Loading summary
Transcript45 lines
- [00:00]
A
What's up guys? On this episode of Full Signal, I sit down with Luke Kawa. He is the markets editor at Sherwood News and one of the smartest financial writers in the business. Today we talk about how AI is disrupting the stock market, the software sell off, what the Magnificent Seven are doing and how their valuations have come down and much more. This is a fantastic conversation. I learned a ton in this and
- [00:22]
B
I think you will too. Luke, I'm super happy you're here today. You do a lot of great work on positioning, sentiment, market flows. Can you just start with your big picture view on markets right now?
- [00:36]
C
Yeah. Thanks a ton for having me. Great to be here. I think a very interesting thing about markets is that they've been doing nothing for a long period of time. Hey, when you're looking at it that closely, I think sometimes the lack of movement there can be a lot of stories embedded in that as well. We have hit a fresh all time high this year earlier in but by and large indexes have been meandering and I would say a lot of the story is the biggest issue for markets is return on AI Capex. And right now markets seem to be very much in a place where they want to search for safety. There's two kind of ways I would say safety has been exhibited in this market. The first is your obvious, your costcos, your Walmart. Things that no matter how bad things get, people are going to go there. People are going to spend a lot of non discretionary purchases, some discretionary as well, but hey, they're in staples for a reason. Then second is there is an AI cash flow that represents safety I think in the minds of investors and that people are very willing to pour into. That's the memory side. That's effectively the amount of price hikes you've seen in the space. The longevity of the supply demand imbalance that's expected to persist. The fact that a lot of these stocks were trading at very low double digits or in some cases single digit P Es that's a lot safer in terms of investing than perhaps, perhaps getting into one of the bigger chip makers. So you have a name like Sandisk, which is up over 100% this year and the PE has gone down. So you know, you don't need AGI to tell you what's happened to the earnings estimates when you have a move like that. So for me the main story in markets has been a lot of looking for hidey holes in kind of expected and unexpected places. As we continue to debate how we feel about AI Capex and returns on Capex. And I really think it is a feel in as much as it is a proof.
- [02:32]
B
I mean sentiment has been everything this year, you know, even the Citrini piece causing the huge sell off in pretty much everything, especially in tech. And you have this great chart here that you shared. The Magnificent seven forward free cash flow estimates have been coming down. Why do you want to highlight that chart?
- [02:52]
C
I think one of the things I was thinking about and pondering coming into this year was we all knew and kind of saw that the Mag 7 and the hyperscalers in particular, we had seen their forward PE estimates come down a lot price to earnings. Everyone was saying, hey, Magnificent seven looking cheap, looking cheap, looking cheap. I'm sitting there going, okay, well if we're talking about an AI bubble, if we're talking about repeating the.com, repeating the 90s, then let's see some inflation, let's see valuations go completely bonkers. If we're going to have a bubble, let's see that instead. I think what the market is really keyed in on and the reason I like to bring in the whole mag 7 is to include the one who's getting the recipients of the hyperscaler cash flows. In terms of Nvidia getting the lion's share, the biggest line item for data center CapEx is going to be the GPUs. And the fact that that is rolling over to me just tells you the extent of the CapEx that's happening and the fact that it's really not being made up or not expected to be made up in a very fast manner. Investors for such a long time have been accustomed to all of the Mag 7 being just besides Tesla, Tesla's own kettle of fish, but being these hyper giant cash flow generators that would then buy back shares, buyback shares, buyback shares. I feel like this capex binge has been like if you had a teenage kid or something and you're watching them get into heavy metal and goth music and going okay, it's just a phase, it's just a phase. Well, this phase is lasting a lot. Still a lot of black eyeliner going on and a lot of falling cash flows and not necessarily a return to what you've expected these stocks to always be.
- [04:31]
B
You have this follow up chart here that basically shows all these hyperscalers across the board. You have Meta, Amazon, Google, Oracle's in there. Their cash flows are all falling for the coming years and I think that's also been reflected in their stock prices. So all these stocks have pretty much stayed flat or sold off in the last few months. Is that tie going to become increasingly obvious, let's say.
