 plus percent. The street was at 30 when Jensen had those comments. And the stock over a 3-day period was down 8%. The NASDAQ was down 4% during that period. So, it significantly underperformed based on higher expectations. And it comes back to this math. And I'll just kind of very high level say that as we've talked in previous episodes I've talked about these numbers being way too low for what these hyperscalers are going to be spending. There still needs to be this significant jump up. I think we're getting to a point especially related to the 2027 but even potentially in some of the commentary about the June quarter and the capex that it's going to be relatively stable consistent with where the street's at. So, we're not seeing these big step ups in part because there are some natural limitations to how much these companies can spend specifically related to their free cash flow. Now, there's other ways that they can raise money or potentially increase revenue growth around AI that they can further spend, but just mathematically it gets harder and harder for these numbers to really surprise to the upside. And so back to your original question, just what's going on with Broadcom, it's kind of what's going on with the AI hardware companies more broadly. And I think that that's what investors are kind of coming to is these numbers. So I'm just going to kind of finish and put it into a very clear view, which is the hyperscalers. This is the kind of the four big ones have basically guided to about 60% capex growth for calendar 26. If we look at the numbers for calendar 27, the street's at 10%. And you know, a year ago that was the case and then they ended up being 60%. I don't think that 10% number for calendar 27 is going to be 60%. It's probably going to be more like 20%. But so I think that dynamic has really kind of caused investors and I think the easiest way to understand how someone feels about a company is just to ask how they do they own it. And I can say we still own in our public portfolio Nvidia but we have reduced the position in part because we think everything is great but there's this psychological piece around where the hyperscalers are going to be spending that could negatively impact the multiple.
And that's where this Anthropic Google Broadcom deal kind of fits into any trepidation that you might have especially I mean listen the thing about Nvidia here is if you just cared about positioning if you cared about expected growth versus valuation, right? If you care about how they get to that EPS growth year-over-year with an acceleration or reacceleration of gross margins because that would be the story if it went from 71.5 basically last fiscal year to the expectation of let's say above 74% this year and you're looking at EPS growth this year of expected to be 70% still and then next year of let's say call it 35% and then revenue growth expected to be, I mean it is astounding based on this base that they're able to or at least the street thinks that they can grow sales this year 70% and then 35 next. You know, you look at it trading at 16 times next year. Like how do you not load up the truck on this name even if you're going to have, you know, competition?
Yeah. But I mean it again like we're trying to figure out is we still have a position. We still have a large position. It's half of a position. And I think that it's from our perspective is that we're kind of looking for when the psychology on this whole infrastructure thing is going to change. I mentioned that the street's at 10% for kind of the hyperscalers for next year and that the revenue growth that Jensen talked about is 40%. So I don't want to geek out with numbers here, but they do matter. Is that if 50% of your revenue comes from four companies and they're growing at 10% next year or maybe let's say 15, how do you grow your business at 35, 40% plus for Nvidia? And the answer is you have new customers that step up and they've talked, you can just look in the language how they talked on their last earnings call how Jensen talked at GTC about other customers starting to become bigger and bigger and that's so to answer your question like why not be more aggressive like trying to figure that piece out that kind of the conversation now is shifting a little bit to the other players the other 50% of their revenue what's going on there and I think there's a case that you know we may be back and buying more. We don't have plans to do that right now, but I just want you to know like where our mind is at is like this still I think is we're still early in AI. I think that the need despite custom silicon, the need for the most advanced thinking is still going to be at a premium and I think that that's good for Nvidia's business.
Okay. And going back to Broadcom for a second here. So the stock is basically trading back when it reported its I guess August quarter I think in early September or maybe it was some sort of, you know, sort of design deal or whatever. But you know the stock again is not too different than Nvidia. It's just gone sideways, you know, on a PE to growth. It's definitely more expensive than that of Nvidia with, you know, lesser margins. Is Broadcom interesting to you here? Especially if we're going to start to see more competition on Nvidia and more demand for Broadcom services amongst, you know, some of Nvidia's biggest customers.
I think custom silicon is important. I think Broadcom's in a good place for that and so it is one that is on our watch list and one that we have periodically owned too. And so I think Broadcom is important. Again, what comes back, you know, the kind of our most important piece that we're focused on is where are we in AI? Not only just on the infrastructure buildout, but on the utility of it. What does all that mean? And when we put all that together, I still think we're in the second inning. And then I realize that that's been a kind of a common message. How is it that we're a year later and we're still in the second inning? How is that possible? And I think the answer is that this is still unfolding. Look at what's happened with Anthropic's growth rate. Like that's a big jump. That doesn't happen. Going from 9 to 30 billion over a four-month period. It doesn't happen unless something bigger is going on. So to answer your question related to Broadcom and all of this, it comes back to, you know, this belief and I hope that we're not like blinded by confidence in that we can still objectively ask like what is really going on here? But objectively, if you look at what the latest data points, the news today from Google and Broadcom, the update from their revenue expectations, I think that it's still pointing that we're early.
Yeah. But Gene, it also doesn't happen if OpenAI and Anthropic aren't burning, let's say, $3 billion a month. Let's be frank, right? So if you think about it in that perspective and you know given how much capital these public traded names are doing on a capex basis I mean a lot of really great things have to happen in inning four, five right for these valuations to hold and you know we've talked about this I mean OpenAI is not meant to be profitable for at least I don't know three to five years or something like that and then we're already talking about, you know, IPOs which are expected to come this year. I do think it's interesting and you've read this I think it was the Information reporting this that while Sam Altman and he's gotten a lot of bad press I mean the New Yorker piece I don't know if you saw that but like, you know, and there's, you know, a lot of tit for tat with everybody in the industry Elon most importantly but as Sam keeps, you know, hinting at that, you know, the going public this year is a priority Sarah Friar the CFO is suggesting that they're probably not ready and the Information article suggests that, you know, Altman kind of putting her in a box a little bit, not including her in a whole host of just kind of strategic decisions. And then on the other side of that, the strategic decisions don't look particularly great. And they keep kind of pushing out the whole idea that this company has the proper business model, you know, relative to, let's say, Anthropic, which has gone from that $9 billion revenue run rate to let's say over 30 is what they just suggested. So I guess this all comes back to a little bit. Are the strategies of these companies, these two in particular, which seem to be very different, right? Are they likely to yield the sort of I guess, you know, earnings three, five years out that would justify a their spend, but b the spend of the ecosystem and validate their business models?
