Transcription
Uh, I will be flying out Friday. So, these there's going to there's not I'm not going to recap the week since I won't be here for the the end of it, despite where things are trading this morning.
Uh, but this is an important one in my opinion for you guys because I am going to spend time on some themes that I've talked about but try to put them into context for the most part around one or two podcasts that I listened to this week. One with Craig Fuller, one with Dylan Patel. That I know I give you guys the podcast info. You can read the summary. I give you the YouTube side, but I highly recommend listening at least to the Craig Fuller one. The Dylan Patel one, if you've listened to other ones in the past, it's very similar.
So, but this is really about the physical upgrade is here now. So, that was the um the theme of my my piece for this year was the physical world upgrade, which was all about leaving this world of the cloud, software, the hyperscalers, them spending money and needing, according to Jensen Huang, 90 trillion dollars to upgrade the entire physical world with artificial intelligence. And we started that process early. It's very semi-heavy. It's power-heavy. It's optical fiber-heavy. That's what the themes are in my thematic portfolios with all the names. And by the way, thank you to everyone who has reached out, especially after last week when I went through a lot of the different details so that you guys can be involved in these names. They the thematic basket continues to just rip. I'll go through some of the slides, obviously, but this theme is in the very early innings.
So, when you see these parabolic charts, yes, there's going to be pullbacks and they will be severe. But, if you continue to waste your time trying to pick bottoms in software, if you continue to waste your time trying to wonder when when is Nvidia going to outperform Marvell? Or when is it going to outperform Texas Instruments? Or when is it going to outperform Intel? It's not going to happen, guys.
Um, this side is so massive. The concentration of the cloud, which was a 17-year trend, and then the IQ raising part of the pre-training of GPU, this stuff is so underinvested. And I through the the webinar I did this week for the subscribers, I highlighted that the semiconductors are now the largest weight at a level two gig in the S&P 500 at 17% and growing. That is not a bubble. Last year at this time, software was bigger than semis. Software's gone down, semis have gone. And at 17%, that's about 10 trillion dollars. But of the 10 trillion, Nvidia, Broadcom, and Micron make up 7.5 trillion. You are going to see the other names in there grow rapidly over the course of the next at least five years, but probably decades. So, as I go through this, really spend the time on making sure that you're prepared for this because the numbers are massive and the money is coming from the Mag 7.
So, if you haven't read this, and again, if you listened to the Dwarkesh podcast, Jensen again reiterated the five-layer cake. If you didn't know what the five-layer cake was cuz I had some people reach out. March 10th, this is on a blog on Nvidia where he goes through this is where we are right now, guys. This is it. There's five layers to this with energy and chips as the base of this. This is where these three is where all of my attention is on making money. The models and the applications are more for the agentic side in terms of the use, the profit margins, the digital biology, the robot. All of this stuff, but this is where the money's being made. And again, this is all underweight cuz this is industrial. He said it would be 90 trillion. 90 trillion dollars again, the global economy is about 120 trillion. That is a massive infusion, and that is why when you go through and look at some of these semiconductor charts, and I'm going to highlight some more today that have just started, you start to realize why this is going on. And that is why I created a thematic basket, which is all related to this movement.
So, here is my thematic basket now up 34% since November 24th. The reason I chose November 24th is cuz that's when Opus 4.5 was released, and that was when software started to truly get disrupted because we entered the agentic age. This entire basket is the agentic age. That's what it's for. It's for the buildout, it's for the infrastructure, it's for everything that you're going to need. These lines here, well, this is software, still down now 16%. It had a nice bounce. It pulled back. I still, if I have to guess, these lows will hold for the IGF. I don't think it will hold for all software names, and I want to make sure I I make that clear. I do not want to own a single SaaS software. I think you are wasting your time trying to spend time on software playing for this mean reversion. There is no mean reversion. Software is a dead thing in terms of built-on code. There will be some names that can survive, but just like Ford has survived for 35 years without the price going up. I think it's a waste of time to try and think that there's winners.
Now, the problem with that is the S&P 500 is in here. Up 4%, 5% for the year. Heavily weighted into that. The white line is the Dan Ives ETF. Like I said, love Dan, got to know him in the last month. Unfortunately, a lot of the names he has in there are related to the web 1.0 playbook, the 2.0 playbook, the software names. And again, that just acts as a dead weight just like for the S&P. The hyperscalers right here, dead weight. I want to use those as a funding vehicle. They're the spenders. You want to be long semis, long the infrastructure side, and you want to be short the hyperscalers. It is going to be very similar. If we have a problem where we run out of compute, the hyperscalers are going to get hurt as well because they're depending on the ROIC and the adoption. These are directly linked. They need to spend money on the compute side, the semis will have 30% corrections just like we saw with Micron from 460 down to 330, but then Micron went right back up. They won't always go up that fast, and I'm sure we're getting closer to a time where they start to sell off just because when Marvell goes from 95 when I write a paper, it closes that day at 87, and pre-market today it was at 170. Those types of moves are building in a lot of what is going to happen over the next year in earnings.