- [04:59]
C
I would say the, the main reason I look at that and the main reason I worry about that is it comes down to something that Jensen Huang was talking about on Nvidia's most recent conference call. The first question he was asked was basically, hey, We've seen the CapEx budgets for 2026. We know you're going to be raking it in this year, but at a certain point is there going to be some ROI on your customers ends that's going to limit their capex? Jensen Huang's answer was their cash flows will improve. Because in this world, compute equals revenue. The last half of that sentence can be completely right and it can also be completely irrelevant to the first part. Revenues can improve without cash flows improving. One thing that we've seen is throughout the AI boom, if you look at for any hyperscaler, what's expected to happen to cash flows this year and what's expected to happen next year, what's been expected is always an improvement. That improvement has never come. And it's never come because the capex keeps going up so dramatically. So to me, to kind of square the circle or put a bow on it. Nvidia, which is clearly the most upstream stock you could get in the AI boom, it's the epicenter. The world revolves around its chips and their use in kind of developing and helping LLMs be trained and inference and so on and so forth. However, it's not being treated like an upstream stock anymore. It's being treated and valued as a play on the success of its customers. That's I think, a very unique relationship that's changed this year. It reminds me of Henry Ford. He said he wanted the people working on the assembly line to be able to afford his cars. It's kind of the same thing. You want the GPUs to be able to earn the return that you're paying for them. And so far the, you know, the jury's out to the evidences is not supportive of that so far.
- [06:54]
B
Wow, that's an amazing way to think about it. With these falling Mag 7 and Big Tech stock prices, do you think there's a point where the valuations get attractive enough? Where suddenly maybe we get an inflection point and investors start going back in just because suddenly they look cheap? Are we getting close to that?
- [07:17]
C
In a word, yes. In another word, I would have said yes, probably two, three, four, Weeks ago. I think when you get to valuations, and especially when you're talking about bigger companies, you need oftentimes valuation plus catalyst. One thing that's been on my mind this week as we're watching markets react to the outbreak of a kinetic conflict in the Middle east is, okay, what have people really been thinking about and betting on heavily? What are the real big narratives this year? I would say one big narrative has been memory over everything else and especially software and consumers of memory. Another has been rest of world stocks over US Stocks. And then another one has been the many over the few. So equal weight over S&P 500. What I would say we've seen in the week since the conflict has started is that all of those trades have started to reverse. And there's some fundamental reasons you can kind of tie into that. You can also just say, hey, this is the excuse to right size or unwind some of the bets that had become very consensus and common. And in that world, one of the bets that gets brought along up as we would discuss would be the Mag 7 finding their footing a little. So that's something I'm keeping an eye on right now pretty closely.
- [08:37]
A
Some of you may not have heard this, but our partners at Public just launched something called generated assets. It brings AI into investing in a way I've honestly not seen before.
- [08:46]
B
Here's how it works.
- [08:47]
A
You type in an idea like AI powered supply chain companies with positive free cash flow, or something like defense tech companies, growing revenue over 25% year over year. Publix AI then dispatches a swarm of agents that can scan every single stock, evaluate them, and instantly build a custom
- [09:06]
B
index around your thesis.
- [09:08]
A
What really stands out is how clearly it explains why each stock is included. And before you invest, you can even backtest your idea against the S&P 500. So you're making decisions with real context
- [09:19]
B
and not just guessing.
- [09:21]
A
Beyond generated assets, Public lets you invest in stocks, bonds, options, crypto, all in one place. They'll even give you an uncapped 1% match when you transfer your investments over from another platform. If you want to build a portfolio that actually reflects your thesis, visit public.com openingbell that's public.com openingbell now let's get back to the conversation.
- [09:44]
B
Where do you fall on the AI bubble? Fears versus AI is gonna get so good it's gonna destroy everything. Like in that kind of spectrum of sentiment, how are you thinking about that?
- [09:56]
C
Oh man, I'm such a weather van on that. My view on that, I'm almost swinging as wildly as the market seems to in terms of the temperature. On that there's been things I've read and days I've woken up where I go, I better come up with something good today or else my bosses are going to figure out that an AI could probably do a good 75 to 80% of what I do in two weeks. There are other days where I'm like, this thing can't even scrape the damn chart I'm looking for. Gosh, I could get anyone out of high school who would be able to fill in these numbers correctly, that sort of thing. From a market perspective, I'd say that one thing we know is that the hyperscalers have told us that a lot of this capex is defensive and the bigger risk is not spending enough rather than spending too much. When you commit ex ante to spending too much, that does tell you that some of the ex ante returns on investment will not be probably as bright as the most optimistic scenarios would happen. On the other hand, I think you have to look at probably a pretty rich environment for startups and companies that we haven't heard of or aren't on the radar to succeed. I've seen some really nice charts about revenue to employee picking up pretty dramatically at a lot of smaller companies and there's a pretty big AI link there. So it's a lot of, I think again, what you can say about what's happened in markets over the past five months. You could also frame it as, hey, there's been a lot of as public markets have stalled out, there's been a lot of value that's accrued to private markets, both in terms of the kind of the big ones we know about in terms of OpenAI valuation rounds, Anthropic valuation rounds, but also probably value accruing and potential disruptive companies that I couldn't tell you their names right now.
