The real risk in AI infrastructure spending isn't hyperscalers cutting capex — it's the circular financing web of Neoclouds, GPU makers backstopping purchases, and private-equity-backed data center buildouts, which could unravel and leave many funding commitments to vanish when the cycle turns.
Josh Brown lays out how GPU makers like Nvidia effectively backstop purchases from unproven 'Neocloud' data-center startups funded by sovereign wealth and private equity, warning many of these interlocking commitments could evaporate when the cycle turns — a dynamic Thomas confirms is the real mechanism behind an 'earnings bubble.' ✦ AI generated
Josh Brown · The Compound · 2026-07-17 · original ↗
starts at this moment · 56:01
“Can I ask something? I keep hearing about the circular investing and the lack of free cash flow and everything.”
Take that example times a thousand and everybody's lending and/or equity investing in everyone else's projects and the concern is when the music stops all these people go for the chairs and a lot of people going to be left without a chair because some of those funding commitments are going to vanish as though they were a vapor.
verbatim transcript · starts at 56:01
56:01>> lack of free cash flow and everything. Let's just say Amazon and Google and Microsoft and all the big spenders stop spending. Doesn't that just shoot free cash flow right back up for them? But that's not the worry. The worry is I start a Neocloud. I say I am building a data center. I have money from Saudi Arabia and maybe the Norwegians and everybody uh private equity Apollo's in
56:25blah. So, I'm gonna build this data center and I'm going to start my own NeoCloud. Jensen Wang calls and says, "I heard the news. I think it's great. We would like to sell you GPUs." Y >> and by the way, we will financially backs stop those purchases. You don't have the capital yet. We know you're good for it because we see that you have all this backing and therefore Nvidia is
56:48going to book that chip sale to me and I'm either using money that indirectly comes from them or I'm backstopping that to raise money from someone else. Take that example times a thousand and everybody's lending andor equity investing in everyone else's projects and the concern is when the music stops all these people go for the chairs and a lot of people going to be left without a
57:13chair because some of those funding commitments are going to vanish as though they were a vapor. Am I explaining that right? >> No, I think that's exactly right. But I mean >> not saying that will happen. I'm saying this is the thing that will make this actually an earnings bubble, right? because these earnings will go away, >> right? And people confuse the economy and the market a lot, right? So the
57:33economy can be doing one thing, the market can be doing another. Economy, as we were just talking about, it's doing very, very well. But, you know, a stock price and the valuation is the discounted value of its future earnings. So, if you kind of look at this chart, there was an expectation that AI was really going to go. uh it got some traction, multiples got super high and
57:55now I think there's a view that as more competitors come in and more capabilities are built out. I mean 18 right you're 18 is below the S&P multiple right now >> 18 18 times for Nvidia. >> Yeah. >> Yeah. Pe people are acting as though it's been disrupted already >> right. Yeah. But it's just it's just the going forward expectation. But I keep coming back to what I said earlier. If
- ·Risk lies in circular financing among Neoclouds, GPU makers, PE
- ·Nvidia effectively backstops purchases from unproven Neocloud startups
- ·Neoclouds funded via sovereign wealth and private equity money
- ·Interlocking commitments could unravel when cycle turns
- ·Everyone lending and equity-investing in everyone else's projects
- ·When music stops, many left without a chair
- ·Some funding commitments will vanish 'as though they were a vapor'
- ·Confirmed as real mechanism behind an 'earnings bubble'