My mission is simple: to make you money. I'm here to level the playing field for all investors. There's always a bull market somewhere, and I promise to help you find it. Mad Money starts now.
Hey, I'm Cramer. Welcome to Mad Money. Welcome to Cramerica, my friends. I'm just trying to make you some money. My job is not just to entertain, but to do some teaching. So call me at 1-800-743-CBC. Tweet me at Jim Cramer.
Currents cross, narratives get scuttled, truth gets obscured, all because of hedge fund managers and big mutual fund trigger pullers occasionally getting on the wrong side of the trade.
One of the things I try to teach you on the show is how much stocks can diverge from the underlying value of the companies. Maybe your stocks were only going up today because some hedge fund imploded last week after eight straight days of the Philly semiconductor index being down, and we just didn't know it. With the Nidec Dwell hedge fund out of the way, the stocks were free to lift. Maybe today's rally, with the Dow slipping 36.38 points, but the S&P gaining 0.4% and the Nasdaq jumping 1.11%, has more to do with some loser fund being blown out last week, a so-called clear event, than it does with how the businesses underneath its stocks are really doing.
This weekend, I told members of the CNBC Investing Club that I thought Friday, last Friday, had all the hallmarks of a hedge fund blowing up, kind of like what happened with Situational Awareness in the summer, but a kind of mini Situational Awareness. You didn't know Situational Awareness was falling apart when it happened; you only knew the day after. But it knocked dozens of stocks off their flight path and caused tremendous havoc, as many analysts thought something had to be wrong with the underlying companies, not just the stocks. Now, I bring up this manufactured, non-fundamental selloff because today is the 39th anniversary of the crash of 1987.
I know that 39, not a big deal, not a round number. Still, it's incredibly instructive. Right now, and I recall it well, a few months into the year, my great friend Larry Levy and I left Goldman Sachs to form Cramer Levy, a hedge fund. It was a very novel thing to do. Larry was more of an outside guy raising money. I was the inside guy with a terrific relationship. But I think that once the money was raised, Larry wanted to move on, and I just wanted to settle down, pick some stocks. That was hard enough. I had a playbook of buying stocks in the best companies. So I moved the fund to Steinhardt Partners, where the legendary trader John Latane schooled me. But good, I was hanging on by my finger, and I was down almost 10%, where I would have had to open my fund, and I didn't help that I didn't have Larry to keep things together.
I sat down with Latane, who promptly called me a Harvard-trained idiot and a disciple of the research department of Goldman Sachs, which didn't know jack about how the market worked and would recommend a bunch of bonehead blue chips regardless of the tape. He told me to sell everything, start over, stay attuned to the market. I had no idea what the heck he was talking about. Tune to the market? I thought I was making the most subjective analysis of each position, but he told me that stocks had nothing to do with companies. They were just pieces of paper blowing with the stock market wind. Wouldn't you know it, he was right.
At the time, we had a market led by a few stocks that the Japanese would buy every night, and they'd keep those stocks up in the morning. Back then, Japan was on par with us for the biggest and deepest stock markets. They were challenging our wealth, and in many ways, their economy seemed a lot better than ours. The Japanese loved the waste management stocks. They thought 3M was terrific. They had an appetite for the drug stocks, especially Johnson & Johnson and Merck. They thought GE was really good. The whole thing was insane because their favorite stocks just seemed to go up every single day. Their buying knew no bounds. In fact, the Japanese were responsible for most of the marginal buying of the whole move upward. They took the indices up to about 29 times earnings by the time October rolled around.
The whole time the market was being taken up by Japan, there were these weirdo salespeople selling you basically something called stock market insurance, portfolio insurance. Yes, insurance. You could ride up the indices and be stopped out down 5% for a fee if something bad happened. If you were up big, you were able to lock in the gains with portfolio insurance, or at least they told you you could. Most people don't remember that time or weren't alive to know it. But Larry mentioned it to me this weekend, and it reminded me just how you can't take all the moves seriously.
I say that because what happened next, what happened today 39 years ago, is that the portfolio insurance didn't work. The insurance agents were shorting stock futures to stop you out. But the selling from the futures pits in Chicago, still a relatively young pit, overwhelmed the stocks in New York, and everything cascaded down.
