Brian: We started this podcast eight months ago talking about convergence—the idea that private markets and public markets were touching in ways they really hadn't before. And then dispersion—that despite things looking calm on the surface, there was a lot of noise underneath. We did touch on the CapEx needs for AI hyperscalers and data centers. What's happened in the last couple of months is that issuance has hit a fever pitch. It may be the only thing people are speaking about, but it is the theme driving markets broadly—and certainly fixed income. John, have you ever seen anything like this?

John: A year ago we started saying public and private markets are going to be indistinguishable. And it's because the biggest public companies in the world are now issuing public and private debt interchangeably, and the largest private companies are staying private longer while doing the same. Everything is blurring—across asset classes, asset types, industries. It totally changes the way you have to think about origination, how broad you have to be, and how you manage risk in your portfolio.

Brian: We are in the middle of the biggest IG issuance cycle ever—the biggest July ever, the biggest August ever. Oracle, Meta, Google, Amazon: big public names doing big public issuance. How is the market structure of IG changing because of what these well-known issuers are doing?

John: I started in the industry during the dot-com bubble, right as it was bursting. Maybe the scale of chatter and discussion was as high then, but I don't think it was the same level of the largest, most established, unlevered companies in the world suddenly turning asset-heavy at this pace. Google was a new company then. Amazon was a new company. It wasn't those same 20 or 30-year stalwarts in the market all of a sudden coming fast and furious at the credit markets. The quantum of credit risk entering the system, and the quantum of portfolio management change happening—this does seem different.

Brian: In the last year, just four companies have done $187 billion of issuance. And they're not issuing in two-year maturities—they're issuing across any maturity, any market. Australia, Switzerland, the UK. What's interesting is that for long-duration buyers—pension funds with long liabilities—it's been a famine for years. The same 20 names to buy forever. Now you have some of the best-rated companies in the world becoming bigger issuers. Spreads are widening because investors are saying, look, I know Google is a great company, but if you're going to issue 30-year paper every quarter to the tune of $187 billion a year, I'm going to charge you for it. There's more risk to that paper just because I have to take it down. And there are mathematical limits to how much of this can be in an index.

Brian: Everyone sees the 30-year public issuance. But the same companies—and a few others that haven't tapped the public market yet—are also issuing off-balance-sheet privately. Why does that matter?

John: Corporate net issuance is higher than treasury net issuance this year. That same dynamic is happening in commercial mortgage markets, asset-backed, public and private alike. There's a crowding-out effect: the more this happens on and off balance sheet, the more room you have to make in your portfolio. And these companies are being advised well—they're not dumb. If one market is tighter, or has looser covenants, or is less saturated, they'll tap that market first. That's why they're issuing in every currency, every region—trying to find pockets of capital that haven't bought them yet. It makes complete sense when you have to finance $5 trillion of spend by 2030. You have to understand where all these risks are clearing, and that's genuinely hard to do.

Brian: Think about what we did with Valor and xAI—essentially GPU financing. Before the most recent wave of mass issuance, we looked at those deals and saw what had to come down the public and private pipeline. Those were conservatively structured deals with no assumption of residual value, and they still came at pretty wide spreads. So when we saw SpaceX was going to come, and saw the quantum of what Google was going to issue—even if you're not buying the private debt, you have to be aware of those trends. If you can buy the equivalent debt and the private market is a hundred basis points wider, it's going to put pressure on public market spreads.

John: The first version of SpaceX to trade was private—those chip financing deals they were guaranteeing. That traded before they even brought a broadly syndicated deal, and it was tradable. So you have private companies issuing debt in the market that is liquid but also private. It's the market's first taste of something liquid, which is interesting—as opposed to the usual dynamic of pricing the big benchmark deal first, then the private deal after.

Brian: We're at an inflection point where investors are being forced to acknowledge that whether or not you want to buy it, whether or not you believe in the liquidity, you have to understand it. That's not going to change. By our estimates, we're looking at $3 to $5 trillion of issuance—and it can't all be public. If you look at the Bloomberg Ag, fixed income benchmarks broadly haven't changed in our lifetime. We're not going to see more than 2 to 3% of any single issuer—investors will cap the index or move on. That leaves roughly $800 billion of room for the large names I mentioned, which means a couple of trillion dollars of issuance that's simply not going to be in the index. We've gone from 0% AI-related exposure in the Ag to four and a half percent in the last year. That's going to keep changing.

