Loading summary
Transcript138 lines
- [00:00]
A
Jack, over the last month we've seen semiconductors take a hit, memory stocks take a hit. Micron is looking pretty red these days. What stands out to you about this sell off?
- [00:09]
B
Well, I'll start by saying I think it's a buy right now. So I'm long and I am not selling. And I don't think that it is wise for people to sell unless they're ridiculously leveraged, which, let's be honest, I think some people are, which I think is bad. But I think the sell off, the absolute top in the semiconductor index probably was like in the middle of June. So June 18th or June 22nd, whatever. I think the weakness actually started a little bit before on probably like June 3, I think, which is a Wednesday when Broadcom reported they were not very strong in terms of the guidance department. And I think that's what started a sell off in specifically I guess let's call it like the, the, not the, the not core chip part. So like the, the interconnects and the switches. So like the Broadcoms, the marvels of the world and I. And, and then of course the, the earnings from the memories has been tremendous. And memory stocks have been coming down since like the middle of June. So I think it, there's a lot of themes within semis that have been doing really well that have faltered over the past two to three weeks.
- [01:20]
A
Do you think that the memory sell off is mostly sentiment driven?
- [01:24]
B
I think it is leverage driven. And I think that, you know, you can't just blame leverage on the way down. I think leverage drove it up and leverage is now driving it down. But I think the fundamentals are intact and strong.
- [01:34]
A
Wait, so explain that a bit. Why would leverage be the thing that unwinds a rally?
- [01:41]
B
Because people own too much. People own more than they have in cash. So they've borrowed money either themselves or they've relied upon these leveraged ETF to do the borrowing for them. And as buying begets more buying and selling begets more selling. And in these, in these leveraged ETFs, which I think are moderately toxic, I think that they can create these compounding effects to the downside.
- [02:03]
A
Okay, I have seen the reports of investors in South Korea taking out loans or borrowing in some capacity to buy stocks, something like SK Hynix or even just the broad market in South Korea, the Kospi. Right. That's been looking, I would say a little frothy at times over the last couple months. Generally though, I am pretty optimistic on South Korean stocks, but also memory Stocks, the AI trade at large. Does anything that you've seen as far as the bear case on this industry stand out to you short term?
- [02:41]
B
The fact that everyone at least claims they own it. The bank of America fund manager Survey said that 83% of fund managers think it's the most crowded trade, which I think is one of. It's crowded close to the record book. So it's, it's crowded so that, that could impact short term returns on long term returns. I'm, I remain a bull. I think, I think there's, there's several things that are going to stop the relentless growth in AI capex and I think it ultimately depends on the returns for Frontier Intelligence. I think that if the, if the, if the model companies are not making money and they can't pay their bills and the hyperscalers are not making money, then the hyperscalers are going to stop spending. So I think that it's probably how this will like end, but I think it could be in like two to three years, not two to three months.
- [03:33]
A
So if the hyperscalers stop spending, then every level of the AI ecosystem goes out.
- [03:40]
B
Yeah. Yes.
- [03:41]
A
Okay.
- [03:42]
B
And it really starts at the model. So OpenAI and anthropic. I know we can talk about Meta. I think I'm, I'm a little more skeptical of Meta's AI strategy maybe than you are. But really all the big spenders have been, they, they've been spending like they're spending moving away from Nvidia. So like the one theme that has not worked really is Nvidia, which is, I'd say probably one of the stocks I'm the most bullish on. So Meta has mainly been sticking tried and true with, with Meta, with Nvidia, but, but like, you know, Amazon and Google, they designed their own chips so they've been relying on Broadcom. There is a little weakness that I have some, some worries about Broadcom going on. I'm not like short the stock and I think that's not a good idea at all. But they basically the, the question is Marvell and, and Broadcom, they design chips for other people. How good of a business is that really? Like you're basically just a consultant which, you know, consultancies shouldn't trade at 60 times earnings.
- [04:35]
A
So to be clear, I do own Marvell. I think it's a great company. I like the stock a lot, even though it's pulled back quite a bit.
- [04:41]
B
I own my own Merv. I owned to be. Can I brag on your show?
- [04:44]
A
Please?
- [04:44]
B
I, I Bought call options on Marvell in late January, early February that were up 1700% and I sold them at the day of the top, June 3rd.
- [04:54]
A
Would you own the stock outright right now?
- [04:56]
B
I do, I do via. Via Basket, but I don't know. I think. I think there are better places in the semiconductor stuff. So I own it like via Beta Semiconductors, which I'm very bullish on. But in terms of like alpha stock picking.
- [05:08]
A
No, as in an etf.
- [05:12]
B
Yeah, just like a general basket. But do you want to know why I sold Marvell?
- [05:16]
A
Tell me.
- [05:17]
B
I was bragging about how much money I was making and how much return I was making to my dad and he told me to sell, so my dad told me to sell and I did. So. Thank you, dad.
- [05:25]
A
Well, I think anytime that you find yourself bragging about your returns, it's probably time to sell.
