원문 보기
When the nine suddenly spikes and high flying technology names start sliding into a weekend, Wall Street usually scrambles to find a scapegoat. CNBC anchor Leslie Picker sat down recently with Fundstrat's Tom Lee and market strategist Dan Greenhouse and Ed Clissold to debate whether this latest sell-off is the beginning of a genuine unwind in the artificial intelligence trade or just another panic-driven dip.
Tom Lee is treating this market drop as a coiled spring and where the real fault lines in tech infrastructure are actually starting to show. ; Let's bring in CNBC contributor Fundstrat's Tom Lee. So with us Dan Greenhouse and Ed Clissold of Ned Davis Research. Thank you all for being here. Um, so Tom, let's start with you.
You know, the market today kind of has the contours of an unwind here. You've got the VIX significantly higher as well. What do you make of the the market as as we've seen it today? Well, um, thanks for having me. You know, the market has been kind of climbing a wall of worry. AI fears escalated this week.
And I think because it's Friday and there's I think concerns that, you know, with oil rising, you know, could tensions or the war escalate over the weekend. So I think it makes not to doesn't surprise me that markets are unwinding today. But I think before viewers think that this is, you know, the start of a bigger downturn.
I mean, markets are at support. Uh, our tech technical strategist Mark Newton thinks we actually probably could rally. We either bottom today or on Monday. And I think from a fundamental perspective, I mean, as much as we should, you know, respect the idea that like AI debt markets uh, may not be that confident, you know, the fundamental story and the and the drivers of AI are intact.
So to me, when you have the fundamentals holding in and and markets getting fearful, it's actually a pretty good risk-reward setup for investors. So, you'd be, Tom, buying this pullback everywhere, or are there specific pockets that you feel like are kind of reaching that that technical area that makes them attractive for a bounce?
I think the setup is for a buy on the memory, the semis, you know, the the the the neo clouds, uh you know, tech broadly, crypto. I think that they've all pulled back to support levels at a time when people have become maybe very fearful of fundamentals, but as you know, if the fundamentals remain intact, then it's it's a coiled spring.
Dan, what do you make of of what's going on with the neo cloud issue right now? Do you feel like the market is unfairly punishing it, or do you think that it could be this canary in the coal mine, and that's why it's become such a a point of focus this week? The problem with the canary in the coal coal mine argument, Leslie, is how many canaries have we killed in the coal mine over the last 3 years as everybody has latched on to every new sign of weakness, so to speak, as evidence that the the bull market was about to end, only to see the broad market, both tech and otherwise, go on to make new highs.
So, I don't know that I would read any more into what's happening now than I would have or did any of the other headlines. The issues appear to me to be rates are higher, and so to the extent that everyone's already nervous about the extent of the AI buildout and it getting over its skis, everybody is starting to knock on the doors of some of the riskier parts of the buildout, and and CoreWeave and the neo clouds are on the riskier end of the spectrum, and even within that context, what's happening at at uh CoreWeave is they're not knocking off the secured debt, they're knocking off the unsecured debt, which is the riskier of the riskier AI buildout play.
So, I think you have an issue here where people are looking to sell something in the face of higher rates, but we heard from ASML, we heard from TSMC, we've heard from any number of other companies that the broader demand environment remains pretty robust. And I I would want to hear from CoreWeave that there were some issues with their business before I I read anything larger into what's happening right now.
But, Dan, demand is one thing, capital is another. And And so far, you've seen the capital chasing demand. And I wonder, as interest rates rise, as there becomes more attention on what's going on with credit right now, do those those two elements kind of bifurcate a bit? Yeah, so I'm making two arguments.
One is rates are higher, no doubt, obviously. And the issue is do I want to believe that everybody in the in the build-out chain is able to finance themselves ad infinitum at higher rates. And the answer to that is probably no. I don't know whether this is the breaking point or or another 25 or 50 basis points higher is the breaking point.
But, as rates move higher, the more speculative tails, so to speak, of the build-out are going to be called into question. And right now, that's CoreWeave. The point I'm making about demand is for the broader market, for investors out there who aren't long Nebius or CoreWeave or and are are are levered, so to speak, to the broader AI build-out, I don't know that you could read anything to what's happening right now in a larger context either for tech in general, although obviously selling off, or the market more generally.
I'm just saying rates are higher, riskier ends of the spectrum are getting knocked, and right now, that's the Neoclouds. Yeah. Ed, what do you think? Is there a capital issue right now? And And how big of an issue is that for the broader market? I mean, you have the cost of financing going up for a lot of these names.
You've got IPOs becoming more challenged with a lot of the data center exposed ones getting pushed out. You're not seeing as much in terms of equity capital fundraising, either. So, you know, before we were talking all about permitting being the biggest hindrance to the the CapEx trend in the AI build out, it feels like capital's kind of moving up the chain of priority of things that are becoming more of a a thwarting to this this build out that we're we're, you know, so much of the economy is pinned on.
