7-Sigma Shock: Is This the Start of an S&P Crash?

I guess we've all been wondering when we were going to finally have something to talk about in the markets, as I've done these for the last, uh, probably four months. It's been less and less of a market recap, more and more about AI, Bitcoin, and just what I've been using AI for. So, we're back into the market, dominating the conversation. We're back in turbulence, uh, rare earth threat. I've mentioned rare earth continuously since early April in these videos in terms of its importance. Well, now we're going to spend more time on it, and know that it is critical for the AI buildout.

So, for anyone sitting in AI trades, which we all are, because anyone making money is involved in AI, uh, the rare earth threat is a real threat and something that needs to be paid attention. The entire video will be on that, along with, uh, the market structure showing stress before the event. The immune system of the market was very vulnerable to something like this at this time of year. Uh, a lot of times the immune system, it doesn't matter. Uh, but if you get an event of this magnitude with stocks at all-time highs and sentiment back up to the highs of the year, and at the same time, the underlying conditions within the factor world, what I call the immune system, showing weakness, uh, we'll go through that. Handicapping the risk going forward with China and the US, whatever is going on, do not lose focus on AI. Even if this is a one-month event and the stock market falls, and I'll go through numbers, it could fall a lot if this lasts for a month. Uh, you still have to be doing your homework on this and paying attention to what's going on. I go through a lot of things. Uh, there were a lot of important podcasts for the week, and you're going to need some distraction because I think spending your time trying to handicap the, uh, the China-US situation will be challenging on a day-to-day basis. Hopefully, what I give you today will give you some framework to deal with, and the debasement trade goes mainstream.

So, uh, at the end of April, I wrote this paper. The light, or not the end of April, April 8th, a couple days, or a few, two trading days after the, uh, Liberation Day collapse. I just wanted to write a paper that basically said that we had reached a level where the market had fallen enough, sentiment had gotten bad enough, uh, and I think yesterday was a pivot day for the market, a time and a time for offense, especially during the inevitable retest. This was with just the horrible news that had gone, or the fact that people thought there'd be a complete collapse. Uh, we're on the other side now. Uh, it was only four months ago that investors were too optimistic. That was me writing this again, back about what had happened in December. Uh, for what it's worth, I believe the shakeup in global trade actually needed to happen. I prefer this to a failed auction in a 10-year worth 10-year rates. Absolutely still believe that way. I think the slowdown that occurred in the economy, the change that everyone getting out of this, the deficit in debt is not, uh, being viewed the same way it was before. That could change at some point, but it was important back then, but it did make people forget the acceleration of AI. Do not forget the acceleration. We're in a very different position when we were back then because every six months in AI is a leap. Uh, and I'm going to go through some things now that are going to get more into because the bubble talk still remains. So, people are still doubting it. And so, you guys know, for whatever reason, I seem to be having problems with my video. Every now and then, it seems to be going black. Not really sure why. Not an expert in, uh, audio video on my Macintosh. So, I'll figure it out the next week we go through. Uh, positioning dramatically different from how the year started in AI progress, making each month feel like a year. I'll bet this three-month sentiment shift will look very different by the time Labor Day ends the summer.