Um, just as a reminder, and as we get into this industrial shakeout, this is the reason why I wrote this paper back in July. So, this is not a new theme for me. You can go read what it is, but basically, this was about not being like traditional demand cycles around real estate and autos, but by a new and powerful force, paranoia, across both governments and corporations, a deepening fear of falling behind in the global AI arms race. That's what is driving PMIs. I still read tons of people focused on real estate and autos and saying that we're going to have an issue because that's where growth is. That is not where growth is. And if you want to go find out, listen to this Craig Fuller interview on Adam Taggart Thoughtful Money back in November. This was before Opus 4.5 came out. And his thing, Craig is a freight transport expert. He was saying the real economy is very, very, very weak. And he was very bearish. And just to go through, he was describing that freight is a leading indicator is a better real-time leading indicator. The goods economy is much weaker than the headline data suggests. Current CapEx wave is still mostly hope-driven.
Now, I love when you get smart people who are wrong, and they turn around and they do the exact same podcast five months later. Miracle turnaround. The US industrial economy is now booming despite high oil prices. And that's one of the things I want to make sure you guys recognize. If you've been listening to the oil doomers, we obviously are going to have an impact in inflation, which I'll go through. But in terms of the global economy stopping, AI is more important. I said it last year with tariffs. I said AI is the only story. AI is the only story. It is the only story story you will need. It's the reason why I'm doing what I'm doing and trying to bring you guys information that is not only real-time. There's a lot of information in this week's video that is actionable because it just happened this week. So, just keep that in mind is that it's moving so fast that you can't depend on sell-side research to stay on top of stuff. You have to be on the podcast. You have to be listening to Jensen Huang. You have to be going through X. And you have to be reporting it.
So, here are the highlights. I'm not going to go through everything, but he said November, he flipped from November to now as bullish as he's been in years. The goods economy is now roaring. Rail shipments up. Chemical shipments at record highs. Remember, chemical moat is part of this. I'm going to go through chemicals again in this because that is a late cycle part of this. You are going to see chemicals. You need to be involved in petrochemicals, particularly US ones or ones that are going to benefit from lower natural gas prices versus the rest of the world. Truck tonnage at a three-year high. Truck loading postings at strongest level. I've read people that said that this was all about Avis in terms of transports. You have to stop listening to permabears and people that get caught in this situation. People like Craig Fuller that were bullish in November, and I could give you a list of people that were bearish in November that have not changed their view. They just move on to the next thing. Bonus depreciation is helping. Cheap natural gas is helping. It's no longer being led by imports but by domestic production. It means the middle of the country is what's driving GDP, and the coastlines are suffering. Keep that in mind when you're reading the surveys. Nothing in the freight data suggesting the war is hurting industrial demand. Insensitive, non-cyclical AI has the capital coming from free cash flow, and even if that free cash flow goes negative, these companies have massive debt, they have massive equity, so debt to equity will rise. It's not good for them. That's why I say be short the hyperscalers, all of them. I don't care which one it is, including Google. Will they go up? I'm sure they will go up. Are they going to go up as much as the S&P? Not in my opinion, because semis are going to drive the S&P performance. Optical fiber is going to drive the S&P performance. Caterpillar is going to drive the S&P. You have to be long industrials, materials, energy, and semis. This is the 1970s again and again and again. This is a physical upgrade.
Flatbed rejections hit the highest levels ever recorded. So, flat they have no capacity. There is zero slack. So, again, if someone has told you that trucking is not happening and this is fake, they're just wrong. And if you want actual factual data, here are the imports of capital goods relative to consumer goods. So, I want you to think about what this is. This is, "Hey, we need tons of semi, we need tons of uh uh of parts for the data center. We need things from Switzerland. We need gas turbines. We need all of this stuff." This is this is back to 1991, guys. This is the ratio of them. This is all about the opposite of what was going on from this period, where there was no capital goods relative to the consumer goods. We are not buying furniture. We are not buying things cuz there's no housing market at this point. Deal with we'll deal with that later, but there's confirmation that the industrial boom is happening. And if you don't believe that data, and you don't believe that um oil isn't hurting the economy, how about we go to Asia? Taiwan machinery, all-time highs. Kospi machinery, all-time highs. And then obviously you have the S&P machinery. They're all moving together. This week, United Rentals, a good proxy for not just the industrial economy, but also when the turnaround is starting. We had the largest day uh on Wednesday for United Rentals. This is in coming out of COVID. This is coming out of the worst of the housing GFC time, and this is coming out of 1998 with LTCM. Yes, Avis. Here we go. This is year-to-date. Avis is up. Who cares? The entire group is up. And this is after Avis had collapsed 70%. This is not about one stock, and don't let people uh try to direct you uh misdirect you with the bearishness. Prices are moving higher. There's no other way to say this.
