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 Cramer, friends. I'm just trying to make a little money. My job is not just to entertain, but to educate. So call me at 1-800-743-CBC. Tweet me at Jim Cramer.
Come in to work and you'll learn that the Chinese are targeting all the frontier AI models, the Anthropics and the OpenAIs. And you wonder if they can get their IPOs off. Maybe not, because the Chinese also have cheap memory, which we think they do, or maybe they don't. But if we can get some hands on it, it'd be great. But if American companies truly avail themselves of the Chinese tech, won't they be targeted by China's open source models? And what does that mean for the price of tokens? Jevons paradox and national security.
Did you understand any of that? I don't. I can't figure out how this will ultimately impact Intel or Nvidia, two of my Charitable Trust core holdings. Could it delay an OpenAI IPO? Impact their deal with Oracle? Will keep repelling AMD. After a day where the Dow shed 307 points, S&P declined 0.19%, NASDAQ dipped 0.05% after what was, by the way, a very strong opening.
All this hand-wringing over AI competition from China makes me want to turn my back on tech and just find high-quality companies like Goldman Sachs and Wells Fargo, or dig my teeth into FedEx and FedEx Freight. Maybe scoop up some Honeywell Aerospace and Boeing. You know why? See, because if you own too much tech, you're going to be slaughtered and you won't even know what hit you.
Okay, not today. And we've been hit pretty hard here, as you will find out if you stick around. But I don't want to be some journo who only piles on when things are down. We might have some momentum here. But if you own these kinds of non-tech stocks that I like and they do go down, you know what you can do? You can confidently buy more of them because Goldman Sachs or FedEx are comprehensible and they're not subject to rumors or Chinese ripostes.
In so many ways, that's what's wrong with this market. I mean, right now, for example, if SK Hynix, the high-bandwidth memory maker from Korea, or SpaceX, the must satellite company, or Oracle with its flagging bonds and stock really break down, then believe me, they're going to take everything with them, including many tech stocks that do not deserve to be hammered.
Now, I've told you how my Charitable Trust has diversified away from tech with new money, with the exception of Intel. More on that in a second. For the club, I've been content to own industrials, not necessarily data-center industrials, as well as healthcare, because we don't want to be the next SKH, the stock of a tremendous semiconductor company that's been hijacked by desperate traders who really should be sticking with betting on the World Cup.
The Bulls remember the early 2025 DeepSeek affair, where stocks were crushed because of the introduction of a Chinese AI program, then bounced right back. Now, the Bulls are arguing that this new AI threat, Kimi K3, brought to you by the Chinese company Moonshot AI, is similar to the DeepSeek affair where everything else was brought down. But if you bought tech, especially data-center tech, you're going to make fortunes, just like when DeepSeek turned out to be second-rate.
I'm not so sure that we should swoop in and buy this time, because the stocks that could be rocked are stocks that, even after the decline, are still incredibly high, like a Micron, like an SK Hynix. They're ridiculously cheap on an earnings basis, but they're high. We will examine some of the most hard-hit stocks later in the show. You can make up your mind. I'm thinking some will have come down so much there is some real opportunity here, and I'm not blind to it.
That said, let me tell you what we're doing. First, we are not touching the Trust's long-standing Apple position. The sharp knives are indeed out for Apple, let's say, management from the tech intelligentsia. The intelligentsia say that Apple missed the most important story of our lives, artificial intelligence. They lambaste Siri. They regard the lost opportunity as critical to Apple's failure for the next generation.
Yet, you know what these same people are posting from their iPhone 17 Pro Max? You know what? They'd rather slip their wrist than switch to Samsung. They remind me of when I sat next to Steve Ballmer at my 35th college reunion when he was still CEO of Microsoft, and he pulled out some sort of gizmo that made me giggle. It was a Windows Mobile. Does anyone else remember that? I doubt it.
So, in its core business, Apple's bulletproof, and they never spent much on AI in the first place. I want you to think about this as you look at Apple down seven bucks today. Did anyone switch to Samsung during this travesty of an AI issue? Did anyone say, 'That's it. I am sick of it. I'm sick of the 17 Pro Max. Get me a Lenovo.'
Perhaps the cognoscenti wish that Apple would spend $200 billion, maybe 70% on the data-center structure and power, maybe give the rest to Google TPUs, maybe buy some expensive Nvidia GPUs. As someone who owns Apple for my Charitable Trust, I actually like that they let Alphabet spend those billions of dollars on AI and then they've had Alphabet pay them a nominal sum. So, it was in the iPhone. The pros have a term for this Apple-Alphabet deal. You know what it's called? A great deal.
