When you get as many things going wrong as there are right now, you do not think about opportunity; you think about safety. I get that. We have the most uncertain of times in the Mideast after the U.S. and Israel attacked Iran this weekend. We have a man-made inflationary event — the closing of the world’s chief oil artery, the Strait of Hormuz. The skyrocketing of oil prices will make it so the Federal Reserve can’t help us, even under President Donald Trump ‘s soon-to-be Fed chairman, Kevin Warsh. We are no longer in the “canary in the coal mine” stage when it comes to private equity; we are in the recognition stage, recognition that things are going awry and losses, big losses, are coming. And, we have an assault on the world’s largest company, Nvidia, from its own customers. So, what’s the plan? Sell, sell, sell? Get out now? Moments like this are why I wrote my new book, “How to Make Money in Any Market.” The easiest thing in the world is to sell. Many of you right now might say that the only stocks that you can own are so-called safety stocks. If you look at the Club portfolio, you would be drawn to owning Procter & Gamble, Bristol Myers Squibb , and Eli Lilly — because those are the three that will hold up if we go into a recession. Oh, and believe me, having been in the news business long enough, the operative question is a by rote one: Ask all guests, call all sources, and ask when the recession will begin? Don’t even ask them if there will be one; just pinpoint a date. Look, that’s a fine scenario. It’s important to warn people of the dangers that can occur. It’s valuable. But it’s only one view. I like to be more constructive. I like to be more constructive because history says to be more constructive. Every time there has been a severe downturn, the stock market has come back, not always with the same stocks, but it has come back. I like to think about “what could go right” with the situations I mentioned just now and figure out ways to stay in the market, so I can take advantage of the chaos. You can pick individual stocks we recommend. You can buy index funds. But there is no hurry; we aren’t even oversold yet. Profit from the chaos First, you can’t profit from the chaos if you don’t have any money. You can’t be on margin — buying stocks with borrowed money — because we don’t know how long these pain points might last. All we know is you don’t want to be in a situation where the problems can last longer than you. Take some hits. Get off margin. Let this be your lesson. Second, there are two kinds of stocks I would sell. I would sell the oils because they are up artificially. The world is awash in oil, and there will be more produced by those who have extra to take advantage of the moment of supply disruption. It will be awkward, and it will feel terrible to sell Exxon Mobil here. I get that. But those are the stocks that are most inflated versus their fundamentals. Anyone who remembers the Gulf War in 1990 knows these are the ones that rollover first after the initial spike that gives you a chance to sell. Oil will come to the market. The quick spike is a lucky moment to say goodbye for those who had the foresight to own these stocks. Well played. I would also sell the consumer staples. I don’t want to take a quick gain in Procter & Gamble because we are not traders. However, at this moment, it is the most overvalued stock in the portfolio on 2026 fundamentals. These kinds of stocks are not going to get the multiple expansion they used to. There are more competitors, higher raw costs, and less heft versus the retailers than at any time in history. P & G does not tell Walmart what to do anymore, and it can’t boss around Walgreens or Rite Aid — because the former is going away and the latter already went. Sure, P & G gets a gain from the weak dollar. But that’s been the spur. I have no next spur other than fear. We — Jeff Marks, our Club director of portfolio analysis and I — will debate selling some Procter & Gamble as soon as we are allowed to sell per our Club trading rules. Third, the problems in finance are real but were exaggerated by the “Apocalypse Soon” memo that came out on Sunday, Feb. 22. It talked about the white-collar workforce being wiped-out because of AI. Monday, Feb. 23, was a brutal moment. The S & P 500 dropped more than 1% that session enroute to a lower week and capping a terrible February, the worst monthly decline for the index since March 2025. This shrewdly predictive tale of woe, entitled “The 2028 Global Intelligence Crisis,” is from an outfit we never heard of called Citrini Research. It was cogent. It scared people. I don’t know whether it was dystopian fun or a real jeremiad, maybe both. But suddenly, it crystallized the crisis that’s going to be predicated on superfast, self-written code, the likes of which Anthropic pumps out with the ease of a fourth grader. It foresees a domino effect: The white collar class will rapidly become jobless from the “Anthropics of the world” and start defaulting on credit, and stop spending altogether. We are a country that runs on white-collar employment, now superfluous under this vision, and a service economy, which would disappear in two years. Welcome to the depression of 