Closing Bell - Closing Bell Overtime: Strategy's Michael Saylor; Is AI spending one giant circle? 9/23/25

Episode Date: September 23, 2025

Wells Fargo’s Ohsung Kwon joins to analyze the market action and what he likes into year-end in his first interview since taking over as Chief Equity Strategy. Our Angelica Peebles breaks down Kenvue’s volatile two-day stretch. William Kerwin of Morningstar reacts to Micron earnings.Michael Saylor, Strategy Executive Chairman, breaks down bitcoin's recent price action and the trend of Bitcoin treasury companies. Plexo Capital's Lo Toney breaks down who is spending and who is buying in the AI infrastructure buildout.    Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

The Dow breaking a four-day wind streak. Energy and real estate, the winners, consumer discretionary, and technology lags. Some moves in commodities today with gold hitting an all-time intraday high, silver touching 2011 highs, and platinum hitting its highest levels since July 21st. Crude, higher for the second day, up 2%. Well, that is the scorecard on Wall Street. Welcome to closing bell overtime. I'm Morgan Brennan, along with John Fort.

Ahead, Fed Chair Powell says consumer spending has slowed. Uncertainty is weighing on the business outlook and economic growth has moderated. So why is the market hovering your all-time highs? We're going to discuss that. Plus, Micron set to report earnings in just a few minutes. Will the company's results justify its 53% run in the past two months? And Strategy Chairman Michael Saylor, his take on the move in crypto, specifically Bitcoin, the future of a Bitcoin Reserve, strategy's recent stock volatility and more.

Well, let's start with Christina Parts of Nevelis, tracking the biggest moves today. Christina. We're seeing somewhat mixed signals across the market today as AI. The A.A. Tread specifically cooled after days of record levels. Let's start with Vertive Holdings, tumbling about 7% after Microsoft unveiled breakthrough cooling technology that removes heat more efficiently than current systems. Eaton also fell on disruption fears from that cooling innovation. In terms of other big AI names, NVIDIA, Oracle, closing in the red, really weighing on the S&P 500.

But, NVIDIA did hit an intraday high just yesterday, so context is everything. Not all tech, though, did sell off. Quantum computing really bucked the trend with IONQ jumping about almost 6% or I should say 4.5% to a fresh 52-week high after announcing a quantum internet breakthrough that could connect quantum computers over long distances. We're getting computing, I should say, also surging 18% in sympathy. And then energy was today's standout sector with high. Aliburton jumping about 7% its biggest gains since April, Exxon and Chevron rose. Chevron almost 1%.

Exxon almost 2%. The oil rally really follows President Trump's U.N. speech, where he criticized renewable energy policies and really pushed Western Europe back towards fossil fuel. And so you really see the divergent moves showing how policy shifts and tech breakthroughs continue driving volatile sector rotation in today's markets. Yeah, he also made some pretty hawkish comments more so than in the past regarding Russia and the war in Ukraine. Christina Parts-Nevilus. Thank you. Now let's send it over to Rick Santelli for a check on the bond market. Rick. Well, this morning we had some data points out

and they weren't particularly strong. We had PMIs for service sector, manufacturing, and composite. And even though all three were above 50, none of them were sequentially higher, none of them were higher than expectations. That really sets the stage for a lot of the types of data we've had. Not too bad, but certainly not stellar. not falling apart. Look at a two-day chart of twos and tens together, and a couple of things should jump out at you. The fact is, we've seen bigger price appreciation on longer-dated on longer-dated treasury yields today, like tens, 20s, 30s, and that has pushed their yields down more than the short end, flattening the yield curve. And if you look at a chart since

the beginning of the month, there's something very important here. That's very easy. We saw the Fed lower rates on the 17th. You see the 17th there, a little past the midpoint. That is the point where rates started to go higher. So everything on the curve is actually higher in yield that won the Fed ease prior to the Fed ease. Why is that important? Because things like mortgage rates are considered really important to the U.S. economy. Maybe the Fed can't quite get at making those particular rates lower. And it isn't just here.