Oh, I also want Nvidia. I know people think I've been unkind to Nvidia of late. I have been critical that they're not buying back more than $80 billion worth of stock, given that the shares are insanely cheap on next year's estimates. Do you know this is actually more of a love tap? People throughout this period, there have been companies that claim that they think the world of Nvidia, but they want to make their own chips. You know what I say? Go ahead, make your own chips.
But if the customer wants Nvidia, what do you do? What do you say to them? 'No, no, you see, like, we don't have Nvidia, you know, we've got the deli worst.' I mean, they go somewhere else. Nvidia is at the heart of the data center. Practically invented the data center. It has no Chinese competitor for real. The fact is, these machines, and they are machines, by the way, packed with hardware and software. They're the envy of the world.
Indeed, AMD comes close, but there really is nothing like Nvidia, and that matters. Yes. And we want Intel, which reports this week. Intel is a triple play: CPUs that you need for the new AI agents, the foundries to manufacture chips, and the packaging of chips with their high margins. Something CEO Lip-Bu Tan knows better than anyone. Why? Because he saved Cadence, the best at packaging. And then he gave you a 50-bagger while he was at it.
His foundries are going to be the best hope we have for an American semiconductor renaissance. Intel is a national treasure. I hope the stock sells off on Thursday. Please, please sell off. Please sell off. You know why? So we can get a bigger position. Understand again, I think the world of AMD, but I can't own both Intel and AMD. Too much overlap.
We think ARM can be perfect, but it needs foundry space. We own Broadcom. Why? Because CEO Hock Tan hates to lose. He's a competitor. Hey, by the way, though, same goes for Matt Murphy at Marvell Technology. But again, the concentration is too much for me because these stocks, at the very moment, mistakenly trade with the Korean stock market overnight, and then they get beaten down by the American press in the morning, which loves to praise the Chinese competition.
Oh my, the American press loves China so much. They're so smart. They're so smart. We're so stupid. We're so stupid. Clowns to the left of me at Caltech, jokers to the right at Stanford, stuck in the middle at MIT.
In the end, today was a day to buy FedEx. It was a day to pick up some Honeywell Aerospace. Hey, how about buying some GE? It was down a ridiculous amount even after it got the biggest order for it ever, 1,000 Leap engines to power the Airbus A320neos, part of a joint venture. If that wasn't enough, I was shocked that the stock wasn't up on that news. Instead, it fell nearly eight bucks. Kind of a crazy market.
But to buy more tech, only if we get a washout sell-off first, where all the margin mongers and the option ogres just get blown out, both here and in Korea, where the NASDAQ, of course, truly had to make a buck bringing them over. I should add, we have plenty of hyperscaler stocks for the travels, but we don't talk about them much unless it's the month meeting.
When we get that speculative washout sell-off, it'll be all over tech, and the values will be self-evident to anyone who still has cash. But the bottom line for the moment, it's time to go to other sectors. They can make you money without the volatility. You know what it makes me think of? Pharaoh's Fury. The last time I threw up at a theme park, but I missed my daughter. I mean, I didn't hit her. Bill in Indiana. Bill.
Hi, Jim. I'm a club member, and I thank you for the help and advice you've been giving me over this period of time.
Thanks, buddy. Thank you. Thank you. Thank you.
My question is about AutoZone. I've held the stock for about five years while there's been a big consolidation in the auto parts industry. So, it's been a very profitable trade for me, but it has now lost about 30% from its high from the last several months ago. So, I'm thinking about selling it. What?
No. No, Bill. I mean, look, they did have a problem, Bill. They had a problem, you know, like they have a lot of Chinese, you know, they had some tariff problems. I don't... This stock sells only at 19 times earnings. They continue to buy back stock. I wish you would not sell it down here. I think that they can make a comeback, but I hear you because I did think this was going to be a big year for them. And so far, I am wrong. Thanks for being a member of the club. I'm looking at the stock and thinking, 'Oh, wow. Don't sell it here.'
