The Hidden Crisis: Private Credit, Hyperscaler Leverage, and the Software Reckoning

All right. Uh, you can tell by uh 15 things on the list. Uh, we got a lot to go through. Uh, it's been a uh, a historical month from a dispersion basis and uh, a lot's going on and I think it's going to be important going forward um for people to stay on top of things just because there's a lot of moving pieces.

I'm going to try to take you through what's happened this week, how it connects a little bit into prior trends that were also in place. Uh, but for those of you on the subscriber side who who saw the webinar um this week, so for those of you who are not subscribers, I did a webinar this week where I really spent a lot of time just going through what I expect uh has occurred, what is going on in particular with credit financials, uh, the spread that has led from the software unwind, but how this is uh turning into a much bigger story and regardless of whether you want to be bearish, you're bullish, whatever the case is, uh, to me the risk-reward has changed in the market significantly and we have a lot of things to get through. Uh, that webinar I'm going to do probably um more than a couple times a month if the current situation continues.

The headline index hasn't moved much, which I'll go through, but under the hood, whether it's the dispersion that's happened, the credit markets, which continue to weaken out, financial stocks, which are the worst performing sector, a whole bunch of things. I covered it in the webinar. If you haven't subscribed yet, uh, you'll be able to get the link, go subscribe, um, and you'll be able to go through and see it, but more importantly, I just want to make sure you know th this is really kind of what the webinar was. So for for these weeklies, for the stuff I do with Anthony Pompiano and any other podcasts I do, I'm really just talking about the bigger picture, uh, that is a long-term thought that could change along the way.

What I did in the webinar this week and the way that I approach things is very similar to what is in here. uh, at some point there's a reveal where the market starts to based on the correlations and the trends that are happening and how much it's spreading and the contagion you get a clearer sense as to what's happening. Uh, these are all of the pieces of the puzzle that I cover in these weeklies but at some point everything starts to gel together and it becomes a bigger story. It doesn't mean anything is going to happen, but what it does mean is that the risk-reward has changed. And what I went through in the webinar is not only what I'm doing in my portfolio and I made some shifts and remember I'm long the physical upgrade trade. A lot of the things you guys have seen here and I still believe in it, but I also believe that the risk-reward has changed because of how big the dispersion is and really the contagion that is going on. So that's what was covered this week.

Um, and if it's uh, if we're in this stage basically during these time periods like in here when energy fell in here when the S&P fell 20%. In here when we had COVID in here when the rate hikes were happening and then obviously uh in here with SVB and then last year with liberation day this is CDX for IG right now we were sitting at very low levels. Um, I started warning about the risk when they were at low levels when I got my turbulence model. And so, it starts with the turbulence model and then it starts to spread. Credit is still tight and the S&P is only 2.5% off its all-time highs. So, it's not like we've done anything. This line in the sand is 5%. uh, if we get a CDX move which I believe the probability is increasing every day uh, you're going to see a move that gets 15 to 20% which I talked about at the beginning of the year and right now the dominoes are shaping up that that's getting closer and closer or at least the probability is increasing for that to occur again it was February 3rd that I got the turbulence model basically warning signal I showed how many of these uh events had occurred uh to warn for those of you who've had uh not going to this is where you go for the subscribing side. So the AO Macro Nexus is where all of all of the stuff on 22V's account is for me. If you didn't get the link to the webinar or if you missed the webinar, everyone got the link. Uh, just go to contact and send it. It will be up on the website uh Monday or Tuesday. Uh, but if you want to see it, just go there. If you want to subscribe, go through it.

By the way, the main uh outreach I think uh and where this is going to be the most hopeful as I go through stuff today is going to be with financial advisors and RAAS. Um, we're getting more and more um groupings of people from RAAS in there. And I think this year because of what's happening in private credit because of what because of what was happening in private equity which is also going to be impacted by private credit. Uh, also the amount of ETFs that were created that have a lot of AI quote unquote software in them, which I think I just continue to see the people pound the table that this is a mistake in software. Uh, I think this is where you at least want to have someone on the other side. I do not think the software uh fall is a mistake. In fact, I think it's gotten to the point where people that are saying that are trying to fight the market, which is never a good thing to do. And as I go through this, I'll highlight it.

From an idea basis, the freshest idea, I finally did release the chemical side uh with about 15 to 20 names, which again you can get from 22V. So the big story this week, at least to start out that got everyone freaked out a little bit was the Catrini global intelligence side. You've got people going on the other side of it and then Gavin Baker put something out that basically I'd say covers one of the angles that I agree with. I do not uh believe in the catrini side. I also believe it is part of the distribution. So instead of saying whether it can happen or not which is what we seem to become a society of everything is binary. That's not the way markets work. Is there a possibility that we have an unemployment rate going higher? Absolutely. It is a possibility. Is it going to happen from AI alone? No, it's not going to happen from AI alone. And the reason that Gavin Baker gives, which I agree with, is that it would take a thousand times more compute. Now, whether or not that's the number that it would need, actually he may have said 10,000, but I'm not going to pull this down. Uh, the main point is we don't have enough compute to replace, but that's not the only issue. And this is the thing I want to make sure you guys realize because this gets into the risk with the hyperscalers that I've been debating people on all week.

This was part of what I went through in the webinar. The webinar covered everything starts with AI. So what has changed in AI since November and this is why I have strong disagreements with the people who are out there promoting like this is not correct what's happening in software. They're also not acknowledging all of the changes that have happened in AI since the end of November. And there have been dramatic shifts that I went through. There's about six of them that I covered that are all new news. And if you don't change your viewpoint or at least say the market is discounting the possibility of something in the future, you're just not doing right by retail investors in particular, but you're not doing right by anyone to sit there and just continue to say the same story like nothing has changed. Unlike Liberation Day where it was a government action that was then pulled back, that one makes sense. COVID where it was a a a virus with tons of money printed and then a vaccine. This one is not something to get away. AI is not going away which is why I called it a supersonic tsunami.