Let's add in the boon from September 1st. You can see the pattern's basically the same. So whether it's the boon, whether it's the guilt's in the U.S. whether it's JGB's Japanese government bonds, we see that the big economies of the globe all share a characteristic. And that characteristic is longer dated yields are stubbornly high because issuance is stubbornly high, which brings me to today we had a two-year auction. Didn't go particularly well, $69 billion. Tomorrow will be $70 billion, $5s, followed on Thursday by $44 billion, seven-year notes. Morgan, back to you. Rick Santelli, thank you. All right, let's focus even more deeply on

the markets. Major averages closing lower today as the AI trade loses some steam after a big run-up following major announcements across the industry. Despite some fears that the trade is now stretched, our next guest says to own AI and AI enablers like utilities. Joining us now is Osang Kwan, Chief Equity Strategist at Wells Fargo. And Osung, it's great to have you back here on set with us. Welcome. Thanks for hanging me back. All right. So what do you think of equities here? Because we've been hearing that valuations are very rich. And certainly it seems like there's a wall of worry for investors to climb. Yeah, we're still bullish equities. We think the market's going higher. We're targeting 7,200 on the SMP by year and next year. And if you think about the drivers

of equities, we identify four drivers, which we call the PRISM model. So PRSM, profits, rates, sentiment, and macro. So I just want to interrupt you real quick because we do have those micron numbers. Christina, parts and nevelace is ready with them. Christina. We are seeing a beat on the top and bottom line, 303 adjusted, which is higher than the estimates on $11.32 billion. And this earnings beat comes when estimates had already increased because the company pre-announced just at the beginning of August. So these are even stronger numbers than expected. For their Q4 non-gap gross margins, 45.7%, also higher. For the guide, there was a lot going on this guide. $3.75, I should say, adjusted for their Q on EPS guide. That is high.

higher than every estimate out there, according to street accounts. And then the Q1 Revenue Guide of $12.5 billion, also higher. And then the big one, too, is gross margins. They're guiding at 51.5. There was concerns that they'd be cutting prices, competition with Samsung, et cetera. And so that is why you're seeing shares up a little bit over 3.5% right now, teetering around there. Guys?

Christina, thanks. Sorry. So let's go back to Prism. Yeah. So Prism, PRS, profits, raise sentiment, and macro. The profit cycle is still very healthy. We're forecasting EPS growth of 11% this year and next, followed by another 12% growth in 2027.

Rates, basically neutral, slightly positive. Sentiment is neutral as well. Macro is negative, but has been improving for the fourth straight month. So if we combine all those four, we're still in a pretty healthy backdrop for the equity market. So we're forecasting about 9% annualized return for the next 12 months. How much does AI continue to drive this? And we just got these micron results,

beat across the board, stronger than expected guidance, after a strong pre-announcement, shares are higher. The name had already doubled before this report, and obviously we've been seeing this across the ecosystem. Yeah, I think this is a AI-led bull market, and I think this is likely to continue. First of all, it's not a bubble. The entire outperformance of the NASDAQ since the end of tech bubble has been driven by better fundamentals in the NASDAQ versus the NFDAQ And I think that's likely to continue. Second, we still think we are in the early innings of the AI investment cycle. So if you look at tech equipment spend, so basically semis and hardware, as percentage of total

U.S. GDP, it peaked at around 2.6% during the PC boom in the 1980s, 2.9% in 2000. Right now, it's only at 2.0%, which means that to get to that 2.9% in 2000, tech equipment span has to grow by 47% on top of GDP. growth, which is obviously very significant. And I think the way this plays out is, I think as long as the equity market continues to reward companies' CAP-X outlook and the growth outlook, this is likely to continue. So if a hyperscaler comes out next earning season and cuts CAP-X outlook for whatever reason, I think the stock goes down.

And if that's the case, companies are going to continue investing AI. So what could go wrong? clearly you're bullish 7,200 by year and next year. There are some things, we could have some choppiness, certainly in between. What could happen where you've got to have to rethink your thesis? Yeah, I think AI CapEx is probably the single biggest risk to the equity market. We don't see that turning anytime soon. Like I said, I think this investment cycle continues.

The other risk could be potential further slowdown in the economy. I think good news is good news right now. for equities. I think further rate drop from here would be bearish for equities. That's essentially how we're positioned. So we're overweight AI, sort of market weight defensives, and we're actually underway cyclicals. Well, this week's news certainly suggests that AI CAPX continues. Osang Kwan, thank you. Thank you. Well, meantime, Kenvue shares higher in today's session, making up some of the losses from yesterday's sharp drop. Our Angelica Peebles is with us here with more on today's move.