- [05:30]
B
Yes.
- [05:30]
A
Okay. Marvell. To me, I think they're just very well positioned because they seem pretty embedded in the broader AI ecosystem. A lot of people go to them for their services. And I think chip design is pretty resilient of a business right now, at least in the current setup. Would you. I mean, what do you think of that?
- [05:50]
B
Yeah, I just think that the chip design talent is at Nvidia and amd. I've been a little bit of an AMD skeptic, not a bear. I'm just. Sorry. Sorry. And AMD Skeptical skeptic Baron. Like, I just like Nvidia so much more. I think. I think that they, they do well. But there are things that I'm way more bullish on. So like I'm. I'm very bullish on semicap equipment. I'm very bullish on Nvidia. I feel fine about memory. It's. It's where the good revenue growth has been. It hasn't been. I'll be honest. You know, talk about bragging. I wish I could brag and say I've been long micro on SK Hynix. I considering buying them now just in terms of individual stocks, but that's where I think I feel the best. And that's also where I'm most familiar. And I feel like it's at least the most simple to me. There's a lot of stuff that's really complicated in terms of interconnects like photonics, like Computer Express Link, which is theoretically going to get rid of this memory squeeze that I feel like there are engineers who at least feel. And then there's intel, which I'm a little skeptical about. Intel.
- [06:55]
A
How come?
- [06:57]
B
I Think that their foundry business is bad. They lose money in their foundry business. They lose billions of dollars in their foundry business. They make a lot of money in actually I don't, you know, in chip design. I think that Taiwan Semiconductor is genius for being like we are just, just going to be a pure fab. We are not going to compete with our customers so all of our customers can trust us and come to us. Intel has a bad history of being way far behind and basically invented like they basically invented extreme, you know, EV technology with asml but they, they totally bungled it. And their new CEO I do like a lot Lip Bhutan is very good. But I just think this, you know, when you see a stock run up 600% and the revenues haven't run up 600% or at least are very looking very likely to run up 600% as is the case of micron. Like the pes of micron have gone down. This has been an earnings driven rise, not a valuation driven rise. But with Intel I get a little bit cautious. That being said, there are some great investors and great semiconductor analysts who know way more than me. Bubble boy is a guy I've interviewed. He's very bullish on intel and you know, shout, shout out to him. He's been made a great call. I'm just not as comfortable. I like the semi cap.
- [08:08]
A
When you think about the bearish takes on AI right now, the Michael Burry's of the world and others who are not as famous what stands out to you as okay, that is the bear call that I could get on board with that. I think we are actually at the biggest risk of right now.
- [08:27]
B
Okay, so I think Michael Burry's argument about chip depreciation, I think he's first published that maybe November or December of 2025. He's a legendary investor, way smarter than me. But Phil, nothing over the past seven or eight months supports that idea. Every piece of Data shows that GPUs are valuable and that they have continued to appreciate. Now I think he could be right, like that could end tomorrow. But I just think we should acknowledge that like literally everything that's happened has not say proven him wrong, but over the past seven months has not been accurate according to his narrative.
- [09:05]
A
Okay, so explain the chip depreciation thesis for step back for a moment and then tell me why you think it's wrong.
- [09:12]
B
So if you and I are investors and we build a building and it costs like 100 million, you know, it's a big building, $100 million that doesn't come out of our costs. We, that comes out from the cash flow statement. So that's $100 million as capital expenditure, and then the cost is depreciated over the next 3, 5, 7, 10 years as a cost. So that's true anytime you really build something and all of the, all the hyperscalers, Microsoft, Google, they are spending, I mean, in some instances over $100 billion per year. And that's not coming out as a cost. They're spending the money now and the cost is going to be taken out over the next three, five, seven years. Michael Burry says that the chips are going to lose their value faster than what the accounting rules allow. So he correctly, you know, I think, has pointed out that GPU depreciation years used to be like three to four years and they've been raised to five years in 2021 or something. And so his argument is that if the payback period and if the chips become useless over a period of three years, not five years, then that depreciating depreciation expense should be higher every year.
- [10:23]
A
And you think that's wrong?
- [10:25]
B
I think that every single piece of data over the past seven months has just not been true for the past seven months. I think, I think like what will truly think this? I actually think that there are people who are way more bearish on AI than Michael Burry, who I actually think they have arguments that are slightly better just in terms of the model companies losing money. Like I interviewed at Zitron, who's a very, very prominent AI skeptic, and I think about OpenAI, like, I think there's a chance that he could be right. And semiconductor companies, in terms of their earnings actually still do really, really well. And what I mean is like a trillion dollars of profits for the semiconductor companies and trillions of dollars of losses for the LL companies. And it's just going to kind of happen, you know, like, I think that
- [11:11]
A
way explain Ed's argument for people who aren't familiar with it, that AI is
- [11:16]
B
being massively subsidized and that when we have a wonderful experience on Claude, like I did literally this morning, that is being charged at like 1 to $200 per month and anthropic is costing them like 3 or $400 per month. So it's similar to Uber, but just at a much massive scale. Like it's not fundamentally profitable.