So, here we are now. We're just past Labor Day. Sentiment did what I thought it would do. We were all the way down here at the beginning of April when I wrote this. We ended up up here, highest for the year as of this week. And Charlie McKelikott from Namora, who I love to read, don't get his stuff anymore, but, uh, we're in the 100th percentile of systematic flows. So, you look at where we were back in April. Now, you look at where we are, actually the biggest percentile, or the biggest C systematic flows in terms of longs. And then you look at the equities, we're at the 88th percentile in terms of the risk parity numbers, and that's because we've seen V come down and we haven't had a correction. So, basically, this is a drawdown for the S&P. We're coming out of this. This is obviously Liberation Day. Now, when you get out of this, um, and again, we hit a 20-plus percent correction back then. I'll go through how, why this is important. Uh, we haven't had anything. When you haven't had anything for a long time, uh, you're due for something. We hadn't even had a 1% correction since August 1st. So, we did get that on Friday. Uh, China curbs rare earth. This was on Thursday. So, Thursday, this started. Market didn't care. It went to new all-time highs. But then just after Pomp and I finished the video, uh, Trump threatens massive China tariffs because of the rare earth export. That just shows how important they are. I'll go through some of the details on this, uh, later. Uh, in terms of the importance of rare earth, is it possible to build out the data centers without rare earth? No, it is not currently feasible to build the scale of hyperscale or data centers planned. So, rather than focus on the fact that you need them for the size we're talking about, which is massive, you cannot do anything close to it. They're also important in military, which I went through last week. I'll go through that again. Uh, Michael Hersan wrote a great piece. He's the only person I've really seen that has continued to warn, uh, about this moment during the entire time. Uh, I had a very different view. I didn't think we were going to get to what happened on Friday. I still don't think it changes anything. But here's what Michael wrote, and he's the only person that I know that really was still calling for this, only three weeks ahead of the APEC. This is a very dangerous moment for global supply chains, including those powering AI. But it is important to note that neither side has yet implemented its threatened measures. There is still a window to back down, and Trump faces significant political risks if he follows through on the threats. He just did the TikTok deal, which is what makes this a little bit, you have to include that in, kind of how could we do a TikTok deal with them less than a month ago, uh, where Trump and Xi agreed to it, and then all of a sudden be at this moment. This doesn't sound like something someone as someone, unfortunately, has been through a divorce. Uh, divorces take a long time to settle for reasons, and even in mediation, there's a lot of back and forth. Someone will step away from the table. You go through this. I went through this. I saw it. So, this is not surprising. I think the person to watch, Scott Besson, has been the person at all these moments, including in April 9th or April 8th, when I wrote the paper, I mentioned that Besson would have to take control, and as you'll see later, the area that Howard Lutnik is in charge of seemed to be the place that created the problem in the first place. Uh, America's soybean farmers are panicking over the loss of Chinese buyers. The farming side is an issue also for Trump. Uh, I wanted to do something with AI, so I went and applauded. Um, they've had a big release. Uh, I wanted to see how I could build an app. Uh, for those of you who've known me a long time, I've always had a data scientist, even back in the days, back in in Brazil at Morgan Stanley, that person's job was to build things for me. I like to create things, but I didn't have coding skills. So, I would use someone who did have coding skills to build things. It may have been in C++ uh, back then, and now it's, uh, you know, was most recently in R. Uh, but I asked Claude, "Are you able to build an app for me? I'd like to have a sigma, uh, a sigma measure on a chart form without going through it." This is something I've asked for Bloomberg for a long time. And there it was. It took three minutes to build. And those three minutes ended up with this. This is a daily sigma move for the S&P using 20-day realized V. And the percentage move, we had a 2.8% fall. Uh, the V 20-day realized V for the S&P before yesterday was 6.4. So, about 40 basis points a day. Seven sigma move, the biggest one we've seen in the last year. So, Friday was a big move. And again, the reason I like to look at sigmas is because it builds in part of that immune system side, uh, in terms of the warnings, but only at the index level and the VIX level. There had been no volatility, which again, think of your immune system, there hadn't been anything, but under the hood, there had been tremendous immune system issues. This was going to be the focal part of the beginning of this week's piece. Uh, and really, it was going to focus on Jared Kubin, who's been putting this out all week. So, these things were all before he's been highlighting what he's been calling liquidation nation, which again, is another way for saying there's been deleveraging that clearly is happening, and it's been going on for the past five months, but really stalled again in the summertime and then picked up pace. Usually, deleveraging, the main thing I want you to focus in here, and he was showing this all day. This was again on Friday or Thursday, sorry, in the morning, the Goldman Sachs hedge fund index, um, and again, I don't know which one he's using. I use the long-short side, which I create myself. Uh, again, one more on this, just different ways to look at it, but the last three weeks' acceleration is what really stood out to me, and there's been carnage since the end of June, but a lot of this is garbage, things that have been moving higher. This is the way I measure it. So, this is the Goldman Sachs crowded hedge fund longs overlaid with the most short positions, the hedge fund, or the most shorted positions in the market. So, you had the longs underperforming the shorts in here. That, to me, is very typical when you're coming out of something like, uh, like, uh, what we were going through, Liberation Day, and we kind of paused here midway through July, started to go the other direction, and then we violently fell down again, even faster than we had fallen here. And this is the point that Jared was making, and others were making, that I was reading through all week, and to me, it stood out. This is a very, very sharp fall, and the line there is basically right around the midpoint of September. Uh, in the VIX market, Goldman had been highlighting the fact that we had seen a zero correlation between the S&P index moves and the VIX moves. Just use one-month implied V and just suppose S&P index versus the VIX. Basically, we had gotten a zero correlation, which means that if the S&P was up, you didn't know whether the S&P was going to, the VIX was going to be up or down. Normally, when the S&P is up, say, 50 basis points, the VIX is down. We had lost that correlation. That's why you see that chart back there with negative correlations. Ben, don't know if his right is the right way to pronounce his name, so I'll just say Ben from Wellington, uh, who I follow as well. Heard him on some podcast, has some interesting views, independent thinker. The two of these guys are, I love independent thinkers. Don't care whether they're right or wrong. They're bringing their own views, and they're bringing unique views, and they're looking at things that are hard to find in the paper and they're hard to find in Bloomberg. And he was calling for the same thing, and he had highlighted, and it's hard to see here, but he's highlighting warning signals for