Things Steve Says on Podcasts

The 10-year is at 5.3%. It hasn't been here since around 2002. It feels like we are getting closer to something breaking. FICO had a very bad week. FICO's stock price was down over 25%. This change probably puts VantageScore at a competitive advantage over FICO. This Aspocalypse story has not gone away, it has just morphed. And I think these software investments are largely deb money, meaning investors are never going to see any return. The days of easy money are over. Private Equity's Software overexposed to software companies. Soon, some number of investors who piled into private equity deals offering a great potential return will find their investments are worthless. Here is the problem. Hi, this is Steve Eisman, and welcome to the weekly wrap. This is for the week ending Friday, October 2nd, but recorded Thursday night, October 1st. We have an exciting announcement to share. Starting this coming Monday, October 5th, we are offering early access for premium subscribers. Monday and Friday will be released early exclusively to paid Substack premium subscribers. All Monday episodes will now be released at 12 p.m. Eastern Standard Time every Sunday. The Friday rack will now be released at 12 p.m. Eastern Standard Time every Friday. For those of you who are not premium members, the Monday and Friday episodes will be released at the usual times: Monday at noon and Friday at 4:15. On this week's wrap, we will discuss one, the war in Iran, oil prices, and interest rates. Two, more negative news for poor FICO. Three, an analysis of private equity and private credit exposure to software. And four, free mailbags. Let's get started. In war news, the week began with President Trump projecting an offer from Iran. That drove both oil prices and interest rates higher and caused the market to decline. We are in an odd period right now where the only two variables that matter are oil prices and interest rates. The 10-year is at around 5.3%. It hasn't been here since around 2002. It feels like we are getting closer to something breaking. Moving on. FICO had a very bad week. On Tuesday, FICO's stock price was down over 25%. Why? Bill Pulte, the head of the FHFA, which is the regulator of Fannie Mae and Freddie Mac, posted on X that he was combining the LLPA grids for FICO and VantageScore, thereby removing the 20-point haircut on VantageScore. This change probably puts VantageScore at a competitive advantage over FICO. As a result, Rocket Mortgage, one of the largest mortgage lenders in the United States, announced that it would be defaulting to VantageScore. I expect that over the coming months, the adoption of VantageScore will accelerate by a lot. The FICO short thesis has been that FICO risks losing its mortgage monopoly. It raised prices 1,600% over the last five years and has angered the entire mortgage ecosystem. The lesson here is don't piss off your regulator. One more point on FICO: losing a monopoly because of bad strategic decisions is as bad a result for a company as imaginable. The strategic mistake was to be greedy and to raise prices by an insane amount. FICO has lost its monopoly in less than a year. Astonishing. The stock was down over 25% on Tuesday and is down over 60% year to date. William Lansing is the CEO of FICO. He has been the CEO since 2012. He should be fired immediately. This is a company in crisis and it needs new leadership. That he has not been fired is evidence that most boards are window dressing. Most board members don't take their jobs seriously. They just take a check and smile. Earnings season is two weeks away. So I thought now would be a good time to do a deep dive on private equity and private credit. There was a time not long ago when we would discuss private credit. Literally every week. The news was mostly bad with massive quarterly redemptions. Things have gone temporarily quiet, but the big question to be answered is whether equity will be wiped out after the debt is refinanced. This deep dive will set the stage by looking at the sheer size of private equity investments from 2017 to 2023. We will then look at how and why and where these software buyouts occurred and which were mostly SaaS, which is software as a subscription. And finally, we will ask: why did private equity have so much money to invest? Throughout the past decade, many private equity firms place a major emphasis on SaaS companies due to predictable revenues, high margins, and high growth. 90% of all tech buyouts were SaaS companies, and SaaS deal volume made up a quarter of the whole private equity buyout market. Massive specialist funds such as Thomas Bravo, Vista Equity, and Silverlake made up about 60% of the total SaaS buyout. From 2017 to 2023, private equity grew exponentially and averaged between $100 billion and $400 billion in transactions a year. The exact amount of private equity buyouts is difficult to completely figure out, but to give you a flavor of the annual size of the buyouts, let's look at 2021 to 2023. In 2021, which was the peak year, U.S. private equity buyouts of technology companies accounted for more than $284 billion. $256 billion was for software. The business model of steady revenues driven by technological moats, strong client relationships, and a low-rate environment, don't ever forget that low-rate environment made these companies highly attractive. Again, don't ever forget the importance of low interest rates in buyouts. Private Equity's Software back up the debt truck when rates were really low. In 2022, there was something of a correction to the unbridled enthusiasm by private equity firms to all things software because the debt tightened as inflation post-COVID grew for the first time in many years and rates rose. This slowdown really showed up in the 2023 numbers. Now, before we get to how these deals were financed, let's first think through the nature of the problem. Private Equity's liked to buy software companies because they fell in love with the SaaS model. The SaaS model gives private equity comfort because it is easily modelable. Multiply the average price charged customers by the average number of customers, and you've modeled revenue. Easy peasy. Now, make assumptions about the growth in customers and how much you can increase prices, and you can see a glorious future. Then lever it up with a nice amount of debt, and you are in business. Two things have happened since 2023. First, interest rates, obviously, are much higher, as therefore interest rate expenses have exploded since these deals were all funded with variable rate debt. So higher rates have already negatively impacted the earnings and cash flow of all of these companies. Second, the SAS Pocalypse has hit valuations hard. The original version of the SAS Pocalypse was that AI will wipe out all existing software. That narrative, which caused software stocks to universally decline, lasted from last summer until early this summer. Since then, software stocks have had mostly a nice rally. But the SAS Pocalypse story has not gone away. It has just morphed. The current narrative states that some software companies will be fine. Which ones? Those