Um I will keep saying it. We are going to get CPI above 4%. If this holds, this little thing here, we're going to be above it very soon. When we get CPI above four, I don't think the Fed is raising rates, but that will be above three-month bills. So, we will have negative real yields again. I think that's an important story, but the most important part, this is from this week's S&P uh data. This includes the service PMIs and the manufacturing. This is not just manufacturing. It will spread a little bit into the service side, because technology is going higher. So, remember, semiconductors are in everything. When you walk into a doctor's office and you're getting a service done, if they need new computers, if they need new machinery, if they need new MRIs, it's all being impacted by commodities and especially by semiconductors. Here is the services part of the uh PMI. This is the ISM PMI for the service side. And again, I just want to highlight, we ripped to the highest level since 2022, when year-over-year CPI was up around 5 and 1/2%. Here you can see the long-term trend of year-over-year CPI. Again, I find it hard to believe we won't get above 4%. There's fiscal stimulus that just continues to show up. Warren Pies put this out and basically said, "Tax receipts much lower than expected. So, we are seeing tax refunds tracking higher versus last year. Tax received for the government's much lower, so we're getting a fiscal push from the one big beautiful bill that's coming at the same time the gas at the pump has gone higher. So, for people that getting refunds that are much bigger, those are probably the same people that live paycheck to paycheck that are being offset by the oil prices. So, that's one of the reasons why we're not some seeing some big collapse. The other one is, as I've highlighted in previous videos, we are an exporter of energy, so it benefits the economy.
Here's the raw industrial uh uh commodities. If the Iran situation was having an impact, you'd see the metals going down. They're not going down. In fact, Robert Friedland uh from Ivanhoe, the entire copper industry just returned from last week's leading copper conference, and basically uh saying they're even more bullish that all-time copper highs could be tested in the coming weeks. Uh there's all kinds of uncertainties. There's all kinds of bottlenecks. Uh if you want to get a sense as to how much people are investing now in trying to avoid the energy situation, uh solar batteries, EVs, through the roof from China in terms of exports. If you want to know what China is buying, remember, silver is part of my uh thematic uh portfolio. Uh it's lagging here, which means it's a great entry point in my opinion uh based on what I see and have written about on silver. Well, China just imported the most silver ever in March. And remember, they stopped exports December 31st. Surprising move here. I think this has implications if you're looking for something that could slow everything down. Uh this is the ADP weekly employment. We've had four weeks in a row now of kind of moving higher. If for any reason we get an unemployment report at any point, which is greater than 150, when everyone is now moved down to zero is okay, I think with inflation moving higher and a stable labor market, you will start to freak out in terms of whether the Fed may go.
Um John Roque uh has been pounding the table on this chart, and this would argue for this being a risk, which is technically he does not like two-year uh notes. Uh and this is from a price perspective, this would mean yields going higher, and his first level is up at 4%, but his target is five. We're currently in the upper threes, uh around three and three quarters, so anything that moves in terms of the Fed possibly going could be something that kind of freaks out the market, even though I don't think the Fed is going to go.
Texas Instruments reported. So, real-time data week. Uh they said, "Industrial demand keeps improving while data center segment flourishes." So, we are now getting the physical upgrade spreading into the industrial side. Texas Instruments is a demand barometer for the industrial economy. They cover so many different things, so recent commentary notes that industrial demand is recovering broadly. Comments about AI-related sensing and power needs suggest that AI workloads are starting to penetrate industrial and edge form factors, meaning robots, test equipment, building assistant, EV, industrial vehicles, not just GPUs in the cloud. This is critical, guys. Remember, we just increased the uh spending for defense. We are getting into the edge. The edge is everything. It's drones. It's humanoids. It's autonomous vehicles. It's computers. It's phones. It is enterprise edge. This is the beginning, the very beginning. This isn't even the first inning. I'm not even sure they've gotten through the national anthem yet. Texas Instruments expands microcontroller portfolio and software ecosystem to enable edge AI in all devices. And again, I bring this up and mention improve energy efficiency when processing at the edge and neural processing units. If you guys have my paper from NPUs, I would go bring it up and go read again. This is all happening now.