the S&P on the same thing between the correlation between the VIX and the S&P. So, these guys were on it. And again, Goldman brought up the, how many days in a row this happened last week. So, the first one I showed you from Goldman was actually from September 29th. This is a redo from October 6th for the prior week. And in that case, you had five days in a row, the S&P was up for 1.09%, and the VIX was up 1.36%, also notching five consecutive days. So, you had the VIX up every day, and you had the S&P up every day for five days. This had never happened before. These are the types of things that suggest that the immune system has weakened. Now, another thing that suggests that here's a differential between VIX. I sent this out in the 22V chat that we have in Bloomberg, just highlighting that I thought all of the immune system, the market was weakening. I'm going to go through the reason, the catalyst, why, as to what I think, uh, is also contributing. But the main point is here, this is the VIX for the equity. It's VIX EQ for Bloomberg. You can see it right there, overlaid with the VIX. We had been rising consistently in equity. So, what was driving the VIX to go higher while the S&P V, while the S&P was rallying, is the fact that we were seeing a lot of vol in single names, and so you had this break that was happening, and the VIX was not moving yet. To me, again, that is turbulence that is happening in the market, and we're likely to see an increase in index volatility. Now, Tricolor's busted money machine, Wall Street rethinking risk. This happened around the exact same time as that arrow, and that's the reason why I bring it up. Tricolor, for which hasn't got the attention as First Brands had, a sudden bankruptcy in September, and again, fraud. This is the, unfortunately, getting into First Brands now, an emergency court filing for an independent investigation into 2.3 billion tied to the company that had simply vanished. The collapse of First Brands has raised questions about the risk of private financing, private debt, and private credit. Those, those two, uh, phrases, have been a worried part of the market, and in particular, it's been for at least four years now that people have worried about this. Uh, I would say it was actually, you know, before that, the signs were there, even in 2019, I would say that this was growing, but it was really growing rapidly post-2020, and we obviously had COVID, so there's been a lot of businesses that were hurt, but then you add in the massive rate hikes that occurred, uh, in 2022, and I think people have been looking for the private debt market to possibly be a problem. The issue with this one is, is it related to the K-shaped economy? Yes, First Brands is being linked in financial commentary, symptoms of a K-shaped economy, and that's because their customer base was predominantly lower-income drivers who repair and maintain older vehicles. At the same time, when you go to the other side and you say Tricolor, well, they do a lot of lending to subprime. So, this is kind of on the subprime, the bottom side. Uh, remember what Paul McCully said? I, I always remember this back from 2008 when the housing market first saw the problems in the subprime. If you go look at that on The Big Short and watch the movie again, the plankton, as Paul McCully would say, matter to the ecosystem. If you get rid of the plankton, then it starts to trickle up. I don't think we're going to run into that here. Uh, that is to me, not, not necessarily important, but you can't ignore it and say that it's not a, a probability. So, business development companies are generally seen as a proxy. BDCs, if you've heard that this week, business development companies are generally seen as a proxy for the 1.7 trillion private credit market, and now their strugglers are indicating a weakening industry. Shares of many public funds, BDC funds like Blackstone Secured Lending Fund, are down around 20% year to date. And then you've got the actual companies that are down too. So, what does that look like? Well, here are the private equity companies related to the S&P, and you can see how sharply they have fallen again in that same time period as when I showed the hedge funds. This is why when you start getting this, you start looking, you're like, okay, the private equity guys fell violently. This is a chart of the private six private equity companies, public related, relative to the S&P. That's what this priv SPX is. Um, you can see straight line down since, uh, since exactly at Tricolor, basically started, and then followed by First Brands. If you want to do your own on this, here are the names. Apollo, Ares, uh, Brook, uh, Brookline, uh, I think it's Brooklyn, Brookfields, uh, Blackstone, KKR, and Owl. You can go through it on your own, but those are the names basically that have been brought into the mix on, on the side. Now, there is a BDC in Bloomberg. There is a BDC index, uh, which has a bunch of funds in it, and you can see how this has collapsed. So, this is not a minor thing. I mean, you're talking about something that was at all-time highs in March, bounce rallied back, and now has fallen off violently again from kind of 64 all the way down to here. You've seen the Jeffries stock price come down as well. Uh, if you do an overlay between those two, uh, the public equity companies are the white line relative to the S&P, and the other line is that index I just showed you, which is the BDC funds. So, they've all been breaking down. This is their importance historically to junk. Every time we've seen junk spreads start to widen, we've seen the BDCs involved. Well, we finally saw junk spreads. This is junk spreads inverted. Uh, they were weakening as well. Had been down every day for the week, or spreads had been wider every day for the week. They had been wider over the course of the last month, again, since September, and this is the fall that ended up happening. So, you end up in a situation where I wonder what junk spreads are going to do. Well, the VIX overlaid again with those, and this is the reason why I was sending the stuff out. So, you already had stuff going on the immune system with factors. You had stuff going on with dispersion in terms of single-name options were showing volatility while the index was not. Looks like some hedge fund risk, maybe some deleveraging. The question is, maybe there were hedge funds that were in the market of the credit side that ended up losing money with Tricolor and with First Brands, and they had already had weakness going on in terms of their book in general, and they wanted to take down leverage for whatever reason. And you end up in a scenario that eventually this starts to feed into the VIX. You add a little bit of a fire in terms of a match, which was Trump on Friday, and you can see where this would go. This is what I talked about exactly with Anthony Pompiano Friday morning. Uh, it just so happens that this, the Trump thing occurred Friday. If we would have done this video there, it would have looked a little bit different. But all the things I'm showing you right now, I only did them through October 9th. So, these are not including the Trump stuff. Uh, one interesting component, no correlation with this one, but now all of a sudden, violent correlation. Another sign to me, I look for things that are anomalies. Those usually mean something. So, what this is here, again, this is the BDC stuff falling off the exact peak. Uh, here is that white line, which is again right around early September, and then the violent fall started at Tricolor. This orange line here is exactly the short side of the Goldman Sachs short basket relative to the S&P, and again, this is obviously inverted. So, if you had a deleveraging scenario where a hedge fund had lost money on the credit side, and they happened to have a quant book on the other side, and they tried to get out of the stuff because this is very quanty to me in terms of this, the, uh, the orange line here, it would have been a short covering that was going on, and you'd get a scenario like that.