2028, as the Citrini report protends. Cembalest to the rescue After the vibe coding H-bomb started ticking, we got a bit of a reprieve from cooler heads, like that of Michael Cembalest, the top JPMorgan strategist, and the best thinker on Wall Street. (“Vibe coding” is a term that has sprung up to describe software writing by AI models such as Anthropic’s Claude.) Cembalest shrewdly questioned the entire Citrini thesis, urging us to accept that humans are more than friction — a main portion of “The 2028 Global Intelligence Crisis.” The vibe coding bots — the all-powerful-economy busters — are not going to wipe out real estate salespeople or credit card companies or DoorDash, or a host of other day-to-day activities that were the ground zero of the memo. American Express will survive, under Cembalest’s view of the future. If anything, he made clear that AI is going to create more but different jobs, while others will disappear. There could be stagnation at some workplaces and growth at others. The beauty of the now infamous, or famous, Citrini memo, as it came to be called, is that it looked at all the amazing friction that’s out there and logically had the AI coders destroy it. The overarching mistake? We are humans, and humans don’t think of themselves as friction. Many times, it looks like the machines can do it all. But they can’t. For example, we don’t want the cheapest flight that can be found by bots; we want the best one for us. We don’t want to pay a hefty couple of percent to Mastercard, but the system needs Mastercard , and it needs its fraud protection, which is why blockchain hasn’t taken off. Too much fraud that can’t be undone under blockchain, the decentralized digital ledger technology that underpins bitcoin. Take that, American Express and Capital One shorters. Can AI wipe out the edifice that is real estate with its surfeit of salespeople? No. Real estate has already been brought down to the lowest common denominator — a website entry with a price. But the amazing thing about real estate is that it has to be sold, as always. We are not going to be able to have every industry be overrun by AI. DoorDash will turn out to be ingrained and not wiped out by the fly-by-night vibe coders because you need onboarding and customer service. Most importantly, for a day, at least, we realized that while white-collar jobs might be stagnating, the survivors are using AI to advance themselves and get more done while not needing to hire as many people. Put simply, the Citrini memo storyline is going to be dead wrong. Millions of jobs won’t be wiped out. But it has made us realize that we are not going to make as much money if we work in some service enterprises like Salesforce , or, perhaps, Visa . Those companies will still be with us. We will see more and more AI, but it can’t wipe out so many companies by 2028 that we will have a huge downturn. The memo demonstrates some possibilities for AI, but it can’t be the monster that’s depicted. Even if it were to be so, it will be a slow-moving monster. The memo demonstrated what could occur, however unlikely. It just had the wrong, super-fast timeframe. Dorsey says, not so fast It was all way too fast because we like what we have. Sure, there is friction, but it is meaningful friction that isn’t easily destroyed. Cembalest will ultimately be right, and sanity returned, for a bit. Then, when the Citrini memo was still warm, we — out of nowhere — got instant verification of it when the elusive Jack Dorsey decided that he was going to lay off 40% of Block staff because of AI. The stock on Friday soared 17% on the news. Shareholders have to hope that Dorsey, who also co-founded Twitter, can make this transformation happen. Revenue has been going up for this service company with a good lending arm. But the market capitalization of Block has been going down. It was $47 billion in 2023. Even after Friday’s rally, it’s now $38.7 billion. Maybe what can save them from irrelevance is a total overhaul, with software taking care of many jobs. It might happen. Maybe the place is bloated from Covid-era hiring. Maybe it’s bloated because it is poorly run. Or maybe it does have a lot of superfluous jobs in a new, AI-driven world. We don’t know. We don’t even care. We just saw the future, and it is Block. How about that 40% number? As someone who has run a business with 300 people, TheStreet.com, that needed to slim down to 200, I had to make this kind of wholesale job reduction, too. Dorsey pulled something that only a lunkhead would possibly do: He has created an amazing reason for all of the good people to jump ship. Who is going to stay at a company where you have a 40% chance of being laid off? You think that everyone in charge of a company where the market has not endorsed your model, despite its hefty EBITDA (earnings before interest, taxes, depreciation, and amortization) doesn’t want to fire 40% of the people that work there to get its market capitalization to soar? Only a bozo would announce that kind of layoff, though. You might as well pick the names out of a hat, good names and bad. Why? Because the exodus is so urgent, not from the people who are fired but from the people who may be fired. So, 40% is needlessly