So, if you don't like Texas Instruments, how about we go to Intel? They emphasize that AI workloads evolve, the importance of CPUs infrastructure is growing. I spent a lot of time on Intel last week, going from dead just seven months ago to breaking through all-time highs, and now we're ripping through them. Up this morning, 20-plus percent. You cannot put a GPU everywhere. We're moving into inference. Again, this is the very beginning. Intel and third-party analysis highlight that CPU demand is being pulled not just by cloud, but also by enterprise AI deployments that re- rely on CPU-heavy systems. Again, this is part of the edge device side. They mention Intel has flagged signs of improvement in PC-related CPU demand, with prior quarters already showing a pickup. One of the mistakes people are making, even if the prices go higher and PC overall PC demand is, let's say, lower than it was last year. Let's say handsets are lower than we were last year. Think Qualcomm. I'll be talking about Qualcomm more and more in the coming weeks. Remember, if you only do a certain this slightly less than last year, but of those, the percentage that are AI-related goes up, that increases all of these companies like Intel. It is the rotation. It is the, quote-unquote, as I wrote at the very beginning, the physical upgrade. Upgrade, upgrade, upgrade. You don't need to have GDP to have an upgrade. You can have zero GDP, and it can be a rotation. And when I mean zero GDP, I mean you can have zero growth, like you do in autos. You could sell 16 million autos this year, 16 million next year. And if this year none of them were AI, and next year 4 million of the 16 are AI. You had massive growth in AI. Well, that has enormous impact for semiconductor chips. That is what everyone is missing is they are focusing way too much on nominal GDP growth and not enough on the upgrade cycle. The Edge Awakens.
So, this is going out Monday morning. I just wanted to give you guys a preview. The Edge Awakens. It's all on news that happened this week and it's not just the Intel and Texas Instruments, which you can see I highlight here, but it's also Nvidia's reported 800 V DC power push in South Korea that was going through the media this week. I wrote a paper on the whole rack. Intel is in the whole rack. So, it's one of my names in there. The Edge AI investment [clears throat] universe, which I highlighted last week and went on the subscriber website, which has 68 names in it. Many of them are in the overall portfolio. But in particular, the power semis, which is what this is on, there were six of them in there. If you didn't see what they've done this week, and again, I put this out the Edge thing last week, this is what the power semis did this week and this is the chart. So, 5 years ago, they were unchanged as of last week. Those six names have gone through the roof, of which one of them is Texas Instruments.
So, I just want to remind you again because this was brought up in Craig Fuller. I highlighted this on the webinar this week. And again, I say this for all the FAs and RIAs that are out there. We've seen a lot of a lot of you guys signing up. Tell your other people that work for the firm, you should be paying attention to this. It's going to be very hard to get this kind of research. This is something that's worth the money and the reason is because I'm covering the space in a way that's real time. Power semis is a real-time story. It's in the very, very early beginnings. There's a lot of names on there that as far as I'm concerned, they haven't even seen their earnings grow in a meaningful way. You saw it with Texas Instruments, but the edge devices are going to keep coming because as I go through, we haven't even started humanoids. Like this literally is the beginning of a long-term process and last week I covered Terafab. Terafab is just your confirmation that Elon Musk believes we're going to run out of chips. You want to invest in scarcity.
6 months ago, post Opus 4.5 and open class. So, before we got there, these were all the things that people were worried about. This is what Craig Fuller was talking about. Here's where we are now. It has only been 6 months ago. All right, let's go through this quickly. The spending numbers, they're being revised higher. So, this is from BlackRock this week. Just look at these numbers. And again, this is the hyperscaler. These are still for the main companies. Remember, you've got other companies that are spending like OpenAI and Anthropic are each targeting now 30 gigawatts of compute. They're not included in that hyperscaler number. Then you've got SpaceX not included in that hyperscaler number. They're going to do 25 billion or so 25 billion or so. Then you've got Tesla this week announcing increased to do 25 billion. These numbers just don't stop. They are massive. And the way you can see them is ERCOT electricity demand. This is basically what we're talking about. So, here's where we are now. We're already at a problem. This problem gets worse and worse in terms of peak demand. Based on everything, and if you want to know what's causing us to be, look at these numbers. I mean, how are we going to get enough power? This is all related to data centers. This is the number out of this number total in terms of what is needed. And this point is being made. Remember when they screamed that Bitcoin mining would boil the oceans? Here's Bitcoin mining's expectations and this is where they were. We're dealing with numbers so massive. And again, look at the growth rate. This is all AI related. If you're not investing in these kind of CAGRs, what are you doing? That's entirely what my thematic basket is. Who cares if Salesforce is going to survive or not when you've got the opportunity to invent invest in those type of CAGRs. You just need to know the names.