The other issue with uh the hyperscalers is the adoption. Enterprises are not going to adopt AI. And this is the part that I want to make sure I'm clear to everyone. They want to adopt, but there's a lot more issues that go in with big enterprises using AI than someone like me sitting at home or any entrepreneur sitting at home. The I can use Open Claw, they cannot. I can use AI agents today, they cannot. I can use Chinese model. They cannot. As the year goes on and we realize that agents for the eight billion people on the world who can afford a computer like a Mac Mini or a Mac studio or as these things go on the cloud can use AI agents every single day. More and more Gentic stuff is being done which means the 8 billion people have tools that the enterprises don't aren't able to use yet. Data's having worked at a Fortune 500 country 500 company at a very high level having run an office for the firm you have to understand that the amount of friction with inside there at so many levels will prevent complete adoption to equal all the firings that Catrini had. It will also make the SAS growth side not a story uh and I'll cover that more. So the real friction slowing enterprise AI adoption. Um, this does not get brought up enough. But I just do not see them being able to do adopt at the pace. If that does uh slow down, it has huge implications because that means the revenues for Anthropic and all of these companies are going to be in trouble. If you add in the data center delays and they can't get the cloud up and people move off the cloud and they move to devices like Apple Mac minis, you end up with a situation where they're not getting the inference money that they thought they'd get or at least not yet.

If you want to watch something that goes through the friction with inside enterprises, watch this podcast. And again, for people who subscribe on the link, all the podcasts, there's about five or six that I'm going through today that I use for some of the information. Uh, there is a recap with the links, so that way you don't have to go memorize them or write them down. Um, this uh in that podcast, the speakers argue the problem is not the models. The models are good enough. Completely agree. The models are already good enough. Claude is already good enough. The failure stems from poor enterprise data quality and scale issues, not LM capability. This is one of the reasons why I love Palunteer. It is a very very difficult situation to get the data organized. So what everyone believes is, oh yeah, we'll just pay more money for all of the SAS companies agents. That's not the way this is going to go. AI simply exposes the long ignored garbage in garbage out problem at scale. Completely agree, lived it, saw it. Uh, McKenzie, these are all of the places in AI risk by enterprise risk category that these enterprises need to solve. They move slow. They make decisions slow. AI's progress from November to now is exponential times whatever you want. It is amazing how much more you can do. The enterprises can't move that fast.

Dario Modai was on with Darkh. Uh, you don't have to listen to the whole thing, but he again went through the bankruptcy risk. And I just want to give you the quotes here. If my revenue is not $1 trillion, if it's even 800 billion, there's no force on earth. There's no hedge on earth that could stop me from going bankrupt if I buy that much compute. This gets into the debt and the compute purchases which I go through in the webinar are getting and have gotten more expensive due to memory. For those people who want to debate what the memory costs are in a data center, to get this correct, you have to spend a lot of time on where memory prices are right now compared to where they were in September. Go through the numbers and realize how much of a blackwell and how much that has changed. We are around the level now on the current prices of somewhere of 45%. Compute as a whole is somewhere around 50 to 60% still and especially with the memory cost going up. So when you go through the numbers and people want to debate whether it's 10 billion for memory, 15 billion on a uh, a 50 billion data center, the main point is the numbers have gone up exponentially and these guys are having to hoard compute if you end up without the power coming through and they've prepaid for all the compute. All of these things fit into the bet. I think the most important thing about this is they're making a bet on revenue coming in and as I mentioned between Chinese models between a lot of things going off the cloud where they're not going to get this you have to start paying attention AI agent openclaw adopts Chinese models for cost edge over US rivals this is becoming a bigger and bigger theme 80% of USA startup US AI startups this is from Andre and Horowitz rely on Chinese open source models ro pal had Emmod Mosto. If you haven't heard E-mod before, he's an interesting guy because hedge fund guy who then went and did a AI company. He always has a lot of um fairly extreme views. I agree with most of them um down the road, but I think he's on spot with bunch of things in here. He talked about crypto. Uh, but most importantly, this uh interaction Claude scores top in terms of human interaction and naturalness of speech and writing quality. Claude is the best for doing Excel documents. could not agree more. His point, the only issue with Claude is that it's too damn expensive. Miniax Chinese model which has the same performance as Claude is 20 times cheaper for the same performance. He's an AI per. So anyone who's going to argue with me that this is not important because enterprises won't use Chinese models, I agree with you. But the friction will prevent the revenue from the from the enterprises of getting there. Any hackings that happen this year, we'll go through hackings as time go on. I've done the cyber side. Will slow down the process. I think on premise is the only solution for the enterprises. I do not believe the cloud is a potential solution for them, which means they're not going to fire as many people. They're going to have to spend more money because the cost of inference is going higher. It causes all kinds of bottlenecks. So remember, bottlenecks, bottlenecks, bottlenecks. That's what we're seeing.

Um, Dan Ives put out the claw event suggested that fears of AI supplanting the software industry may be overblown. I don't have a problem with saying AI supplanting the software industry may be overblown. I think for enterprises that story is correct. But as I go through things I think you're missing a bigger picture here. And that's why I I think this is really not a uh I don't think this is the right thing to do to totally call this a ghost trade. that there's no reason for this going on and this is purely fear as I go through this. This is not purely fear or this stuff. I I don't know if this is just an ETF sales pitch or what, but I I put this through and ask uh Claude to be a forensic uh uh accountant and to take a Quantico FBI approach. And I won't show the results, but let's just say that kind of stuff of saying this is a goodbye when nobody's embedding any of the conversation of the the themes that have changed since November. Uh, I've mentioned Andreas Steno. I like his work. Uh, you know, he's been skeptical on software, but I disagree with this part too. Uh, making the rounds that Bloomberg uh can be replaced with perplexity. Uh, it can't completely be uh replaced by it. Obviously, some of it can be, but I think the more important thing is what's going to happen to the seats of Bloomberg. Is it a growth side anymore? Instead of saying that AI is a binary thing and it won't replace, everyone should agree with that. We still have taxis. Uber never replaced all the taxis. But the problem is, have taxi medallions increased since Uber came? Is there a problem with the fact that these things aren't going on? This is the issue. So, it reminds me of something I've used in a bunch of papers over the last 20 years, which is this famous Sher I'm a big Sherlock Holmes fan, and this famous uh scene from one of his books when he asks Watson, "What do you see in the sky?" And he tells him, "I see galaxies. I see stars. What do you see?" And Sherlock Holmes says, "You're missing the main point. Who stole our tent?" This is the problem with right now just saying software is not going to be replaced with AI. It is a very very loose story.