aspirational, and the announcement of it all is rather lunkheaded. I happened to have been in a firing war room as we fired the good with the bad. The Grim Reaper, however, anticipated our executions, and many good people left. Why, then, did Dorsey do it knowing that’s exactly what would happen? For precisely what did occur: a callous jump in valuation. It is entirely possible that he had hired too many people in his absentee CEO-dom, and maybe the company can afford a 40% reduction. Maybe it is so bloated that it should be a 50% reduction. Whatever. It was an instant verification of the Citrini memo. The software stocks, especially the software-as-a-service (SaaS) stocks, gave up the ghost. Who can afford to be in the stocks that offer expensive wares that can make your company more efficient, which is what these white-collar services do. The software companies that had been so valuable to your institution may be totally unnecessary, as the white-collar workers who run them. I think that we underestimate the survival of these software vendor companies. Many weren’t AI-oriented — but, like the incredibly run ServiceNow , have shifted to be AI-oriented. The most cogent case proved to be Salesforce, which happened to rally hard on its earnings. Salesforce has seen its market capitalization shrink well ahead of when many fear it will no longer be needed because of self-written code. When Salesforce reported its latest quarter , it got credit for its AgentForce platform, which develops virtual agents instantly for any company that needs them to automate processes with little to no human intervention. The product has had a remarkable uptake. But it can’t offset the high single-digit growth of what I call RemainCo, the non-AI, SaaS portion of Salesforce’s business. When Salesforce reported, a short squeeze ensued. The rally didn’t last long. And, by Friday, we were back to thinking that the destruction of the market capitalization of software stocks was a given, and we can’t pay up for these companies. Their stocks still had to be sold; it is just a matter of time before their revenue will slow. Sure, it will take time. But an outfit like Salesforce is just in denial. That means we will see continued market pressure on the software companies’ worth and profitability. No, it won’t happen at lightning speed, as Citrini lays it out. But it will happen. It could mean a rapid fall off of the non-AI portion of Salesforce. Not rapid enough to impact the sales of the company, just the multiple of the stock. Ultimately, I think AI kills call center jobs and low-end white-collar jobs over time. It might make some more efficient and others fodder. The market has spoken: I am too bullish. Things are going to get really ugly, and the market knows it. Dorsey will be right. Salesforce’s upbeat leader, Marc Benioff, with his notion that the SaaS-pocalypse won’t occur, is wrong. It is going to happen. Just you wait. So, the Citrini thesis made it to the end of the week, and it drowned out both Benioff and Cembalest’s counters. They died in Dorsey’s swimming pool. The private equity problem You have to marvel at the perfect storm of the Dorsey-Citrini negativity, dovetailed with the potential demise of the next pillar of tomfoolery, the private equity and private credit firms. These firms had been the envy of the old-line investment banking houses. They burst on the scene with incredibly profitable business models a couple of decades ago and just got stronger and stronger until a couple of weeks ago. Literally. That’s when they were revealed to be nothing more than houses of bad risk. In reality, these firms had been doing OK, not great, because the stock market wasn’t providing the liquidity it once did for their well-tuned portfolio companies to come public. They were suddenly “revealed” to be nothing but bad bankers with tranches of private credit lent to the firms that Citrini pinpointed. Of course, I am using Citrini as a shorthand. It is more appropriate that I use Blue Owl Capital , the principal credit culprit of the moment. Let’s go over Blue Owl, a five-year-old company known for private equity work, even though it’s a lot more than that. Over the last few years, this market has experienced incredible growth in enterprise software, and Blue Owl has helped fund it by providing financing to the companies that went private in the field. The two firms that have proselytized the recapitalization of enterprise software the most in this field are Thoma Bravo, with about $184 billion in assets, and Vista Equity Partners, with about $100 billion. Thoma Bravo and Vista have been bullish on enterprise software as a category, which has been relatively immune to the economy. It’s been a tremendous source of targets for ages. These two lucrative firms are the linchpin of the hysteria we see now because of AI. They have bought billions of dollars’ worth of public enterprise software companies that were doing well, but have not been rewarded by the stock market. They would therefore take them private, fix them, grow them, and then bring them public. Classic example: Anaplan, a Thoma Bravo company that went private in 2022 for $10.7 billion. Anaplan is a general consulting company with