Reid Hoffman came out and basically said started talking more about something I'm I'm spending more time on, which is we're having a problem here with the data centers being built. You've got bottlenecks in in parts coming from China. You've got the politicians and the looking for votes and threatening rising electricity costs. So, they're spreading doomer news in terms of trying to get voters to turn. SemiAnalysis put this thing saying that now we've got the Blackwell shipment showing a slowdown. The issue is not demand, but infrastructure readiness. Some cloud service providers still do not have the data center infrastructure fully operational. This is what's happening. This is why I put the data center thing up. My analysis in the the subscriber website, which you guys can go see. Anthropic and Amazon expand collaboration for up to 5 gigawatts of new compute. Anthropic has troubles because they don't have enough compute. Meta and Microsoft looking to pay for their AI spending by trimming workforces. Trimming 10% of workforces not trimming um 7% for Microsoft. They have to finance this somehow. The spending will happen regardless of whether it comes from bonds or it comes from firing people, but it's going to come. The question is, will they be able to get the ROIC? The problem is for those companies, it's very hard to get the ROIC unless your data centers are built. That's the issue. So, the progress is happening. Demand is there from enterprises, but the question is, is the performance there? And as I go through, the adoption is screwed up as well. Why energy is becoming a boardroom priority. This was a great piece this week came out from Pure Storage, now Everpure, a company that is also on my portfolio thematic portfolio. I would go read it. The potential power shortage for AI infrastructure. Just look at these numbers, the shortfall available supply. Do I think they're going to find a way? Yes, but not in the near term. Data centers growing electricity consumption is set to double. There's just no way to build the infrastructure fast enough, so it's going to have to come through efficiency and things like this. So, this is what I was saying before about South Korea media report. Nvidia has reportedly visited Korean power equipment companies and asked them to design data center infrastructure around an 800 V class DC architecture. This is being interpreted as an effort to minimize power conversion losses. So again, everything is focused on power efficiency. This is the reason why on the optical side, you've got so much demand and this is the reason why the power semis are there. So, optical necessary for inference. This is why we've got Coherent and Light and Corning. It's also why we have Marvell, which is one of the leaders in silicon photonics and has purchased companies including one this week. All of these companies fit on being able to somehow or another find a way to get the compute through without needing more gas turbines, more gas, more everything by being more efficient. There's a race on all of these sides and this is where this fits in and this is why those South Korean heavy-duty power things go, but it's also where the power semis go. So, we've got shortages. The problem is the demand's growing so fast, we have run into shortages. We've got component shortages in many things, but in particular in this, the optical transceiver market. Naphtha has become a big issue because of the Middle East conflict. Critical semiconductor materials including ceramic capacitors, power devices, helium. They're expected to face rising prices, shortages and geopolitical disruption throughout this year. Lead times already there.
Dylan Patel gave an interview with Invest Like the Best with Patrick O'Shaughnessy. Again, highly recommend listening to it. The critical parts in there. The economy has flipped. Ideas are are now abundant. So, get back to the abundance side. You want to be short abundance. Ideas are abundant. Anything that you can come up with an idea like a software, it comes to market immediate now. You can utilize it immediately. The death of software that is built on code. He described SemiAnalysis' own AI spending exploding from tens of thousands to 7 million annualized run rate. So, I want you to think and do the math in your head. If a little company like SemiAnalysis has gone to a 7 million annualized run rate on AI spending, think about what a Goldman Sachs, a Morgan Stanley, a JP Morgan, Eli Lilly, how much money they have to spend and you start getting to the 90 trillion dollars, guys. You have to start really doing the math and realize how big this is. On token demand, he says the key story is not falling cost per token, but exploding demand. He expects meaningful robotics breakthroughs in the next 6 to 18 months. The whole AI stack is tightening. Remember, the whole rack is the whole AI stack and that gets back to the five-layer cake. So, margins are expanding, memory is severely constrained, and even upstream items like copper foil PCB materials. Again, PCB material you're going to run into the same thing in chemicals, guys. He does highlight a public backlash against AI and they need to have the This is going to be an issue all year. The bottlenecks are going to be an issue. Um People will say, and I've heard this, is why I wanted to highlight this is, "Oh, the memory story is old news. Everyone gets it." No, no, no, you don't get it all. DRAM capacity will still double or even triple because the market needs such massive production capacity and they have to steal capacity from somewhere else. In a capitalist economy, the only way to steal capacity is by driving up prices to destroy demand. Elon Musk is building a Terafab. It's all related to the same thing. We have years ahead of a shortage, guys. Years.
Here's another way to look at it. This is just the AI training and what is needed or the change that is going uh uh basically from inference. So, we're going from pre-training into AI inference. This is where we are now. Every year it gets bigger and bigger. That's why when you see these CPU charts and you start realizing why we need so many CPUs not just today. The agentic AI shifts the latency bottleneck from GPUs towards CPUs, just what I said. So, again, chatbots, that's what most of you use at this point. Well, let's go through everything and realize that complex orchestration, which really gets into the enterprise side, especially we need an enormous amount of memory. Morgan Stanley's report on agentic CPUs has a TAM expansion forecast in line with mine from a month ago. I thought it was crazy for saying it would reach 100 billion in 2030 and now it's becoming consensus.