companies that have moats because their software is deeply embedded in enterprises, like perhaps ServiceNow and Salesforce. How other companies could be at risk? Which ones? B2C companies that interact with consumers like Intuit, Bookings, and Expedia. And some B2B companies are at risk as well. It is still very early in this story, and we will be monitoring it carefully. However, AI does not have to destroy a software company for it to have an impact. It could simply slow seat growth and make it more difficult to raise prices. That would hurt future profitability and valuations. Now, before we get to the full implications on software buyouts, let's look at some of the largest software transactions and how they were financed. Number one, Citrix Systems allows employees secure remote access to their company's secure system. I think this business model is vulnerable to AI, as AI might be able to do the same job, but much cheaper. The buyout. Was led by Vista Equity Partners. The purchase price was $16.5 billion, and it was funded with $8.55 billion in syndicated bank loans and junk bonds. That means 52% of the purchase price was funded with debt. Number two, McAfee Corporation is a cybersecurity company. Now, given all the hacking that AI models have been doing of late, I hope you've noticed, cybersecurity has a good shot at surviving the SAS Pocalypse. Public cybersecurity companies have done very well, and the deal was led by Advent International and Palmyra. The purchase price was $14 billion, and the deal was financed with $8.96 billion in junk bonds and levered loans. That means 64% of the purchase price was funded with debt. Number three, Zendesk is a customer service platform. I would say this business model could be vulnerable to AI. The deal was led by Hellman and Friedman and Permyra. The purchase price was $10.2 billion. The deal was funded with $5 billion in private credit loans in a deal led by Blackstone Credit. And that means that 49% of the purchase price was funded with debt. Four, Proofpoint is a cybersecurity company. So it has a shot of surviving the SASPOCOLPS. Also, Thomas Bravo led this deal. The purchase price was $12.3 billion. The deal was financed with $4.6 billion in a syndicated first lien loan. That means that 37% of the purchase price was funded with debt. Five and last, Anna Plan is a cloud-based scenario planning platform. This business model feels very vulnerable to AI. Thomas Bravo led the buyout. The purchase price was $10.4 billion. The debt level is relatively low at $2.5 billion. That means that only 24% of the deal was funded with debt. This particular private credit loan was led by a who's who of prior credit, including Blue Owl, Gallub Capital, Blackstone Credit, and Apollo Goal. So let's sum up the debt purchase price of these five large companies and also whether the company, in my view, is vulnerable. Citrix, 52% funded by debt, vulnerable. McAfee, 64% funded by debt. Probably not. Zendesk, 49% funded by debt, vulnerable. Proofpoint, 37% funded by debt. Probably not. And Anna Plan, 24% funded by debt, vulnerable. So three out of the top five transactions look, at least to me, vulnerable to AI intrusion. The range of debt to the purchase price is 24% to 64%. However, if you examine a much broader sample of software buyouts, you will find that on average, they were financed with roughly 50% debt. Here is the problem.

audio, and documents. I'm sure many of you. Have used Adobe. The perception is that Adobe is particularly vulnerable to AI. Now, Adobe hit an all-time high of $630 a share in February 2024. Today, the stock is 240. That is the decline of Get This 62%. Now, Adobe is not the only software company that the market is worried will suffer from AI decimation. But for purposes of our analysis, let's assume that most software companies are down 50-ish percent from their peak, with cybersecurity being a very large exception. The stocks are down not because their business models have fallen apart. They have not. Investors are just worried that they will. The same thought process applies to non-public software companies. The business may look fine for now, but the valuation multiples have been decimating. Note that corporate debt is not like mortgage debt. When you pay your mortgage every month, you pay interest and principal such that when you make your last payment, you have paid off the entire principal. But at a corporate loan, only interest is paid. That means that when the turn of the loan is over, the borrower has to raise new debt to pay off the old debt. Here's the mathematical problem: let's suppose a software company is bought out for $1 billion and 50% of the purchase price is funded by variable-rate private credit debt. Let's also assume that this transaction took place in 2022 and the debt was five-year debt, so that it will be refinanced in 2027 next year. And let's also assume that because of the SaaS apocalypse, the value of the company has declined by 50% since its purchase. Given what has happened to Adobe, that's not a crazy assumption. The fundamentals of the company that we're talking about, this hypothetical company, might still be okay, but the valuation multiple has been cut by a lot. So now, the total value of the company, which was a billion, is only 500 million. The debt does not go down in value until after the equity is reduced to zero. Since the debt is 500 million and the new value is 500 million, equity is worthless. And now, when the private equity firm approaches a private credit firm to refinance the debt, they have to come up with new equity to shore up their balance sheet. Like I said, this company's earnings and cash flow might still be fine and the equity could still be zero. Both can be true. In negotiating a new package, private credit will demand that private equity put in another $500 million. Now, will a private equity firm do that? Maybe, but I doubt it. That would mean that the private equity firm will have put in a billion dollars of its own money for a company that is now worth only $500 million. I think most private equity firms will just walk away. Private Equity's will then own the company, and what they will do then is anyone's guess. So, what are the implications? I don't think this private equity software problem is big enough to sink the economy. It is, however, big enough to seriously damage the reputation of the private equity industry. Not too long ago, a private equity fund would become monetized in three to four years. Because of the weak IPO market of the last several years, that duration has extended to seven years. And I think these software investments are largely debt money, meaning investors are never going to see any return. The duration for private equity monetization is going to keep extending, and many investments will be marked to zero. Soon, some number of investors who piled into private equity deals offering a great potential return will find their investments are worthless. One of the reasons why private equity has grown so much since the great financial crisis is because of what I call a sleight of hand. During the GFC, institutions watched the value of their