Now, in three lines, please summarize all the views on private equity in the All-In podcast this week. Private equity is totally baked. This is from the All-In podcast this week, this last Friday, so this is a week ago. Uh, was this week's view on private equity about how AI is disrupting capital markets? AI is rewriting, is rewiring how capital flows, reducing dependence on traditional financing, weakening PE roles, and forcing public markets to reinvent the next wave of growth. The other thing it is doing is destroying all companies in its path. I will continue to say this over and over again. We are going to see more bankruptcies. We are also going to see more business startups from young people. But we are going to see massive disruption from AI, which are disrupting the business. And this is only going to accelerate. If you don't have the money, meaning if you have too much debt, you're not going to be able to reinvent yourself. They go through this from the capital markets perspective. I'm going to keep saying this, that we will see plenty of bankruptcies, plenty of issues. You want to invest in AI and you want to be short the stuff that is not going to do well, and I'll go through some more of those.

Now, the S&P for the week, uh, not that big of a week, down two and a half percent. So, even with the Trump stuff, it's not like this isn't a move that happens regularly. In fact, we were up, you know, one and change percent three of the four weeks before. So, we're really not giving back much, I think, for the month. Down one and change. NASDAQ down 2%. Not a big deal. Uh, the MAG 7 down 2.7%. Not a big deal. SMH, Semiconductors, the AI trade, no, not a big deal, especially since it was up 5% the week before. Basically, it was down 3.3%. Three of the last four weeks were up more than that. So, you're up huge in September. We're giving back some of the gains at this point. The Russell 2000 down 3.2% for the week. Again, not a big deal. The retail favorite trade, retail favorites, they were hit hard. So, retail is finally being forced out. I would think that given I just don't see us going right back up to the highs that whether it's the retail favorites or this meme index, I think there'll be some more weakness in this just because I think there has to be an airing of this. So, on Friday late drops the 100% China tariff bomb, but Trump keeps door open to Xi meeting. This is from SCMP. So, this is the Shanghai, uh, uh, morning post. So, you're at a position right now where the Chinese papers like, "All right, he's saying this, but it's not like, like when Michael Hersan said that there can't be discussions." Um, the S&P, I wanted to just show this again, the immune system side. Liberation Day happens here. S&P had already fallen, excuse me, about sneeze 10%. Um, we had already weakened the immune system completely. Uh, but positioning was already significantly changed. The tariffs had been built in. There wasn't a lot on the books here. That was one of the reasons why I was more positive. This time, you were at all-time highs just before Trump came out with that. Market didn't have any time to go. So, I wouldn't underestimate that this, if we don't see a change quickly from this rhetoric, I think we're going to have to reposition things because we have a lot of MO investors that didn't have a chance to get out in a tape that still had strength. They've now have to deal with this. As a reminder, the MAG 7 relative to the S&P, not the MAG 7 on an absolute basis. The MAG 7 relative to the S&P, and this is why I was saying I had started to warn in here early December. I have a YouTube video that says we're starting to see factor unroll. I'd be careful of what's going on. Now, we had the deep seek moment somewhere in here. The MAG 7 relative to the S&P was already significantly weak going into this, underperforming by 15%. Now, we've been weakening slightly, but nothing like what's been going on. So, I think you just have to again pay attention to the immune system. If I had to pick a spot, if the rhetoric keeps up, I would think that we would retrace all the way back to here. And I say this for a lot of reasons. I, I'll, I'll go through them, but I want you to remember here is April 9th. This is when Trump went out and said, "I'm putting a pause." But he was still isolating China in here. So, we were starting a retest, a violent one. During this period in here, rates were falling, or rates were going up at the same time that stocks and the dollar was going down. So, we had stocks, bonds, and equities going down, and we saw Trump flinch again. But in this case, it was Besson that at a dinner said, uh, there's room to negotiate with China, that started the rally. That was the low since that point. We never took it out. We never retested that. That is the point. So, if we are going to have a full round trip, the whole reason to me that this thing happened was because this is the point that China said no rare earth. Here is where the US flinched. So, in this case, as to this, this whole rally, in my opinion, has been related to the AI trade. There's no AI AI trade without rare earth. So, rare earth is the critical component, and I think that's where the negotiation is, and I think that's where it's going to be. If we get an unwind, it would be, in my opinion, because you've got a lot of funds that are trapped. So, the first level would be this 66, sorry, 6229, which again, is significantly lower than where we are. That would be another 3% at this point. And then the ultimate fall that I could see is back down, uh, to the 200-day moving average, which also happens to be this gap higher that occurred in June. So, those would be the levels that I kind of think about, uh, as you're, as you're handicapping this. Depending on how the rhetoric goes from here, it is all rhetoric. If they get an agreement, we're going to have a V bottom. So, you have to trade things accordingly. And