>> [sighs] >> I mentioned the Intel, which happened last night. Also, Open AI released GPT 5.5, bringing the company one step closer to AI super app. On many metrics better than Mythos. China's Deep Seek released a preview of the long-awaited V4 model. It looks like they used Nvidia chip, so that reduces a risk for Nvidia. Oh, yeah, that's right. Uh nothing has been solved in Iran. The only thing that's not happening is we're not seeing bombings right now, but I wouldn't be surprised if that picks up again, but we do have the disruption. The oil shock that dwarfs history won't have a rapid recovery. This to me is the story. The doomers saying that this is like COVID have already been proven wrong. Um this is not a US, you know, stock bubble thing. The rest of the globe is going through this. This is an AI story. Again and again and again, I'll keep saying it. If you keep getting caught into these and go. Now, that doesn't mean that the S&P 500 won't finish down this year. That doesn't mean we won't see a 10% correction over the course of the next 3 months. I think it's very possible we will. Again, I think the issue is the inflation and the possibility that rates will go higher. The bottlenecks are there. We could have disruptions where semi production doesn't meet what was expected. We miss earnings. There are tons of things that can go on in a 1970-style bottleneck. Do not get caught in looking at parabolic charts and not realize they can fall 30% and still be within a parabolic move. That is what you have to get used to. We are used to these nice linear staircases and not these elevators up, elevators down type moves, but that's what you're going to see. And the way that I know that is if you go back to the 1970s and you look at oil, if you look at it on a quarter basis where it strips out the daily moves, you're going to see something that's almost a straight line, but if you look at it daily, you have plenty of big corrections in this stuff. In this case, you can see the numbers. It is completely different. Forget the price side, the regional supply, I mean, these are big numbers. The demand has not been impacted and that's because again, most of the demand that's happening globally is happening in AI, which does not need oil.
Kuwait declared force majeure. Force majeure is going to be a story for this year. It's going to be across many, many items. Do not think this is just going to be on the oil side. This will be in a lot of places. Russia extended their fertilizer export quotas until December as the global deficit deepens. Here comes the food inflation. I mean, these numbers speak for themselves. You don't think people are going to be complaining and there's no way to get around this. The shipping has gone higher in terms of diesel prices. Fertilizer prices have gone higher. Everyone forecasting or anyone forecasting that we're not going to see any kind of an increase in inflation, if we don't see increase inflation, I will be shocked. UAE asks about a wartime financial lifeline. I think the disruptions that have macro implications from what has gone on, people need to pay attention to.
Okay, I put this slide in the subscriber thing. I'm going to emphasize it twice here. Again, what I want you to think about is just the massive nature of all of this in terms of the data center infrastructure moving into the Blackwell side and the Vera Rubin and where we are and the build-outs of more data centers necessary and what this all leads to. Again, we need more tokens per watt. We're going to have grid overload, so we could have blackouts. We could have situations. We've got power capacity limits. When you're running things hot and you don't have the investment in the physical infrastructure, expect there for there to be problems. At the same time, if the models get so powerful, you're going to have hackings. We don't have the security ready. When you move too fast and you run things too hot, it is a different kind of scenario than after the GFC where unemployment was 10% and we were gradually taking it back to four. Now we're fully employed. We have rates below inflation and we have way too much demand relative to supply and the money to fund it. It is not done by debt. It is not going to be a systemic event. Even though I'm still negative on private credit, I'm still negative on software and I think the credit cycle is in the early stages of deleveraging. It's not going to be fast because there's too much capital flying around and too many industries that are fine. This is where the edge is. We haven't even gone to the edge yet, guys. That's why I wanted that visual there.
So, let's go through the analysis on the S&P. We've seen the earnings amazingly go higher despite the situation Iran. Well, if you want to break down where they're coming from, they're coming from technology, energy, and materials. This is the entire estimate revisions. And out of that, it's basically one company is 51% of 100 of it. Micron Technology is 51% of the earnings revision. Now, what else do you have on the list? Exxon, Chevron, Broadcom. These three alone make up 85%. This is not a broad thing. This is isolated to those areas that are part of all of those names, by the way. Those four names are part of my thematic portfolio. Those have been there all year. That's what we're seeing is the earnings growth again from this side. I thought this was uh I showed this this week in the subscriber webinar. I want to make sure that everyone gets to see it because the S&P earnings growth for the Mag 7, 23%. Here's the rest of the S&P 500. But, if you strip out Nvidia, then all of a sudden it goes to 6.4. So, if Nvidia were excluded, the other 495 would be required would be reporting higher earnings than the remaining Mag 7. So, the Mag 6 uh or the Mag 7X Nvidia is only growing earnings at 6.4%. Think hyperscalers, guys. Think funding.