- [11:39]
A
And you think that is a realistic risk right now? If something like an anthropic or OpenAI tell the world, okay, this is not sustainable?
- [11:50]
B
Yes, And I think, um, I also will say that a lot of revenue, like Anthropic has seen huge revenue increases and this year in terms of their ARR, probably about $50 billion right now in terms of an annualized revenue rate. But there is a chance, and I'm not saying I agree with this or disagree, but there is a chance that that is pissing a lot of executives off and saying where we're, we are being charged so much, like where are we getting the return on investment? I, I mean, I definitely think that is the case that there are in, you know, the C suite is kind of pushing back and there's some people who are saying this is, this is unworkable.
- [12:30]
A
Internally at Anthropic or external?
- [12:32]
B
No, no, no, at the clients.
- [12:33]
A
At like clients clients. Got it.
- [12:35]
B
You know, JP Morgan, for example.
- [12:37]
A
Okay.
- [12:37]
B
Jamie Dimon is saying we're spending how much on Claude?
- [12:40]
A
Well, I think the big narrative at the start of the summer was token maxing. And then suddenly all these companies realized, wait, we're spending more on tokens than headcount, so we need to somehow figure out the economics so we're not shooting ourselves in the foot as a business. And I think you have a lot of companies that are looking for ways to one, be more efficient with tokens, but two, just find cheaper tokens in some way, either coins, different models, or figuring out something in the AI stack that lets them save money.
- [13:10]
B
Totally. And I think that you asked what began the semiconductor sell off from its extreme, you know, bull market. I think the end of the token maxing era probably coincided with roughly the peak in early June or mid June in, in semiconductors. But I think that if clients are actually getting value out of it and they're not being ripped off and they're, they're actually saying, okay, for the things where I need a genius, I'm going to use Claude or Codex. And the thing where I is kind of just data entry, I'm going to use a very cheap model.
- [13:38]
A
I mean like Deep Seeker, Deep Seek.
- [13:40]
B
Yeah. A Chinese model, a, A lesser, more commoditized model then I think the, the customer benefits. And that is long term good for AI adoption and long term good for revenues. Even though maybe short term, I mean, short term it's kind of good to rip your customers off, let's be honest. Well, okay, but long term, no.
- [13:58]
A
Today's episode is sponsored by Direction Funds. They just launched an ETF based on SK Hynix's new ADR listing. The Direction Daily SK Hynix Bull 2x ETF Ticker SKHL seeks 200% of the daily performance of SK Hynix through swaps referenced to the company's Nasdaq listed ADR. Whether you're trading on Q3 earnings, the high bandwidth memory narrative or AI capex cycle, SKHL may be a precision tool to amplify your Catalyst convictions. Trade SKHL from Direction Investing in the fund is not equivalent to investing directly in SK Hynix. An investor should carefully consider a fund's investment objective, risks, charges and expenses before investing. A fund's Prospectus and Summary Prospectus contains this and other information about Direction ETFs. To obtain a fund prospectus and summary prospectus, call 866-476-7523 or visit their website at direction.com significant risk involved Distributor Alps Distributors, Inc. I think it's smart to point out that most people don't use AI to its full capabilities. But to me, if you acknowledge that it feels a little bearish that something like Anthropic or OpenAI is going to maintain such a high revenue rate if they need to eventually raise prices. Will people be able to pay Those prices beyond 20 bucks a month if they're only using it for search? That's what I wonder.
- [15:30]
B
That's a good question. I mean I definitely use Gemini a lot for search. I would say it's Search plus, but those are not like Albert Einstein and a data center that I'm willing to pay a gajillion dollars for.
- [15:42]
A
And I think most people would agree with you is the thing. So if right now people are happy to pay 20 bucks a month or even 100 bucks a month for these various models once those prices start going up, which I think at some point they will, unless they figure out the token thing, we will see probably a drop off in adoption and they'll have to try to regain their tam, let's say, because they're going to figure out okay, who is willing to pay at this price point to make us a sustainable business.
- [16:12]
B
Yeah, well I don't think the consumer has enough of a of a wallet share willing to spend that it justifies the basically a trillion dollars being spent. It has to be the enterprise. And so I think Anthropic was the first to realize that OpenAI realized they were right and now is, you know, has raced to to grow in the enterprise. So it ultimately is going to be companies and corporations, the JP Morgans of the world, that are to justify the spend if it is justified.
- [16:35]
A
So I think Anthropic was Right. To lean into their enterprise mission statement. Let's say instead of going for retail, they're really trying to chase companies. Even I think about my own use. I have. So I pay for Gemini, I pay for ChatGPT, I pay for Anthropic. I don't know if I'm going to keep paying it forever for all three because in theory, if one becomes the most powerful thing by a landslide, you should only need to buy one instead of paying for three different subscriptions. But right now I use Chat for images, I use Anthropic for various types of deep research work and then Gemini use more day to day stuff. So I have different use cases for each one. So at some point even I have to ask myself, which one am I going to cut the payment for?