Angelica. Yeah, John, it's certainly been a wild couple days here. So Kennedy shares closing up nearly 2% today after falling 7% yesterday. And the stock's down about 15% from earlier this month when the Wall Street Journal first reported that the Trump administration was planning to link Tylenol use during pregnancy to autism. So the stock is up today, despite the president saying some version of don't take Tylenol a few dozen times during his announcement yesterday. And when you look at what the administration's really doing, it's recommending that pregnant women. take acetymenifin sparingly during pregnancy, which is actually in line with what the current guidance recommends. And in terms of regulatory changes, the administration is really just making

people aware of a possible link and one that's shaky at best. So medical groups like the one that represents OBGYNs standing by the use of acetaminopin during pregnancy. And the FDA itself pointing out that while there have been some studies suggesting an association, a causal relationship has not been established. And the FDA, including that statement in a letter it said, out to doctors. The agency is also saying that it's going to update the label of Tylenol and other acetaminofin containing products to warn of the possible link. So again, you know, you saw some strong rhetoric yesterday, but in terms of action, not as much there. So how different is this, Angelica, from the way the federal government has communicated

about these kinds of things before? It sounds like certain things about the guidance aren't changing, but we just had the president talking about something that a company, involved is disputing. You know, it's certainly different, right? You know, it's hard when you have the president of the United States saying one thing, but even his own medical advisor saying something else, right? So you have the president saying over and over, don't take Tylenol, don't take Tylenol. But then when you look at the guidance, it's not quite that strong.

And, you know, I saw some reporting today that that was not the plan necessarily. They did not want to have such a strong statement. But, you know, the president, of course, is going to say whatever he wants to say. But, you know, I really think it comes back to looking at what the data shows. When you talk about the fact that there hasn't been an established causal link, has anybody actually tried to establish that link and look at it directly? You really can't because it wouldn't be ethical. You know, I talked to some research who said they would love to run a study like that.

But, you know, how many people are really going to sign up for a trial that's looking at whether taking something causes a serious disease for your child, right? And so that's the problem. And, you know, nobody would conduct that study and how would you get people to participate? And that's really the unfortunate thing with everything. Everyone wants answers, right? Does doing this cause X? And it's hard because that is really what's going to give you the definitive answer of whether

something does or doesn't cause something, but it's just not feasible. And so that's why we have to rely on these observational studies, which aren't great. And so you can't say definitively one way or the other, and that leaves all of this confusion. and it's really, you know, confusing and frustrating for people who are forced to make decisions. For sure. Well, Angelica, thanks for giving it to a straight, Angelica peoples. Well, up next, much more on that move in Micron. We're going to hear from a top analyst with his first reaction and how he's navigating the chips right now. And later, Chattagy's Michael Saylor.

Standing by with his outlook for the crypto space as Bitcoin seems to be chopping sideways, at least right now, over time's back in two. Welcome back to overtime. We're getting some news out of Washington. Amon Jabbers is here with the details. Hi, Amon. Yeah, Morgan, this is a new reporting from Reuters just out within the past couple of minutes

on the Trump administration. According to Reuters here, seeking to take another equity stake in a major American company, in this case, a 10% stake in lithium Americas. According to Reuters, the Trump administration is renegotiating the company's $2.26 billion energy department loan to develop the Thacker Pass lithium project. And as part of that deal, the U.S. government is now, according to Reuters, seeking an equity stake in the company.

The Trump administration supports the project, and the project is seen as sort of one of the keynote ways for the United States to wean itself off of dependence on Chinese supplies for lithium, which obviously is so important in EVs, batteries, and other electronic companies. A quote here in Reuters from a White House official unnamed saying President Trump supports this project. He wants it to succeed and also be fair to taxpayers, but there's no such thing as free money. Morgan, back over you. All right. Amund Javers, thank you.

Lithium America is up more than 50% right now. But John, we should note that market cap as of the close was a little less than 750 million. This is a small cap company. So perhaps not surprising to see such a big move here in overtime. Yeah, it's a big move, especially at that market cap. Well, yeah, I mean, that was a huge move. Let's get to another check on Micron, fourth quarter numbers out moments ago. Stock's higher by about 2.5% as the company beats on EPS revenue and the Q1 Guide.