So, so capital has a price and and the market's trying to figure that out right now and it's quite possible that you you could see a bit of a slow down in the the build out. Now, there's so much demand still. We're going to have to see what the next earning season looks like and so we're going to have to wait a couple weeks to get a really good answer.
I think that's part of the reason the market's selling off today is it is it you don't know and and so you have to wait until we get those. The last couple earning seasons have really saved the day where the numbers have been so strong and the beats have been so great. I mean, last earning season S&P 500 expectations were for a pretty good 25% earnings growth and they came in at 32.
And here we are again around 25, 26% expected earnings growth for Q3. We'll see if they can if they can beat there. And so we're going to have to just be a little patient here. The market typically is weak in September. We got that with small caps. We didn't get that with the large caps. So, maybe this is a little bit of a catch up from what we normally would have seen last month.
Lee did not flinch. His take was that investors was simply climbing another wall of worry using Friday afternoon to hedge against weekend escalation in energy prices and international conflict. But, rather than conceding that the artificial intelligence trade had topped out, Lee pointed to the technical framework laid out by his firm's chief technical strategist, Mark Newton.
Newton's models indicated that the broader market was sitting directly on critical technical support making a decisive bottom likely between that very session and the following Monday morning. Lee argued that while the bond market might be showing jitters around technology debt, the actual fundamental demand driving the computing revolution has not budged.
When the underlying business drivers remain fully intact while market sentiment turns outright fearful, you are almost always handed an asymmetric risk-reward opportunity. My take on Lee's initial reaction is that he is diagnosing an emotional liquidation rather than an institutional exit. When you manage risk professionally, you do not hold unhedged high beta positions over a weekend if you suspect oil could surge 10% before Monday's opening bell.
Traders dump what is liquid, and that invariably means selling off high multiple winners. But there is a profound difference between a portfolio manager trimming risk to survive 48 hours of headlines and a permanent structural impairment to corporate earnings. Lee understands that liquidity flushes frequently masquerade as systemic shifts, which is why he refuses to mistake a short-term panic for a secular turn.
Picker immediately pushed him further, asking whether he was recommending that viewers blindly buy the entire pullback across the board or if he was hunting in very specific corners of the market. She wanted to know if this was a rising tide environment or a selective tactical trade. Lee responded by zeroing in on a handful of high beta sectors that have absorbed the brunt of the selling.
Semiconductor manufacturers, memory chip providers, specialized cloud infrastructure firms often called neo clouds, and broad technology alongside crypto assets. In his view, all of these groups had simultaneously retreated into major technical support zones at the exact moment investors decided to question their underlying fundamentals.
Because he believes the core growth trajectory has not altered, he characterized the entire complex as a coiled spring waiting to release upward once the force selling exhausts itself. Where I find Lee's stance compelling is his willingness to lean into the most heavily shorted and volatile segments of the computing chain.
Take memory and semiconductors. These are not speculative concept stocks. They produce the physical silicon required to run massive language models. However, calling the Neo Clouds a coiled spring requires a bit more scrutiny. A Neo Cloud is essentially a specialized data center operator that borrows aggressively to acquire high-end graphics processing units and leases that computational power back to enterprise clients.
It is a capital-intensive, highly leveraged business model. When interest rates are low and liquidity is infinite, it prints money. But when rates stay elevated, the interest expense on that debt can eat away at the operational margin faster than the market expects. If you're enjoying this so far, hit like and subscribe to the channel and let's keep going.
That vulnerability is precisely what prompted to pivot the panel, turning directly to Dan Greenhouse to interrogate the Neo Cloud issue. She asked point-blank whether the financial turbulence hitting these specialized infrastructure firms should be viewed as an unfair market overreaction. Or if it is the proverbial canary in the coal mine signaling broader distress across the entire technology complex.
Greenhouse met the canary metaphor with open skepticism. He pointed out that Wall Street has spent the past 3 years identifying supposed canaries in the coal mine, proclaiming the death of the bull market at every quarterly tremor, only to watch the benchmark indices roar back to all-time highs. But he did not dismiss the underlying mechanics entirely.
Greenhouse conceded that the problem ultimately comes down to the cost of money. Because benchmark interest rates have climbed and remained elevated, nervous investors are scrutinizing every link in the capital expenditure chain. When anxiety rises, market participants naturally start pounding on the weakest, most speculative doors, and the private neo cloud players sit right at the far end of that risk spectrum.
aara. Narrata. t;> . What made Greenhouse's analysis so sharp was his breakdown of where the pain is actually landing. Looking at the balance sheets of these infrastructure companies, the distress is not hitting their secured debt, which is backed by the physical hardware sitting in server racks. Instead, the selling pressure is concentrated almost entirely in their unsecured paper.