All right. For the moment, when it comes to the market, I think you have to look outside of tech to get the best opportunities, but that'll change if we get a real washout. On Mad Money tonight, I'm digging deeper into this hell, often anything related to AI. Don't miss my rundown of where the worst spots are and where to find opportunity. Then, is it time to press play on a position in Netflix, and why the stock has been beaten down so much, and whether now's the time to buy? I'll find you. And this earnings season hasn't been too exciting, but there are some things to keep your eye on that I'm pointing out. 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.
After an ugly close to last week's trading, it's clear we're witnessing a vicious meltdown for many of the key components of the AI infrastructure trade. I mean, this was the hottest part of the market until about a month ago. Today, many of the AI infrastructure stocks were able to bounce. Others tried and ultimately failed, but they're still down dramatically from their highs. And that's what matters. And that's why I want to spend some time focusing on what's happening here, why it's happening, and how we'll know when this meltdown is truly over. Not the kind of thing I saw today, but truly over.
See, the crazy thing here is that the AI infrastructure collapse is barely reflected in the averages. The Dow and the S&P are only down about 2 to 3% off their highs. The NASDAQ indices give a bit more of a clue, with the Nasdaq Composite, the IXIC, and the Nasdaq 100 down 6% and 7% from their early June highs, respectively. But that's nothing compared to the declines you're about to hear about in the AI infrastructure plays.
Let's start with the memory and data storage stocks, formerly the single hottest part of the market. Now, the Roundhill Memory ETF, symbol DRAM, you know, like DRAM, which launched in April, is now down an astounding 35% from its late June peak. Micron Technology, Sandisk, Seagate, and Western Digital are all down anywhere between 30 and 41% from highs set about a month ago. They all went parabolic, and this extends to the broader semiconductor industry.
The Philadelphia Semiconductor Index, or SOX, is down roughly 20% from its high, and the VanEck Semiconductor ETF is off nearly 17%. Looking at some of the top individual names, the semis, we see varying degrees of weakness. AMD and Intel, which investors have piled into as we realized the simple CPU chips that they make would see an increased demand as AI agents exploded, have come down 14% and 32% respectively just for the last day of June.
Broadcom's down 24%. Marvell is down 41%. Arm Holdings, a top chip design company, is off 40% from its high. And Teradyne, a semiconductor testing company that had been a real darling, is down 32%. Even the semiconductor capital equipment makers have been hit. ASML, the Dutch company that basically has a monopoly on the equipment used for leading-edge AI chips, has been least damaged, down 13% from its high as it just reported a tremendous quarter.
But Applied Materials, KLA, and Lam Research have all come down between 28 and 33% from their highs, their highs set less than three weeks ago. That's breathtaking. Beyond the semiconductor complex, many of the leading networking equipment plays have come in significantly. Corning, which I own for the Charitable Trust, has come down 44% from its end-of-June high. It's taken my breath away.
Astera Labs is off about 38% from its high set the same day. Coherent and Lumentum, the two leading photonics companies, have come down 35 and 30% from their highs set in either June or May. Now they are just, you know, trying to bounce. Some older networking plays that have been seeing a bit of a renaissance due to the AI infrastructure trade, they've been hit too. Ciena down 41%. Oh my god, that was a good one. And Cisco's down 15%, both from early June highs.
Hey, speaking of hardware within the data center, the leading makers of servers have been hit. Hewlett Packard Enterprise, HPE, is off 31% from its early June highs. Dell, arguably one of the best quarters of the year, has fallen 19% from its high set at roughly the same time. It's not just that. The great AI data-center buildout has been a boon to many classic industrials, at least until a few weeks ago.
Vertiv, which makes power and cooling equipment for the data center, has come down 23% from its mid-June highs. Modine Manufacturing, an old maker of thermal management systems that had found new life in AI, is off 29.5% from its late May peak. Caterpillar, it had been on fire both because its equipment is being used to build data-center structures and because its power generation equipment was increasingly being used to power these sites. It's now down over 19% from its high three weeks ago.
And Bloom Energy, oh my, a company that makes distributed power systems, had been one of the darlings of the AI year, but it's now plunged 44% from its high in late June. The biggest winners from AI infrastructure are the biggest losers of the last few weeks. Now, that includes the construction and engineering companies that we've been, you know, I've been talking a lot about these. They've been making a mint helping to build data centers or power infrastructure.
Sterling Infrastructure. Oh man, we had them on. What a hot shot. Quanta Services. These have come down 35% and 20% respectively. And of course, the companies that actually build data centers have seen their stocks get slammed too. Take the NeoClouds, which exist to rent out their computing power to other companies. So loved at one point, the leading name in the space, CoreWeave is down 47% from a recent high set in May.