Joining us now is Morningstar Senior Research Analyst William Kerwin. He has a neutral rating for the stock. William, welcome. So is this perhaps sustainable AI momentum enough to get you off? off a neutral rating? Well, we'll have to see when we digest more of the release and the call tonight. But the strong guidance is really the story of the release so far here. You know, expectations were strong coming into the print, and the results were already

really pre-announced back in August when Micron raised its guidance mid-quarter. But this guidance blows the expectations out of the water, quite frankly, especially when you consider profitability. And really what we see this showing is that not only it's AI demand strong, but it remains significantly supply constrained, and that's really a boost for the pricing from a company like Micron and its high bandwidth memory chips. Well, it seems like that's the part that you've got to worry about if you're an investor here, right? Is the memory group more than any other I can think of is known in technology for boom bust, and there are questions right now about the sustainability of this AI-driven demand.

What will it take, given those questions, to convince you, and considering that this, the supply constraint conditions won't last forever. Well, the way we see it is that AI is driving a particularly strong and particularly enduring up cycle for memory chips like those from Micron. But it is not eliminating cyclicality in the long term in our view. So when we look over the course of five to 10 years, we expect those cycles to manifest, even as AI demand may remain strong over that time, the cyclicality results in a mismatch of supply and demand.

And that's a lot harder to forecast from the chip makers like Micron. But over the next two years, it looks pretty strong. I mean, stocks up more than 40% over the past month. It's basically doubled since the start of the year. Do you buy in at these levels? Do you wait for a pullback? Well, we think buying in at these levels is assuming even more upside from here. And we think a lot of that upside is already baked in.

You know, as I said earlier, this outlook for the first fiscal quarter is really exceptional and well above estimates coming in. But the relatively modest stock pop shows that I think a lot of that upside was already being baked into investors. And this really met maybe some of those by-side expectations, even if it exceeded the sell-side consensus estimates. All right. William Kerwin, thank you, on Micron.

And don't miss Micron CEO, Sanjay Marotra. That's tomorrow exclusively on CNBC, 9 a.m. Eastern Time. Well, coming up right here on overtime, investors don't seem to be playing defense. fence in this market. That could be leaving some without a safety net. Mike Santoli digs into that. Plus, Nvidia gives to OpenAI and Intel. Oracle teams up with Open AI. Corweave and Nvidia shake hands and Amazon and Anthropic signed deals. The list goes on and goes with what's starting to look like in AI circular money flow. Is this a reason to worry or has overcapacity become a thing not to worry so much about in AI? We're going to debate when overtime returns.

Welcome back. Take a look at Palantir. Today is the one-year anniversary of the company joining the S&P 500. Since its inclusion, it's up more than 370%. It's one of the best one-year returns for any stock in the index. It's also the third best-performing S&P stock in 2025. Now, the company's on track for its seventh straight quarter of double digit gains. Among the Mag 7, only Amazon matched that streak. Back in 1999, the stock has gotten expensive, though. It's got a forward PE of 240, and certainly that valuation has been a big talking point and continues to be for Palantir. Yeah, you love it if you own it. Well, despite rising chatter about caution, the market's still not leaning into traditional safety trades, but with signs of life in the VIX volatility index and defensive stocks near historic lows,

could that be starting to change? Well, let's bring in senior markets commentator Mike's in for Yeah, John, always hard to kind of pick up these tremors and decide if it's the start of something bigger. But just the setup here in terms of the S&P 500 exposures, the top line here is the aggregate for the four traditional defensive sectors. That would be health care, consumer staples, energy, and utilities is down around 19% of the total index. Two stocks alone, Nvidia and Microsoft are around called 14%. This is from Todd Stone over at Stategis, just kind of aggregating this stuff, just to highlight exactly how skewed the index has become in the direction of those huge growth stocks. You know, there was an earlier version of this where invidia was almost as much as all of the health care sector. Today was a countertrend day, by the way, all those defensive sectors were up and the mega cap tech was down.

Here's the volatility index, though. And, you know, it's hard to see here, but it's kind of got this very subtle pattern of kind of potentially basing. We were down on 14s not that long ago. We closed just below 17. As the S&P was making new highs, we were not making new lows in the VIC. So it could be some of this sort of seasonal traders picking up the possibility for a little bit of chop here as we have pent-up selling maybe unleashed a little bit in the broad market, John. Mike, it seems like we've been talking about different versions of this.