Here's the argument I want to make. Friction. Again, all of these conversations with the suits that are hanging up at a high level. Uh, one of my great achievements of leaving Morgan Stanley was not having to wear a tie anymore. Uh, occasionally I have one on. on the website there's a photo of me with a tie on but trust me I don't even know where my ties are right now. Um, the big thing is this for any startups as I said will they ever use salesforce.com I see Daario Modai saying of course they will look at these 500 billion800 billion dollar companies I agree if you're an 800 billion company like OpenAI or XAI or any of these will you use Salesforce sure but if there's 300 million entrepreneurs building bespoke things. If you want to change your CRM because you have another business line, you just do it. I do this all the time on the software that I build for my business. I change it on the fly. That is what is the power of AI. So the other part with going against this, again, let's assume that of the people I just showed, you tend to agree and you want to step in and buy. Well, there's two arguments you're making. One is, okay, I I think this is all fake. Okay. Well, the market's telling you it's not because this chart to just jump into this and say I think we're going back up all the way to the highs. This is the IGV versus NDX. So, I agree. I'm I'm in the same camp. We're at the point now where I think software has done has done as much as it needs to do. Would I buy a name? No. Because anyone can miss a name and fall 20 to 40% at this point. Do I think there's upside? Yeah. We've moved Salesforce down to the same PE as Ford. Is there upside? Sure. But remember, Ford has unchanged for the last 36 years. If you ask me, there's a big similarity between Salesforce.com and Ford. I'm not saying they're going out of business and they're not a zero, but I am saying if you wasted your time trying to pick the bottom of Ford and buy it, you want to go with the flow. Go buy some silver, go buy uh some analog semiconductors, go buy some chemicals, go buy some energy stocks. We're going to need more of that over the next three years. We just don't know if the software names are going to be in business in three years or which ones will.

If you believe the equities are wrong and you're saying this is a ghost trade, then I will go to the argument that I've always heard since I was at Morgan Stanley. I'm from the equity side and everyone who's watching this that's a macro person from the fixed income side knows what I'm about to say. every fixed income person during the 90s especially while the equity market was in a raging bull market that made no sense and the bond people were sitting there both having equity envy but also thinking everyone on the equity side was an idiot um and that the sharp pencils were on the fixed income side there's a reason for it bond investors have to do a lot more homework and pay attention to things on a detailed basis because the fear was they wouldn't get paid where equity people it was trading on hope so the math behind bonds is showing up. So this is the leverage loan side for technology. So if you don't believe the software trade, just look what's happened over the course of the last 5 weeks. So if you're a a financial adviser or an RAIA and you have private credit and you have technology stocks and software stocks are a big portion of your portfolio. When the debt and equity are moving the same way, it's a big issue. This was the liberation day. It was barely a budge down. It didn't move that much. This is a huge move. The bond market is matching up right now on the lever loan side and it's spreading into other places. So option traders pile into bets against the software exposed loan ETF. This is follow it real time BKLN. Tons of puts being purchased over the course of the week. Has a lot of software in it.

Here we are month to date. S&P only down 1%. So for all the things I'm saying again you can leave the market going it's fine. and utilities, staples, energy, material, industrials. And then down here, basically, you have the MAG 7 and financials. We'll get into that. Uh, down 1% for the month. The S&P again, you wouldn't know anything was going on. Russell 2000 up 70 basis points. Even the NASDAQ, the QQQ, only down 2%. Not a big deal. The big deal under the hood was we had 150 names. So if you take this, this is 106 of the S&P 500 were up over 10%. And at the same time, you have basically 46 that were down at least 10%. So 150 of the 500, 30% of the names moved at least 10%. That is why you've had such a massive dispersion. That's why my turbulence model was up. That is never a healthy sign, but we know what it is. It's a rotation. The utility and staple side. Just remember this. And I covered this on the webinar. We are at a point where I want to buy things that I know are going to be here in 3 years. I don't know if Salesforce will still be growing in 3 years, but I know McDonald's will still be a business in 3 years. I know that CVS will. I You can go through the list and go through the companies. They are overpaying now for multiples in companies based on who's going to be here in 3 years. When will that end? I don't know. This has a rounding side to it. Is it possible that we could make a new high and then come back down? Yeah, I think the risk is growing that there's going to be a correction that is big. But outside of the US and outside of the dispersion, MSCI World X the US looks fantastic. Again, I showed this before, we barely broke out middle last year, but this chart is going parabolic the other way of IGV. And again, this is because they don't have software in there. It's low tech. Here are the monthly numbers for MSCI world. If you're looking for something to buy, this is the place where I would go. Foreign stocks, foreign stocks, foreign stocks because the software thing is leading to more rotation.