Unsecured debt represents the riskiest tranche of an already risky buildout, making it the very first asset class traders liquidate when liquidity dries up. At the same time, Greenhouse highlighted that bellwether equipment suppliers like ASML and Taiwan Semiconductor Manufacturing Company have reported exceptionally robust global demand.
Until the core operators report that clients are canceling compute orders, he refuses to extrapolate stress in speculative credit into an economy-wide crisis. My take here is that Greenhouse correctly separates balance sheet engineering from end user demand. If an over-leveraged infrastructure provider runs into trouble servicing high-yield notes, that is a capital structure problem, not an indictment of artificial intelligence adoption.
The physical servers do not stop processing queries just because the debt used to purchase them trades at a discount. In fact, if a speculative provider stumbles, the hyperscalers, the trillion-dollar balance sheets like Microsoft, Alphabet, and Amazon are waiting in the wings to absorb that capacity.
The computing power still gets consumed. It simply migrates to entities that do not rely on high-yield debt markets to keep the lights on. Picker was not entirely satisfied with that answer, and she pushed Greenhouse on the tension between sheer demand and available capital. She argued that demand is one thing, but capital is what enables companies to chase that demand.
She wanted to know if rising interest rates and mounting scrutiny in credit markets are creating a genuine wedge, bifurcating the appetite for technology from the financial ability to build it out. Greenhouse acknowledged the validity of the challenge. He conceded that no industry can finance an unprecedented infrastructure build out ad infinitum if borrowing costs continue to creep upward.
Whether the market's breaking point lies at current yields or requires another 25 to 50 basis points of tightening is something nobody can predict with certainty. But he reinforced that as capital becomes more expensive, the speculative tails of the market will naturally get cut off. His core thesis remains steady.
If you are an investor looking at the broader market and you are not holding leveraged speculative credit instruments, the localized repricing in private cloud debt does not warrant dumping your entire equity portfolio. This is a critical distinction that casual market observers often miss. When interest rates rise, capital does not vanish.
It becomes discerning. The era of cheap funding allowed secondary players to raise billions with minimal covenant protection. Now, as the price of leverage normalizes, the cost of capital separates sustainable business models from speculative experiments. That is healthy for the long-term health of an industry, even if it creates painful volatility in the short term.
The speculative fringe suffers, but the foundational layer of the economy continues to march forward. Picker then broadened the scope of the conversation, turning to the panel to ask how deep the capital constraints might run. She observed that the primary obstacle to the artificial intelligence boom appears to have shifted.
Months ago, the conversation revolved almost entirely around physical bottlenecks, like delays in securing utility power and navigating local municipal permitting. Today, the conversation is increasingly focused on the tightening cost of capital, marked by stalled initial public offerings for data center operators, and a noticeable slowdown in equity fundraising.
Clissold weighed in to summarize the core reality. Capital has a price, and the broader market is currently in the middle of a messy price discovery process. He pointed out that it is entirely plausible that we see a measurable deceleration in physical construction timelines as companies recalculate their financing assumptions.
Yet, the saving grace throughout this entire cycle has been corporate earnings. Clissold reminded the desk that heading into the previous earning season, analysts had modeled a healthy 25% profit growth for key technology names. And the reported numbers blew past expectations to deliver 32%. With current quarterly estimates once again pegged around 25 to 26%, the upcoming earnings prints represent the real arbiter of whether this expansion can withstand higher capital costs.
Where I land on this entire exchange is that the panel laid bare the real battleground for the months ahead. The question is no longer whether generative technology is useful or whether enterprises want to deploy it. That argument has been settled by the sustained capital spending of the world's largest corporations.
The genuine issue is whether the earnings generated by this infrastructure can outpace the rising cost of the capital required to build it. If corporate America continues to report earnings growth above 30%, the elevated cost of debt becomes entirely manageable. The cash flow simply covers the interest expense.
But if revenue growth slows down toward the baseline cost of capital, the market's tolerance for high valuations will compress rapidly. Watching the panel work through this dynamic makes one thing clear. The current market volatility is not evidence of a structural collapse. It is the friction that inevitably occurs when an aggressive debt-fueled expansion collides with higher borrowing costs and geopolitical nerves.
The speculative edges of the market are being forced to clean up their balance sheets, while the heavyweights with fortress financials continue to deploy capital into essential hardware. As upcoming earnings reports roll in, the market will finally get hard data on whether enterprise cash flows are expanding fast enough to justify the sheer scale of this infrastructure buildout.
That realization is the ultimate stress test for this bull market. If those margins hold, the current dip will look like a brief pause in a much larger narrative. But if they crack, the conversation in the boardroom will shift from expansion to survival. Either way, the next few weeks will be the final word on whether the AI trade is a structural shift or a temporary overshoot.
I'll be watching the guidance closely to see who is actually making money and who is just spending it. Thanks for watching. Leave your take in the comments. .
Comments 0