- [11:45]
B
Yes, I think, well, when I think about my own use cases for AI, it doesn't feel like a bubble because I'm using tools and I'm being able to do things in such a short amount of time with these tools that I definitely couldn't do a year ago, definitely not two years ago. So for me, if I just look in the mirror I say, wow, this is definitely not a bubble because these tools are so effective. And maybe that's a separate conversation from let's say, a valuation bubble, but at least in the use cases it seems very real.
- [12:18]
C
Yeah, there's value that will accrue here from it's a Technological advanced value will accrue. Whom does it accrue to after all this capex? Not sure if it's the hyperscalers, maybe it still will be. We know in the near term it is accruing to memory stocks based on what's laid out for some of the major chip makers seems to be accruing to them as well. But yeah, no, the big question is what's the deadweight loss for the company spending the most versus what we as consumers and as users? This tech end up reaping, that's going to be a very fun one to track. But the technology to your point I think is definitely helping. AI has been a fabulous editor and sparring partner for me. Love to get it to fight me and tell me why I'm wrong about things. It's great, love using it for that. So definitely agree with you that it's definitely had a positive impact on my day to day.
- [13:13]
B
Okay, so you have this other chart you shared AI disruption appearing in the rates market and I think everyone talks about how it disrupts different sectors and we have all these sell offs and trucking and real estate and financials. But I think you're the only guy I've seen point out this connection here. Can you explain what's going on?
- [13:33]
C
Yeah. So I can't take credit for this. I got to pass it off to you to Peter Williams at 22V Research, he had a really, really good note on this that I thought was fabulous. Had to write it up. So what he flagged was effectively I would say since the start of the year and particularly since post government shutdown data has started to come back out, say for most of the past year or two years. I would say the general way to describe the U.S. economy and the labor market has been it's slowing and softening. I would say more recent data said it's stabilizing the stabilization has resulted in effectively pricing out of near term rate cut odds. So I believe at the time of that chart I think a full rate cut wasn't priced until July and I think it's the timing has probably I believe been pushed back since then in light of the just energy price shock and the way the markets, the front of the rates market at least has been treating the outbreak of war over in the Middle East. But on the other hand even as you've had kind of those very near term expectations of what the Fed will do get curbed, go down, you've had if you look at effectively what are the expectations for where rates will Be at the end of 2027, those have gone down. It's very rare for something that is encompassing effectively an 18 month period for those two to be heading in dramatically different directions by about, I believe it's 40 basis points at that time what Peter wrote and I agree with a good chunk of it, but what he wrote is that this is effectively US pricing in that things are fine now, but all of this investment in AI capex is going to have an adverse impact on labor income in the short term. There will be disruptions, there will be job losses and that will be something that the central bank might have to respond to in the form of easier monetary policy because of lower consumption. What I would tack onto that or what I more see as a more causal possible mechanism is that AI is going to disrupt industries and companies. That is going to cause credit stress and the credit stress which your Blue Owls, your Apollos or whatever you want to effectively point to. There seems to be a headline today about issues in private credit, but that that will be more of the mechanism for employment weaknesses. It's the fact that there will be losers and losers will come with it with job losses. So that's more how I think about it in terms of that chain running through credit, running through possible asset market weakness, feeding through to consumption rather than just straight kind of like the advent of the technology and the adopters are the ones shedding a lot of jobs.
- [16:06]
B
I mean these are a bunch of second and third order effects of AI that I think most people have not thought that deeply on. I know you have. What do you make of let's say the recent credit tremors? I mean you touched on it just now a little bit. But I think even for myself I try to keep up with it but it's a little opaque from just trying to read up on it. What are you watching?