We had Nim to Leb warning on the software bankruptcies. Traders rush to dump software loans. So again, if you think this is just an equity problem, it's not just an equity problem. The private credit story is a real story and people need to start being on top of it. This is not some ghost thing. This is a real thing. This is something that has a lot of the eerie part of the subprime side where retail ends up getting trapped in things and we're seeing debt widen. The question is, will the contagion spread from tech and go into other places? And as I go through this, it's already in the equity market in this. So, watch the private credit side and just be aware of it. And if you've got investors that are in there, you want to stay on top of this thing so at least you have a sense as to what's going on. If this happens, UBS now sees private credit defaults reaching 15%. I highlighted the fact that this was this week. The reason I highlighted is because three weeks ago they said it would hit 13% in worst case. They've already upped it. Uh, you've got uh hedge fund managers trying to buy things at a discount basically out there saying private credit's going to worsen. Apollo bad loans were not just software. They were in industries like healthcare, transportation, manufacturing. They had a bad week this week. Um, you start treating as volatility in hiding. This is someone who went through again the great financial crisis. We have here uh we have a lot of things going on that remind me of that. It also reminds me of parts of the dotcom bubble. And remember the unwinds that were happening there were not just in equity. It was in telecom. It was in Enron. There were a whole bunch of things that led that as the tide of liquidity went out, which is what's happening in private credit. You have investors that are trapped in vehicles now. This is becoming a bigger story. They want their money back, but unfortunately they signed up for something where they can't get their money back. So then you have gates going up. So once you start getting into this, you have to make sure that this is not an isolated problem. Leverage didn't go away. It just moved off the bank's balance sheets. I said this before, I'll say it again. The seeds of the current crisis were always planted in the prior crisis. When you stop the banks from being able to make loans, it leads to private credit and you get growth in that. And then when the tide of liquidity goes out there in an opaque situation, it's a little bit harder.

New credit blow up in London. This was on Friday. MFS is collapsing in London with themes similar to those of the auto lender Tricolor and First Brands Group. Banks such as Banko Santandere and Jeff are scrambling to recoup money with other entities like Apollo and Barclays. The collapse of the UK property lender sent shock waves through Wall Street. Again, this seems to involve fraud. Tide of liquidity goes out. You start to see who's naked. Apollo holds 20% of MS MFS senior debt. Boaz Weinstein again, we own a lot of CDS on life insuranceers in price. In past crisis, they went from trading at life insuranceers are becoming more of a story. I'll get into that also in a little bit. Uh, here's the chart of credit as we end the month. So the white line here is the option adjusted spread for technology. Basically we are approaching levels if you remember that CDX. We are at levels where historically the VIX goes higher. The S&P has a draw down. The main reason we haven't had it yet is cuz credit hasn't spiked here. If we do see the red line which is all sectors which is moving higher and is at the highs basically since liberation day. The green here is junk cash bonds. So this is the option adjusted spread. We're tracking that way and it's spreading. Here is the IGCDX. Again, this is when my turbulence model started to give me a signal. I had a lot of people in credit say, "I'm not seeing anything." Well, now you're seeing something. Here's the leverage loan total return index. This is the BK BKLN, but as a total return index. You guys can see the symbol in Bloomberg if you want to track it. This is the 200 day moving average. Basically, since the Fed pivot in October of 22 when the tightening cycle ended, leverage loans have been in a bull market, not so much anymore, and it seems to be just starting. Here is the monthly returns. Normally, if people are trapped, this stuff tends to get worse. Now, I will have you go back to a report in 2024 from the IMF on the global financial stability report the sector could experience. This is about the rise and risks of private credit. It could experience large and unexpected losses in a downturn. Okay. Well, we're starting that right now. Liquidity risk could rise with the growth of retail funds. That is where we are right now. So, this was again from the IMF.

Julia Lar Roach had on Chris Whan. Uh, Chris Whan was uh one of the people I love listening to during the SVB crisis situation. and I thought he was the most rational and uh actually had a a good uh a lot of good insights in terms of both sides. Uh, I wouldn't care that much whether he's right or wrong. I think he's been in this a long time on the financial side and I think he does good work. Uh, I'll actually be on Julia's show coming up if we can agree on a date in the next couple weeks. But this is probably the main thing that I took from the interview. I would tell you right now, Julia, probably half of all the managers out there that do private equity are going to end up having to go out of business. I don't want to say it again, but I'll just finish it. Because the results on their portfolios are so poor, they're not just going to be able to raise new capital. So, he's not saying this will be a domino of collapses. What he's saying is they're going to go out of business. Their portfolios may be purchased and they won't be out of business. So don't take it as the end of the world. I am not a systemic bear. I do not believe this can't be controlled. But I do believe private credit is big enough and has been a big enough issue that has been a worry on everyone I know probably for the last 3 years at least just because of the size and how quickly they had grown.