that's why to me, the VIX ends up being the thing. And if you were lucky enough to put VIX calls on, I'd be rolling them as we go. I don't know, every seven points, maybe eight points in, in, in the VIX. We've already gone up five, six points since the lows of, uh, of Friday, and more since the lows of last week. Um, based on this statement, options markets reflect. This is the thing I put out on VIX calls. This was two weeks ago. Why vol calls, VIX calls, uh, VIX convexity trades can work better in this, and this is an illustration of it. I don't think this is the last time. So, I would continue to focus on this. If you're a mutual fund or people, you have to build something into your process to where you don't have to sell out of things at the low and then turn around and buy them back. If you can't handle drawdowns because you're measured on a quarterly basis, this has to be part of it because I think we're going to have these on a regular basis. In a concentrated market, especially one with Donald Trump as president and all the things that are happening, you have to expect that there's always going to be something that can happen. And if it happens at a time when the market's not ready for it, you're going to get these types of moves. I want to remind everyone, I showed this last week. If you haven't listened to it, go listen to it. This interview, um, with, uh, Andrew Roll's CEO, is really important now for the things that I mentioned. Every US-China-Taiwan war game shows the US exhausting key munitions in six to eight days with two to three years to rebuild stocks. That's critically important because of the rare earth side. So, some supply choke points are outside of defense demand. Rare earths, magn, these are all the things on the list with China. China specifically mentioned that on the US side, they think they're hurting us from a national security basis. They're focused on things like this, which are for a national security basis. So, the US edge is software. What did Trump announce? He's thinking of cutting off some software to China. So, these guys covered it in there. This is why these podcasts I bring them up here because I think they're really good. Uh, and they're the way that I'm figuring things out. I don't show you guys the other 20 I listen to a week because most of them get thrown out because they're garbage, and I don't make it through, and I come in and I look at to see whether I should read them or not. Uh, for the size of the buildout plan, by no rare earth. I showed you this before. Without rare earth, you can't do the buildup. So, what had gone on? Let's handicap this. Now, China hawks grow queasy over Trump's push for deals. Again, this was in the last six days. So, this occurred over the weekend. You started seeing that people were starting to worry that he was trying to make a deal. David Sax spoke about the criticism, uh, a day later, "US must win this AI race." He also signals caution. We shouldn't sell our latest and greatest chips to China, which would give them an advantage. So, if America's AI stack captures 80% of global market shares, and the US wins, if China gets there, the main point of all of this is that we had reached a point where this is not an easy divorce. They want stuff from us, we want stuff from them. It's very difficult, and Trump is trying to thread the needle. I still believe he will get a deal done, partly because of this. Trump says US and China approved TikTok deal. This happened September 19th. Now, why would the Chinese be mad enough to do rare earth? If you guys don't listen, or I'm sorry, don't have the expose of Hugh Xiin, you should. He, he put this out 17 hours. Well, this is from this morning. So, this was yesterday after the announcement. He has always been viewed as a voice piece for the CCP. So, if you're looking to see what the government is saying over there, at least in real time, it's not going to be President Xi. It's not going to be people inside the, the, uh, Communist Party, but this person has been a good person to pay attention to to at least see what they're thinking. On September 9th, 29th, the US Department of Commerce, that's Lutnik's area, announced a new rule that expands its entity list export restrictions to subsidiaries at least 50% owned. A move clearly aimed at Chinese companies. This was widely seen as Washington's attempt to create new leverage ahead of the next round of the China-US talks. China also holds powerful leverage to counter them. Well, I guess rare earth is it now. That rule that was put in place on September 29th, this is a, a labor, I'm sorry, a trade lawyer, global trade lawyer, Gibson Dunn, a watershed moment for export controls, the risk and complexities of the Commerce Department's affiliate rules. The amendment, one of the furthest reaching changes to BIS regulations in years. So, this was not a minor event. It did not get press, but it was big. And again, this is just highlighting this is a credible institution. Gibson Dunn is a powerful firm. This is on their website. So, it's them, you know, some of these guys are saying it, but they put it in. But regardless, this was being viewed that way. So, let's compare that announcement to China's this week on rare earth. Are they similar? Both moves are strategic and timed ahead of a planned high-level Trump-Xi summit, signaling escalation in their. They employ the same logic: weaponizing choke points and global supply. The two announcements are structurally similar in their strategic intent and mechanism control. So, from when you read this from my side, the Chinese responded in kind. Now, Poly Market, what is the probability of a US-China tariff agreement by September 10th? It was up at 90-plus percent. It fell 270. It's back up to 80 after Trump did the 100%. So, I just wouldn't go too crazy on this. I still think a bargain's going to happen. If that does happen, you want to be doing your homework on wire.