John: To be clear, they're saying they'll fill that by 2030. But you might get there by 2028. The question is: how much sooner do investors feel like they're full on public IG corporate issuance from these companies within a benchmark-aware framework? If you're just running today's playbook and filling the coffers of IG money, you're probably getting there before 2029—maybe before 2028.

Brian: Do you think the market gets it yet? At the beginning of the year spreads were very tight. Inside the index there's some dispersion—some of the names are wider than they were a month and a half ago. Around mid-June to early July it felt like peak fear. People were saying, this is real, it's happening. Are we at peak fear, or is there more pressure on the IG public side to come?

John: Fear relative to the rest of the market is high. I'm not sure it's peak, but it's pretty high. One reason it might not go much higher: a few things have to happen. Either spreads continue to widen and money comes back in at more exciting levels, or the companies start to show the revenue and earnings that justify continued buying at leverage levels that are improving over time. The third thing is you can't keep issuing at this pace without the broader market starting to move. So far, every time someone has bought hyperscaler debt wider, they've bought something else tighter. The rest of the market has stayed remarkably benign. But there's a limit to how far that goes. If spreads on the best companies in the world go to 150 or 200 basis points, that's going to impact the broader debt markets for large, frequent users—financials, cable companies, highly CapEx-intensive businesses. Those are going to be impacted. So relative to the rest of the market, this is probably as high as the fear gets—because everything else is going to have to converge a little.

Brian: Let's talk about high yield and loans, because that space is really interesting right now.

John: Similar price action and portfolio construction dynamics, but a different end game. In the high yield and leveraged loan market, software is at the widest tick of the last five years. Non-software is at the tightest. CLO managers are capping their software exposure and buying everything else tighter. The question is whether the same can happen in IG—can you just keep buying the rest of the market tighter while this goes wider and compartmentalize it? Probably not at its extreme, because in levered land you're worried about losing principal. That's not a serious concern when you're lending to Amazon or Google or Microsoft. So I think over time you'll have to bring it back into context with where the rest of the market is pricing risk—which means these names can ultimately be a much larger percentage of your portfolio than leveraged software or asset-light services businesses.

Brian: Right. And what we're seeing in high yield is the dispersion inside the AI sector itself is very wide. You can see bonds moving two or three points every week just because a new deal is coming and people are comparing structures. We've seen data center deals with wildly different residual values and documentation—a deal that looks great today could be a hundred basis points wider tomorrow when something with better structure comes. And we're seeing fits and starts: a couple of weeks ago one deal had price talk between 98, then as low as 92, then back to 98. I think you can see windows in high yield where nothing gets done for two or three months because the market just stops.

John: The pipeline management piece is really critical here. Understanding the pipeline and where it's coming from—that's actually harder to see through in high yield than in IG.

John: How do you change the way you're managing portfolios and risk in this environment?

Brian: People don't like to change, but this is going to force them to think about markets differently—because even if you don't want to buy AI-related issuance, you're going to have a chance to buy different risks at unique spreads. And if you're a pension fund that has bought a lot of long-duration bonds, you're going to have a wider variety of things to choose from than you've had in a very long time. The Ag today is basically 70% government risk and 30% IG at all-time tights, so you're getting no spread for buying it. I think this AI issuance factor is going to force people to think about their allocations differently for the first time in 30 years. It's really hard to pass up an extra 1% of yield from an investment-grade issuer on a 30-year bond, especially if you're an underfunded pension fund.

Brian: We saw this in equities: what blew up active equity managers was factor-based investing. In fixed income, the AI factor is going to drive more than just the hyperscalers, and you will see more dispersion. Active management is going to be forced to grow up, take different types of risk, and diversify. My personal view, having sat in this seat using both public and private securities: you have to take alpha by any means necessary in a market with this much dispersion. We're going to see the equitization of fixed income markets. This AI factor and the associated issuance is going to force people to change what they own in fixed income for the first time in a generation.