Black Monday had arrived. Stocks totally divorced themselves from the fundamentals. But we didn't know that at the time because we figured there had to be some fundamental reason for the meltdown. It was impossible that the market could lose a quarter of its value in one day just because of weird insurance policies. The Dow Jones Industrial Average doesn't plunge 22.6% in a single session for nothing. Not possible. Or so we thought. Yet it turns out there really was nothing fundamentally wrong of any consequence at all. Rampant automatic selling destroyed stocks even as the underlying companies were quite strong. In retrospect, it was a spectacular buying opportunity.
I had by that time started working with soon-to-be fiancée Karen, and she was from the Latane school of thought because she worked with him. She hated the market right before the crash and urged me to keep selling until I owned nothing. She did that because for weeks in advance of Black Monday, she said there were no real buyers, but a wave of sellers that were knocking down stocks for no reason. So after selling all my stocks earlier in the year and then buying back all different stocks, I just sold everything back again except for some puts in Johnson & Johnson. I was in cash for the crash for no reason other than a trader's judgment that something was very wrong with the market, not the companies underneath, which were fine.
She told me that we could buy all we wanted after the crash. So all I could say was, 'What crash?' She said, 'Oh, the one that's about to happen.' Then Black Monday hit, and we looked like geniuses. Best call ever. Anyway, as people searched for reasons why stocks would go down, was it interest rates? Was it recession? We pinned the blame on the actual stock market machinations, not the companies. Our recognition that the selloff had much more to do with the mechanics of the market, not the fundamentals, cemented us as people who must know something, and not much else. But it was enough to get my hedge fund to be able to raise a ton of money, albeit very small by today's standards.
I think we're seeing the same thing happening right now: stocks having little to nothing to do with the fundamentals in some, but not all, the data center stocks right now. They all go up in bunches, even if they shouldn't, at least not this fast. And when they go down, they go down in unison really quickly, with nothing happening at the companies, just like '87. There is a big difference between now and then, though. These data center stocks have nothing to do with the S&P 500, which is doing pretty terribly anyway. They're out on their own, and they're vulnerable to any kind of hedge fund blowout because the size of funds these days is immense, and they play with insane amounts of borrowed money.
Here's the bottom line. On the anniversary of the great crash, remember on this green Monday that you may be seeing huge moves based on the machinations of money managers who are out over their skis or crushing the slalom. All totally divorced from the fundamentals without a shred of alimony. Doreen in New York. Doreen.
Hey Jim, thank you so much for taking my call. Very...
Of course, Doreen. Thank you. Thank you for calling. How can I help you?
I need to know. My question, I need help on is SoFi. It has been sliding ever since I bought in around $17, and at today's levels and last week's 52-week low, I need to know: should I be buying more to lower my cost basis, or should I just hold the shares that I have?
Okay, SoFi is, in the end, an advisory bank. Okay, but it is a bank, and it sells for much more than almost all the other bank stocks. So as the other bank stocks get pushed down, it gets pushed down by them. Should you sell it? Look, here's what I think. It's not going to go up. It's down 45%. It's not going to go up until we see the other bank stocks go up because that's what it's trading with. And the bank stocks right now do not trade well. So that's my admonition. I would not expect any big rally now unless we get some sort of huge move up in the big banks, and I don't see that coming right now. Let's go to Dale in New Jersey. Dale.
Jim, how are you? Thanks for taking my call.
Of course, man. I'm glad you called. How can I help?
Yeah, I'm calling about Allstate Insurance. I already own a position. Um, you know, it has a PE of like five, pays a small dividend. It's been doing pretty good, and I was wondering if I should increase that position.
I don't want you to. Look, the stock is incredibly cheap, okay? I mean, amazingly cheap. It's just had a really nice move off the bottom. And I just think that you got to just... it's moved so much just in the last couple weeks or so that all you're doing is betting that you know what's going to happen. You're going to have a multiple expansion from six to seven. I don't like to play that game. Let's see how the earnings go.
All right. Now, sometimes it's the mechanics of the market, not company fundamentals, driving the action. Keep that in mind the next time you see a big move. On Mad Money tonight: after last week's tech wreck, I'm assessing the damage and hunting for buying opportunities on the road. Plus, what are small businesses telling us about the job market? I'm getting a read from Paychex. And could rising interest rates derail the data center buildout? I'm explaining why the biggest threat to the AI boom may have nothing to do with the AI boom itself. So stay with Cramer.
