The Most Hated Sector Is Where the Value Is

Peter Boockvar (0:00)

It's not a question of if it hits a wall, it's a question of when it hits the wall, because all you need is a 25, 30% order rate to go to 10 to 15, which is still growth in the double digits, but your stock price can fall dramatically. The market is pricing in, saying that over the next five years, there's a 55% chance that you're going to default. If you slow the pace of construction, then the whole picks and shovels trade falls apart. This is gut check time for this trade.

I think anybody that's gotten over their skis with it, needs to pull it back.

George Noble (0:45)

Hey, Peter, how are you?

Peter Boockvar (0:47)

George, good to see you.

George Noble (0:48)

Good to see you. It's been a while.

Lots going on. Every day I wake up, I have no idea what the hot topic of the show is going to be. And so today is Thursday. Who knows, we'll do something to do this tomorrow. Something else is going to be happening. But there are a few big things, things that need to be addressed. You and I were talking on the fly a little before we started. So let's get into it. What's top of mind for you, Peter?

Peter Boockvar (1:13)

Well, to me, it's right now, it's the two biggest things is the AI trade, of course.

And not just the trade, but the economic reliance, I should say reliance, the reliance that the US economy has on it. And of course, interest rates with the global rise. And last couple of past week, a little bit of a pullback in yields. But those are the two big picture things, of course.

George Noble (1:42)

Great minds think alike. So let's start with the AI trade only because we had Meta yesterday, we had Tesla, we had SpaceX. Core-weave bonds go down every day, Oracle CDS goes up every day.

We got a situation here. So what's your take?

Peter Boockvar (2:05)

So I try to break it up into a couple of things. First is, of course, the actual infrastructure build of these facilities that create the computing power that these model makers use and to hope for, sell it to the rest of us.

So on the infrastructure side, we know all the numbers, we all see them, the extraordinary nature of it in terms of size relative to the economy, the heavy contribution to the stock market and the economy at the same time. And any prudent investor, when you have a cycle that is as enormous as this and how it's driven GDP growth and how it's driven the stock market, everyone should basically have a dashboard of things that can say, okay, what can disrupt this?

Because this trade, this thing, it's not a question of if it hits a wall, it's a question of when it hits the wall. And it's important to look for the signs that tip you off. Then of course, on the other side, we have the users of the technology. And what we've seen so far is that, I think, there was a stat that I saw yesterday that 80% of the users make up 1% of... No, the stat was, most of the people that are using AI are a very tiny cohort. And a lot of it is VC-funded companies that are using 80% or paying for 80% of the models. So it was maybe 1% of the users paying 80% of the models.

George Noble (3:55)

I saw the same. Yep. The other thing is, we're doing a conference. I want to mention it again. October 27th, 28th, Tuesday and Wednesday 27th, 28th. 20 of the best technicians on the street. Sean Rowe, Dave Mikoski, Brian Shannon, Rick Benson, and you are. Just go down the list. 20 of them.

Each coming with three ideas, you're going to get 60 ideas over three days, 27th, 28th, 10th, to 4th each day.

Replays will be available immediately if you can't make it. This is going to be the best conference ever. Thousands have attended these conferences.

Peter Boockvar (4:34)

So I know over time, this iteration of AI, because I want to highlight iteration because AI has been around for a while. It's this form of it. We'll be game changing in many ways, and we'll all use it.

But that is what is going forward. But right now, so much is dependent on the actual construction of these data centers. So we've had a couple of things here, and we know that the initial money tree that financed this came from the free cashflow of the hyperscalers. Then that tree became bare, and they said, okay, we have to find other sources.