little to no AI that has now been loaded with AI by Thoma Bravo, and it is in the re-IPO (initial public offering) stage right now. It was meant to fit this current stock market perfectly. It is amazing how much better the company is than it was when it went private. If we didn’t have this morass, I thought an Anaplan IPO would be well-received as a kind of pure-play AI consulting company — kind of like SAP or an Accenture super-charged with intelligence. A genuine code-writer for the new world. It doesn’t matter right now, though. It is, in a word, wrong. The wrong time, the wrong place, even as it is AI. The stock market doesn’t seem crazy about it because no one trusts anything software right now. Thoma Bravo is a terrific company. When I look at most of its portfolio companies, I marvel at how well they are doing and how much better than they were before Thoma Bravo took them over. Of course, Thoma Bravo has some clunkers like any other private equity firm. But its record is pretty impeccable, and, most importantly, the companies it owns are much, much better when they were public. It is a beautiful portfolio with only a handful of companies that are in arrears. Some would argue that Vista, its competitor, is even better. Both have better records than almost any company in this take-private and then bring-public-again industry. They are the envy of this world and, until recently, regarded as the best of the entire sector. It doesn’t matter. They are the nucleus of the bear case. That’s because, in order to take companies private, Thoma and Vista have to take down debt. That debt has been syndicated to a host of private equity companies. Most of those private equity companies, like Blackstone , KKR , Carlyle Group , Apollo , and Ares , have really exceptional credit analysts, with whole groups of technology people who can examine and decide whether a slice of high-risk debt is worth taking on. These companies can wrap up a host of these high-risk loans and put together products that, when loaded up with debt, can offer attractive yields both to institutions and to individuals. As good as these wraps might be, though, they are the locus of fear. Wall Street is convinced that these indebted companies are going to fail. The fear could be justified if Thoma Bravo and Vista are weak in their work. But, as I just said, they aren’t. They are money good on most of their properties and are very shrewd when it comes to introducing AI into their organizations. This is a fearful market, though, and these loan packages are now viewed as being weak. The marketplace thinks Thoma Bravo and Vista didn’t do their job well. Again, that’s wrong. But that doesn’t seem to matter right now. To most of the smart people I know on Wall Street, these loan packages are just a bad agglomeration of private credit. That’s another despised portion of the market that used to be loved because it generated high yields at a time of Federal Reserve-mandated low yields. The thought, the incredibly pessimistic thought, is that those who bought the debt of too many high-risk enterprise software companies may not make it because of the amazing AI products out there now, even as the portfolio companies of the two biggest private equity firms are about AI. They embrace AI more than most public portfolio buyers. When I look at their portfolios, I see a host of companies that could come public if they had to, just not at much of a premium or perhaps a discount to where they went private. Let’s skip from the older, go-private companies — call them the Apollo-Blackstone faction, and venture into the riskiest private equity-private debt outfit, Blue Owl, a $300 billon manager that, out of nowhere, has become the riskiest company on Earth. The company embraced enterprise software like no other. It is thought to be a terrible buyer of loans, even as that’s not clear. To be blunt, rather than being the smartest guys in the room, they are thought to be knuckleheads and bought everything to do with heavily indebted loan packaging, both public and private. Their loan portfolio is said to, well, stink. They had two funds that had regular redemptions where institutions and individuals wanted out in far bigger numbers than the illiquid portfolios could allow. So, Blue Owl foolishly shut down that method of redemption. Instead, they tried to merge some of the ne’er-do-well firms. That failed. So, instead, again, they came up with this complicated idea to sell some of the loans at what they said were near par and then started giving the proceeds to the owners of this private credit rather than have traditional redemptions. They sold a gigantic $1.4 billion in loans. Now, here’s the real issue. Blue Owl made it clear that these loans were pretty regular and that it had a lot of good loans to choose from. It made it clear that most of the loans were sold to good arms-length outfits, and one was sold to an insurance company that Blue Owl managed. The management at Blue Owl might have thought this was smart. The marketplace thought it was ill-advised. Rather quickly, the bears came out and presumed that Blue Owl cherry-picked the best loans and sold most of them to the three arms-length investors and the rest to