If I took you through some of the private um equity managers that have been uh especially aggressive in the last year and you look at their AUM growth and you realize that a lot of that is coming from a combination of the insurance side and retail a lot of these companies bought insurance companies and they use retail books. So again, for RAAS and FAS, if you're not getting the story, go subscribe, call up Mark Whailing at the firm and we can talk about a bigger package and I can come speak to you guys on a regular basis uh to at least keep you up on this so you can talk to your clients. If you haven't followed this story, this is in Connecticut. Um, and this is a problem of an insurance company that basically had to run into a liquidation process and there's a lot of losses happening on the variable annuity side. So, I just go read it. Um, this kind of slipped through the cracks this month, but I'm bringing this up because when people ask me how are the hyperscalers connected to what's happening in the software and the debt side, they are absolutely connected to this. um 100% connected to it. The reason is when you go through the details and you realize who has been helping all of the AIS in their SPV offbalance sheet financing for the data centers, you start to get back to the same names that I've highlighted in multiple points already from the software stuff that is not from the data centers yet. The problem with the data centers is not that AI is not going to be a boom. It is but everyone underestimated memory going up to the level it did at the same time as the bottlenecks happening that are happening and the bottlenecks are happening for a variety of reasons on the data centers which I can go into detail with people if they're interested you've got all of those bottlenecks and then like I said you've got competition from the Chinese open source which is caught up which is preventing a lot of the revenue that would come to these companies if they don't get the revenue in as I go back to Dario Modi this is a levered bet. So, if you have a levered bet, and I bring this up because Meta Auditor Ernston Young raised red flags on data center accounting, this is a big story. Um, and the reason it's a big story is because you're starting to get all of these things. You've got lawsuits going on with Oracle, lawsuits going on with some of the private equity companies by people that are trapped in things. And again, this is part of the doing business. This is not like some horrible story, but the fact that Meta's auditor raised red flags on this at their earnings report is just something you need to pay attention to because in the story, this means it was one of the hardest riskiest judgments the auditor had to make. Such a warning label is rare for a specific high profile transaction and major audit client. Meta owns 20% of the venture. Funds managed by Blue Owl, which we've seen has has sold off sharply, own the other 80%. A holding company called Beignet Investor, which owns the Blue Owl portion, sold the then record 27.3 billion of bonds to investors. And here are the senators asking the Financial Stability Oversight Council led by Scott Besson to investigate the risks posed by this to the AI system. Uh, you can't hide from these things. The strange case of Meta. Did Meta pay 270 million a year to potentially keep some of its AI assets and liabilities off the balance sheet? Now, this is something you've seen Jim Chenos talk about. You've seen Michael Bur talk about. The only reason to me this matters now is because of what has happened with software stocks. Again, I did not for the life of me expect to see software fall off as fast as it is. When I was writing paper after paper about a rerating in software, I did not expect it to go this fast. That's why the webinar went through what has happened to justify how quickly software has gone down. I thought it was a three-year process. I didn't think it would happen in 2 months. By happening in 2 months, it creates the credit cracks because believe it or not, guys, equity is part of the capital structure. That's why you're seeing the bonds come down. If the bonds are coming down, that means the bond investors think there's a risk that they're going to get their money back. Well, one of the ways to hedge that is to go short equity. We're in that cap structure sign that happened with the fracking situation with oil. that happened with the banks and the mortgage brokers. If it's a game of deleveraging, remember there is massive leverage in the system. Leverage did not go down since the great financial crisis. It just shifted from the banks to other places and now we're seeing the tide of liquidity go out. We've still got commercial real estate. We've still got private equity in terms of where did Harvard have to sell things off last year? How far down below par? All of these things are in there in here. Fun fact, this was the fifth largest bond issue ever done in the United States. The Meta deal, the accounting privilege of looking somewhat asset light and liability is a plus with the private credit deal with Blue. Of course, Meta gets to keep the associated debt of 27 billion off its balance sheet. S&P issued a statement clarifying they will not consolidate the MetaBlue debt with Meta's debt for this accounting privilege. The Metabl JV is paying 6.6% 6% to borrow money from blue L funds. Again, insurance companies, retail, and whoever else is in these funds. To be fair, the Meta press release talks about the speed and efficiency in which private credit operates. So, this is the reason we needed to get this done quickly and private credit allows us to raise the capital quickly and get the deal done quickly.

Here is the overlay between private equity stocks and the hyperscalers relative to the S&P. You cannot separate the hyperscalers from the risks that are happening. Here is the hyperscalers relative to the S&P. And again, this is an equal weight of Amazon, Google, Meta, and uh Microsoft. You have the worst month over the last decade except for 2022. So I leave it to you guys to make your own decision, but I think there is a clear sign of where the risks are associated. You had coreweave down big on Friday. Oracle CDS is still at the wides. There is a direct line right now into all of these deals and the fact that the software thing has unwound means there's an issue. Now, if it were only the software stocks and the software debt and the private equity firms, I I wouldn't be completely worried about it. But now on Friday, financial shares waloped by AI credit wos hit three-month low. So, as the S&P is spinning its wheels near all-time highs, the financials broke under the 200 day moving average where they closed the week. In the webinar, I went through how bad of a sign this is historically. Uh, again the last five days the two worst sectors meaning last week the banks so this is not the private equity this is the banks commercial banks so that includes the KRE which was down big the regional banks and includes other names with inside there and then this one consumer finance down 8% this includes MX which was down big last week again it's spreading guys so now you've got financials that were hit and you're even getting to the cream of the crop On Friday, Goldman Sachs relative to the S&P had its worst days since the great financial crisis. Here's just their outright move, worst move since the liberation day.

And again, if you want more contagion, and I said this on the webinar, this is one of the things I want my portfolio right now is duration. So 10-year yields went below four. This is a fairly crowded view that we would have a steepening of the yield curve and rates would go higher. I talked about this on the webinar as well and we broke the 200week moving average for the first time since the tightening cycle started to be built into the market. So you remember how much people have been bearish on treasuries. It shows up in almost everything. You'll never hear anyone say I I love long bonds particularly macro people. Uh, I think it's a very crowded short to be short fixed income.

Here is the major thing that I want people to have in their mind and why this this disruption is so real and so important. This is the gamecher that happened. I wrote this in Substack. If you didn't read it, AI agents change time. We are repricing the capital structure because of time. Time has shifted, guys. We are no longer in a world that moves slow. We are in a world that moves fast. And all of this disruption is because AI agents allow things now to be built not just by humans, but by a billion to a trillion bots, infinite bots running on Chinese software. This is something you just can't imagine would happen. I didn't envision this going on to the pace. Open Claw was a gamecher. Valuations are falling for a reason. AI is repricing the future.