Don't miss a second of Mad Money. Follow Jim Cramer on X. Have a question? Tweet Cramer #MadMentions. Send Jim an email to
[email protected] or give us a call at 1-800-743-CNBC. Miss something? Head to madmoney.cnbc.com.
Last weekend did a pretty number on the market, especially tech, turning into a full-blown panic selloff by Friday. Hey, today we got a reprieve from that meltdown, but it's extensive enough that we got to go over exactly what happened last week, why it happened, before I can start getting truly excited about the AI data center space again. Yeah, it was that bad. So, how did last week become such a disaster for tech? Well, it started slowly on Monday. The Dow actually managed to gain ground in part because the bond market was closed for Columbus Day, but parts of the AI data center cohort still got slammed, like the memory stocks. Didn't help that the weekend before last, Hock Tan, the CEO of Broadcom, that's a Charitable Trust name, explained that capital might be the biggest constraint to the AI buildout rather than chip shortages. In a world of rising interest rates, the cost of capital is a major problem when we're talking about a multi-trillion-dollar cycle.
On Tuesday, most stocks were down, including tech. But it wasn't what led us lower. Instead, people were worrying about the banks. At the same time, we heard more negative chatter about the financing side of the AI data center buildout. By Wednesday, things took a hideous turn. Oh, it started with, yes indeed, CoreWeave, a company that builds data centers that rents out their computing power. This thing has been a great growth story, but they've had to accumulate a mountain of debt to pay for these warehouses full of servers. And until last week, we seemed to be okay with that.
Why did investors start keying on Core in particular last week? First, I point to the fact that Elon Musk, who has a similar business contained in SpaceX, drew attention to Core's borrowing costs on Tuesday night, like the fact that some of its recently issued bonds have yields approaching 15%. And Musk then, what happened in response to that fact? Musk tweeted simply, 'Yikes.' Wow. If 15% is your cost of capital, then it's going to cost you an arm and a leg to build a data center. Don't—words to the wise—don't ever have anything that Musk could say yikes about.
Separately, there was a poorly timed article in the Financial Times last Wednesday featuring CoreWeave's co-founder and chief development officer, Brandon McMahon, who talked about a desire to tap insurers' unlimited pools of capital for his company's financing efforts, making the insurers sound like they were brain-dead marks. Unacceptable. That's not how you raise money. Regardless, investors started panicking, and the stock plunged more than 15% on Wednesday. Broader Nasdaq 100 down 2%.
Thursday seemed like it might be the beginning of a recovery of the AI data center cohort. After Asian markets had followed U.S. stocks lower overnight from Wednesday to Thursday, we had a down open in the U.S., but then we stabilized in part. Why? Because of some reassuring comments from Jensen Huang, the CEO of Nvidia, about, yep, CoreWeave. He gave CoreWeave his seal of approval. Makes sense, as Nvidia is already a huge shareholder, and said it didn't need any help. That was really important. It did not need any help. In fact, he said it was doing better than people seem to believe. The Nasdaq finished the session in the black, with the Dow and the S&P both roaring higher.
But on Friday, the bottom fell out of the entire market, with tech leading the way lower. To me, this felt like pure fear taking hold, with the CBOE Volatility Index spiking from 16 to 22. It closed under 20. The Nasdaq Composite finished the day off 2.15%, the worst performance since July 23rd, when that overleveraged hedge fund, Situational Awareness, when that thing blew up.
Now, over the course of last week, there was an escalating sense of doom about the AI data center boom. Anthropic released some updated financial metrics as part of its early roadshow, because you know they want to come public someday. While the AI lab is still growing nicely, it's not growing as fast as the bulls hoped. Believe me, if it had been better, we wouldn't have had such a tough week. Insult joined injury. There were also some rumors swirling about another AI-focused hedge fund possibly blowing up, like I alluded to at the top of the show. That would definitely fit with the action.