How long can the big names stay so big compared to the rest of the market for a long time? So historically speaking, is there a comparison? Is there an analog? Are there times where this just continues for a considerable amount of time? There are. I mean, I think you'd have to look at two very specific eras. One was the so-called nifty-50, which was kind of late 60s into the early 70s. It was multiple years. Very expensive group of stocks. The thing is it wasn't just seven. It was more like a few dozen stocks of that type. And then, of course, the late 99, 2000 period, when you did also. get almost this concentrated. The difference here is that it's, again, just that six or seven stocks covering a third of the S&P 500. The rule is when the fundamentals or the macro break, that's when the trend breaks. Usually it's not happening on the fly just because investors decide to rebalance and rotate around. It happens because either there's a massive disjunction in the earnings expectations for those big stocks or we get a macro shock. All right. Mike Santoli,

Thank you. It's time now for a CNBC News update with Bertha Coombs. Hi, Bertha. Hey, Morgan. The man accused of trying to assassinate then presidential candidate Donald Trump last year on a golf course tried to stab himself in the neck with a pen just after he was found guilty on all five counts against him this afternoon. Ryan Ruth did not appear to hurt himself in any way and was back in the courtroom a short time later in handcuffs to complete the proceedings. Ruth faces life in prison. House and Senate Democrats opened investigations into the Trump administration's decision to close an FBI probe into border czar Tom Homan. Last year, Homan was recorded accepting $50,000 in cash from FBI agents who posed as businessmen looking to get border security contracts. And Major League Baseball has approved the use of robotic umpires for the 20th, 26th season. The system uses technology similar to the line calling system in tennis. Human umpires will still be used to call balls and strikes, but teams can make two challenges

per game in which the robot's reviews will be shown on the video boards. Very interesting. Back over to you, Morgan. All right, Bertha Coombs, thank you. Up next, Strategies Michael Saylor, breaks down his forecast for the crypto space. His recent meeting, with lawmakers as industry advocates for a strategic Bitcoin reserve. We have a lot to talk to him about. Over time. We'll be right back. Welcome back to overtime. Check out the price of Bitcoin. It's down to about $11,000 after hitting nearly $118,000 after the Fed cut rates last week. Despite the recent volatility, it's up nearly 30% since Trump's tariffs announcement back in April. There 180 public companies in the world now with Bitcoin on their balance sheet. That's according to

Bitcoin Treasuries.net. Strategy is the largest, the most prominent of these so-called Bitcoin strategies, really sort of coined the concept. The company was one of a number of industry leaders in D.C. last week as well, pushing for the strategic Bitcoin Reserve bill. So joining us now is Michael Saylor, Strategies Executive Chairman. Michael, it's great to speak with you. Yeah, thanks for having me, Morgan. Okay, there's a lot to get to. But first, I do do want to start with the fact that there are reports that you were in Washington last week meeting with policymakers about this strategic Bitcoin Reserve bill. How are you advising lawmakers right now and why is the case there to create a strategic reserve? Well, I think Bitcoin's

good for the nation and the country would benefit. We did buy something like 78% of the entire United States for about $40 million over the past 100 years or so. So I think the United States ought to own a large part of cyberspace, and Bitcoin is the next frontier. What does it do to prices, especially as they've been chopping sideways, even coming off a little bit recently? You know, I think the real story in Bitcoin right now is corporate adoption. We've got a consistent acquisition of Bitcoin by the big ETFs, like Black,

And that's taking up all the natural supply, and they're doing that on behalf of institutional investors. But these 180 companies that are capitalizing on Bitcoin are buying even more than a natural supply being created by the miners. And they're putting upward pressure on the price. And I think that as we work through the resistance of late and some macro headwinds, we'll actually see Bitcoin start to move up smartly again toward the end of the year. At the beginning of this year, you and I in January sat down, had an hour-plus conversation at the ICR conference about this very idea, Bitcoin Treasury, what you were doing at strategy, and where it made sense for others to potentially adopt a similar concept as well.

What are the types of companies that you're seeing lean into this? And at what point do you think it starts to accelerate if it does? Well, I think you've got two types of companies. You've got operating companies that would otherwise be returning their capital via dividends and buybacks, or they would just be holding low yielding money markets. And they're choosing Bitcoin as a Treasury Reserve asset, so that actually improves their capital structure. It strengthens those companies. There's a lot of those.

