AI’s Credit Boom Could Be the Market’s Next Risk

SPEAKER_1 (0:00)

Thanks so much, Sam. Joining me here in studio, like you mentioned, Zed Francis, the co-founder and CIO of Convexitas. Zed, always great to have you with us. Now, you've got another quarterly note. You know I love their titles. This one is Starting to Wobble. So tell me, where are you seeing the wobbling and what is helping us stay balanced right now?

Zed Francis (0:20)

Yeah, so, you know, I've been talking about for a while that we're financing growth, and there's three main pillars, governments, corporations and consumers. And the pillar, unsurprisingly, that's getting a little bit issues in that stool, and why it's starting to wobble is obviously governments. We've had a significant sell-off in rates, global developed countries, but here in the United States. And I think everybody understands the story that we're running large deficits, that means there's lots of supply, that's going to continue to happen for the foreseeable future.

But I think the why over the last nine-ish months, we're finally starting to see that move in long-run rates that kind of stalled out for kind of range-bound three years, is actually on the demand side, not the supply side.

SPEAKER_1 (1:10)

And so as we look at it, I mean, the tenure has recovered some. I think we were around 511, 512 when I opened the show. Certainly a difference from the 536 that I was recently talking about.

But equity has largely shrugged this off. So I mean, how are you looking at the disconnect between what the bond market is signaling and how the equity markets are responding, and how long realistically can these higher borrowing costs persist before we start to see that pressure growth overall, as well as these valuations?

Zed Francis (1:44)

Yeah, so again, I think that's why it's actually coming from the demand side or lack of demand, rather than necessary all the supply. So if you think about folks that hold treasuries, they're kind of a spectrum of price insensitive versus very price sensitive. And your most price insensitive folks are basically central banks, like they're buying and selling based on balance sheet issues, nothing to do with price. Sovereigns, kind of similar, hey, we have a lot of dollars, we want to invest them in a safe place, we need to utilize dollars, we're going to liquidate those treasuries to utilize those dollars where we need it. Again, pretty price insensitive. The middle group, which was providing a lot of the demand for the last handful of years, were insurance and pension companies. And we'd say they're mildly price sensitive. Mainly they're doing liability matching, so what they care about is what's all the money that they have to pay out over the next 10, 20, 30 years? Can I buy a safe security to facilitate those distributions? So it's less about the specific price, but more can it match what I actually need to outflow? And insurance companies are growing at less than the increase in supply, and pensions have basically become fully funded. They already did their de-risking from the corporate side, meaning they've already sold equities, they've already bought fixed income. So there's not that new demand from that group, which basically kind of leaves us, it leaves us as taxable investors have to sop up this additional supply. And to tie it into your question is, I think us as investors are still saying, I'd rather own equities than rates of 5% because I'm a taxable investor. And after taxes, that's more like three and change. And I like equity returns better than three and change after tax.

SPEAKER_1 (3:28)

Well, especially with some of the returns that we're seeing in some of these names, if you're playing them correctly. But another part of your central argument is tied to some of these names that people have done. Well, I mean, the hyperscalers, they had, I see you say issued roughly 250 billion in investment grade debt just in 2026

How fragile does that then make this AI boom?

Zed Francis (3:51)

Yeah. So if you're a regulated entity, you mainly have to have a significant amount of your balance sheet invested in investment grade paper. So that can be treasuries, which I guess US has technically ranked AA, or it can be the hyperscalers, which are pretty much all investment grade. With the amount of new issuance, those names have widened and have been giving concessions at new issuance opportunities. And so I think that is sopping up some of the demand for treasuries as it's shifting to other IEG issuers, which happen to be hyperscalers. What I do think is interesting, this potential earning season that I'm very much paying attention to, is if the hyperscalers kind of step on the gas. And what I mean by that is do they say we're actually getting more aggressive on AI infrastructure build out with the concept of creating a duopoly in the new sector.