It’s not the expense, it’s the return. Friday’s plunge in technology stocks was proof of that. This is probably the most significant decline in technology stocks in over a year. We need to know what we are up against, potentially. We need to know whether the multi-trillion dollar spout is drying up or just drying up. In the portfolio I use for the CNBC Investment Club, a charitable trust, we are consolidating traditional technologies (semiconductors, software, data centers) and gradually winding them down and moving into other types of technologies: technology-infused pharmaceuticals and aerospace. We were concerned that Nvidia didn’t yet have enough customers to meet their new needs, so we tried to center our portfolio on Intel rather than Nvidia. No, I’m not giving up on Nvidia. Still, it’s great and I think it’s going to be a great quarter. However, the “behavior” of stocks is too eloquent. Apple’s move screams that its deliberate or de facto decision not to spend hundreds of billions of dollars on AI is great. We are having the best month in three years. Many critics are second-guessing Apple’s decision-making. That’s wrong. Apple decided a long time ago that if they made the best mobile device, a lot of stuff would flow to them. That’s right. This allows us to choose which companies to partner with: hyperscalers and chatbots. Because those companies quickly turned into a commodity. Google effectively had no choice but to let it go, at least on a net basis, because while Gemini is no Claude of Anthropic, the resurgence of Google Search is questionable. So I was drawn to Intel, a now slow new company, because I saw the rapidly changing ratio of graphics processing units (GPUs) — Nvidia’s big, expensive chips — and central processing units (CPUs) — Intel and Advanced Micro Devices (AMD) — and possibly Arm Holdings (if it could get foundry time, despite its partnership with Intel). When Lip-Bu Tan became Intel’s CEO, the ratio was about four GPUs to one CPU. Now, last Thursday he told me that it’s about 1 CPU to 1 GPU. Data centers will soon have four CPUs for every one GPU. GPU gross margins are much richer than CPUs. Intel, if managed well, can change that. This is a well-run Intel. Additionally, CEOs are probably the most avid semiconductor investors and know how to use their money wisely to build foundries (factories that make chips) when supplies are in short supply. Most importantly, he knows the packaging equivalent of bundling CEOs together to make them more powerful, as it becomes increasingly difficult to make nodes smaller and more powerful. Jensen Huang says Moore’s Law (Intel founder Gordon Moore’s concept) that we can keep building ever more powerful and smaller chips may have reached its limit, or has definitely reached its limit. When in doubt, choose Jensen. So we bet on an unexpected turnaround and chose Intel. As always, I will buy slowly for the trust. We had just over half a position betting on ourselves that the company would report great numbers. We lost the bet. It’s spectacular, perhaps even better, and was up more than 10% in after-hours trading last Thursday. High fives all around. We had a good look. Now, I was off to Long Island to hold a rehearsal dinner for a special wedding. My son-in-law, Will Detweiler, who has been in the job for 21 years and raised many children, and his wonderful current wife, Caroline (Wynn With) Wilkie. It was a fairly extensive prelude to the actual incident, and was such a large rehearsal dinner that rehearsals for the rehearsal dinner were held. I wasn’t confident in the market, but I was confident in the market’s reaction to Intel. It’s a little difficult for a president to go on at length when he’s talking about saturation bombing of an insidious enemy that appears to have more missiles than its people. Who knows if we’ll be able to get oil to the point where we’re back to war, war, or bears can confidently talk about 5% to 6% inflation. My confidence was misplaced. Overall. I woke up early on Friday so as not to disturb the participants and took a call from American Express telling me it would cause selling pressure. And out of one retina I saw Intel trading at $106.30 instead of $109. Then $106.28, then $106.20, and the pace was sickeningly relentless. Few stands were built before market hours. Count the rise in several hands. After rising and changing to $103 just before Friday’s open, you can calculate how much it will fall during the day. I switched gears, went into thank my lucky stars mode, and left myself with plenty of room to buy. If you still want to do it. After a picture-perfect wedding, I was forced to attend nonstop, something I’ve only been able to do with one wedding and two Super Bowls. So what exactly happened here? Why did Intel close by nearly 8% on Friday? We’ve heard a lot of reasons for the technology sell-off. Most notably, the market has decided to exercise its power to block spending. With Alphabet’s stock now trading nearly 40 points below what once looked like a scary-value secondary stock (at least before we heard about Google raising its capital spending again), many sellers appear to have decided enough is enough. It does not reward spending with higher market capitalization. This was a cruel decision, especially considering Google Cloud had an incredibly strong quarter. Given my predilection for the Philadelphia 76ers in the wake of an interview with Thomas Kurian, head of Google Cloud and sometimes referred to as the LeBron James of technology, this statement took on new poignancy. That Alphabet phone call made me wish Ruth Porat was back in the CFO role, and I wish she had made the 80-plus phone calls explaining why expenses keep going up. Yes, “to meet demand,” as CFO Anat Ashkenazi has said multiple times while sprinkling