Same story with IREN, which is down 43% from a local high set last month. Nebius, arguably the second-best NeoCloud name after CoreWeave, is down 39% from its late June high. And smaller players like Applied Digital, Hut 8, Terawulf are down anywhere from 28 to 45% from their highs. You can argue the same sell-off has hit new IPOs with AI exposure. SpaceX, Cerebras, the former is down 47% from its high set on the third trading day, while the latter's down 54% from where it peaked on its first trading day.
I remember when that was the one that was supposed to wipe out Nvidia. SK Hynix, the Korean memory chip maker, has seen its shares whipsaw around after debuting on the NASDAQ 10 days ago. They're currently down about 22% from their highs just set last Tuesday. From what it's worth, though, the company's Korea-listed shares are down about 39% from their late June high. By the way, Samsung's down 34% from its peak in mid-June. If it's connected to the data center, it's getting crushed.
Other 2026 IPO names with AI infrastructure exposure have also come in. There are a few industrial rollups that were popular because they were selling equipment into the data center, like 4G Power Solutions. We got a question on that last week. Madison Air Solutions and Inno those are all down anywhere from between 22 and 41% from highs set last month. Wow.
Just last week, we saw evidence that the weakness of the AI infrastructure trade was impacting the IPO market. One new data center operator, C Square, oh, that would have been loved a month ago, was priced its deal at 21, down big from the proposed range of 23 to 27, and it traded even lower since then. Oh my, nuclear play. Oh boy, remember those? Standard Nuclear nearly cut the size of its offering in half and priced its deal well below the range.
But the stock still fell from its offer price of $15 into the single digits now in the first three days of trading. That's awful. Now, I haven't mentioned the Magnificent 7 yet because this is a diverse group of vaguely tech companies with enormous market capitalizations. But the Mag 7 names have varying degrees of exposure to the AI infrastructure trade. For Nvidia, it's almost everything. For Apple, it's barely connected at all.
But even Nvidia is only down 14% from its mid-May all-time highs. It's holding up much better than the other data center supporters. Thank heavens. Because, you know, I think that even though it's a supplier, I think it's head and shoulders above the others. Let me give you the bottom line here. When you look at the carnage in the AI infrastructure space, once the hottest part of this entire stock market, it is clear we got a problem here. But is this merely a temporary detour? Or are we looking at something much more worrisome? Stick around after the break. I'm going to give you the answer.
Coming up, is the sudden AI stock sell-off just a blip or something you should seriously worry about? Cramer's doing the calculus next.
Like I mentioned before the break, the whole AI infrastructure complex has been melting down for weeks now, and we need to know what to make of it and whether now that everything is real negative, there might be some stocks worth buying. So let's talk about why this is happening. The truth is there are a bunch of overlapping reasons. First, we've got the macroeconomic factors. Now that the war with Iran has flared up again, oil prices are rising, and it'll be much more difficult for the Federal Reserve to cut rates.
We've seen this movie before. These AI infrastructure stocks were able to roar after our government reached a kind of détente with Iran in early April, but now that's over. And the price of crude has jumped from $67 on July 2nd to $83 and change today. That actually impacts these stocks very hard and not in a good way. On the interest rate front, after the benchmark 10-year Treasury yield peaked at 4.7% mid-May, it then fell to a low of 4.36% in late June, but now it's back to just below 4.6. So 4.3 all the way back to 4.6.
The market wasn't even able to enjoy a cooler-than-expected June CPI report last week because we knew that it was just likely to be ephemeral. That makes this a tricky market, especially for stocks that had huge gains. Those are the stocks people sell when they're worried about economic turmoil and inflation.
How about more AI-specific issues? The big worry with the AI infrastructure buildout is that it just seems to be costing too much. And now the bears are wondering who the heck's going to be able to make any money from all this investment. We know that the hyperscalers are paying increasingly large sums to build these data centers because many of the components have skyrocketed in price. Labor costs are getting mighty steep.
It's gotten to the point where even Alphabet, which spews cash, needs to raise money, roughly $85 billion in total, consisting of $45 billion placed in an unwritten secondary offering last month, followed by a $40 billion at-the-money offering that's likely going on right now, and it's gumming up the works. Who else is footing these bills? Well, there's Oracle. After a nice run through April and May, this stock's fallen an astonishing 51% from its high on June 1.