We even saw signs of stocks getting hit by AI displacement worries for the first time in months. When Perplexity announced they'd release an open-source project to help combat rogue agents, some of my favorite cybersecurity stocks got slammed. It won't hurt their businesses, but people have to go about selling first because the stocks are up a lot, but also secondly because the owners reflexively can't resist selling when they see any of the generative AI companies say something about providing their own cybersecurity. These companies are believed no matter what. But whatever the exact reason for the negativity, this all ended up being a rather rough week for the tech stocks. They're not necessarily the entire market. You know, the Dow finished the week up 0.67% last week, even as the Nasdaq 100 was down almost 5.5%. It was the worst week since April of last year, when the Liberation Day tariffs hit. I bet you felt it. I know we felt it at the Trust.
So, let me give you a quick damage assessment of the hardest parts of the AI data center story. Let's start with, well, you tell me. That's right, CoreWeave and its troubled brethren. This is what's known as the neocloud, and all the so-called neoclouds were crushed last week, with most down more than 20%. GIGA down 25%. Even after some decent bounces today, these names are still well off their highs. Nebius is the best off, down 43% from its June high. But the rest, CoreWeave, IN, Hut 8, Applied Digital, Core Scientific, have all been more than cut in half from their 52-week highs.
The second worst subsector: memory and data storage. When we start hearing lots of chatter about how expensive it'll be just to raise money to build data centers, then the demand for these components that go into the data centers gets called into question. The memory and data storage group peaked in late June, sold off hard during that Situational Awareness-driven crash, and although these stocks rebounded dramatically, they never got back to the previous highs, and that is a real bad sign for you charters out there. Micron, SK Hynix, SanDisk, Western Digital, and Seagate were all down 15 to 20% last week.
Even after bounces today, the highest quality memory stocks, I'm talking about Micron, SK Hynix, they're still both down over 25% from their highs. The others are down more than 40%. These are major, bad bear moves. In the case of Western Digital, it's down 50%. Now, we like Micron for the Travel Trust. Why are we gluttons for punishment? No. We bought it after spending some time at their Boise headquarters. We're situationally aware that the stock's not acting as well as we would like, but the business is great.
We saw the same dynamic with the previously red-hot fiber optics companies. Lumentum, Coherent down 25%. Corning dropped more than 15%. How about the broader semiconductor industry? After a vicious beatdown last week, the Philadelphia Semiconductor Index, or the SOX for short, finally broke its eight-session losing streak today, and the VanEck Semiconductor ETF broke a nine-day losing streak. Regular viewers know that these losing streaks can get real ugly for the semis. But man, this group was just hammered, hammered mercilessly.
Similarly, when you look at the main semiconductor capital equipment makers, they don't paint a pretty picture either. These are really great companies: Applied Materials, KLA, Lam Research, ASML. They were down anywhere from 8 to 19% last week. ASML, by the way, reported a really strong quarter. I thought it might turn things around. That was wrong.
Finally, let's look at the hyperscalers. Now, those are the true tech titans, and they fall into a different category entirely. They are the ones doing the lion's share of the spending on the great data center buildout. These stocks were by and large spared from the worst of last week's selling, with the exception of SpaceX and Oracle, which are the most financially tenuous. But Alphabet, Microsoft, Amazon, and Meta only had modest low- to mid-single-digit declines. And by the way, Amazon stock rallied back more than 12 points today on a Morgan Stanley upgrade, showing these stocks have a very technical term going for them: mojo.
Oh, and I can't resist showing you that Apple, the only tech titan without major exposure to the AI data center theme—remember, Gemini is basically their partner, and Alphabet pays them—was actually up 2% last week. Some of that is because, yes, the Duo, own it, don't trade it, has worked well. And I think the Duo is going to be terrific. I actually heard some people say things negative about the Duo today, which only made me emboldened to say buy Apple.
Here's the bottom line. Last week's AI data center meltdown was so severe that I had to use a whole segment just to cover the extent of the damage. Stick around after the break, and you know what we'll do? We'll tell you what needs to happen before the tech wreck can turn into a buying opportunity. Notice I didn't say it is a buying opportunity. Mad Money's back after the break.
Coming up, are the tech stocks out of the woods after Friday's beatdown? Cramer is breaking down how they can find their way back, next.