All right, this is where the adoption problem gets in. AI is everywhere, the agentic organization isn't. I think you should read this report. It's a McKinsey report. I'm not going to show you the details, but basically, having worked at Morgan Stanley, knowing how archaic the entire infrastructure is, to make a wholesale change is impossible. And Aaron Levie put this out this week. It's remarkable how often you need to be dramatically upgrading your AI architecture given the pace of progress in AI models right now. For the fourth time I upgraded my open claw today. Now, not all today, but during the time since I started it, GPT-5 is now been uploaded into my open claw. So, again, and I'm working on getting more Mac Minis and more Mac Studios so that I can use either Deep Seek or Kemi K 2.6, which came out this week, which is equivalent to Opus 4.6 or Opus 4.7 as an open-source model that is anywhere from 70 to 90% cheaper. We're moving too fast for companies to know what to do. Amazon boom is creating a massive duplicate tools and data inside the company. As someone who uses this all day long, I don't know how you separate the two. You've got two separate components, the personal workflow, and that's what this duplicate tools inside the company are. Meaning, if I build something, but the person next to me doing the same job builds a completely different tool, which one do we use? Why do we have two tools? We're using more tokens. Maybe one's more efficient than the other. Then you've got the orchestration level, which is more from a top-down basis. How do we replace 80% of the people when we need to make sure that AI can do It's a very complex thing. The adoption is going fast from a place of zero, but don't be surprised. I still don't think the ROIC is going to come in as fast and that's the reason why I want to be short this. So, many of you on the RIA and FA world, this is going to be a problem.
So, this is the chart. The white line here is Russell growth. The Russell 1000 growth. So, the overweight position that many, many people, including my stepfather, have is in growth. Uh there needs to be a reweighting and it needs to be not necessarily towards value, but just have to know that growth is not happening the way it was. This is about semiconductors. This is about energy. This is about materials. That's where you need to have everything. This chart is overlaid with the hyperscalers relative to the S&P. So, since 2016, and if I took this back, and the reason I haven't is because of the hyperscalers, but if you took this back to 2007, this has been all of the performance. You just wanted to be long growth relative to the S&P or the hyperscalers. Well, now I think that's going to be in a permanent reversal the other direction. Salesforce came out with their headless 360 and then immediately uh were billy clubbed uh and down 8, 9% yesterday on the back of IBM's earnings. This is something I talked about a couple weeks ago. The bar for the software companies is, if you beat earnings, which is what's been going on, that's not a surprise. Unless you can actually show that your AI stuff is going to offset the rest of your seat-based model, that's the only way your stock's not going to fall. If you miss for any reason or it's slightly disappointing, you're down 8 to 10%. That's why it's a waste of time to spend time in the software side. This spreads into the private credit side. It's a very large default uh this week in Medallia. Probably didn't read people's radars, but wipes out 5.1 billion in equity, imperils the lenders, too. Blackstone's public BDC has massive size, nearly 400 million, and complies as 5% of the net assets. If you're in private credit, good luck.
Uh again, this chart just kind of bring it in home with the themes. Here, since the lows in October, I wanted to show that this is my thematic portfolio. It's been up. This is the daily numbers every single day so far this month except for one. Uh and it was down slightly that month and slightly that day. Here is the basket relative to the hyperscalers. This is the trade that I want to have on. So, for long short people, for hedge funds that are looking for the basket and looking for access, this is the way I would trade it. Uh I think it's going to have You can see it's a good sharp ratio. They are correlated. It It's going to matter. This is it relative to the S&P, and there's two things I want you to realize. Yes, it's had a breakout, but it was unchanged 6 years ago or 5 and 1/2 years ago as of last week. So, this is the beginning of a trend. My portfolio is not something that is late stage. This is the beginning of a trend.
Uh thanks to the people who reached out on the technical side. There were a couple RSIs that were wrong. I ran everything. I will put this up every single week. And again, for people calling, this is not what I think people should do. This is not some scoring that is done by other people. This is purely something I put together. There is an explanation that's on the subscriber site. This is giving you the ability of getting all the technicals for them in one place. What I did was say I did a video this week that you guys got how to use Claude to turn your portfolio list into trade ideas. You can do whatever you want. If you like to find names that aren't working on my list, where their RSIs just went from 30 to 35, but they're down 20% for the year, go for it. I show you in this video how you can put different prompts and how you can do it. It's a short video. Watch it. It's 5 minutes, but it'll give you some ideas on how to do this.
