CRWV v. NBIS: Examining Neocloud Approach & Capital Awareness in AI Buildout

**SPEAKER_1** (0:00)
And Brian Mulberry, Senior Portfolio Manager, Zax Investment Management is with me. Thank you so much for being with me. I want to get all your big picture, but some of your thoughts here on CoreWeave, ahead of the real report to out after the bell.

**Brian Mulberry** (0:15)
I think a lot of really high expectations here, because you saw the monetization coming from SpaceX, Microsoft, and Google, and so I think what they're going to really have to do is beat expectations by a lot, showing that their business model is actually tracking the same way their competitors are.

**SPEAKER_1** (0:31)
What do you think about what you heard from Google and Microsoft? How does that fit in to what you may or may not hear from CoreWeave?

**Brian Mulberry** (0:39)
Well, I think monetizing excess compute was something that we didn't necessarily expect to happen. And so, since they're directly in that business of leasing out Nvidia GPUs, what does the demand structure look like? What is their margin? How are they doing in terms of pricing? We would expect a lot of capacity out there and a lot of demand, but what does the monetization really look like? They should hopefully still have about 100 billion in a backlog of orders. If they can break through that and do even more in terms of a future backlog, that's better for profitability. These are really high bars for the stock to cross, but that's really what the market expectations are, given the strength in this area so far.

**SPEAKER_1** (1:16)
But we may, I mean, we are waiting on some big numbers when you think about cloud and AI overall, but we are waiting on some huge numbers when it comes to the year over year. And in fact, the earnings per share may fall over 300%, but the revenue on the other side expected to grow.
They're somewhat traumatic, aren't they, when you think about the year over year? Is that what you expected at all?

**Brian Mulberry** (1:42)
Given their business model where they build up the demand first, book the orders and then go out and spend the money to actually build the actual compute, this is how they go. And it's a little bit inside out, certainly from a traditional business model, because that EPS number could be a little bit shocking.

**SPEAKER_1** (1:57)
Yeah, the revenue is expected to jump over 100% with the earnings per share falling over 300% year over year.

**Brian Mulberry** (2:04)
And that's just to spend the capex that they have to actually spend to build out the compute that they've already leased.

**SPEAKER_1** (2:11)
Overall. And the earnings predictions markets, of course, are also expecting some volatility. I mean, what would you do when you see this stock maybe making a big move? I mean, what if it moves 10% in the next 24 hours, let's say?
It was the recent low was 60, 55, not that long ago here. And right now, trading at 88 and change. Some of your thoughts on volatility, does that provide opportunities?

**Brian Mulberry** (2:36)
Yeah, if they can show that backlog of orders at 100 billion or more, then I would buy on any dip at this point going forward. The volatility that's out there is certainly some macro themes. We're seeing obviously interest rates tighten. The 10-year treasury at 4.7, that's retreated a little bit today in the afternoon. But that still is the real cost of capital. So anybody that's financing this growth is going to experience a squeeze in earnings in the future. But that future earnings growth for CoreWeave comes from that backlog of orders.

**SPEAKER_1** (3:02)
Right, that nearly 100 billion. We turn our attention over to Nebius. And some of your thoughts there and where that fits into this story.

**Brian Mulberry** (3:12)
Nebius is a little bit different animal, kind of in the same space, but they're more of a full stack provider where they have a whole bunch of stuff in line vertically integrated they can monetize, where it's just compute for CoreWeave. Nebius is more of a full service provider where they offer a little bit of a layer of software to help customize what that data stream looks like on the backend. They can do networking solutions. They can help put it all together in one stop shopping. So they have various business units that they can show growth in. And that's what the market's gonna be looking for is not just monetizing compute, but how is everything else doing in that full stack solution.

**SPEAKER_1** (3:44)
Right, full stack versus one thing. And I mean, does one seem better positioned based on demand and AI growth and demand going forward than the other, just fundamentally?