assets decline every day. It was a painful experience. After the GFC, many institutions realized that if they invested in private equity, the volatility of their assets would be reduced. I call this a sleight of hand because the volatility was still there. They were just not provided a daily price. Private Equity's Software took those assets and invested the proceeds in an exceptionally low rate environment, which was perfect for a strategy employing leverage. Times changed. Rates are now much higher. The days of easy money are over and private equity is overexposed to software companies, which has become extremely problematic. It's going to be years for private equity to work through these problems. We shall see how they do. And by the way, some software deals have already failed. So, for example, in 2024, Blue Owl and other lenders took control of a company called Pluralsight, wiping out Vista's entire equity SIG. This August, Medallia, a Thomas. Bravo deal was taken over by lenders as equity was wiped out as well. It's important to point out that neither of these companies was in cybersecurity. Plural site is an online learning and workforce development platform, and Medallia is a software platform that helps businesses improve customer and employee experiences. I actually looked up what Medallia does, and that's what I found. And frankly, I really have no idea what that means. But it sounds like something AI can do a lot cheaper. And now for the mailbag. First up from Jeremy E. Steve, first of all, a longtime fan, I was one of the random Twitter accounts begging you to start a podcast back in 2023, 2024. Glad you listened and or independently arrived at the logical conclusion to do so. Thank you for your advice. Secondly, the 10-year treasury and inflation question. CPI X energy was 2.4% for August, which was good. Seems most movement, if not all, is from energy, maybe tariffs to a far lesser extent. In any event, how does the Fed raising rates address the core issue? Is this yet another example of the Fed suffering from a man with Hammer syndrome? Hiking seems especially blunt against this context. No? Interesting question. The Fed has zero control over oil prices. So when oil prices are rising and that creates fears of inflation, the Fed raising rates has no impact on oil prices. Now, one could argue, as this viewer does, that the Fed is acting as if everything is a nail. Where it gets more complicated is when, like now, oil prices remain elevated for an extended period of time. Then, higher oil prices start to creep into the overall economy and into overall prices. For example, higher oil prices raise the price of fertilizer, which in turn raises the price of food. There are myriad other examples of this. That's when, like now, the Fed feels compelled to act in hopes that by raising short-term rates, it will slow the economy and bring inflation down. Admittedly, it's a very blunt instrument, but the best we got. Next up from Teddy. Quote: Regarding your comments on CNBC and in the sub stack about AI leaders wanting to create a duopoly through regulation, don't you think that two things can be true at the same time? They do want that, and some type of industry guardrails are needed to protect AI from getting out of hand. How can this be achieved without the Frontier Labs having ulterior motives? Fair question. It's very tempting to say that AI leaders want regulation to foster a duopoly, but also want regulation to create guardrails. I wish that were so. I just, I just don't believe it. The problems that have arisen so far are largely hacking, and that is something that I believe can be dealt with by management and by better oversight, their own oversight. But the idea that AI is going to destroy the world, I think, is just a subterfuge. I really believe they don't believe it at all, even though they have been saying it for a long time. Next up from Wayne. I worked for European Bank during the GFC, and I would skip to the train every day thinking we had zero exposure to the U.S. real estate because I knew it was going to blow, just to know how. I got a lot of calls from old workmates to make sense of mortgage-backed securities because I did that early in my career. And the first deal I was in, we made the bank keep the garbage loan so the MBS could get a good rating and that someday, 15 years later, the banks with the garbage would blow. My Clean European Bank put capital with the investment bank Yahoo's to get a return. They loaded up on mortgage-backed securities. Either way, the head Yahoo that got the bank suck left or was fired and started a hedge fund in Connecticut. There is no justice. I always wondered how you spotted the bad mortgage-backed securities. Was it in the notes of the financials? Two parts to the answer to this question. First, in the 1990s, I was a sell-site analyst covering the financial services sector. Part of my coverage was the first generation of subprime mortgage companies. For reasons we don't need to go into, most of the industry went bankrupt in 1998. The second generation of subprime mortgage companies went public in 2002 and 2003. Funny thing, most of them were run by the same people. They just changed the names. I had seen this play in 1998, and it ended in tragedy. So I was waiting for the same script to unfold. I just did not know when, but felt it was inevitable. Second, I knew through our research process that underwriting standards had deteriorated dramatically. But what gave me the confidence to do something about it was Moody's securitization database. For a fee, anyone could get this data. For every month, every month, every securitization reported all of its credit metrics. We could compare the trajectory of delinquencies of recent securitizations against older ones. What we found was that by the summer of 2006, loans securitized in 2006 were going delinquent and in Incredibly rapid pace, far quicker than older securitizations. That's when I knew. This week, on September 30th, on our premium subscription service, we interviewed Edwin Dorsey, creator of the Bear Cave newsletter.

If you're Microsoft or Meta, Google, well, not Meta, but Google, and you're building a data center, that's a long-term project. You're not raising overnight debt. You're raising long-term debt. You're raising 10-year debt, something like that. So the Treasury has a $6 trillion deficit this year. Now I add on top of that $500 billion in AI debt. And there's a bit of a competition, which didn't exist before. So I think one of the reasons why the 10-year is as high as it is is because, like I said, there's a crowding out effect where AI debt is crowding out Treasury, which is a crazy statement because nobody ever crowded out Treasury before, but now they are. And that's why Scott Besson tried to pull a rabbit out of the hat by trying to buy $6 billion. Dollars worth of long-term treasuries in something called Operation Twist, where he would issue short-term debt to buy long-term treasuries. And it's failed miserably. So the Treasury Department has lost a lot of credibility because when he started, I think the tenure was like at 4.75 and now it's at 5.21.