- [16:31]
C
It seems very interesting because when you contextualize it with okay, the economy seems to be steadying, stabilizing, nominal growth seems to be fine. When nominal growth is fine, generally you don't have that hard of a time paying back your obligations. You have to weigh that against what did private credit really like investing in software in a lot of cases because both it's a relatively high margin business and also the cash flows can be generally pretty predictable. ARR in terms of recurring revenues is the big stat and go to number in a lot of software. The fact that that is what is most directly seen as imminently under attack by AI it leaves you in a situation where okay, software companies who we've loaned a lot to seem to be disrupted by this. The companies that are trying to disrupt them are also need a lot of capital. So it seems like the loans I've had are getting riskier and there's still a lot of ones I apparently need to or being asked to make by the industry that's disrupting them. So that seems to me to be a very delicate or odd situation where it doesn't seem like both of those legs should be taking place at the same time. Like either the AI disruptors get funded and don't face a lot of stress, are successful and software companies are disrupted or software companies effectively get to hold up a little better because AI companies don't get the funding that they need to be able to grow and offer that disruption. That's again, I think another balancing act to look at and think about in
- [18:07]
B
the weeks to come see you and I will pay attention to this stuff and maybe people on Wall street, but outside of that really small ecosystem, I think general population does not pay attention to this stuff, which is makes me a little nervous because there are definitely issues as you're laying out and those are bubbling up more and more. We're getting more and more headlines about various credit companies that are having some tremors here. So let me pivot to macro. Here you have this great chart that says essentially the. The job market and The S&P 500's dynamic, their correlation, as goes the job market, so goes the stock market. What do you mean by that?
- [18:49]
C
Yeah, I think this is something that I'm very willing to revisit and I think in the next if the promise of AI is quote unquote real. I think this is a worldview and a thesis that will become increasingly pressured. But for now I think it's the kind of playbook you have to go to in terms of analyzing the economy because it's worked for at least all of my life. What this chart shows effectively is that the six month change, if you're looking at the green bars there or the red bars there, the six month change of the stock market and the job market is the same 75% of the time. I like that track record. If your timeframe is anything more than a day, more than a couple months, the stock market and the job market are going to go in the same direction. The big discrepancies we've had where they don't are one the biggest generational inflation we've had, 2022 we had effectively job market still adding while the stock market was Fall. And we do see some kind of divergences at times when the stock market is picking up at the end of a recession as we've delivered over a policy response that still just hasn't fed its way through the system. But that to me is. I think there's a really big idea that you can sound smart by saying, oh, the stock market is not the economy. I'm like, hey, 75% of the time. I'll take those odds all day. In terms of what to set your watch to, I think it's a good reminder that the American economy and the American stock market might not be. The magnitude might not align or make sense all the time, but you need things to be going up, you need spending to be going up. That's what keeps credit stresses down and that's what keeps earnings powering higher. I do think that the relationship between the two has shown to be pretty strong. And it's something that people should think about and be willing to set their watch to, knowing that 25% of the time it might not work, but most of the time it does.
- [20:37]
B
Well, you have this, this kind of follow up chart here where you clearly label it the stock market is the economy and it shows how bear markets have coincided with recessions or high inflation. And I think that's a bit. Part of me thinks it's intuitive that that would be the case, but maybe not. I mean, why, why did you put this chart together?
- [20:58]
C
Yeah, again, because I just, I think a very dominant meme or reference point for investors is that the, you know, the stock market is not the economy. And people kind of use that as a backdoor way to justify or pay credence to, I think a lot of reasonable concerns that people have about the economy and about their personal welfare. Obviously, during a high inflation period with negative real income growth, people aren't going to be very happy about the economy. But you know what? Hey, that's the time when the stock market reflected it and it was falling a lot. Again, I think that's just another chart that shows that drawdowns have big drawdowns, ones that we care about, ones that investors that you should look to respond to if you had the crystal ball. Those are ones that are associated with real economic fallout. It's the same thing as I think we can go into with just the idea of earnings drawdowns pretty much every time The S&P 500 is as a bear market. It's also coincided with forward earnings estimates falling at least 10%. The idea again is that the economy. Corporate America's earnings power is also very directly a function of your spending power and ability to grow. That given the degree of interconnection between those two things, we should care a lot about what the stock market does, not the least of which because the stock market, we can have that reflexive effect where a falling stock market makes us feel less wealthy and causes us to then spend less. I do think that's been an underrated point about why consumption was able to do so well last year. It's that this was the most retail bought dip ever. Was probably was the Q1 dip and Q2 dip as the momentum stocks broke down and then tariffs kind of weighed on the market. So, you know, the beneficiaries are people that are kind of more disposed to spend than if, you know, it had been hedge funds and investment firms that were really reaping the benefits of having bought all the way down to the early April bottom.
- [22:53]
A
Real quick, we'll get right back to the interview. Just wanted to pop in and say if you like this content, I read a newsletter every single morning called Opening Bell Daily. I cover macro, the stock market, asset
- [23:05]
B
prices, why things are going up, why they're going down.
- [23:07]
A
And if you want to get that for free, you can sign up at the link in the description.