- [17:18]
B
Yeah. I also think that you and I are thinking very practically from a business from a revenue perspective and certainly there are people at these companies, OpenAI Anthropic who are thinking that way. But there also is an element that probably you and I don't have that is is prominent in Silicon Valley of we are building a digital God and we are creating artificial super intelligence and this is going to rule the world and if we don't create it, someone else is going to rule the world and we're the good guys. So we need to build it and we need to build it before China. We need to, need to build it before China. And so I think that the AI capex could continue well in excess of the return on invested capital. So that's why I think that maybe where we are right now in terms of how much spend is justified, but that we're approaching a frontier and that in six months it won't be justified, but the spending is going to continue because we have to be China.
- [18:04]
A
So.
- [18:04]
B
And we have to create the, you know, we have to create the machines of loving grace, like Dariama Day says,
- [18:10]
A
and we might get there. I'm optimistic about generally the smart people we have working in these AI companies. If we get to the point where the market starts to act like, okay, we're not seeing the ROI on these hyperscalers, where do you think that capital rotates to from an investor perspective?
- [18:30]
B
Oh, everything that's been losing. So consumer staples software, by the way, I have to say, like, if there are people who are bearish on AI that is totally fine, be bullish on software, be bullish on the things that are going off. Like if, if AI is a total flop. Salesforce is a screaming buy. ServiceNow is a screaming Buy Intuit is a scream, but basically everything in the IGV ETF maybe other than Microsoft and Oracle is a screaming buy if AI is fake, which some people believe.
- [19:00]
A
Okay, so, so I'm saying that if
- [19:02]
B
you're, if you're told skeptic, you say, oh my God, you're. These people are pumping these totally worthless AI stocks is a total bubble. Don't just like be sitting in cash, like be, be buying software stocks that have crashed because the market thinks AI has turned the software stocks into a bubble and it's popping.
- [19:16]
A
It's like vote with your portfolio. Yeah, yeah, essentially. What do you make of the. We saw IBM sell off 25% in a single day. I think it was the worst day ever for this 100-year-old company. What did you take from that?
- [19:29]
B
I think it shows that there is a massive unwillingness to spend at the enterprise legacy software. So I think technically, like, you know, we've never worked in a company that is like spending at this level, but it's really old school stuff, literally powering mainframes. So like literally like not even a data center, but just like a giant computer, a computer like the size of this room and the software that runs on that. Interestingly well actually I'll just, I'll just say that. So I mean IBM CEO said we are, we had some trouble doing some sales because all of our clients, they wanted to buy chips, they wanted to buy Nvidia, they wanted to buy memory, they wanted to link up to the data center. They wanted to do this in this whole new AI economy, not the old economy. So I actually think that enterprise, it shows that enterprise software is being challenged. And Phil, as you know, these enterprise software stocks have been absolutely decimated. And because of a fear of AI and the actual results, the earnings results haven't been that bad. They've actually been good. But we haven't seen the A result of. Oh my God, companies are actually stopping their spending on enterprise software. This is the first time we actually have seen that. By the way, there are companies that are publicly traded that do what IBM software does. And actually one of them is, is Broadcom. So like roughly 15% of their revenue is in software for enterprises specifically most mostly for. For mainframes. And they bought in like 2018, 2019. They bought a company called People can Look it up, but I think it's called ca.
- [21:03]
A
So are you saying that if IBM sold off, why didn't its peers sell off that do the same thing?
- [21:09]
B
Oracle does the same. Microsoft kind of does the same and Broadcom is the same, but they also have so much AI exposure. So if, if the, that part of Broadcom was a publicly traded stock, then I'd be like, why didn't it sell off? But Broadcom, I mean they're making tens of billions of dollars designing chips for that are for AI, for, for Google and Amazon. So yeah, I'm not a bear on, on Broadcom, I'm bear. I'm a bear on Broadcom only relative to semiconductors broadly, which I'm a wild bull on.
- [21:36]
A
Okay, is there anything that looks compelling to you in the software basket? I personally, I don't buy any software. I'm not in it right now. I haven't been in it since the AI trade really kicked off a couple of years ago. Do you like anything there?
- [21:54]
B
I own ServiceNow.
- [21:56]
A
Interesting.
- [21:57]
B
I think that there's a chance that they actually are going to do very well in the age of AI agents because it's so embedded in the all the Fortune 500 companies, like the biggest companies in the world that those companies are going to manage their AI agents through the ServiceNow platform. So I own ServiceNow. I also own a company called Verisk that is an insurance data provider and software provider. That was an extremely high quality business and in an age of AI it may kind of be not as high quality. So that's an example. Like if AI is fake, Verisk is going to do so well and is so cheap right now. But I think that Verusk could be flat to down over the next 10 years if, if this machines of loving grace AI superfuture is here, which I'm, you know, 50 50.
- [22:43]
A
So Steve, Steve Ho. Who?