By the way, the price of credit default swaps on Oracle's debt, you can insure it, have risen to fresh highs, up more than fivefold year-over-year. That's not a sign of confidence. We need Oracle to pull some sort of rabbit out of a hat, make some money here, anywhere. Meanwhile, the two leading AI labs, OpenAI and Anthropic, are going to be paying a fortune for these data centers one way or another.
In late June, again, right around the time these stocks peaked, it was reported that OpenAI was thinking about delaying its IPO until 2027. They raised $122 billion, the largest private funding round ever, earlier this year. And you got to figure that it'll tide them over until an IPO next year. But with the amount of spending they're committed to, that's an open question. Here's what they should do. First of all, they should listen to me. They got to shoot for a fourth-quarter deal no matter what.
As for Anthropic, at one point in early June, we heard that a late-July IPO could be in the offering. Well, we aren't hearing that anymore, though. An IPO this year still seems likely for now. That should come in September. But the, well, that's my, again, that's my view. Okay. And by the way, I'm going to be right. But that's okay. I mean, my wife said I've been arrogant lately, so I just decided to go all in. All right. Right here. She said, 'You know, you really have been arrogant, Jim.' I said, 'You know what? All right.'
But the AI labs are part of the problem. We've seen more and more grumbling recently about how much it costs to run some of these leading models, like those from Anthropic, in an enterprise setting. That's how we got to the latest issue du jour for the AI infrastructure trade: the emergence of a new Chinese model called Kimi K3, an open-source model from an outfit named Moonshot AI that's supposedly just as good as the top frontier US models.
Sure, that's what caused the latest fall for the AI infrastructure stocks last week. But given that these stocks were already getting hammered before the China news, it's clearly not the whole story. There are other similar smaller issues that have contributed to the weakness here. For example, over the last few weeks, we've been realizing in real time how much of an impact bullish Korean retail investors have had in pushing up the memory stocks.
Last week, Korea's top financial regulators suspended new listings for single-stock leverage ETFs. Got to wonder how much those Korean leverage ETFs were responsible for the final extreme push higher that we saw in the memory stocks before they all peaked in late June. And either way, leverage is likely to be magnifying the moves of this broader group, pushing them up excessively and then making the down days all the more painful.
Sector and thematic ETFs that package all these stocks together compound the issue as well. They really make it difficult to find a bottom. Let me give you another reason for the AI infrastructure meltdown: too much stock. I mean, this is something I've been warning you about all year with new IPOs and secondary offerings. We have been flooded with ways to play the AI data center theme, and eventually an oversupply of stock can and will push down the entire group.
I think we need a breather here. I think it's happening right now. Keep in mind, we just had three of the largest offerings in history: the $75 billion IPO of SpaceX, the monster secondary offering from Alphabet, and that dastardly $26.5 billion listing of SK Hynix on the NASDAQ earlier this month. According to the stats from IPO experts at Renaissance Capital, we have had 86 IPOs that raised a collective $142.4 billion this year.
Okay, that figure alone is up 791% from this point last year, and it matches almost exactly the amount of proceeds from all of 2021, which, by the way, was a wild year for IPOs. We don't have enough money sloshing around to pay for much more stock at these levels. It's killing us. So, at the end of the day, while we don't dismiss each of these individual new factors that might have caused or contributed to the AI infrastructure sell-off, we also wonder if we're just running into the wall of supply that we always knew was going to come, always knew was going to present us with problems. And it's happening right now.
Put it all together, these AI data center stocks have turned, let's just say, they've been sunk by a perfect storm of negativity. I think the main issue is the flood of new stock, which makes it hard for these names to keep rallying. But the resumption of the war with Iran definitely hasn't helped. So how does this all end? Well, a couple things need to happen. We probably need the IPO market to calm down, but that's a problem that tends to fix itself by lower prices.
Last week's two AI-related deals flopped. That tends to dissuade others from coming public. We need that to happen, too. I'll also have to be watching the earnings reports from the hyperscalers closely. They need to give us tangible evidence that their huge capex bills are actually worth it. We've seen in recent quarters that companies with cloud infrastructure businesses have seen accelerating growth, and we need that to continue for the AI theme in general.