Before the break, I walked you through the details of last week's nasty tech selloff. Even though the AI data center stocks kind of rebounded today, the whole theme has been called into question, right? And I think that this group will struggle to find its footing for the moment, especially with this sword of CoreWeave hanging over us. So what needs to really change before we can start bottom fishing with some degree of safety? And we know I say some because it's never safe when you bottom fish. First and foremost, it sounds simplistic, but you know what? The charts need to get better.
The AI data center stocks peaked in June, then rebounded until last week when they got obliterated again. That does make for some ugly charts. When you look at some of these groups represented by ETFs that have been at the core of the AI rally earlier this year, I'm talking about the VanEck Semiconductor ETF or that newer round memory ETF, the aptly tickered DRAM. They sure look like they've made double tops, and DRAM is actually DRAM DM. That's one of the worst patterns out there, a double top. It almost always leads to lower prices. So they need to break that double top pattern, which means we need a nice rally, more than we got today, that proves there's still a plausible path to a pattern of higher highs and higher lows.
Worse, the ones that don't have double tops have the most pronounced head-and-shoulder patterns that some people think is even more negative, to the point where the only way out seems to be lower. To the seasoned chartist, these stocks have no bottom. You know, if you're an optimist, the good news is that as bad as things got last week, we did not take out the lows from late July. Now, that's when we got that crescendo of AI data center selling as Situational Awareness, a highly levered hedge fund, finally got liquidated.
Today's bounce, though, I'm going to call it a good start. How about that? With Friday's puke session, now known to us as get-out-now selling, both the Nasdaq Composite and the Nasdaq 100 fell below their 50-day moving averages. They haven't lost touch with those moving averages, and today's bounce kind of helped us get back closer to the key levels. But it would be very helpful if the Nasdaq indices would just rebound above the 50-day moving average, ideally by the end of the week.
Beyond the charts, which means many of you are about to turn the show off because, like, all you care about is the charts, but I think the fundamentals still matter. The next big test for the AI data center complex will be hyperscaler earnings reports. Now, SpaceX, you ever know that guy Adam Jonas at Morgan? He writes a piece every day about SpaceX. SpaceX is going to give us some color when it reports tomorrow night. I'm going to be sure to read him. I think he matters the most of anyone. It's going to be huge. Meta Platforms, with fewer moving pieces, might be even more useful when it reports on Wednesday night. After that, we'll hear from both Alphabet and Microsoft on Wednesday of next week, with Amazon reporting the following night. Yeah, we got some biggins.
What do we need to hear from the hyperscalers? Well, if you're hoping for a resurgence in the AI data center cohort, these companies need to maintain their gigantic capital spending. That's right, need to maintain, not cut. See, if they start to signal that they're preparing to dial back their investments, then the whole group's going to get whacked. Why? Well, because, you know, we need to see that they believe in it. What do we want to hear? How about this: 'We have line of sight to hundreds of billions of dollars in revenues. It would be foolish to cut back, and we don't have to, because 2027 will be the year of the bountiful AI harvest.' Hyperscalers, please take that down. It's really good.
On top of that, we want to see the hyperscalers continue to make serious money from the data centers. We've started to see this from some of them, like the excellent cloud infrastructure growth numbers from Alphabet, Microsoft, Amazon. SpaceX has...
shown it because it can build data centers then lease computing power to others for significant sums of money. It's the best example yet of exactly how much money you can make when you create what we call compute. Meta has been less clear because it doesn't have a cloud business. But maybe it moves in that direction or perhaps it will use this compute from Muse. That's the personal AI agent platform that is going so incredibly well. In short, we're about to run a high-stakes gauntlet of tech earnings. And if the hyperscalers blink on AI spending then the whole data center edifice could be like...
As long as they don't blink though, I think we can get through this especially if they put up strong cloud infrastructure numbers. Third and maybe the most important, I don't think the AI data center cohort can put in a sustainable bottom until after the midterm election. That means we're in limbo until people go to the polls two weeks from now. Opportunities for tomorrow. We know that data centers have become politically toxic, even in states that have been previously supportive of these things. So far, the Democrats have emerged as the party that's more opposed to data centers, and the polling suggests they'll likely do pretty well. Very high odds of taking the House, decent chance of taking the Senate.