positive comments about Alphabet’s balance sheet. I found both statements hurtful. That’s because Alphabet’s balance sheet, with its income statement showing negative cash flow, is no longer so glamorous. Okay, not great. If you owned Alphabet because of its rock-solid balance sheet and stock buybacks, you’d be thinking of a different, older Alphabet. But the real pain comes from this phrase we hear so often: “meeting demand.” I’m not sure Alphabet understands what “meeting demand” means in a market that’s starting to lose a lot of money on these hyperscalers. We are by no means looking for companies that “meet demand.” We want them to “make money”. I haven’t heard anything like that. The only company in Alphabet that makes money here is Apple. I’m tired of this “meeting demand” thing. It feels like the companies owned by the trust are losing wealth in everything they make, but are trying to make it back quantitatively. What has happened is that we have finally come to accept that these companies are not ignorant of what they are doing. Alphabet must feel it can make a profit by building it, but there are better stories out there. What do you need for this horror show of spending when you can support the club name Johnson & Johnson? The technology behind 3M’s materials science doesn’t have to cost billions of dollars. It just makes money. Again, I do not belong to the idea that the emperor has no clothes. One of these emperors will probably be human because it’s business-to-business (B2B) and we love the tenacity of B2B. The fickle nature of OpenAI’s business-to-consumer (B2C) paradigm is believed to be the cause of this store’s leaky bucket. Oh, and OpenAI has to go public in the worst possible way. It’s just a little bit, but thank you very much. In a sense, it doesn’t matter whether you sell the technology because you don’t see a profit in it, or because you think these companies lack discipline. Either way, hundreds of billions of dollars will flow out of this group. That’s why I’m wondering if this damn seller full steam ahead mentality is nearing its end. So let’s get back to Friday’s general selling. If you’re looking at Intel’s stock price falling, you could say it’s because the company didn’t get the price target increase that Wall Street analysts expected. They were disappointed. But AMD is meeting at the same time, and the only companies that can benefit Intel at the moment are Nvidia, which will be announced late next month, and AMD at the moment. AMD’s stock price also fell, but not as much as Intel’s and at a faster pace. The sales pitch was fierce. By the end of the session, with wedding flowers flowing, Phosforo (my wife’s agave spirits company) flowing, and toasts ready, we could almost see the sell shift from being about capital spending discipline to rumors of capital spending cuts from hyperscalers. Someone blinked. I’m not sure if anyone actually blinked. I think we’ll find out on Monday. But unless someone other than the suppliers comes out soon and says, “We’re beating the numbers,” I think we’re going to see $1 trillion worth of market capitalization jump out of the top of every giant company that’s still using AI. There are no illusions here. Tech companies can’t sustain sustained gains if their eyes bleed. There are too many companies desperate to see that happen. We know the Dow Jones Industrial Average has been able to resist and the S&P 500 has only slumped slightly, but the Nasdaq is back in charnel house mode. They say we continue to move away from traditional technology. We’re watching closely to see if someone projects a profit and it might lift all ships, but we recognize that, at least for now, there aren’t enough companies left that still need the hardware. Conclusion So why Intel? Because the only way to amortize all your spending right now is to use an AI agent, and the agent runs on Intel’s CPUs. And robots. We know that robots is such a huge market that it can’t be left to Tesla alone. I think all hyperscalers should offer them. These aren’t as much GPU products as the CPU market, they’re packed with robots. There are many use cases in and out of the data center right now, but I believe the next big wave of demand will come from robots from B2B to business to B2C. The current demand for all kinds of technology keeps the CPU going. But I think robots provide a use case where companies can maintain some demand for their stocks even after they embrace a prudent spending pause. I think we’ll find out soon if it’s just a rumor or an actual hiatus. In any case, keep moving towards new and different technologies while only supporting Intel when it comes to older technologies. Pauses can be very unpleasant, so you may not buy a pause device at first, but don’t get ahead of yourself. And remember, a beat-and-raise from just one hyperscaler changes the entire equation. And three more of those companies will report earnings this week: Amazon, Meta Platforms, and Microsoft. Oh yeah, our friends in the Apple-holding club are also reporting this week. (Jim Cramer’s charitable trusts are INTC, NVDA, AAPL, AMZN, META, MSFT; see here for a complete list of stocks.) As a subscriber to Jim Cramer’s CNBC Investment Club, you will receive trade alerts before Jim makes a trade. After Jim sends a trade alert, he waits 45 minutes before buying or selling stocks in his charitable trust’s portfolio. If Jim talks about a stock on CNBC TV, he will issue a trade alert and then wait 72 hours before executing the trade. The above investment club information is subject to our Terms of Use and Privacy Policy, as well as our disclaimer. No fiduciary duties or obligations exist or arise from your receipt of information provided in connection with the Investment Club. No specific results or benefits are guaranteed.