And then there's a second sort of company, the true Treasury companies. They're capitalizing on Bitcoin. Bitcoin is digital gold. The world ran on gold-back credit for 300 years. The world's going to run on digital gold-back credit for the next 300 years. So treasury companies, in essence, are holding digital capital and creating digital credit instruments on top of that digital capital. And there's, of course, a huge demand for equity and credit instruments

in traditional capital markets. And Bitcoin is emerging as the ideal form of digital capital to back those instruments. Yeah, and I want to dig into that a little deeper here, Michael. But first, I mean, gold's having a huge run right now, too. It's $3,800 an ounce. It's basically trading at record after record this year. But Bitcoin hasn't sort of moved lock steps. So how to think about that analogy in this environment?

J.P. Morgan said gold is money. Everything else is credit. This morning I said Bitcoin is money. Everything else is credit. The success of gold is indicative of the demand of mainstream investors is to have a bearer instrument that's a long-term store of value that's not a currency derivative and doesn't have counterparty risk to a corporation. Gold fits that bill. Bitcoin fits that bill even better and it's got the digital upside. You know, you can't teleport gold and you can't

program gold to vibrate it a million times a second on a computer. So Bitcoin is the technology version of gold. It's digital gold. It's going to be, in my opinion, 10x bigger than gold. But they're the same idea. How do you compare that to Ethereum, which investors seem to be very hot on? And we're starting to see ether treasuries built out, company, you know, holding companies built out as well, like Bitmine. And then you've got stable coins as well, which seemed to be surging. Just a few minutes ago, headlines that tether are seeking between $15 billion and $20 billion in exchange for roughly 3% stake through a private placement, for example. You know, Bitcoin has emerged as digital gold. There's consensus. What do you want to do with gold?

gold, you want to issue credit, you issue credit against the underlying capital asset. It's going to be the primary monetary crypto. Ethereum, Solana, the other tokens, they're emerging as sources of digital finance. Their use cases is networks that circulates and tokenize stable coins, tokenized currencies, tokenized securities, tokenize real world assets, and implement defy. And so the future of those networks is going to be determined by technology, by regulation, by competition, and, of course, by the execution of those management teams. How should investors think about navigating an investment in strategy specifically, which

I know you've described as a levered bet on Bitcoin versus the other BTC-backed fixed income instruments that you've launched in recent months? Yeah. The credit instruments we're offering, like Stretch, are meant to compete with short-term money market funds or other sorts of fixed-income instruments. So if you're a fixed-income investor and you believe in Bitcoin and you're interested in something that has higher yield or perhaps longer duration, then you would look at one of our credit instruments. If, on the other hand, And if you're a Bitcoin maximalist and you actually want more Bitcoin, then our equity is actually got credit amplification in it.

So the equity is amplified Bitcoin. So people that want more volatility, more upside than Bitcoin will go for the equity. People that are interested in having the volatility and the risk stripped away will look at the credit. Okay. How do you ensure you're not over your skis with any of these instruments if Bitcoin plunges again? Well, we've got about $70 to $75 billion of Bitcoin. We've only got about $6 billion of these preferred credit instruments outstanding. So it's not even 10%. And of course,

the preferreds never come due. And so they're not really leverage. I could say that we've got a PhD in leverage at strategy. We live through the crypto winter. So we've constructed a balance sheet that could handle any conceivable drawdown and continue to perform. So how do you respond to some of the critics out there? Jim Chano's a couple days ago saying that your tactics to boost returns are quote-unquote financial gibberish. As I would say, for 300 years, the entire Western world ran on gold-backed bonds. What we're doing is distilling yield, duration, you know, and we're offering those in digital, credit instruments. We're stripping away the volatility. We're stripping away the risk. And there's a

huge number of people in the world that want some sort of digital credit instrument. The ones we're offering are higher yielding. They may be more liquid and they may be longer duration than all the comparable credit instruments out there. The only way to create them is using digital capital and Bitcoin is digital capital. I think the short sellers don't understand the use case of digital credit built on digital capital because they've never seen it before in the history of the capital markets. Well, we appreciate you coming on to talk about it. Michael Saylor of Strategy. Great to have you. Thank you. Well, I had tens of billions of dollars in AI investments are announced seemingly every day, everything from infrastructure funding