I also want to see evidence that the end users of AI, the companies that pay for Claude and ChatGPT, are getting, at least, like their money's worth. Someone other than the suppliers must start making visible amounts of money here, or at least forecast a profit. What else? How about a fresh cheese fire in her in that help? Here's the bottom line. It's very tough to own these AI infrastructure stocks right now.
But if you want to stick with this group, I recommend circling the wagons around the highest-quality stories, especially profitable companies with reasonable valuations, like I said at the top of the show, or at least historically ones that are coming back. Those are the ones that will make it to the other side relatively intact. The others, for those, we can wait until the new stock is digested, and then we can pick at the rubble. Hey, how about we go to Evelyn in Florida. Evelyn.
Hi, Jim. I'm at The Villages in Florida. I watch your show all the time and tell my grandkids to watch and learn.
I previously was in a ladies' stock club in Newtown, Connecticut, where I researched and bought CLS, Celestica, at $31 a share. Today it's over $300. And but where is it going now with the new Canadian pair?
Okay. Well, look, the problem with Celestica is it's part of the buildout of this entire data center movement. So, it's going to head lower. It's the 29 rings. These stocks are all coming down. It's not necessarily the Canadian, the real issue here. I hate to say it like this, but it had such a run. I mean, you bought it at 30, it's at 307. There's just going to be profit-taking, and that's what it is, because it is a really good company. Let's go to Stuart in Florida. Stuart.
Booyah. Jim, how you been, my friend?
I'm doing fine, Stuart. How about you?
Doing great. I want to buy a dominant long-term compounder. While the stock is down 36% and trading at its lowest PE multiple in five years, bears are focused on transient macro headwinds like inflation, compressing restaurant margins down to 23.7%. But I'm looking past that. Their massive expansion plan of 370 new locations and 38% digital sales mix protecting throughput and positive transactions. Is this time to get in Chipotle?
I think, you know, look, it reports on the 29th. I doubt it runs away, but if you want to put a quarter position on only right now, because there's always a chance that, frankly, they have this lettuce issue, and we just can't have the lettuce issue get in the way of building a new position. All right. It's tough to own anything in the AI-related space right now. There's too much supply. But if you want to say stick with a couple of high-profile quality companies, I'm in favor of that.
Now, much more Mad Money ahead. Are investors still watching the stock of Netflix? I'm taking a deep dive into the company amid a big day for media news. Then I'm running through the top stories of this earnings season so far and all your calls rapid-fire in tonight's edition of the lightning round. So stay with Cramer.
Has the stock of Netflix finally become too cheap to ignore? Over the last 12 months, this stock's now down 44%, including a 9% decline just since it reported last Thursday. It's gone from a beloved market darling to a complete pariah. Of course, Netflix is not alone here. The entire media cohort has been struggling. But in the old days, Netflix rarely traded just like another media stock. This is the company that invented streaming video.
But now everybody in the business has their own subscription streaming platform. Now we got to start asking ourselves if Netflix is still unique, or is the competition for your eyeballs at last catching up to this amazing business. When the company reported last week, the sales missed expectations slightly and earnings per share matched the consensus assessments. These were not awful numbers, but they certainly weren't what the bulls hoped for.
Worse, Netflix's free cash flow was much, much lower than expected, just $1.53 billion. Wall Street was looking for $2.67 billion. That was mostly from higher tax payments, including taxes on termination fees they got when Warner Brothers walked away from that merger. So, this wasn't necessarily a collapse in the underlying cash-generating power of the business, but when a stock has already lost the market's confidence, a free cash flow miss is not what you want to see.
The worst part, though, is the guidance. For the third quarter, Netflix guided for weaker-than-expected revenue and earnings while also narrowing their full-year forecast. Of course, the company isn't falling apart. It's still on pace for 13 to 14% revenue growth this year with a 31.5% operating margin, $12.5 billion in free cash flow. Most companies would kill for those numbers. The problem is that the direction of travel is a lot less exciting than it used to be.
In the second quarter of last year, Netflix grew at a 16% clip. The growth here is decelerating. Management blames some of the deceleration on timing and tougher comparisons, but it's harder and harder for the bulls to dream up ways for the company to accelerate its core subscription business. Netflix also announced that it will begin releasing its What We Watched engagement report annually rather than twice a year.
Remember, Netflix stopped giving us quarterly subscriber numbers in the first quarter of 2025. Now it's reducing the frequency of engagement disclosures. Whenever a company decides to give you less transparency, Wall Street tends to assume management's hiding something. Again, not ideal. At the same time, Netflix hasn't been producing the kind of franchises that used to drive dramatic subscriber growth. Where's the next Squid Game or Stranger Things? Maybe they're coming. We're certainly ready for it.