I start going through Bitcoin more and more. If you guys haven't done your homework on Bitcoin, by the time we get through this, if there is an event where the private credit situation does deteriorate, the Fed will have to do something. It'll be another liquidity facility to help. We have a midterm election coming up. We've got lawsuits going on. You've got the Democrats saying there's an issue with the SPVS. If people get spooked out, you're going to have to see liquidity facilities put in place, especially ahead of the midterms. That's when Bitcoin will start to separate itself because at this point software is going down creating a deleveraging phase on the back of deflation. We need AI which I'll get into as well. That is the other powerful part. Ben Horowitz was on moonshots. There were so many great nuggets in this. I highly recommend going in there. I'm not even sure they realize how many good nuggets were in there. But one of them was his statement about open claw. He basically said as you go through this that openclaw changed the way Silicon Valley views AI they're looking at this different this changed everything he goes through the startups getting more involved this was something that shocked people Carpathy had talked about AI agents being as as much as a decade away and for enterprises which is where I think everyone made the mistake it's true because there's so many things that have to happen to avoid the mistakes, the hallucinations, all of that stuff. A lot of it has to do with the data cleansing. The enterprises are going to take a long time. But for someone sitting at home, even with the mistakes, you can deal with that. I can deal with them making mistakes, with them doing something I don't want them to do. But I can run a business that way. I just have to make sure that they're not in my files that matter and I have them in a sandbox. This can't be done in enterprise. So the problem is again if you have too many people you can't handle the speed. I don't care what goes on. Too many people mean too much friction. Humans mean friction. Everyone has different incentives. So if you go back to the show me the incentives I'll show you what happens. The incentives at an enterprise are to not blow up. It's to be risk averse because they get paid a lot of money at the seauite level to not make mistakes. AI will not happen quickly which means in my opinion there is a high probability the revenues for the open AIs for the anthropics for the enterprise level are going to be much slower than what people had expected and the debt or the capex is going to either have to slow down or they're going to have to cancel buybacks in the case of the public companies. It doesn't really matter. Founders are excited about productizing openclaw instantations. He talks about how there are just tons of startups coming in where the people are like I can build something on open cloth. So this is having an impact. So what you're going to get is the bot population bomb. There are 8 billion humans on the planet. If we start using agents in any meaningful sense, you'll get to a trillion agents very quickly. That is where this problem comes in on so many levels. Kilo Claw again, another another ability by this stuff to happen. Allowing anyone to deploy hosted OpenClaw agents into production in 60 seconds. OpenClaw users are allegedly bypassing antibbot systems. This is where the hacking stuff starts to become an issue. When you've got billions of Einsteins running around and then trillions at the end of the year, AI compresses time horizons. If three years becomes uncertain, long duration equities repric first because their value depends heavily on distant cash flows. Credit cares about survival, not story. Multiples compress before spreads widen. That's what we've seen. Duration shrinks across the capital structure. In an AI accelerated regime, balance sheets matter more than narrative. Stocks price the future. Bonds enforce whether the future can be funded. That is what is happening. And that's why we've gotten to an important thing.

This gets back into equities or opt optionality on the future. So, especially for financial advisors, retail, anyone watching this that doesn't think of equity and fixed income in the same context. There's a big difference particularly in the cap structure. So, I'm writing a paper about how equity is now becoming debt. Meaning, you're at a point where if you don't know if an equity is going to make it, the PE goes down to 10 or 12 like it did with Salesforce, what's the probability of them being in business in 3 years or at least growing in 3 years? What's the probability of getting money back? Salesforce is a big company. They're not the issue. They're still announcing buybacks, but you have issues for other companies. So, we're making a bet that margins will stay high. This is the risk that I think is going to be an issue this year based on all of the things that I've mentioned. Uh, the cost for a data center, the cost for cloud, the inability to give all of the capacity that's needed means if people want capacity for AI agents, it's going to cost more. Um, I don't know what's going to happen, but I don't see the data centers being built in time to supply what's necessary. If anything, I do for for a few clients, I do a a monthly data center recap and the the the terowatt hours, the gigawatts by 2030, they just keep declining based on all of the data center delays. So, you can't price growth in the future.

If you don't know what's going to happen. Fixed income, you either pay the coupon or you don't. So that's just the reality. That's why bonds and and equities are just different. And you see where it is. Uh and in stress regimes, it becomes obvious. There's secured debt, unsecured debt, preferred, and then there's the equity. Equity can rally while bonds quietly deteriorate. But when the bond market cracks, equities usually follow. So watch credit. Credit is smarter. Credit's cracking.

Um, the other thing that has changed this month in this Substack that I wrote uh about agents and time, it also destroys the Black Scholes model. I'm not going to go through all the details here. You guys can read it if you want. But basically, all of these things that the Black Scholes model assumes, we are at a point now where none of that's there. We are trading like prediction markets. A stock can go from being a 25 PE, it can go down to an 18 PE the next day and then be at a 12p by the end of the week. That was never possible before. Um, now you're getting movements that are jumping faster and out of nowhere. So, you're getting multiple standard deviation moves. And so, we're going to have times where things are quiet. You don't get a lot of movement and then it's going to go and market V. This is where it gets counterintuitive. May not increase proportionately. So individual equity vault is unambiguously higher because you keep getting these oneoff gunshots. But market vault, this is what we've seen so far. But the dangerous part is that we are probably episodically explosive index moves. That's why I said my whole thesis on this is volaval VIX call options or other points. You just have to have these in your portfolio to some degree.

This is the paper that I'm releasing which is basically going to show people that the leverage in the system is in the equity market. The equity market's the most important thing. I've highlighted this before but just so we see it. If this were debt to GDP, which we have at the government level at very high levels at over 100%, but this is the equity market relative to GDP, the quote unquote Buffett indicator, market cap to GDP is at 220%. Equity is where the leverage is in the system. That leverage is predominantly tech stocks. The 22V team, Dennis Debush and the strategy team, this is a a must-see uh chart. uh they broke this down and said, "Let's look at the S&P cash return relative to net income." And again, this is going down because the capex is huge. So, they're spending money and you have this down at levels which historically has been when you've come out of a recession or the dot bubble here. This is massive amounts of spending. And the reason this gets important as you break it down, first of all, fair value You could make the argument right now that we're 17% lower than we are just based on history. If you go through the impact to this of what I put in here, rising memory costs, private competition meaning from China, data center delays or the open- source, sorry, this is the private competition meaning open AI and XAI and anthropic uh relative to the S&P companies. So, they've got competition from very very big companies. the enterprise friction, meaning the adoption and buyback productions. So, people keep asking me, will the hyperscalers cut capex? I don't think they will. Do I think Anthropic or OpenAI could? Yeah, I think Open AAI technically did. They said they were going to do 1.4 trillion. Now, we're talking 600 billion by 2030. Uh, didn't hurt them because they're a private company with a valuation and no one seemed to care. Uh, but you could cancel buybacks if this becomes an issue and I think that would not be taking well. Um, if you go through what has when this has happened before and what has caused it, you get into these things. And again, like I said, it's mainly been times coming out of a recession. If the cash growth fails to keep pace with the current record high capex spending, we enter period often referred to as finance as the capex hangover. It's already consuming over 90% of their operating cash flow. Free cash flow goes negative when you start to get into a bad situation. A shift goes from cash to debt financing. We're already seeing that. You're crowding out shareholder reserves uh returns. Buybacks disappear. Dividends stagnate. You get a value a valuation derating. I just want you to read this because this is something I believe investors will stop valuing them as software companies and start valuing them as utilities or industrial firms with a lower multiple. That's what we've seen, guys. You spend too much money and you're being disrupted by something that is structural. if there there's a question on whether you're going to be able to do it and no one has proven yet that they can make money off AA agents. Nobody. Uh the hyperscaler pivot to capex or at least relative to their costs I should say. Um so this is just the reasons they've fallen down to there. So again I I I talked about it. I think you should just keep it in the back of your mind because even at the S&P level you can see how this problem is growing.