Now, I want to make sure we summarize at this point. The immune system, the market was weak going into this. This doesn't mean we're not going, even though I still think there's going to be a China announcement. I think for the next couple weeks, we are in a back-and-forth market. I think you're going to have a chance to look for, uh, for signs to get in. If you've had some names that you've been looking to buy that had run away from you, you might get a chance to buy them at levels that make more sense. I'm going to go through some of the AI stuff in here. Uh, the main point is we are in a less certain period. Let's call it a mini version of what happened, uh, after, uh, Liberation Day. The lows could have been Friday. Uh, I will see what happens when we get into next week, but I think you have to have more of a trading mindset. Uh, you know, as I told Anthony Pompiano when he asked how I deal with these types of things, when I trained, or trained when I taught my son how to drive, and I was giving him advice, I said, "Hey, you live in an area with a lot of deer at nighttime. If you're not paying extra attention and you're not slowing down, you're making a big mistake. You're more likely to hit a deer. You have to pay attention. It's the only way you're going to deal with it. You will run into one in the wintertime." That's the way I think the market's going to be. It's going to be less certain, which means it's dark. There's going to be deer running around, which is Trump's tweets. Uh, I think you just have to be more careful now as we fill this out. The AI trade will give you plenty of times to get involved.

So, the, the bubble talk, it was another week of bubble. I'm not going to show you all the stuff, but Wired had one on there in terms of our week, and then this was going around in this. So, again, I, I think these memes are funny. Every person who is bearish sent this out saying, "Yeah, this is a clear sign of a bubble. Just stop, guys." Uh, all right. The must-listen-to. So, Dylan Patel is the only person, and I'm going to say this because I've had lots of, uh, hedge funds talk about Goldman Sachs people and Morgan Stanley people. No offense to the places that I grew up at. Zero chance that they have anything over Dylan Patel. This guy lives only on the entire supply chain and everything going on. And he does plenty of interviews. He has a reasonably priced subscription for SemiAnalysis. This is the only person, in my opinion, to listen to for the supply chain and where we stand. He understands the compute side. He understands the power side. He understands the model side. So, he is heavily respected. However, he got to this level. I love listening to him. He did an interview with A16Z, uh, recently. I think that was earlier in the week. Maybe it was last week. Uh, and then he did this one, which was with Patrick O'Shaughnessy, Invest Like the Best. Phenomenal interview. Everything you need to see. If you're going to listen to one thing this week, listen to this. Uh, I'll just give you some of the highlights. Dylan begins the interview by framing the AI buildout as the highest stakes capitalism game of all time. I actually think they're wrong on this. I think that's what Patrick O'Shaughnessy called it, but regardless, it's in the interview that way, and I agree with it. This is the highest stakes capitalism game of all time. That's why people can be bearish. They can sit there and say this is a bubble because we've never spent this kind of money. Now, he stresses that AI's economics are inverted relative to traditional technology cycles. I've talked about this. I've written about it. I wrote about it this week. In a normal market, you first build a product that finds customers, then you scale the infrastructure to meet the demand. This is in reverse. We need massive clusters ahead of the products. And the reason is, as I get in, they can't actually release the models. So, the models you're using are not the models, guys. Hate to tell you this. Not sure if you understand it. You don't have the real model. I'll get into that. But the ability of releasing, you can't use the videos you want to use. You can't do anything. So, for anyone to say this is a bubble, we don't have enough infrastructure for even the AI we have today. The models we've released today, we don't have enough. The models they have behind the scenes, who knows how much extra billions of compute they need, but it is in making it. They have no ability to release what they want to release, or they'd already have the revenues coming in. And oh, by the way, Oracle's RPOs are revenues. Regardless of what everyone says and everyone goes through, they might actually be getting the money, but as I've said on interviews and as I've said it in public and as I've gone through it, if you have a restaurant and someone orders a billion cheeseburgers, you're not going to be able to make them. That doesn't mean you didn't get the order in. It's an RPO. It is something that people want. So, unless the models slow back, it's a problem. And that's what he says. Models. Patel's warning is stark. If the compounding slows, the feedback loop snaps. Now, this is what people have gone on. So, this is why I'm showing you things to go through it. He says that's not happening. Everything that he sees on there is that the models are ahead. The same thing I'm seeing, and as I'm going to go through this, he's involved with every company in terms of talking to them. So, the highest stakes capital, this is why Patel calls it a capitalism endgame scenario, which I've called it too. Um, the belief in infinite improvement sustains equity valuations, credit, national industrial policy, chips. AI progress isn't just a technological race. It's a macroeconomic necessity. Again, there's no way to go backwards. This has to be spent from a government level and from these guys. And they're spending this. If it wasn't for them, we wouldn't be able to do it. Meaning, if they didn't have the concentration and the size that they are, they wouldn't be able to do this. They wouldn't be able to get the AI to where it is. The fact that they have the money and have the ability of raising the money because their equity is so massive allows them to do this. OpenAI is a different situation. They don't have the equity to do it. They don't have the revenues to do it. They don't have the free cash flow. So, they're the ones that people pick on, but you're dealing with massive companies with massive valuations. So, his framework, wave one, language and cognition. This is the LLMs. Wave two, power and materials. This is where we are. We need all of this stuff. This is where we are right now. This is the investment side. He talks about this company, Periodic Labs, what they need to do. These are all the places you want to go through it. The company's involved at these levels. This becomes the embodiment side. Okay, we're getting into the embodiment side now. From compute to labor is the opportunity. He makes the point that every major investor in VC he speaks speaks with sees labor costs as the single largest addressable market for AI. If the global labor spend is roughly 50 trillion, then anything can perform. This is the revenue that they are looking to get. This is what they're going to get as AI agents and as humanoids and as autonomous vehicles come. This is where it is. These are the jobs. So, regardless of kind of where your mindset is on this, remember this is what they're after, and this is what all of this stuff does. So, this is the ultimate scaling law of capitalism. You're finally getting to labor. This is why I say this is the end of capitalism the way we know it. The AI will cannibalize everything. And that's what I believe will happen. That's why I'm the most bearish person as we get to 2030. But right now, you're investing in AI. I can't worry about 2030 until we get closer and see what the market's going to do. We're trading day-to-day here. We're trading what's going on. And right now, if your job is to make money, I'd be focused again on AI and looking at the next stage of it, which I'm calling the proactive stage, which involves all of the things that he mentioned here. A global neural network of edge devices. This is the proactive stage. Each learning locally, each contributing to a shared intelligence layer that spans both bits and atoms. The true TAM for AI isn't the $1 trillion chip market that we're seeing now. It's the $50 trillion global labor market. Every factory robot, delivery drone, and humanoid assistant becomes a node in the world's most valuable network, the network of learning machines. That is the next stage. We haven't even gotten there yet. And we're going to get there starting this year or next year. We don't know. We've got Elon Musk doing something. I had to cancel my whole Tesla thing for this week because I don't have enough time. Um, this section of Dylan Patel's interview on on-device reinforcement is one of the most forward-looking in the whole points. I'm not going to go through the whole thing, but this is all on NPUs and all on the edge. The next intelligent frontier is not in the data center. He talks about this. It's not in the cloud. It's not in the things that you've invested in so far. It's at the edge. So, as we build up the data centers, we're going to be doing this at the edge. Meaning involved in the physical world, involved in human, um, in, uh, AI agents. Everything is going to be happening on a reasoning basis. This is the part that disrupts the labor side. Uh, data center AI is brilliant for training, but it's too slow for reaction. Again, the reactive stage where all of the stuff is going, uh, you guys can read this on your own, but he starts to get into the need in the autonomous vehicles. He made a comment in this interview about Elon thinking until we get AI interacting with the physical world, you don't have ASI. Please expand on this point. I'm not going to read this. I just want you to understand that until we get to the point of actually having AI interact with the physics of the world, Elon Musk's point is we can't have ASI. It makes perfect logical sense to me. You can have it be really smart. It can do things, but until it can actually look at something and then react to it and be involved in it, we're not there yet. We have a long way to go, but you're going to start seeing some of this with autonomous vehicles. That's why they are so important to the front loading of that TAM is because once you get to humanoids on wheels, you're getting closer to humanoids. I was going to do a big thing on humanoids. I'll do that next week. From playbook to play, why NPUs are AI's championship moment. If you guys are not subscribers to 22V, I wrote this this week. It's a long deep thematic piece on NPUs. It goes through including the companies and things that you should be involved in. And again, so you can see where this goes. This is just