Ed has become a real expert in prediction markets, and we spend a great deal of time discussing the value of prediction markets and their problems. Also, we get occasional requests from subscribers to purchase our coffee mugs, pillows, and other merchandise. Go to steveeisman.com, which will bring you to Substack and our merchandise page. On this week's wrap, we will discuss the war in Iran, oil prices, and interest rates, troubling news from Oracle and Blue Owl Capital, Paramount cleared for takeoff, softbank junk borrowing to fund an equity investment, some aspects of AI circular financing that I find particularly disturbing. Meta's beignet, Meta's Muse, now that Agentic AI is cute, is it deflationary and will it make a real difference? And one mailbag about shorting against the box. And now let's get started. I've said this before. This is a tough market, and that the only two variables that matter right now are oil prices and the 10-year yield, neither of which, in my view, is predictable and both of which move the market. War news flips like a pancake, and the 10-year keeps climbing. Earlier in the week, the 10-year declined below 5% for a short period of time, and the market rallied. But on Wednesday, there was some very strong economic data, and the 10-year has now climbed to a multi-year high of 5.10% plus, and the market declined. But later on Thursday, there were reports of a potential deal regarding the Strait of Hormuz, and the market rallied back to flat. However, for now, it looks like 5% is the Rubicon for the market. Should the 10-year remain above 5%, a correction I think is probably imminent. Moving on. There was some troubling news from Oracle and Blue Owl Capital. Blue Owl is developing an AI data center called Project Jupiter for Oracle. Oracle sent a notice citing force majeure, a legal term, to Blue Owl. The notice is an attempt to put off payments. the data center gets derailed and fails to come online on time. Now, the debt tied to Project Jupiter was already trading at stress levels at around 90 cents on the dollar. I'm sure it's under more pressure today, Thursday. Although this shows that Oracle is becoming increasingly nervous about its lack of cash flow, its high debt levels, and its weak credit ratings. This week, Paramount settled the antitrust lawsuit that was brought by 11 state attorneys general. I've always thought this lawsuit was ridiculous. Netflix has a 300 billion market cap. Disney is at 180 billion. And the combined Paramount Warner Brothers company has a market cap of something like 88 billion. In the streaming wars, size matters. So the merger actually improves competition. This case was never about antitrust. The California State Attorney General wanted to place controls on the company. That's why the settlement involves a board overseeing CNN and a requirement that Paramount release 30 films in theaters each year. None of that has anything to do with antitrust. Frankly, I don't find any of these entertainment companies to be good investments. Netflix has hit a growth wall because the streaming wars have stopped being about expanding the pie and become only a food fight for market share. The only way to increase revenue seems to be price increases, but prices are now so high that raising them more risks angering the customer base. Moving on. In AI news, SoftBank really went out on a limb. Masayoshi-san, the chairman of SoftBank, has a history, a very long history, of loving leverage. During the summer, he got a $10 billion bridge loan from a group of banks to fund a $10 billion investment in chat GPT, OpenAI. This week, he is in the process of raising $11 billion in junk bonds to fund that $10 billion investment in OpenAI. The bonds will cancel the bridge loan. This is all really out there on the frontier of risk. It's one thing to invest your own cash. It's another thing to raise junk-rated debt to make an equity investment. Moving on to AI circular financing and why looking from the top down, which is conceptually accessible, needs to be paired with bottom-up analysis of the accounting and specifically the underlying off-balance sheet financing that we're about to talk about. This kind of financing is designed to obscure. Now, many commentators have discussed the circular nature of AI financing. NVIDIA, for example, invests in a company or lends money to a company that in turn uses that cash to buy NVIDIA chips. Now, you may not like it, but it's all out in the open. Easy to understand, NVIDIA is funding their own growth out of their own magnificent cash flow. This is not what I want to address today. What I find disturbing is some of the off-balance sheet financing using special purpose vehicles that I thought had gone the way of the dodo bird after the Enron and post-GFC debacles. I was wrong. It's still happening and in size, and boy, do I feel foolish. Let me take you back and remember Enron. To reduce its balance sheet debt levels, Enron created special purpose vehicles that were owned, get this, by the company's CFO, Andrew Fasto. That's right. Enron transferred billions in debt to an entity owned and controlled by the company's own CFO. You can't make this up. Prior to the scandal breaking, I remember reading Enron's 10Q, where all this was completely disclosed and wondering how the auditors could allow this nonsense. How could the auditors agree that the debt transferred to the special purpose vehicle owned by FASTO meant that said debt was no longer on Enron's balance sheet? I also remember talking to a sales site analyst about it. He shrugged his shoulders and said that it must be okay. After all, it's disclosed and the auditors signed off on it, which just goes to show how gullible any of us can be. After the scandal broke, the U.S. government sued the auditor, Arthur Anderson, for criminal malfeasance. The lawsuit put Arthur Anderson out of business. Fasto went to prison and his illegally gotten gains were wiped out. Bad off-balance sheet stories come in many flavors. Enron is an example of off-balance sheet vehicles called SPVs, which should never have been off-balance sheet. Arthur Anderson was paid a lot of money both for its audit and its consulting, and they sold their souls. Why go through these complex manipulations? The goal. Is to get debt off the balance sheet so the ratings agencies won't penalize the debt ratings. If the debt is gone, the company is less levered, at least technically. Ratings agencies can be myopic, and in rating the company, often only look at the debt that is on balance sheet. It's like the off-balance sheet debt has disappeared. Poof magic, this works until everything goes sideways. When the debt goes bad, like it did in the GFC and Enron, that's when it lands back on the balance sheet and everyone involved theoretically gets held accountable, like Fasto and Arthur Anderson. It's always good to go back in history and recall relevant stories now that off-balance sheet techniques are back with a vengeance in the new world of AI. Before diving in, I'll explain some of the off-balance sheet manipulations during the GFC. The great financial crisis saw a different but similar kind of off-balance sheet tale. Wall Street created a structure called a CIV, SIV, a structured investment vehicle. CIVs were a type of shadow bank, pools of investment assets kept off the balance sheets of major commercial banks to bypass leverage calculations and regulatory capital requirements. Civs played a major role in amplifying the 2007-2008 financial crisis. The business model of a CIV relied entirely on capturing the spread, the difference between long-term investment yields and short-term borrowing costs. The borrowing, short-term debt. Civs raised cash by issuing to institutional investors low-interest, short-term debt instruments like asset-backed commercial paper. This short-term debt had to be constantly rolled over, meaning reissued, as it matured long-term. Banks used that cash from the short-term borrowings to buy higher-yielding, long-term, structured financial products such as mortgage-backed securities and collateralized debt obligations, many of which, you guessed it, were tied to U.S. subprime mortgages. When the subprime mortgage bubble burst, the value of those long-term assets plummeted. Suddenly, panicked investors refused to buy the Civ's short-term commercial paper. This is classic surfing the curve, buying long-term bonds, funding them with short-term paper, and capturing the spread. Now, I never quite understood how a bank could use a CIV to own assets off-balance sheet, but the ratings agencies allowed it. Shame on them. Still makes no sense to me to this day. In the end, however, once the buyers of the Civ short-term paper refused to roll it over, the banks were forced to finance the long-term assets themselves, and the entire Civ came back on balance sheet. Poof magic again. It was quite embarrassing. The supposedly smartest people in the room no longer look so intelligent. But unlike FASTO, no one went to jail for this fraud. The GFC gave birth to the concept of too big to fail. Are we there again? I certainly hope not.