- [22:46]
B
Nice.
- [22:47]
A
He is with Silicon Data now. He's a quant researcher. He was on the show recently and he said he liked FactSet as a stock because it's taken a hit. It's a sort of data software financial services company. But he thinks that again, AI will make this company stronger and it'll actually deepen its data moat that it has and it's developed over many years. I thought that was a pretty interesting pick. Reminds me of what you're saying here, Jack. Let's talk about what we've seen in the banking sector. We've seen ridiculously good earnings across the board. Morgan Stanley, Goldman Sachs, JP Morgan. Everyone's crushing it. What's going on here?
- [23:26]
B
I've never seen the earnings this good, Phil. I guess maybe they were as good in 2021 when they were booking revenue from reversing credit losses they took in 2020. So maybe they were that good as good then but, but other than that I've never seen them this good on credit. They're doing really well. Their net interest income, they're doing really well. On the Wall street side of course they're doing really well. I think because of the SpaceX IPO like JP Morgan and Goldman Sachs just booked a simply ridiculous amount of revenue. So I, I would not expect it to be as good just because it was so good. You know it can't continue. Of course it can. But also just in equity trading like I don't know, the equity revenues from equity trading are up so much and you think about the inputs from, from that, it's like how much people are trading, the volatility, the level of the markets and obviously all three of those are up. But I feel like the revenues from equity trading is going up way more. I don't know if like the bank CEOs have given, have invested more in that business and they're using more of their bank capital balance sheet. So maybe there. But yeah, it's, it's really good. I have some somewhat niche investments there but I think you know, it's been a phenomenal place to be as an investor just looking at the earnings, maybe the stock prices have gone up slightly less than the earnings.
- [24:44]
A
I've had a lot of investors tell me that coming into 2026 they loved financials and it's been fairly unloved I would say in recent years. But financials have certainly picked up I think this quarter. Obviously they're crushing it and I think that's probably going to continue. We have a bunch of new IPOs coming in but by the way there's
- [25:03]
B
so many, we're talking about banks. So many non bank financials are so good. Visa, MasterCard, S&P Global, MSCI, Moody's, like these are incredibly high quality businesses that you normally don't get to buy at the medium multiples they are. Normally they trade at a huge premium to the market. I also would add ice, you know, Intercontinental which owns the New York Stock Exchange and other things and also CME people think that perpetual futures are going to crush these businesses and maybe they're right, but I don't, I kind of don't think so. So there's opportunity there.
- [25:35]
A
What about stablecoins? Isn't that a big potential bearish headwind for them?
- [25:40]
B
I think the odds that stablecoins replace or displace consumer credit applications, so credit cards are exceptionally low. I don't want to say 0% because then if I'm wrong, you know they're going to play a soundbite like legendary investors like Jim Chanos who've made so many great investments on the, on the bearish side one time, like he said, short Tesla. And then everyone says in the comments, I don't want to say 0% but I just don't think it's, I don't think it's going to happen. And I, I've been arguing with like some of these business school professors on Twitter who like about stablecoins and I'm like dude, you don't understand how these things work.
- [26:13]
A
Like I'm not sure I fully do. What are they saying they disagree with?
- [26:17]
B
Literally it's, it's illegal for banks to make loans with stablecoins. Stablecoins are based off of the Geni
- [26:25]
A
act from last year and, and the business.
- [26:29]
B
So, so the people who would have, if, if in 10 years you and I are buying a Starbucks coffee with stablecoins as credit, not debit, but credit, but credit, it would have to be either they, they'd have to change the laws which just got passed by Trump or there would have to be a decentralized lending protocol that's going to compete with JP Morgan or Citi or any of the big credit cards. So it's not going to happen now on debit cards, maybe on business to business, especially cross border, I think that that is a, that is a play. So I think yeah, I mean there's some, you know, maybe a little circle, maybe a little coinbase there maybe. But I'd say, I'd say I'm a stable coin. I have an open mind, I'm not a hater. And there's a lot of thing, it is growing very rapidly from a very small base but a lot of it is still used for you know, cryptocurrency like speculation. And the percentage that's growing from like remittances and cross border payments is a much smaller percentage than like just looking at the amount of, what's it called, you know, tether outstanding.
- [27:34]
A
Okay, I think that's fair. Jack, I want to go through a few stock picks that you like right now. You sent these over right before this. The first one you have here lam research. It's up 84% year to date. Why do you like the stock?