One of the uh scenarios that I went through. Take from this. This is how you can use my documents. When I do a deep research one where it's 10, 12, 30 pages, uh you bring it into a chat and then you start going through the investor takeaways. You can take it even further. I'll show you some more on that, but that's the way I use it.

So, the MPU one went out this week. Now, in terms of the demand side, we got a ton of things this week on this chip demand continues absorbing all new supply. So there was earlier on in the week which gets forgotten some article on Oracle and renting prices and they don't have the margins. Their stock sold off about 8% and then it rallied all the way back before the Trump announcement. Uh Kobay letter this really good one in terms of uh uh going through you can follow this there but computers quickly becoming the world's most valuable commodity continues on this thread. AI comput demand is now growing over two times the rate of Moore's law, creating a massive shortage. Uh, something everyone needs to be on top. AI has broken this law. AI's compute demand has grown at double the rate of Moore's law over the last 10 years.

Now, Sam Alman interview A16Z. Another one to listen to. Uh, I didn't listen to the whole thing though. I listened to most of it. What does Sam mean when he says capability overhang in the interview? He said this in there. So I wanted to take this further for you guys. He is saying today's models are already capable of much more than people realize are routine use. There's a gap between latent ability and mainstream deployment awareness. This is a really critical thing. So when these deals are happening with Nvidia, when they're happening with all of these big places with AMD, they get to know how far ahead these models are. If you knew what the future held, and we're talking about doubling computing power if you already had something that could do something unbelievable and you saw it, you're going to be willing to invest in it, particularly if you're Nvidia and people that are going to benefit from the compute needs. But also if they have the money and they are concerned with curing cancer with all of the things that will happen that Demis Sabis has talked about that Sam Alman has talked about desiring the education side everything the choices that are going to have to be made because we don't have the compute electricity he says this regularly you would make investments too a backlog of untapped abilities and current models we've never seen this before where the models again to Dylan Patel's point are moving faster than the supply We don't have the buildout. The buildout theoretically should have happened a long time ago, but that's not the way this works. That's why this is not a typical technology cycle. So, if you compare this to the dotcom bubble, you're making a huge mistake. If you compare it to any cycle, you're making a huge mistake and you will be chasing this. The things that can delay this regulations from the government, which are happening, and that's a risk. I was going to focus on that, but that can slow it down. Listen to the all in podcast from last week. Also listen to all the moonshots that can slow it down. rare earth can slow it down. We could not get power. We could have a power problem. It's not going to happen from the models. So, just remember that that the models are moving faster than the the government can deal with it. It's moving faster than the power and the compute can keep up with it. Does this mean the models are being slowed from release due to demand? Yes, you read his comments on this.

So, um the compute and power breaking it down. So, one of the things I also do is I take two transcripts, put them together. Will you compare the Altman and Patel things and go through it um in terms of are they saying the same thing? What are their views? Sam Alman's capability overhang, we just talked about it compliments this Dylan Patel's view. If model improvement were too slow, it would be severe, but he remains confident that scaling laws are holding the capabilities will keep rising. Now, he doesn't get to see what OpenAI is. He gets to hear Scuttlebutt on it, which is one of the reasons why he's saying we're not there yet based on everything he's heard. It justifies the massive capex buildout. They're going to keep building out. This is going to keep happening. This is not a bubble. They are making more advancements. So when people say, "Oh, the model's not as good as it should be at this point." They don't even see the models. They don't even use the models. I guarantee you the people that are complaining about it. Even if they're a tech person, just go back and look. If they've been saying this for a year, nobody changes their views on AI. Nobody. I've gone back and reattended very sel I see someone say, "Nope, I underestimated." Alon Mus said there was no way that OpenAI would ever get the money for Stargate. I now the complaint is he's getting too much money. Nobody admits they're wrong. This is kind of the world that we live in. Altman highlights supply scarcity. Patella firms ongoing demand growth fueled by continued breakthroughs. Together they suggest an AI economy still driven by real technological momentum, not an overbuild.