to chip funding. But is there enough revenue to finance this? And is there a risk of overbuilding? That's next. Welcome back to Overtime. Is this AI data center spending spree a visionary investment ahead of a generational payoff or is it a circular boondoggle that will end in tears? Let's look at two examples from tech history that illustrate the possibilities. Exhibit A, the dot-com dump. In the 1990s, as the consumer internet was taking shape, telecom providers like WorldCom and Global Cross spent hundreds of billions of dollars on servers and networking gear, anticipating a revolution

that would change global business. Venture capitalist funded startups that bought internet equipment without the sustainable revenue to justify the cost. It was too much, too soon. Spending on technology equipment peaked in 2000, along with the NASDAQ, and while the dot-com revolution did arrive, many startups, telcos, and even equipment makers like Sun Microsystems didn't get to enjoy it. They collapsed and got bought out by other firms. But the AI race doesn't have to end that way. Let's look at exhibit B, e-tail elation. In the throes of the same dot-com bus that felled some telcos and equipment makers, mainstream investors soured on the very idea of massive infrastructure investment, so they hated Amazon. It was spending hand over fist on a global

network of e-commerce distribution centers, seemingly duplicating the work of UPS, FedEx, and the Postal Service. Meanwhile, eBay and others avoided those costs, riding on top of existing infrastructure, but this time, infrastructure spending paid off. Amazon investment enabled the broad selection and fast delivery and its revenue surge through dot com winter justified the bet. So will today's AI spending and buildouts from the likes of open AI, Oracle and NVIDIA, and Corweave end up more like dot com dump or etail elation? Let's bring in Plexo Capital founding managing partner and CNBC contributor, Lowe Tony, Lowe. Thanks for having me. Great to have you here. Which is it? I love the old school references because I've got

enough gray hair like we talked about to actually have been there. I had hair before the dot-com bust. It's a little known fact. I cut all mine off in anticipation. I think what we're going to see is a little bit of both, because there's elements of both, right? So we had the vendor finance equipment purchasing that happened during the dot-com bust. And unfortunately, that fiber stayed dark for a long period of time, and it went past WorldCom's ability to be able to benefit from it. So, you know, Verizon, level three, they picked that up for pennies on the dollar because ultimately that demand did come and that became the core part of the infrastructure. So let's look at Amazon. Amazon did spend hand over fist and many did not understand why

they were doing that. Couldn't believe they were doing that. But ultimately that bet paid off initially for Amazon selling their own products and then ultimately opening it up through their marketplace so that sellers could build and sell on or could sell on top of that infrastructure without having to build it. That was amazing, right? I get the demand from Amazon's marketplace, and I get the fulfillment from the distribution centers that they built, not even to mention what happened long term with AWS. So there's good analogies there, right?

Because you can almost think of OpenAI doing something similar with Invidia. Invidia almost becomes kind of like the distribution centers, right? They want to make sure that there's enough demand inside of that compute, and then you can look at chat GPT as being a way that OpenAI is able to monetize from the consumer. They also sell to the enterprise, but guess what? They also have the same analogy back to eBay with people that are building applications that use the open AI infrastructure, both the compute from Nvidia, as well as the access to the APIs to power their own applications without having to be. build that out. So I think about also on a smaller scale Apple. If you look at the amount that they spent on equipment rollout to build iPhones, it was huge and it was unprecedented at the time, but the iPhone also turned out to be an unprecedented hit from a single company as far as

what people are willing to spend on smartphones. So how much of this depends on the size and sustainability of the company that's making the bet on AI here? Are the hypers differently positioned from companies like a core weave or other smaller companies that are saying, hey, we're beneficiaries of this? Yeah, I think you can look at the hyperscalers a couple of ways. The hyperscalers themselves, as you know, are also trying to build out their chips to lessen that dependency that everyone seems to have on NVIDIA. But we're seeing an unprecedented scale. We've never seen this type of scale before for this kind of hybrid vendor financing equity model. So I think it's even hard for the hyperscalers, which have to also finance the capital needs across a multiple lines of business to even look at this deal the same way.