- [27:48]
B
I think that the semiconductor capital equipment space for Micron for Taiwan, Semi or Intel to build enough chips to stop this, to cause prices to go down, to implode the bubble that we allegedly have. They're going to have to order so many machines from asml which reported today from LAM Research. So ASML is in lithography, LAM Research is in etch and deposition of the chip and they also do a few other things and they have a very dominant position. So I think one thing that people may not know is just how oligopolized or in some instances for asml, for EUV lit's a literal monopoly that these companies are. So even memory, which is the technology is not at all an oligopoly. Like it's easy. There are only three companies as you know, that have, that are in this space because, because it was such a brutal competitive industry. So you know, there's actually a fourth company, a Chinese company called CXMT that's actually going to, you know, be ipoing pretty soon, I think. And so that is, that is the bear case of China is going to kill the market. And that's also a potential bear case for Lam is that so much of their sales are to China. So oh my God, Trump is going to pass a law and he's going to ban Lam Research from selling to China. I literally was just listening to a clip of Trump talking about how much he loves President Xi and how much he said presidency is so good looking and so tall that if there was a movie made about him they couldn't cast anyone in Hollywood. Cuz no one is good looking enough of the male actors in Hollywood to play President Xi. So I'm like, is Trump really going to ban Lam Research from selling to China? I don't think so.
- [29:27]
A
Okay, so tell me this, it's already up a lot. Yeah.
- [29:31]
B
I didn't answer your question of why I like it though.
- [29:32]
A
Sorry, none of that was why you like it. I thought you did. Okay. Would you like it buying in at these levels up 80% this year?
- [29:40]
B
Yes, I actually, I bought some last week at what it was probably 5 to 10% higher. So I, yeah, I definitely do. And I think that these, these companies are very, very, very solid and they also have a high recurring revenue base. So because the LAM research as it sounds, etch deposition, it's very, it's a very violent process. So actually these machines needed to, they need to be replaced and had to have their parts replaced a lot. So lam Research, roughly 30% of their revenue is recurring revenue from servicing the machines and selling spare parts to the machines. So there's a chance, let's say the Edron, this is a giant bubble. But let's Say it pops in like 2029. There's gonna be so many lam research machines ordered over the next two years. And in 2029 as the bubble is popping, there's going to be a lot of recurring revenue from these, from these machines. And I think that's also true of ASML which reported today, which you know, by the way these hedge fund guys, they're, they're pretty smart. Like I interviewed a hedge fund investor, Vals latv, probably one of the best in terms of Sharpe ratio semiconductor performances at scale like I think two or three billion dollars under management. And he was like, he actually liked memory and did not like the semiconductor cap equipment names because they could only grow their volumes 30% whereas memory he said could grow more just by pricing. And, and ASML today said we're going to grow our volumes by 30% exactly. So there you go, he checked it out. So that is the bear case of there. How much can they grow their revenues? Probably only like 40% because they grow volumes by 30% and then pricing and then the tailwind of the recurring revenue is, makes it, makes it 40%, maybe 45. Whereas like memory. Look, I, I'm not, I don't mean to insult anyone in the memory business but like they've kind of ripped their entire customers off like that they may be ruining the party. Like I mean if I was at Metta or Microsoft or Google and I was spending $100 billion and I'm like I'm paying all this for what on memory? Like these people need to play ball, these people need to stop ripping us off. And maybe, maybe it wouldn't be wise for me to say that, you know, politically or just in terms of, you know, relationships with the companies. But I do think that is a, that is a risk that like the, the money that is being spent on the Capex, they're not buying more chips, they're not buying better chips, they're just paying more because Micron is ripping everyone off.
- [32:04]
A
Well, we had Apple raise prices, Microsoft raise prices on their consumer facing products, laptops, consoles, things like that. And apparently according to the reports that was in direct response to rising memory prices.
- [32:15]
B
Oh it definitely was 100%. Yeah.
- [32:17]
A
Okay, so I want to get to this next stock here. Texas Pacific land, not an AI play. It's up 40% this year. Tell me about this.
- [32:25]
B
It is an AI play, is it?
- [32:27]
A
Yes. Okay, tell me.
- [32:28]
B
It owns a lot of land in the Permian Basin. It got that land from a bankrupt royalty, bankrupt railroad in I believe the 1870 1873. And by the way, shout out. I interviewed Leoquat Ahamed, author of the book about the bust of 1873. The. The bubble, which was railroads. Okay, so they own their giant landowner. They have royalties on the oil in the land. They also own the surface on the land. So oil companies pay them for drilling the oil and also for the right for the privilege of having their rigs on the land. And they also sell the oil company's water and sell oil companies the right to put the dirty water back into their land. So that's. It's oil and gas royalty play. Optically it trades at a probably 50 to 60 price earnings ratio, which can look expensive, but it is really not. Like I think one thing I've learned is like, you know, a crappy oil company trading at 8 times earnings is so much more expensive than TPL at 60 times earnings. And people just have to, people just have to take my word for it.
- [33:32]
A
But, but that's not a good thesis.
- [33:34]
B
Yeah, well, okay, well, well, I'll just say like look, look at how, look at the returns on equity of revenue. Yeah, I didn't share thesis, so it's not, it's not, it's not a good thesis because I didn't share a thesis.
- [33:45]
A
Why is it an AI play?