when they reported that I still to this day can't understand why no one has commented on this other than a few sub stackers, which is NVIDIA's revenue in the quarter that was just recently reported was up over 100%. But if you go in the queue, it says that 70% of accounts receivable is from five accounts. I mean, think about it this way, that the largest company on planet Earth, basically 70% of its revenue comes from five. Customers is very frightening.

let's look at a different prediction AI leaders used to make. They argue that AI would cause massive layoffs throughout the economy. That's not Armageddon either, but it would be terrible. It just looks like it's not true at all. Now, perhaps there is less hiring in tech because of AI. Perhaps. But there have been no mass layoffs, and AI leaders are not even talking about it anymore. So the AI industry has a tendency to make bold and terrible predictions that just don't come true. Laypeople, reporters, journalists, politicians, and others with decision-making power swallow each rumor whole without a shred of disbelief because it's all so new. They have no scaffolding within which to understand what's actually happening. And here we are with monstrous predictions being given total credibility. So what's really going on? I think this is a subterfuge, meaning a trick. Here's my theory. First of all, the claim that Dario Amode and Sam Altman are making about slowing down is just false on its face. The entire future, the entire future of Anthropic and OpenAI depends on there not being any slowdown. They both have hundreds of billions of commitments to hyperscalers. OpenAI alone represents 300 billion of Oracle's 600 billion plus backlog. If they were to slow down, they could not fulfill those commitments. Also, Anthropic is going to go public this year. Not just this year, maybe in a month or two. It can't slow down. What would it say to investors on the roadshow? Are they going to say that growth has been great, but now it's going to slow to a crawl? A slowdown contradicts their entire growth narrative. The only way to fulfill all those commitments is to not slow down. Elon Musk actually put it quite well. He said, quote, it's certainly some crazy 4D chess to say that there's whatever a 10% chance of annihilating. Humanity. But by the way, how much allocation would you like in our IPO? I love that quote. So again, what's really going on? Business is potentially slowing, or at least getting more difficult. And like a side street hustler moving the shells in a shell game, Amode and Altman don't want anyone to know that business is slowing or getting more difficult while costs are rising and capital is scarcer. They prefer scaring everyone into creating some kind of regulation that will protect their pricing power in the U.S. They are moving the shells fast and furious. Business is potentially slowing because token maxing is ending and open weight models keep taking market share. There are no pricing moats in this business. Today, I have the best LLM and tomorrow yours is better and cheaper. At the same time that business is slowing, the cost of building data centers and the cost of capital are increasing. Anthropic and open AI need moats to defend their businesses because they see, I believe, a price war coming. So they are fomenting hysteria, thereby inviting the federal government to regulate the industry. I think they believe that via that regulation, they can create moats that will foster an AI duopoly. That's what's really going on. It's ugly. Lying has unintended consequences. It's one thing to lie to investors. That's old school Wall Street fraud. Lying to politicians can disrupt the entire economy. AI won't cause extinction, but these two CEOs are creating massive damage. At this point, this gambit looks like it's a complete failure. President Trump stated this week that he has no interest in regulating AI. So I bet the entire doomsday slowdown narrative is going to disappear very quickly. However, the damage is done. Data center regulation is the centerpiece of many local elections. And now people actually fear and believe that the product being created in the data center next door is actually going to kill them. Going public into the distortions they keep creating will be very illuminating. It's unclear to me how to keep a shell game going in an IPO process that requires transparency. And now for the mailbag. The first question relates to our interview with Ryan Tunis, the PNC property and casualty insurance analyst at Cantor. The question is from Axam, who asks: quote, maybe I missed it, but I don't think the investment income side came up. PNC insurers invest afloat, and with rates higher, the bond book keeps rolling into better yields. If rates hold or rise, that's a structural earnings tailwind, independent of the underwriting cycle. Is the market still underpricing it or caught up? Unquote. Let me give some background to this question. Property and casualty insurance companies collect premiums from their customers. They then take that money, called the float, and invest it almost solely in bonds. So if interest rates go up, PNC companies will gradually make more money on this float. Thus far, the viewer is correct. What I think he might be missing is that generally investors in PNC stocks don't care all that much when PNC companies make more money on the float. They care the most about pricing. Is pricing going up or down? That's why if rates are rising, but pricing is declining, PNC stocks will generally come under pressure. The next question is from Atmike, who asks, quote, Steve, how do individual investors diversify away from AI exposure without abandoning equities? And are there any ways to hedge the risk or even place cheap bets against AI? Now, one of the difficulties with this market is that so much of it is correlated to AI. It's not just tech. Utilities are correlated because they provide the power to AI data centers. Industrials are correlated because they build the AI data centers and the power for the centers as well. Large banks, investment banks, and alternative asset managers are correlated as they provide the financing for AI data centers. There are 11 sectors of the SP 500. And as far as I can tell, the only sectors that are uncorrelated are Staples and Healthcare, which combined are only 14% of the SP 500. So buying a healthcare or staples ETF would help diversification. Thankfully, almost every sector of the SP 500 has a subsector that is uncorrelated. So for example, within financials, the property and casualty sub-sector is uncorrelated to AI into the economy. And there are ETFs that have low volatility stocks that would provide some diversification. Here are three. One, the LVHD, which is the Franklin U.S. U.S. Low Volatility High Dividend Index ETF, SPLV, Invesco S ⁇ P 500 Low Volatility ETF, and finally the