In terms of the Nvidia 800-V DC not announcement, but report in South Korean media that I talked about. Um again, as a reminder, this was it and the demand for power firms. What I did was I want to remind you this edge side. So, as a reminder in that first part of this video, I talked about the connection to the edge. This is up on the website. It went up last week. Power semis are one of the verticals that are in there, and there are six companies. That was the chart I showed you. That is this chart. Equal weight of those names. That's what it's up now. These are timely things when I put them out. Now, I did a dual one, and I connected the Nvidia side with the edge investment side using the Claude skill, which I gave to you guys last week as well. For those of you who tried to use it and it didn't work, I can't say this loud enough. It works. All you have to do, if it doesn't work for you, is put it into Claude and say, "How do I do this?" You guys have to get used to using AI, and the more that you use it, the better you will get at it. That is the reason why I'm doing the videos. I can't individually go through every single person and figure out what you did that was wrong. So, please take the prompt that I gave you, upload it, try to run it. If it doesn't run, ask it what I'm doing wrong, and then Claude will take you through and fix it. If you're not paying $20 for it, I don't know if it'll work, because this is something that I know that my son, who's a college student, I gave him the entire thing. He ran all of them. I know that a lot of people have reached out and showed me what they're able to do. This is the kind of stuff you can do is take that news report, upload it with my edge investment universe, and say, "Hey, out of these names, give me some other names. Give me small cap names. Give me mid cap names." That's the way you use Claude connected to the work that I'm doing.
This is what I'm working on now, molecular moat 2.0, chemistry at the interface. We're getting new news where advanced packaging, photonics, and liquid cooling coverage converge. If you guys are interested more in chemicals, I will get some more, but because of the Middle East conflict and what has happened with the the regional oil energy prices being different, uh I think there's an opportunity here, so I would go spend time on the chemicals. I will bring you something in the coming weeks.
Bitcoin. Again, yesterday we got a huge move down in the software side. Remember, this overlay is what I've shown before. This is software relative or software overlaid with Bitcoin. I believe Bitcoin has made a bottom. The MACD weeklies are going. If I'm right, they should start moving at some point and get up to this level, which would get it close to 95 to 100,000. I believe if this is going to happen, it needs to happen soon. Uh and I say that because we have broken the correlation. Uh there's inflation that's coming. I gave you guys the quadrants in terms of if we're where we do phenomenally well for Bitcoin, we're entering that quadrant the next week. Uh sorry, the next month. If you think about PMIs, I've talked about PMIs going higher. Now we've got confirmation, the industrial thing, that PMIs are going to be going higher. We will get up towards my 60 level. So, once we get through the oil shock that's happening, I fully expect that we're going to see it go higher. So, between inflation and PMIs, that is normally a good thing. But the other thing is, by the end of this year, this is my main prediction for the people who are interested in Bitcoin, but also for hedge funds and mutual funds and private wealth managers, I've said this before. I'll say it to you guys. If you want Anthony Pompliano and I to come speak for your FAs, give us a call. By the end of this year, this growth asset basket will be a problem, and people are going to be looking for things they can invest in. And the problem is no one likes to buy the end of parabolic charts. The one thing good about Bitcoin right now is that it isn't a parabolic chart. It is unchanged right now. You're getting the chance to buy something unchanged from 2024. That is a very, very powerful situation, and that reminds me more of things that wealth managers should be doing. I think it's a great opportunity.
Now, I have a Bitcoin proxy. I haven't shown it to you guys. It's three components, and it involves the three things I just said. So, it involves the Qs. That is the software side. It involves gold. That is the store of value side {slash} when everything is breaking down, where do we go hide our money? And includes copper. That is the PMI side. This is the long-term chart going back to 2017. The white line here is the proxy. We have alligator jaws, guys. I like the catch-up trade that's coming, and I think gold is going higher. I think copper's going higher. I think the Qs are going higher. I don't think the Qs are going higher because of the hyperscalers. I think the Qs are going higher because of biotech, semis, all of that stuff. There will be changes in the Qs. I don't think the growth baskets are going to do the same thing, and I certainly don't think software. So, I think by the end of the year, you're going to be wanting to increase your weighting in Bitcoin as a growth thing. And ETFs are coming for crypto. So, I've been asked this a lot. If you've been looking for stable coins and tokenization, you've got STBQ, Amplify Stablecoin Technology, Amplify Tokenization Technology. I'd go take a look.
Uh and finally, for those of you in California, I will and those of you attending this event on the institutional and the public pension side, uh reach out to me. Um you can reach out to me at on Twitter. You can reach out to me on LinkedIn. You can reach out to me on on email if you have my 22V email. Uh you will be able to find me at this event speaking at the institutional investor event. That's it for this week, guys. Good luck trading these parabolic markets, and I'll see you next week.