- [33:47]
B
Because the former chairman of Google, Eric Schmidt, is going to be building a data center on Texas Pacific land. And he said up to 10 gigawatts, which is a ridiculous amount. I think that in the near term, like I think that they could make one gigawatt. This will be over the next few years. So I still like it. Even if this doesn't happen. The stock probably goes down, but I still like it. But this is an added, added tailwind and this really is the shortage. There's not actually a shortage of GPUs. That's what a guy, Ben Paulian, who I interviewed, who knows a lot more than me, who's a giant Nvidia Bull, he said the real shortage is in powered land and Texas Pacific has that.
- [34:24]
A
So as you know, I run something called Best Ideas Club and it's a stock pick every week and we hold it for 52 weeks. One of the stocks we published I think in January was Texas Pacific Land.
- [34:38]
B
Down to them.
- [34:39]
A
It's down since then. So it's not been one of the best performing stocks in our Best Ideas Club, but it gives me a little peace of mind. If you're coming on bullish at current levels, that's a good sign. Last stock you have here S and P global down about 8% this year. Tell me about the stock.
- [34:57]
B
I. Interesting. I would have thought it was down a little bit more. I'll take your word for it. So it's down because they have a software business which the market hates software. The market says AI is going to replace everything. I think even if that happens long term, it is a good idea to buy the stock at this level. It's going to take a while even,
- [35:15]
A
even if that happens, like what is the company.
- [35:17]
B
It is S and P Global. They have a credit rating business. They have a index business. They literally own the s and P500, the rights to it. So they make, I believe, over a billion dollars just from their index business, which is an extremely high margin business. A competitor to it would be something like msci, which I, which I also like. And a competitor in the credit rating business would be someone like Moody's, which I also like. And then there's Fitch, which is private. The. So the credit rating business is also an exceptionally high margin business. And I mean yes, they own, they get more from like securitized rating, securitized products and high yield than investment grade. But there is, so the boom in credit right now is investment grade. Like all these companies, you know, issuing 40, 50, $80 billion of investment grade debt that I think is good for the stock. And then I don't think that they're going to be hurt that much on the software. And also just from an accounting point of view, their earnings, what they report as their earnings is lower than how I perceive their real earnings to be. They're taking massive depreciation charges on intangible assets that I even Warren Buffett, who's the king of like probity and accounting, says that in this particular instance it actually is appropriate to adjust earnings in the. In favor because like if I buy a company for $10 billion and my earnings are $1 billion, but I'm taking like an $80 million charge from depreciating the assets that I bought that my earnings are going to look lower than they actually would be. So here is the rare case we'll actually say EBITDA is actually looking a little bit better than net income.
- [36:59]
A
Fair enough. I've certainly not heard anyone talk about S and P Global that much recently, which I guess is a good sign, right? You don't want to crowded trade, Jack. Let's say we're building a portfolio for me, 30 year time horizon, pretty high risk tolerance. We have $100,000. Where do you start allocating this I
- [37:21]
B
put 5% in ASML, 5% in Lam Research, 10% in Nvidia. That sounds so drastic, but really about the market weight. Yeah, exactly. I put a little bit in, actually. You know, I mean, do I want to put Micron in there? I don't think I do put Micron in there. I don't, I don't think I do it, you know, and then I'm going to add a little. I'm going to add a little Berkshire Hathaway. I think Berkshire Hathaway is a very underrated AI play, even without Buffett. Yeah, yeah, that doesn't matter. It's about the legacy that he built, the assets he built. The insurance company is great. I think insurance is going to get better with AI. I mean, if insurance doesn't get better with AI, AI is fake. Like, because, you know, and then the railroads are actually, they own a lot of land. I think data centers can be made on the railroad land. And I think they could make EBITDA of longer term, like in 10 years, maybe, maybe, maybe a billion. Maybe something like that. And BNSF and Union Pacific are the two best railroad plays for that. Obviously, Union Pacific publicly traded and BSF is owned by Berkshire Hathaway. Berkshire Hathaway. Also. One of Warren Buffett's famously bad deals was buying precision cast parts. They actually make the blades for the turbines in G.E. vernova and Mitsubishi Heavy Electric. So, oh, look, that's like 20 basis points of the revenue. I'm not going to say it's a big thing. And then also they own Apple, which I think, you know, is going to be a giant beneficiary of AI. Just, just because from their services thing, they're going to have to be on the, on the, whether it's anthropic, whether it's OpenAI. They're. They're going to have to pay to play, to be on. To be on the iPhone that 15, 20% tax. Do I. I also like shorting Meta. I think, I think Meta is a nice.
- [39:07]
A
In the portfolio. You're going to short Meta.
- [39:09]
B
Oh, okay. No, no, because, because it's for you and because you like, you like Meta. Okay, so let's see. I'll put some S and P Global in there, some Visa in there, some MasterCard in there. I'll put some Alibaba on there. I'll put some Tencent in there. But I'll do a little. I'll, you know, I'll buy it through Naspers, which is a South African play. So you actually get it At a cheaper valuation.