KBWP, the Invesco KBW Property and Casualty Insurance ETF. One more suggestion. Let's say you have a portfolio that has a lot of tech with a large amount of unrealized gains because the stocks have gone up so much. How can you hedge that risk? Now, one way to do it is to short all or part of your positions against the box. What does that mean? Let's say I own 100 shares of NVIDIA and I've owned it for years. So I have massive embedded gains. Let's also say I'm nervous. I want to reduce my risk. My problem is that if I sell my NVIDIA position, I will pay 20% in federal long-term capital gains taxes plus the long-term capital gains tax rate of my state. Instead, I could short 50 or 100 shares of NVIDIA. That's called shorting against the box. By doing so, I have now reduced my NVIDIA risk, and I can wait until I think things will get better, and then I can undo the short. By the way, for those of you who have never shorted stocks, you need to make your account into a margin account. It's not complicated. You just have to sign a document. Finally, we got a bunch of comments about my premium substack interview with Wolfgang Munschau, author of Kaput, The End of the German Economic Miracle. Wolfgang is largely negative on the growth prospects of Europe, but one viewer challenged that view, or at least partially challenged it. So At Simon wrote, this is heavily focused on the big four, UK, Germany, France, and Italy. I am a bit disappointed not to hear of countries like my own, Sweden, with a debt to GDP ratio of 36%, that's quite low, and probably the healthiest capital markets in the EU. Making claims that entire Europe has not grown for a decade is not true across the continent. Poland, for example, was not mentioned once, and I wonder why that is considering the growth they've had in recent times. Overall, I agree with the complexities around Brussels versus capitals and overregulation, but also felt it was a bit incomplete. Nice to listen, though, and I will give Kaput a read. I saw the book. This is a fair criticism. It is definitely true that parts of Europe have done better than others. However, it is also true that the UK, Germany, France, and Italy dominate the European economy, and there, growth remains very weak. These European countries have small tech sectors, and since tech is what drives growth, European growth is generally slow. On the positive side, Europe has a richer welfare system, but it is overregulated, in my view, and seems almost allergic to risk. Wolfgang was not optimistic that things will get better. This last week, on Monday, September 14, we dropped an interview I did with my friends and big short partners, Vincent Daniel and Porter Collins. We had a wide-ranging conversation about Scott Besson, the deficit, gold as an investment, and why it is more difficult to short stocks these days. So check it out. This coming Monday, September 22nd, we will drop an interview with George Noble, former PM at Fidelity and now a Substack newsletter writer and podcaster. We discuss interest rates, gold, Tesla, and SpaceX, the precarious nature of the AI revolution, and what would cause this market to unwind. So please tune in. The best way to support the Real Eyes and Playbook is to subscribe to Substack through steveisman.com. Subscriptions are free, and we appreciate your support. And that's the wrap. This podcast is for informational purposes only and does not constitute investment advice. The host and guests may hold positions and stocks discuss, opinions expressed on their own and not recommendations. Please do your own due diligence and consult a licensed financial advisor before making any investment decisions.

let's look at a different prediction AI leaders used to make. They argue that AI would cause massive layoffs throughout the economy. That's not Armageddon either, but it would be terrible. It just looks like it's not true at all. Now, perhaps there is less hiring in tech because of AI. Perhaps. But there have been no mass layoffs, and AI leaders are not even talking about it anymore. So the AI industry has a tendency to make bold and terrible predictions that just don't come true. Laypeople, reporters, journalists, politicians, and others with decision-making power swallow each rumor whole without a shred of disbelief because it's all so new. They have no scaffolding within which to understand what's actually happening. And here we are with monstrous predictions being given total credibility. So what's really going on? I think this is a subterfuge, meaning a trick. Here's my theory. First of all, the claim that Dario Amode and Sam Altman are making about slowing down is just false on its face. The entire future, the entire future of Anthropic and OpenAI depends on there not being any slowdown. They both have hundreds of billions of commitments to hyperscalers. OpenAI alone represents 300 billion of Oracle's 600 billion plus backlog. If they were to slow down, they could not fulfill those commitments. Also, Anthropic is going to go public this year. Not just this year, maybe in a month or two. It can't slow down. What would it say to investors on the roadshow? Are they going to say that growth has been great, but now it's going to slow to a crawl? A slowdown contradicts their entire growth narrative. The only way to fulfill all those commitments is to not slow down. Elon Musk actually put it quite well. He said, quote, it's certainly some crazy 4D chess to say that there's whatever a 10% chance of annihilating. Humanity. But by the way, how much allocation would you like in our IPO? I love that quote. So again, what's really going on? Business is potentially slowing, or at least getting more difficult. And like a side street hustler moving the shells in a shell game, Amode and Altman don't want anyone to know that business is slowing or getting more difficult while costs are rising and capital is scarcer. They prefer scaring everyone into creating some kind of regulation that will protect their pricing power in the U.S. They are moving the shells fast and furious. Business is potentially slowing because token maxing is ending and open weight models keep taking market share. There are no pricing moats in this business. Today, I have the best LLM and tomorrow yours is better and cheaper. At the same time that business is slowing, the cost of building data centers and the cost of capital are increasing. Anthropic and open AI need moats to defend their businesses because they see, I believe, a price war coming. So they are fomenting hysteria, thereby inviting the federal government to regulate the industry. I think they believe that via that regulation, they can create moats that will foster an AI duopoly. That's what's really going on. It's ugly. Lying has unintended consequences. It's one thing to lie to investors. That's old school Wall Street fraud. Lying to politicians can