Uh Fed Governor Waller spoke this week. I just want to make sure you highlight that as we're in this situation. And remember, we do have a K-shaped economy. If the stock goes down, the economy is very levered to the stock market. We have very good PMIs. We have very good growth. I think at the end of the year, the growth will come through. I think the earnings will come through. Profit margins may not come through. But I think at some point if this thing doesn't see a significant change in the month of March. And when I mean significant, I mean where everything kind of tightens up the credit markets. I'm not caring for an equity bounce because that's why I think going through software stocks is a waste of time. Remember the job market is weak and I do expect it to get weaker. Um, not in a way that Satrini talked about, but I just think it's much weaker than people realize if for no other reason everyone's waiting to get fired.

Um, there is an increasingly dystopian tone to commentary on AI. Fed Governor Waller say the thing that with AI is that it's coming at us so fast that it's easy to start seeing jobs that may go away before you see the new jobs that are going to be created. Um, I I tend to agree. Um, I do think people should leave and start being an entrepreneur if they can. Uh, this got a lot of this got a lot of press. Again, I I cannot believe what circulates. Um, it got press cuz he laid off 40% of people. I mean, just go look at the stock over the last 10 years. I mean, they overhired. So, in the same way that a lot of the private equity firms have taken in tons of retail money and tons of money uh and purchased um companies like insurance companies, you just get to the point where you got to get rid of people. So, I don't think this is an AI thing as much of it's they needed to do it because their stock was down over the course of the last 5 years.

Um, this didn't get a lot of press, but I just want to say it. I think this again is a read through the lines indication that the reason Dariamod is talking so openly and so often about the risk. He's insinuated it's open AI, but I think the real risk is that their business is doing well and yet they're showing every sign that they're surprised they're not making as much money. So I showed last week that their margins have come down by 23%. This week they dropped their flagship safety pledge and the reason embedded was we didn't really feel with the rapid advance of AI that it made sense for us to make unilateral commitments if competitors are blazing ahead. So this is a competition thing to get rid of something that they were built on. That says to me again, like I said, if my revenue is not $1 trillion or it's even 800 billion, there's no force on earth that will prevent me from going bankrupt. I I think this is the issue to pay attention to, particularly since now they're in a heated fight with the defense department.

Uh this was all week. And of course, the president of the United States, I'm directing every federal agency in the United States government to immediately cease all use of anthropics technology. We don't need it. Um, in the same week, a hacker used Anthropics Claw to steal Mexican data trove. So, this is going at the Mexican government agencies. I just want you guys to think about what's coming in terms of the cyber side and what it would do to enterprises. Would it speed up their adoption or would it slow down their adoption that is necessary for Anthropic to have their money because they've already spent the money on the capex? So, we get back to the whole adoption. Will a massive hack or fears of hacking growing which is becoming more and more of reality because of open claw because of how many people now have AI agents that can sit on a Mac Mini or a Mac Studio and do insane 24 hours a day damage. It's going to slow adoption. There's no way it speeds it up.

Howard Marks uh he wrote an article back in I believe it was November, December, may have been October. Is it a bubble? where he went through to AI and effectively what he put in there is there are definitely signs of a bubble and there always are bubbles in this kind of spending but I kind of took it as he didn't really know so this is the reason why everyone loves Howard Marx he's a wellthoughtout person but also he did more work so his point is by making more powerful tools analytical tools available to everyone not just experts AI democratizes innovation I'm putting this up here because you guys have heard me speak a lot about the things that I'm I've got highlighted here. It democratizes innovation. I just wrote a Substack on this. It allows more people to become inventors, innovators, and entrepreneurs. This is the rise of the 300 billion, 300 million in the US. It's the rise of the 8 billion. This is the reason why Bitcoin will be the thing that comes out of this. I'll go through all the reasons why over the course of the the final 10 slides. AI democratizes innovation. It creates a cycle of self-generating ideas where ideas generate more ideas. This is the whole point about bespoke. This is the thing about building stuff at home. If you talk to someone who says has a view on AI and on software and they don't have openclaw and they haven't used it or they can't give you a long re uh reason on why it's so important, I don't think you should listen to their viewpoints on software at least as a growth. No software company is going to be thrown out of Morgan Stanley right now in any meaningful way. But the question is, will they get more seats? Will they pay for the AI agents? And will there be another Morgan Stanley behind the Morgan Stanley? Will that Morgan Stanley be a crypto AI native company, Fintech, that builds their own software that never gets to 5 billion and builds their own? The democration supports ROR's concept that ideas can be shared repeatedly without being used up, potentially transforming innovation by allowing more people to contribute to economic growth. There's the 300 million people as opposed to the 10,000. AI's emergence is poised to be the latest example of the creative destruction economist. Joseph Shumper, I've referred to him. I don't even know if it's every month for the last probably seven years, but certainly over the course of the last three years, the possibility that job placement may preede job creation such that the unemployment rate may rise in particip.