However, I'm not that worried about Washington here. I think the main issue is optics. No politician wants to endorse data centers ahead of the election because they don't want to drag their party down, but that becomes much less of an issue on November 4th. At this point, I think investors just want some certainty about what Washington will look like at the start of next year. Although, who knows if we'll have certainty after the election. You have to wonder if the Justice Department will sue to block any Democrat from taking a seat from a Republican. Don't laugh. It could happen. Maybe we get one last washout in the group after the results come out.
But outside of campaign season, state and local leaders generally like bringing in new business. We just got to get past the election day. And regardless of the results, our leaders won't be so afraid of the backlash, at least until 2028 rolls around. Data centers may be unpopular, but elected local officials who want to cut taxes can do so by doing some hard bargaining with the hyperscalers who want to locate data centers in their towns. Everyone likes lower taxes. They like them more than they care about big buildings with machines that might or might not cut employment five years from now.
Finally, there's the looming IPO of Anthropic, one of the leading privately held AI labs. Ideally, we see Anthropic come public and trade well after it's open, but I don't even want to get that greedy at this point. I just want to see the deal happen. Before the break, I mentioned that Anthropic's pre-roadshow event last week contributed to the selloff because quite a few people were disappointed by the numbers. They previously said they want to come public in November, although now we're hearing whispers of December. Great. Will you get this done, guys? We need to see that someone is making money or will soon be making money from all this data center investment.
Sure, the hyperscalers have seen great returns from their cloud infrastructure businesses, but Anthropic and OpenAI represent a huge amount of their business from these platforms. Microsoft's new cloud revenue is hostage to OpenAI. Amazon and Google's cloud growth is hostage to Anthropic. So, we need to see at the very end of the spending chain that this could be a sustainably profitable business that's worth investing in. That's what Anthropic's IPO represents.
If there's enough demand that Anthropic's IPO is successful, that will give investors much more confidence that the AI data center theme has legs, although it will also suck a massive amount of money away from the rest of the group just because Anthropic is so darn enormous. Here's the bottom line. Even after today's bounce, I still think it's just a little too early to say that the coast is clear and it's safe to dive back into the hardest-hit AI data center stocks. I just don't think it's safe to be aggressively bullish on the group until the charts improve, the hyperscalers report, and the elections are in the rearview mirror and Anthropic manages to come public. That's what you need to keep your eyes on before circling back to this newly beaten-down group. In other words, don't get too excited about today. If we get the Cosby down today, we'll be right back into the...
the House of Jerry in Missouri. Jerry, hey Jim, thanks for taking my call. My pleasure. Jim, my cost basis is $82 and I am now up over 178% even after last week's 12% haircut. I've taken out over half of my cost basis. Do you think I'm being reckless letting the rest of my CrowdStrike ride?
Just the opposite. We just decided to wet out your basis and say, well, not where you got it come from, but where it's going. I think you have a good chance it's going to go higher. And let's not forget because you took out so much, you are playing, my friend, with the house's money. And that is kaching kaching. Even if today's bounce, I think it's too soon to say, hey, it's safe to dive back. That's very glib. Okay. And we're not going to play that game.
Now, much more women money including my sus with Paychex. I'm getting the state of small business and hiring in the US with the CEO. Then I'm explaining why you shouldn't underestimate the importance of interest rates in the AI-driven market. And all your calls rapid fire tonight's edition of the lightning round. So stay with Cramer.
Little less than a month ago, we got the latest set of earnings results from Paychex. That's America's second largest payroll processor. Big outsourced human capital management division. Stock bizarrely plunged 9% in response. What went wrong? Paychex's core business posted a revenue miss. That said, the stock has now rebounded from its lows and erased most of the decline. I think some of Wall Street's actually re-evaluating that quarter. They delivered a solid earnings beat and their overall revenue was slightly higher than expected, but their management solutions business by their own admission came in a little light. That was enough to overshadow everything else.
Now, Paychex says the customers are simply moving from a la carte HR solutions to their one-stop shop HR business. Still, investors were underwhelmed by the company's guidance for the current quarter, and that didn't help. Last week, the company held its annual shareholder meeting where the management hammered home their new agentic AI technology that's being built into their payroll and human resource system, which is the future. So, let's check in with John Gibson. He's the president CEO of Paychex. Mr. Gibson, welcome back to Mad Money.
Jim, it's great to see you.