disrupt the entire economy. AI won't cause extinction, but these two CEOs are creating massive damage. At this point, this gambit looks like it's a complete failure. President Trump stated this week that he has no interest in regulating AI. So I bet the entire doomsday slowdown narrative is going to disappear very quickly. However, the damage is done. Data center regulation is the centerpiece of many local elections. And now people actually fear and believe that the product being created in the data center next door is actually going to kill them. Going public into the distortions they keep creating will be very illuminating. It's unclear to me how to keep a shell game going in an IPO process that requires transparency. And now for the mailbag. The first question relates to our interview with Ryan Tunis, the PNC property and casualty insurance analyst at Cantor. The question is from Axam, who asks: quote, maybe I missed it, but I don't think the investment income side came up. PNC insurers invest afloat, and with rates higher, the bond book keeps rolling into better yields. If rates hold or rise, that's a structural earnings tailwind, independent of the underwriting cycle. Is the market still underpricing it or caught up? Unquote. Let me give some background to this question. Property and casualty insurance companies collect premiums from their customers. They then take that money, called the float, and invest it almost solely in bonds. So if interest rates go up, PNC companies will gradually make more money on this float. Thus far, the viewer is correct. What I think he might be missing is that generally investors in PNC stocks don't care all that much when PNC companies make more money on the float. They care the most about pricing. Is pricing going up or down? That's why if rates are rising, but pricing is declining, PNC stocks will generally come under pressure. The next question is from Atmike, who asks, quote, Steve, how do individual investors diversify away from AI exposure without abandoning equities? And are there any ways to hedge the risk or even place cheap bets against AI? Now, one of the difficulties with this market is that so much of it is correlated to AI. It's not just tech. Utilities are correlated because they provide the power to AI data centers. Industrials are correlated because they build the AI data centers and the power for the centers as well. Large banks, investment banks, and alternative asset managers are correlated as they provide the financing for AI data centers. There are 11 sectors of the SP 500. And as far as I can tell, the only sectors that are uncorrelated are Staples and Healthcare, which combined are only 14% of the SP 500. So buying a healthcare or staples ETF would help diversification. Thankfully, almost every sector of the SP 500 has a subsector that is uncorrelated. So for example, within financials, the property and casualty sub-sector is uncorrelated to AI into the economy. And there are ETFs that have low volatility stocks that would provide some diversification. Here are three. One, the LVHD, which is the Franklin U.S. U.S. Low Volatility High Dividend Index ETF, SPLV, Invesco S ⁇ P 500 Low Volatility ETF, and finally the KBWP, the Invesco KBW Property and Casualty Insurance ETF. One more suggestion. Let's say you have a portfolio that has a lot of tech with a large amount of unrealized gains because the stocks have gone up so much. How can you hedge that risk? Now, one way to do it is to short all or part of your positions against the box. What does that mean? Let's say I own 100 shares of NVIDIA and I've owned it for years. So I have massive embedded gains. Let's also say I'm nervous. I want to reduce my risk. My problem is that if I sell my NVIDIA position, I will pay 20% in federal long-term capital gains taxes plus the long-term capital gains tax rate of my state. Instead, I could short 50 or 100 shares of NVIDIA. That's called shorting against the box. By doing so, I have now reduced my NVIDIA risk, and I can wait until I think things will get better, and then I can undo the short. By the way, for those of you who have never shorted stocks, you need to make your account into a margin account. It's not complicated. You just have to sign a document. Finally, we got a bunch of comments about my premium substack interview with Wolfgang Munschau, author of Kaput, The End of the German Economic Miracle. Wolfgang is largely negative on the growth prospects of Europe, but one viewer challenged that view, or at least partially challenged it. So At Simon wrote, this is heavily focused on the big four, UK, Germany, France, and Italy. I am a bit disappointed not to hear of countries like my own, Sweden, with a debt to GDP ratio of 36%, that's quite low, and probably the healthiest capital markets in the EU. Making claims that entire Europe has not grown for a decade is not true across the continent. Poland, for example, was not mentioned once, and I wonder why that is considering the growth they've had in recent times. Overall, I agree with the complexities around Brussels versus capitals and overregulation, but also felt it was a bit incomplete. Nice to listen, though, and I will give Kaput a read. I saw the book. This is a fair criticism. It is definitely true that parts of Europe have done better than others. However, it is also true that the UK, Germany, France, and Italy dominate the European economy, and there, growth remains very weak. These European countries have small tech sectors, and since tech is what drives growth, European growth is generally slow. On the positive side, Europe has a richer welfare system, but it is overregulated, in my view, and seems almost allergic to risk. Wolfgang was not optimistic that things will get better. This last week, on Monday, September 14, we dropped an interview I did with my friends and big short partners, Vincent Daniel and Porter Collins. We had a wide-ranging conversation about Scott Besson, the deficit, gold as an investment, and why it is more difficult to short stocks these days. So check it out. This coming Monday, September 22nd, we will drop an interview with George Noble, former PM at Fidelity and now a Substack newsletter writer and podcaster. We discuss interest rates, gold, Tesla, and SpaceX, the precarious nature of the AI revolution, and what would cause this market to unwind. So please tune in. The best way to support the Real Eyes and Playbook is to subscribe to Substack through steveisman.com. Subscriptions are free, and we appreciate your support. And that's the wrap. This podcast is for informational purposes only and does not constitute investment advice. The host and guests may hold positions and stocks discuss, opinions expressed on their own and not recommendations. Please do your own due diligence and consult a licensed financial advisor before making any investment decisions.