Okay. Yeah, that seems like same thing. And to be sure, the AI transition I'm contemplating could have profound implications for monetary policy. I'm saying the same thing. The most significant thing that distinguishes AI is something we've never dealt with in connection with prior technological developments. Listen to Howard. AI's ability to act autonomously. According to Claude, AI was, and again, he did all this work using Claude. AI was at a level one in 2023, level two in 2024, but now it's level three. And the difference is a big one. Something big is happening from Matt Schumer. I went through this. He references it and talks about how well it's written, but also talks about a lot of the things in there that fit into the same piece. So, the bottom line is, is it a bubble? Is the technology a fat or an illusion? I say with conviction that it's a very real thing with the potential to vastly alter the business world and change much of life as we know it. I think it has the potential to I think its potential is more likely to be underestimated today than exaggerated. completely agree. As I pointed out in December and every example of sweeping technological innovation, the headlong rush to build infrastructure has vastly accelerated the adoption of the innovation and caused a lot of capital to be malinvested and destroyed. There's no reason to assume this will be different. My entire webinar, everything we've gone through here is saying the same thing. The adoption of the innovation takes longer than people recognize, but the buildout is happening and the friction is growing. And if it wasn't for OpenClaw and the ability of Chinese models to be used very close to US models, I don't think that altogether would be there. But when you add in the data centers and everything else, it's an issue.

Um, it's important to note that more money is going into inference capex these days than training capex. This is a good thing. This is why it's not systemic. This is why I have it in here. If it was going into pre-training, I'd be more worried. But there is massive pre-training data centers being built. Those are the ones with the massive memory costs associated with which is another thing of the equation that if you compare inference, a lot less cost for memory. If you go into the big data centers with the blackwells and the heavy heavy compute, you're at a different story. So again, I wrote democratization as the input concentration is the output. how it's repricing this, the entrepreneurial label, everything that he just talked about is in this paper if you didn't see it. The one thing that I spent a lot of time on that he didn't is how capital changes in the A of AIA of agents. This is a deeper story. It's about duration. Equity begins to behave less like ownership of a franchise and more like a call option on execution. This is the whole thing of prediction markets. You just don't know whether they're in there. A call option can fall 20% in a day. AI rewards velocity and velocity demands different infrastructure. Public markets, again, I have said publicly, I will continue to say it. I don't know what public companies will be able to grow in a decade. I don't know in 5 years, I don't know any of the non physical ones. So you guys can take it however you want, but this is where the story for crypto and Bitcoin starts to take off. Traditional banking rails were constructed for human speed, for transactions between institutions operating on a quarterly and annual clock. None of this was designed for a world where millions of AI native businesses need capital formation and continuously are transacting. Tokenization enables fractional ownership and capital access. The last 15 years we rewarded the question, who can build the largest moat? The next 15 will be a different one. Who can adapt fastest when moes are no longer guaranteed time? This is why I wrote it. What does the capital architecture look like when the economy it serves runs 24/7 as time is the scarcest asset of all? This is where you can go listen to Raul Palmad. You can go listen to Jeff Curry. You can go listen to Ben Horowitz. All of them talked about this. All of them including Jeff Curry mentioned the role that crypto plays in this world of speed. All three of these gentlemen talked about it. In the case of Ben Horowitz, AI cannot fulfill its potential without crypto. If AI is going to fulfill its potential, like it would help a lot if crypto was a pervasive utility for it. The financial guardrails is one angle which I just talked about but this is the other one and this is the part that I talk about with everyone scared about quantum which is down the road. We are not there yet and we may not be there for a decade. Nobody is worried about agent swarms. I can say this over and over again. But where do you go if you've got deep fakes and you can't tell when someone is calling you who it is? What do you do when you get a video and you can't tell whether it's real or not? And what do you do with agents hacking into your system? Well, you go to cryptography and you go to the blockchain and onchain. That's where we're headed. If you haven't done your homework on it, you're playing from behind. And oh, by the way, the final positive note for Bitcoin, my thesis has always been that the only time that the people who own the world's wealth, the 800 trillion, will go buy a little old asset with $2 trillion is when they start to doubt everything. So, you're doubting everything with AI as I just showed. But the other thing is you have to doubt the investments that you're in. If I'm right about public companies basically not growing ever again, then being tokenized and becoming these just out there things that are surviving like Ford, where are you going to put your money into something that everyone else is putting your money into? Part of it will be gold as it is. But in particular, the biggest negative that could happen is this. The government is fighting with the hyperscalers and Anthropic may not be a hyperscaler but basically what they're saying to the world right now is we are taking over AI because AI is too important for national security. It's too important for each individual. If the government gets involved and it becomes regulated, what multiple should a regulated utility- like company have? That is what I imagine right now for multiple compression. The hyperscalers have a debt issue. I think they have a we've gone too big, too fast, too important thing. Uh, a lot of stories about anthropic basically saying they didn't want to be involved in the Iran situation. They were clearly involved in the Venezuela situation. AI is already a military issue and we already have hackings and involves Claude and anthropic. I think it is very very possible that the government gets too involved. The nationalization of AI threatens innovation in the American mind. This is a libertarian paper. I bring this up again because I think this is a story that everyone needs to have in the back of your mind. The most bullish thing for crypto will be if the most important fiat assets are basically in some way, shape, or form controlled by the government.

That's it for me this week. Um, again, thanks to everyone who's subscribed. Thanks to everyone who's reached out and who's sending me what they've been building. Uh, I love seeing all the open claw stuff. I love hearing what people are doing. the video series on AI and how to use it more and make sure your kids are using it. Uh I'm hoping to get the entire completed series out by the end of next week. That is my goal. Uh and I will see you guys next week. Go check out the webinar.