the captive insurance company. Perhaps Blue Owl didn’t realize the marketplace would view it like that. Maybe it had no choice. Maybe it was a terrific way to deal with the redemptions. At this moment, though, perception is everything, and Blue Owl was viewed as a suspect parvenue to the private-credit world and a total dingbat to emphasize private equity firms. So, within a week, again, the marketplace viewed Blue Owl, a big and recently revered firm, as the point of pain. If AI is destroying traditional software firms and there are many private equity firms taken private and many of the loans were syndicated to outfits like Blue Owl and the ones Blue Owl got were packaged and sold to institutions and individuals and a host of investors decided they didn’t want own highly levered tranches of these companies and they want to redeem them in traditional fashion and Blue Owl would no longer let them, then the private equity world was done. Stick a fork in it done. It didn’t matter if the whole daisy chain was wrong; it’s where we are. Oh, and in case you thought that we weren’t, many of these companies had what they call business development companies in their portfolio, which are considered to be risky, too risky for traditional banks because the bank examiners wouldn’t let them be taken down. Some of these business development companies, including those of Blue Owl, have these. They can be publicly traded. And almost all of them are being slaughtered. Does that make sense? Yes, because, again, they are filled with high-risk loans, packaged and brought public by these same firms. Who wants those? Right now, only rubes. They are the Achilles’ Heel of this process. Some of the companies in these business development packages are going bad. Some of these companies are trimming their distributions. All of them are under pressure. Making matters worse, they are being compared to the mortgage problem of 2007. You can’t put that negative to rest. That’s where we are now. A horrendous spot that looks like we are all going to be crushed under the weight of terrible loans, many of them made by Vista and Thoma Bravo — and many of them thought to go bad soon because of AI. Look, it doesn’t matter if that is a false narrative; it is the narrative, and it is why all of the banks are going down as they, too, are said to be infected by these kinds of loans, as well as all sorts of other loans of companies that may be going bad. Nvidia can’t catch a break Let’s put a hangman’s bow on these and go to the final issue: Nivida . Last week, Nvidia reported a terrific quarter . It happened to do so exactly when all of tech was being roiled by the Citrini memo and a sense that the users of Nvidia chips could be in trouble, even as they aren’t. Maybe they won’t be ordering as many. Not true. But we still had one more, one-two punch. Just when Nvidia reported, both Amazon and Alphabet launched a full-court press hyping their own chips. These chips are far less expensive than those of Nvidia, and they are having great success because they dwell on the inference side of these monster chip platforms, not just the training side. On Friday evening, Nvidia revealed a similar product. If it weren’t for Iran, the similar product would be noticed and Nvidia would be back in the driver’s seat. Nvidia is so big and so important to this market that when its stock fell from $196 to $176, it caused a wave of selling that wasn’t staunched. Perhaps, again it will be, maybe not. Bottom line Let’s bring everything together. A war in Iran will boost inflation. That may cause the Fed not to cut interest rates, even as the soon-to-be new Fed chief has said he will. At the same time, the private credit world is said to be teetering because it is based on syndicated and damaged private equity loans. Sophisticated investors know that stinks. Private equity firms are going to go down because so many companies will be wrecked by AI. That’s conjecture. But that is the narrative. The proverbial canaries in coal mines will then cause a recession. AI is so powerful that the recession is going to wipe out the economy and create a permanent class of unemployed white-collar workers. Plus, the bellwether of AI, Nvidia, seems to be on the verge of peaking. With all that on the table, we seemingly have an apocalyptic moment, the moment that we are in right now. What do I say? I say this is all so attenuated as to be ridiculous. Fear prevails, however, until otherwise dispelled. I don’t know how or when it will be vanquished, but I know it will. But until then, that’s where we are as of Monday morning. Brace yourselves. When what I just talked about becomes common parlence we could see a good amount of selling. To me, it means to look at a host of companies away from this gigantic group. I want you to go back and listen to the Club’s February Monthly Meeting from Friday. You will know what to do, what kind of companies to buy, and how classic this selloff really is. I like the moment. I am alone. (Jim Cramer’s Charitable Trust is long PG, BMY, LLY, CRM, NVDA, AMZN, GOOGL. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
How to think about everything presumably wrong with stocks and what to do about it
Mar 2, 2026