or pay them. Right. We had an incredible conversation and covered every aspect of the AI story, its strengths, weaknesses, the financial shenanigans, and where and how it could fail. So please tune in. In this weekly wrap, I will discuss one, war, oil, and interest rate news. Two, the increasingly insatiable need for commentators to predict a dystopian ending to the AI bubble. Three, why I'm not there yet. Four, AI's insatiable need for capex is the major driver of GDP growth. Five, is open AI the potential catalyst for a recession? And six, some recent events. So let's get started. This week, the war heated back up and all prices spiked, thereby driving fears of mounting inflation. The 10-year yield almost brushed against 4.8%. And there is no question that at some level of interest rates, the market will correct. What that level is, no one really knows. I thought that 4.5% was the Rubicon, and I was wrong. Is it 4.8% or even higher? I don't really know, but we are certainly getting closer. I'd also point out that part of the problem is the enormous amount of AI debt being issued. This supply is putting pressure on rates. And I'd also point out that Treasury Secretary Besson's recent attempt to reduce long-term rates appears to have already failed. One more point: higher rates put pressure on any company issuing debt. However, higher rates are good for consumers who save and put their money in the bank or in a money market fund. Before we get to some of this week's events, I am going to spend some time discussing the desire by many commentators to call for the end of the world and what it would actually take to cause a recession. Now, one of the last great sitcoms on network Work TV was the Big Bang Theory. For those of you who never watched the show, it was about a group of friends who were science nerds. In the show, there was a minor character called Stuart. Stewart owned a comic book store that the main characters liked to hang out at as they all loved comic books. Now, recently, Chuck Lorry, the creator of the Big Bang Theory, created a new show on HBO called Stuart Fails to Save the Universe. In this new show, a catastrophe has enveloped the world, and humanity now lives in a horrible dystopia. Stewart finds a machine that allows him and his friends to travel to alternative universes, constantly searching for a better place to live. Thus far on the show, every universe Stewart jumps to is just another form of dystopia. Show is quite funny, but captures something essential. There is something about imagining a dystopian future that people find captivating. I've noticed that over the years, since the GFC, whenever I am interviewed, the interviewer is almost begging me to predict the end of the world. How do I feel about that? Well, I predicted the end of the world once, and believe me, it was no fun. I am in no rush to predict to the end of the world again unless I am really convinced that it's going to happen. But I'm not going to make such a prediction just because it will get a lot of press. There is no question in my mind that the entire U.S. economy hinges on the success of AI. The amount being spent is just so large that were it to stop, the economy would go into a recession almost immediately. There are several commentators who are making exactly that prediction. They could be right, but right now the data does not support the end of the AI story, at least not yet. So let me point out where the points of weakness reside, because it's time to consider how the unwind could, could occur, what the catalysts are, and what the impact could be. Now, there is no question that the large tech companies have transformed the dynamics of their businesses. They used to run businesses that threw off so much cash, they didn't know what to do with it. But because of the hundreds of billions they are each spending on AI CapEx, their cash flow has evaporated. And in some cases, they are raising both debt and equity to fund that capex. This is a topic of great concern. It gets a lot of press, but it is not necessarily the death knell of the AI story. After all, if, and it's a big if, AI generates new businesses and large returns on this capex investment, then the hyperscalers will be okay. Now, there is also a great deal of circular financing going on, mostly led by NVIDIA. And these transactions are very complicated and convoluted. But essentially, NVIDIA is either lending money or making equity investments. The companies then take that cash and buy more NVIDIA chips. Now, I'm oversimplifying, but that is essentially what is going on. NVIDIA's potential investment in Perplexity and the recent investment in poolside are only the most recent examples of such transactions. This financing circularity is also a topic of great concern. However, here too, if AI succeeds, then these financing techniques will probably turn out to be okay. So, where is the Achilles heel? I think that it resides with Anthropic and OpenAI because they are so central to the entire AI food chain. According to reports from various Wall Street firms, something like 70% of hyperscaler AI revenue comes from Anthropic and OpenAI. 70% equates to around 25 to 35% of total hyperscaler cloud revenue. I can't confirm those statistics, but they sound right, given what we actually know about Oracle. Oracle has a backlog of around $600 billion, and we know, because the information is public, that around 50%, 50% of that backlog is from OpenAI alone. And by the way, OpenAI accounts for over 20% of CoreWeave's total contracted order book. Hyperscalers are spending about $700 billion in CapEx this year and even more next year. And that spend accounts for around half of the 2% GDP growth projected for 2026. So one must conclude that the health of the US economy is extremely dependent on hyperscaler CapEx, and hyperscaler CapEx is highly dependent on the health of Anthropic and OpenAI. That's the chain. Now, between the two, OpenAI is the weaker entity and has lost several key senior employees. Let's list them because the list is kind of illuminating. One, Denise Dresser, chief revenue officer, resigned in August 2026 after only eight months on the job. That probably means she gave up all her stock options as those usually don't vest for at least a year. That's a big data point. For her, I guess, moving on was perhaps worth the lost options, and maybe she never thought they would vest because going public is being pushed further out on the timeline. Two, Brad Lightcap, former COO, resigned in August 2026 after eight years with the company. Three, Caitlin Kalinowski, head of robotics and consumer hardware, resigned in March 2026 over governance concerns regarding an agreement with the Pentagon. Four, Kate Rauch, Chief Marketing Officer, and Kevin Wilde, science research overseer, both departed in April 2026. The most concerning departures are the ones that took place this August, Denise Dresser and Brad Lightcap. Supposedly, OpenAI is getting closer to an IPO. That's the big payday for employees because it means that eventually they can sell some of their shares. That two such senior employees would leave now is an important data point. More importantly, OpenAI's financials look tepid. The Wall Street Journal reported that OpenAI's June quarter revenue reached $6.7 billion, up only 18% versus the March quarter. Compare that to Anthropic's revenue of $11 billion plus in the June quarter, which was up over 100% versus the March quarter. What's even worse than the relatively weak revenue growth is the explosion in costs. Again, according to the Wall Street Journal, OpenAI's costs reached $12.3 billion, up $3 billion versus the March quarter. So in three months, revenue increased $1 billion, but costs surged $3 billion. Things are not moving in the right direction. Now, you might ask, what is the problem at OpenAI? Part of the problem, I think, is that OpenAI made a strategic mistake. It chose to chase the business-to-consumer market, B2C. It did so by giving its product away or underpricing it with the hope that later on it could raise prices. That strategy proved to be flawed, and now OpenAI is chasing the business-to-business market, B2B. The problem is that it's late and playing catch-up with Anthropic, which chose the B2B market at the outset. Again, I think that OpenAI is potentially in trouble, and this bears continuous monitoring.