- [39:35]
A
Let's see you go in sectors, ETFs, like what else fits into
- [39:43]
B
maybe semiconductors? I'd put 5 to 10% in SMH. That's the Vinneck Semiconductor ETF, which I think is a little better than the BlackRock one, which BlackRock one is very heavy memory. So I think it's a little, a little, a little heavy there on the, on the exposure. Let's see, I'd say some, a lot of other single stocks. So I say mostly single stocks. Like, I mean I'll just throw one name. I don't own this name but like Disco is a Japanese blade company. They make the diamond encrusted blades to split the chips in AI and they have a very dominant position and it's also recurring revenue because the blades break a lot. So it's like that. I mean, let's see, there's a lot
- [40:30]
A
of names for one. One portfolio.
- [40:31]
B
Yeah, well, I mean I like being somewhat diversified, you know. I mean.
- [40:35]
A
Yeah, I. Okay. You wouldn't go into VOO or something like that?
- [40:41]
B
That's pretty good. I like that.
- [40:42]
A
S and P or just something like something vanilla in the corner. Yeah, is how I think.
- [40:46]
B
Yeah, sure, sure, that's good. I mean it's what my client wants, you know, if that's what you want. I do think that S and P Global is going to beat the s and P500 index over time with way more volatility, so people say. Oh, so it's lower Sharpe ratio. Okay, fine, whatever.
- [41:01]
A
Okay, I like it.
- [41:02]
B
But yeah, I don't like matter here. I don't like Space X here. I actually think that both of them are pretty good shorts in a market that is very tough to find good shorts. Because also, I mean like, let's be honest, people who say like they find shorts but they're short like Pepsi, that's not a short.
- [41:17]
A
Like why not?
- [41:18]
B
Because. Because it's such a low beta stock, the odds that it goes up like you know, being short 1 meta or being short $100 of SpaceX is so much more risky than being. It's a fake short. You know what I'm saying?
- [41:33]
A
Not really, no. Because it's not a volatile stock.
- [41:37]
B
Shorting it to have the same level of risk, you'd have to short $500 worth of Coca Cola than to short $100 worth of SpaceX.
- [41:43]
A
Got it?
- [41:43]
B
Yeah.
- [41:43]
A
Yeah. Okay, so I want to ask you real quick before we let you go, you don't like Meta. Is it because of the current price or because of its fundamentals or what? Why not?
- [41:56]
B
I think that their AI strategy is, does not make a whole lot of sense to me. And I think AI could be a giant revolution and Meta doesn't win. And I think if AI is a giant bubble, they are probably the most exposed. They're literally spending $250 billion on their data center in Louisiana for. For what? To make better Instagram ads? Or are they really going to be that much better? I mean, and also the Meta glasses, I don't love them and I still think that. I actually think Apple is long term going to win because really what people want is something that's connected to the iPhone and Meta isn't connected to the iPhone. That being said, if you're going to launch such a bold product, which I'll give Mark Zuckerberg, he's very bold, I think Kylie Jenner is the brand ambassador to do it. But I still don't, I still don't think that. I think sales are gonna be good this year, but I think that it's not gonna, like, it's not gonna be an iPhone.
- [42:51]
A
I don't think, I don't think that the bull case of Meta rests on the glasses.
- [42:56]
B
What is the bull case of Meta?
- [42:58]
A
They're advertising. There is no business in the world that if they're doing advertising, they are using Meta. They're running Meta ads across every platform. I think WhatsApp is a giant. They're figuring out the monetization. But if you are a enterprise customer of Meta, you are a lifelong customer, period. So I think it almost doesn't matter. Sure, they got the data center going, they got the glasses, but if you buy them as an ads business, that is a forever revenue generator. That's my view.
- [43:28]
B
The question is how much money they're spending on the data centers now. Is the depreciation expense going to be more or less than the. Than how much revenues are going to go up?
- [43:38]
A
Yeah, it's good. And I don't have the answer to that. And to be clear, I personally do not own Meta.
- [43:44]
B
Oh, I thought you did.
- [43:44]
A
Sorry, sorry. I don't own Meta. I like Meta and I think it's fallen way too much over the last year. And I also like Zuckerberg as a founder. I think he's a very visionary guy as far as what he's been able to build and how he's pivoted every step of the way over a couple decades now. So I wouldn't bet against Zuckerberg. And I also think their ad business is a forever business.
- [44:09]
B
It's a terrific business. I'm not, I'm not denying that like
- [44:11]
A
they'll never, they'll never not make billions and billions of dollars from ads whatever else they want to focus on.
- [44:17]
B
I think that's probably true. I say of, of the Mag 7i, I like Google and Amazon a lot more.
- [44:25]
A
That's fair. And I, I like Google way more than meta. I own Google. I've been in Google for a while now and I think that's also a great forever company. Right. Okay Jack, where can people find your work online?
- [44:38]
B
They can search up the Monetary Matters podcast. I do interviews twice a week and I'm posting on Twitter @Jack Farley96 Perfect.
- [44:48]
A
I love your show. You do a great job and you interview some of the smartest investors and hedge fund people in the world. Thank you for your time. We'll do this again soon.
- [44:56]
B
Thanks Phil.