Personally, I like their American Experiment documentary and I Will Find You series, but those weren't pop culture sensations like Netflix has had in the past. Unfortunately, the streaming business has become increasingly, here's a word you never want to hear, commoditized. Consumers can subscribe to Netflix for one month, switch to HBO Max for the next, then move over to Disney Plus, Apple TV Plus, and Paramount Plus. The switching costs zero.
Apple, Google, and Amazon can afford to treat entertainment as a strategic side business. Apple doesn't need Apple TV Plus to generate Netflix-like margins. Amazon uses Prime Video to make Prime subscription more valuable and more products and gain market share. Which brings me to the Warner Brothers deal. Many investors hated the idea of Netflix buying Warner Brothers. The prospect of that acquisition crushed the stock.
Bulls didn't want Netflix entering a bidding war with Paramount over an old media business filled with declining cable networks. When the transaction fell apart, Netflix got a $3 billion termination payment. Wall Street actually treated that as a victory. But the problem here is that it now looks like Netflix arguably needed Warner Brothers more than even the skeptics thought. This deal would have brought an enormous library of intellectual property, including HBO, DC Comics, CA characters, and decades of television programming.
Still, the stock's almost been cut in half over the last year. So, let's talk about what makes the stock interesting. When they're cut in half or near it, I get more interested, not less. First of all, Netflix repurchased $4.7 billion of stock during the second quarter. That's the largest quarterly buyback in its history. It still has roughly $27 billion remaining under the repurchase authorization.
Netflix expects to produce $12.5 billion of free cash flow this year. It can fund the slate, the slate, invested advertising, sports, live programming, and still retire an enormous amount of stock. If the share price remains depressed, Netflix can keep up the buyback. Don't forget, they've been among the best capital allocators in the industry, growing into a global behemoth while maintaining relatively little debt versus when it first started.
At 19 times this year's earnings estimates, Netflix is the cheapest it's been since 2022. And you know what happened since then? It turned out to be one of the great buying opportunities that I've seen. From 2025 through 2028, JP Morgan expects Netflix to produce a compound annual growth rate of 12% for constant-currency revenue, 20% for operating income, 24% for earnings per share, 22% for free cash flow. If those numbers are even close to right, sign me up.
The stock's trading at a discount to the S&P 500 as a whole. That's crazy. Premium growth expected for the next few years. Then again, many members of the Magnificent 7 trade at similar discounts to the market, and some of them are better businesses. Heck, Nvidia trades at a lower price-earnings multiple based on next year's numbers. So, should you buy the dip? Look, I won't pretend that the quarter was great. It was a disappointment.
The content slate clearly isn't as strong as usual. Meanwhile, Netflix is investing heavily across ads, games, live programming, podcasts, sports, including some potentially big NFL games, and AI. And while it's historically cheap, 19 times earnings isn't exactly a steal. If engagement keeps slowing and revenue growth falls into the single digits, the stock can certainly still get cheaper.
On the other hand, this is not a broken company. It's one of the best companies around with one of the best products, and the numbers are still better than most. I think there's a reason why these guys are buying back stock at the fastest pace in history. Advertising revenue should roughly double to $3 billion this year, and management believes the gap between the economics of the ad-supported and the ad-free plans is narrowing.
Netflix estimates it's less than 45% penetration of its addressable broadband households, captures only about 7% of the entertainment revenue available in these markets, and accounts for just about 5% of global television viewing. So there's still plenty of room for growth. So, here's the bottom line. Netflix may no longer be a hypergrowth monopoly story, but at 19 times earnings, hey, I'm thinking you're getting a well-managed business, one of the largest corporate buybacks in America.
It's not often you get one of the best companies ever created put on sale. That said, you know what? You can take your time. As you saw from today, you don't have to rush this if you're still a believer in Netflix. I'd put a small position here and then gradually add on the weakness in pyramid style, because I wouldn't be surprised if the weakness sticks with us for a while, but that ultimately you have a potential big winner. Mad Money's back after the break.
Coming up, you've got questions. Cramer's got the answers. Get charged up for a fast-fire lightning round next.
It is time for the lightning round. Of course. We play here the sound, and then the lightning round is over. Are you ready to light with Sam?