Uh Moonshots podcast this week. So did they speak about Claude Imagine in the podcast? I'm only bringing up one part because again I don't have uh enough time on this. So claude imagine came out real time software. So remember I just showed you something I did in Claude in terms of building something that took three minutes that wasn't a piece of software but they are talking about claude imagine which allows you to go from idea to startup big SAS firms can cannibalize their own business risk the incumbent dilemma there I'm starting to get into this software thing I get asked a lot of times on the software side I'm going to spend some time on it now the entire 13 trillion labor market becomes the new addressable market this they continue on the same themes that we had heard I like when different smart people are coming up with exact same themes. Now, software is eating labor. I again want to say that for everyone who has tried to sell me on Salesforce.com and Adobe and they're too oversold and this is going on and they're going to benefit. As someone who uses it, I find it highly unlikely that any software company, particular a big software company, where their price to sales are already at levels that are suggestive of being a software company, if your growth of new revenue stops, meaning if you can't find it, you're in trouble. So, they're going to have to meet a very big hurdle in my opinion to actually get out of the doldrums uh and actually survive. I think all of them are under attack. And this is part of the thing that goes on. So this is another A16Z titled software is eating labor. This is focused more on the labor side. Software's new goal, replace labor, not just digitize it. Startups literally apply for job listings. So you have to understand it. Startups literally apply for job listings. They're offering to do the task for less than human wages. So these are startups that are going into job opportunities and highlighting that they're an AI receptionist for 20,000 versus 45. This arbitrage has started, guys. This is a profit margin bonanza in terms of getting rid of people. I hate to say it, but the global scare opportunity. The labor market dwarfs software. The future giants will be those that make software look small by automating labor directly. Points of agreement between that podcast and Dylan Patel's. Again, both believe traditional SAS margins and models will be challenged by AI native entrance. Both see enormous new market opportunities by replacing humans. Jeff Beav's family said your margin is my opportunity. That was the business back in 2007 to 2020. The AI adapt adaptation of this your job is my opportunity. Artificial intelligence entering the MPU area. Unfortunately, that is what's happening.

One more thing in the moonshots podcast at minute 49 Dave talks about other areas outside of semis where the supply chain beneath the surface where we could find companies in that space. Please give me the details on that. That's all I did guys transcript. I'm just showing you how you can get ideas. It goes in from that point reads what he wrote and it gives you a bunch of companies in here. Okay. Next step I do I go through I ask it. It gives me the companies. It gives me the theme. Then now for each of these three verticals only give me public companies that are mid or small. And again I'm looking for a broadening out. So the most of these are large cap. But if you guys want to go through it that's fine. Then it goes through. It gives me small cap businesses midcap businesses. I give that to John Ro. He gives me a technical score. Then I go through and have an analysis. And now I have names to trade that have a great technical score. Oh I mention missed one thing. Sorry. In this one here, uh, I wanted to go through the most recent earnings reports to find out that they are seeing large demand. So again, I go from a podcast to the supply the verticals. I go from the verticals to the names. Then I say, give me small cap and midcap names. Then I go through and say, okay, now only filter the ones through that have actually had commentary that suggest that they're benefiting from this. And then give me the technical ones that match up with it. That's the way that you can come up with ideas very very quickly from a podcast using LLMs.

Uh Eric Schmidt, I've told you before, if Dylan Patel is the only person you need to listen to on the supply chain side, Eric Schmidt's the only person you need to talk about on the future and what's coming, he is the Michael Milin of this day and age in terms of having all these smart businesses and people talking on a daily basis, including the governments. He's the person you want to go. So when he says this is an extraordinary, please read it. The progress is amazing. state of the AI report. Go look at it. Tons of uh uh things. You can go listen to the video. It's 25 minutes. Uh he doesn't tweet or post much worth doing. Almost 100 million jobs could be lost to AI. This is a Bernie Sanders report which came out. So the job thing will be part of the midterms. Gen Z's anger is not just in the US. It's a ruling. It's erupting across the world. This is going to be a big story and it'll be emphasized again by the fact that Mandani is at 89% as we sit here today before the election in a few weeks.

The debasement trade has made the headlines with Groman. Um, and again, this is as obvious for all you that have not moved money into uh, Bitcoin at this point and are still worried about its volatility or anything like that. uh since COVID in dollars, NASDAQ up 165, S&P up 102, home price is up 59. In gold and NASDAQ's only up seven, S&P is down 18, home price is down 37. So gold's outperformed all of these except for NDX. And in Bitcoin terms, NDX down 78, S&P down 85, and your home is down 87. Muhammad Alan the debasement trade the banks are starting to talk about it Charlie blow this was before Friday I mean Bitcoin's down another 8% or something like that over the last two days gold and Bitcoin are the top performing major assets so far in 2025 we've never seen these two in the number one spots the debasement trade that's why people are talking about it that's why Bitcoin's up 100% over one year and it's up a th000% over five years amazing Peter Deiamandis Morgan Stanley Wealth unit advises 2 to 4% in crypto. And then for Bitcoin, later in the week, Morgan Stanley opens Bitcoin and crypto fund access. Finally, if you thought all wealth clients had it, you were wrong. They did not. And Luxembourg claims bragging rights as the first Eurozone nation to basically invest in Bitcoin in a sovereign wealth fund.

That's it for this week. Sorry about the video, but you guys don't need to see me. You get to see everything else in there. out. I'll see you next week.