I mean, what this deal really does is it positions InVIDIA as the core infrastructure provider. So what cements them long term to be part of that AI tech stack, if you think about three layers, the infrastructure layer, the model layer, and the application layer. So whether or not, going back to the comparison of the dot-com days, invidia will have the benefit of being able to wait for that dark fiber to light up, the compute to light up. That said, I believe the demand is already there. Or is this the data center version of WeWork? Meaning right now you've got Class A office that's doing quite well, even post-pandemic. But during that period, WeWork's whole thesis of, hey, we're going to lease this space and then sublet it. It didn't work out, right?

Could it be that there are some players in this AI game right now making commitments, expecting business that if there's a valley of doubt or a period of uncertainty, even in an overall positive AI rollout, they're going to be in trouble. I think we're so early in the innings of this baseball game to use that analogy, we're in the first or second inning. And when we think about where the value is going to recruit next long term, again, with those three layers, the infrastructure layer, I think, Nvidia is showing they're trying to lock that in. The models, you've got open AI, you've got Anthropic.

The application layer is where we believe that there's going to be the most innovation and the value that still remains to be captured at scale. And so we're just starting to scratch the surface with companies like perplexity that are building on top of that infrastructure. So I think the demand is already there based on what we see from the users of the API and the users on the enterprise and the consumer side. But I think we are positioned for an explosion on the application layer.

How does it change the investment landscape when you have Nvidia making all of these types of big bets investments into other companies that ultimately potentially become customers of invidia they're not the only ones i mean core we've announced a similar strategy a couple of weeks ago we've seen it uh across the ecosystem aerospace has been doing this for forever but i just wonder what it means uh when you see these types of deals structured this way and gaining momentum here well i think it means a couple of things with ai in particular number one it reflects the capital needs they're so massive we've just never seen investments done at this scale before and so you

a player like NVIDIA which has both the capital and the expectations and they're locking in a lot of that demand long term. But then I think what we also have is we also just are seeing that we may not still fully understand what the demand potential is going to be. And I think it's going to be massive. But coming back to what this means to the ecosystem, I think that even though this aligns very well with the administration's goal of having AI deal done within the United States so it aligns perfectly with that however I think the regulators are still taking a look I ultimately don't think it will go anywhere but it doesn't portend well for all of the folks especially at that

infrastructure layer that really want to compete against invidia because again I believe this truly cements invidia as being the core of the infrastructure layer all right low Tony great to get your thoughts today thank you and to have you here on set appreciate it well coming up Mining on the moon. Yes, you heard that correctly. Sounds like something from a sci-fi movie, but it may be closer than you think.

And a crucial piece of this massive computer build-out. That's next. Up next, we'll get you set for the trading day ahead, your Wall Street rundown coming up after this quick break. Welcome back. Helium 3 is a resource critical for quantum computing and other futuristic technologies like fusion energy. It's extremely rare on Earth. It's typically a residue from nuclear reactors, but it's abundant on the moon. That's why Blue Force has struck a 10-year, $300 million deal with moon mining startup, Interloon.

Keeleon 3, the commercial market price is about $20 million a kilogram or just under $3,000 per liter. So it is, it still remains the... the only element in the universe that's priced high enough to warrant going to space and bringing it back to Earth. The unit economics of doing it, you know, we project into the future when we have vehicles like Starship and Blue Origins New Glen. We project the unit economics to work out very favorably. Now, Interlun was co-founded by Rob Meyerson, the former president of Blue Origin, alongside James, Mantefev. Blue Force is a manufacturer of dilution refrigerators for Quantum. Chief Business Development Officer David Gunnerson thinks the promise of Quantum begins to be realized,

thanks in part to IBM, Alphabet, Microsoft, even Honeywell, and that demand for infrastructure, and thus Helium 3 will ramp in coming years. I think it's not the numbers that we have in AI today, but I think the whole promise with Quantum is that you should need a smaller infrastructure to sort of, actually solve certain problems that actually people are not solving today because they are not possible to actually solve on traditional high performance computers or with AI. So I think it's not only going to replace traditional computing on old problems, it's also going to let us actually solve new problems that we never thought we could actually use computational power to do. And those

infrastructures, I think they will start modest and it might also need quite a lot of support from traditional computation to actually perform the task. But they are like a booster to the computation we do today. So Interloon will begin delivery in 2028. For more, check out Manifest space. That's going to do it, though, for us here at overtime. Fast money starts now.