- [23:12]
B
Let's get back to the interview. I mean, it's an amazing thing to highlight. The guy who I think has connected the stock market and economy most is probably President Trump because he points to the Dow and he says, look how strong the Dow is. The economy's booming. And I think, you know, someone in markets or on Wall street will always again, sound smart by saying the stock market's not the economy. But when you have the president calling it out the opposite and pretty much saying, look at the scoreboard, it's really hard to separate the two. And I think people start behaving as if they're the same thing as well.
- [23:49]
C
Yeah, no, I think that's, I think that's a very good point. I do. Yeah. It's become a lot of things have become political in this day and age. And effectively when you have the president effectively tying his flag to that, then certainly it becomes somewhat of a political statement. That doesn't make it less true or less valid just considering the source or not.
- [24:09]
B
So given all that, let's say the macro risk, the geopolitical risks, some of the AI concerns, are you optimistic for the rest of the year?
- [24:20]
C
Hey, I think someone we both know very well, and I've learned a lot from Sam Rowe says most famous lines stock market usually goes up. You have to have a really good reason and high conviction I think to be negative on the prospect of growing your wealth in the stock market at this point. There are reasons I don't have a good enough reason to cause me to take my ball, pick up and go home. So I'll keep acting as accordingly. I do think that we've seen in Q1s in a number of years since the AI boom, just the revisiting of questioning of ROI of CapEx. Can this be done more efficiently? We had pretty obviously deep seek last year and then at a certain point the earnings power in the near term takes over. So if, if I had a base case, it would probably be that the earnings power of the companies that are participating the most aggressively in this boom is something that once again recaptures the attention of the market. But what I'll say on the other hand is if AI doesn't work, then I don't think anything works. I think the past five months or so are your proof. No matter if it's a pocket of AI working like memory, no matter if it's financials really working, then not working, there simply doesn't seem to be enough weight for the rest of the market to really give you a tremendous return. When the mega caps aren't working. RSP has gone up. It's hit record highs more frequently than spy, but still I think not something to write home about even relative to a high yield return.
- [25:59]
B
Totally to push back on you for a second. When you think about AI working versus not working, what do you think that means?
- [26:09]
C
The company is spending the most, getting rewarded for spending so very, very narrowly. That to me is what AI working means, the market rewarding investment. So it's probably definitely worth defining. So yeah, good call. But yeah, it means effectively the hyperscalers not going down because I think that's the backdoor big threat to the investment. It was one thing when I think when Mark Zuckerberg started investing in the Metaverse, pretty much everyone was looking at him with some side eye going like what the heck is this? Eventually your stock price goes down. Enough. You say, okay, enough's enough. You cut jobs, you slash Metaverse spending and then the market cheers it. The market has not been cheering AI Capex for some time and that is the big risk to AI Capex. I think the, the hands in this are a lot stronger in terms of the impetus to spend and what they're seeing as the benefits from this tech. But that is the risk And I think it would be disingenuous to not point out that at a certain point in time, capital goes where it's treated best. And the market's lately been telling us that they don't think hyperscaler spending, the amount of hundreds of billions they're spending on capex, is money being treated best.
- [27:21]
B
Well, the counterpoints to that too is maybe it's not that the market's not rewarding the spending, it's just that they're worried about AI disrupting these businesses. Is that something that's also variable?
- [27:35]
C
Gosh, I think that's a very good point. And as mentioned, a lot of this hyperscaler spending in Capex is defensive in nature. We're spending because we can, and if we spend and we're the leaders, then the odds of us getting disrupted go down. However, the odds of margin compression, I would say, also are going up. In that scenario, the more you have to spend persistently over time, depreciation does come home to roost at a certain point. This does set up as a situation where if the spending has a large defensive element and that is a dominant part of it that stays over time, then that's a recipe for margin pressure. Margin pressure, in turn, is a reason why you don't pay a higher multiple for stocks than you used to. That's, I think, something that's completely within the range of outcomes here.
- [28:22]
B
Okay, Luke, where can people find your work online?
- [28:26]
C
Best place to find me and all the folks I work with, great colleagues at Sherwood News. Sherwood News is the best place to reach us.
- [28:34]
B
Okay, and just to clarify, Sherwood's part of Robinhood?
- [28:38]
C
Yes. Robinhood is parent company. We're an independently operated media entity.
- [28:44]
B
Amazing, Luke. I learned a lot speaking with you, and I hope we do it again soon.
- [28:48]
C
Oh, my pleasure, man. Thanks a lot.