John: A few things are top of mind when managing a portfolio that's both public and private across asset classes. One is duration—where are you taking it? We've generally liked the chip-financing deals because you're making a two-to-three-year bet, heavily amortizing, with no bet on residuals. The second is convexity. Many of these deals are structured where they can call you at par over a couple of years, or extend you 20 years if the financing market isn't there. You're taking negative convexity that the IG market isn't typically used to. The third is factor risk. AI is a cross-asset-class disruptor, and it's being financed across asset classes. You have to stress-test your portfolio: if July 2026 lasted three months, what does your portfolio look like? Correlation is a really important part of credit management now in ways it wasn't before.

Brian: If you took a total portfolio approach and asked, what do I want to own in fixed income right now—how does that look?

Brian: Very short duration: treasuries are great for liquidity, no argument there. Past cash, in the three-year sector, the trade for the last decade was short corporates—all the juice is gone, 28 to 40 basis points. Now the AI factor in that maturity is not riskless, but you're going to get a couple hundred basis points for a three-year bond. In the ten-year sector, I'd own a combination—some treasuries if you're a pension buyer, plus a basket of high-quality issuers that are widening. And for long-duration portfolios, I can now build a basket of great companies in 10 and 30-year maturities at durations of whatever you want, with money-good outcomes. In that longer-end basket, I'd include plenty of private bonds—because if you're choosing a seven, ten, or thirty-year duration, you're basically saying you don't need that money for a while. The high yield and loan market in the AI sector is really interesting, but you'd better understand it—don't do that passively.

John: Just be careful of the native convexity—make sure you understand whether you can get called out of a risk at the wrong time, or extended. But otherwise I agree. The public corporate market is starting to get genuinely interesting. Broadcom five-year CDS is trading above 100 basis points today. Oracle is at 200. That's the min-max range of where high-quality high yield is clearing. You now have to ask: do I want to buy the next high yield bond, or put money to work in companies for whom staying investment-grade is essentially existential? They have large equity market caps, they're going to do whatever they can to maintain their ratings. Those are pretty good places to start to deploy.

John: There's a misconception that private is riskier and public is safer. A lot of these deals were being financed in the private markets until private markets started demanding tighter covenants and higher spreads—and then they went public. Look at a lot of the data center financings that were done as commercial mortgage loans and are now being wrapped in 144A bonds and sold to multiple buyers where the lease risk isn't as tight as it should be. The Meta deal brought a few weeks ago—the fact that it says Meta doesn't make it the same as a Meta investment-grade corporate bond. It's a different structure, a different spread. You have to understand the lease risk, the tail risks on the specific project being built. We work with our commercial mortgage lending team and our high-grade capital solutions teams to go through all of those details. You can't just buy it because of the name on the plate. If it's too good to be true, there's probably a reason it's clearing at that level.

Brian: That's one of the reasons we're always looking for things that aren't AI-related. It's so hard to find them—when you go through the index, the big banks, the big telecom companies, everything is being touched by this. We're constantly asking our traders and originators to find us good-spread paper that doesn't have AI exposure.

John: AI is going to be brawn as well as brain—as it moves into physical applications, asset-heavy industrial companies, autos, airlines, consumer, retail, energy, defense—so many of these industries are being reinvented and many of them need a significant capital infusion to get there. Those are all part of a great opportunity set. It doesn't have to be only the hyperscalers. A huge portion of your portfolio can and should still be differentiated across these other asset classes. Pacing is really important.

Brian: As we look ahead to Q4—one thing on my dance card is the global yield picture. It was our thesis at the start of the year that the Fed would ease, there'd be some economic softness, and even in a higher-for-longer world, the market would say that hikes are done and yield curves would flatten toward a new normal. That's not happening. And if the US government is going to issue effectively unlimited debt with the hyperscalers right behind them—and Europe eventually figures out its own deficit issues, Japan continues to be a story—are we going to find a new clearing level for yields? Is capital still just too cheap if people are going to issue this much?

John: The cost of capital has been higher for the last year, and it's not surprising. You start a war, AI issues trillions of dollars of debt to finance massive CapEx, defense spending is being reshored, the treasury is still funding a massive deficit. Yields are going higher if you're over-levered, mismanaging asset-liability, or betting on a business model that only works in a low-rate environment. Those are really, really hard right now. The premium is on growing businesses, high-cash-flowing businesses, businesses that work unlevered. In this environment, where you don't really know where financing rates are going to end up in a year or two, trying to engineer a return through financing structure is a dangerous game.

Brian: What else is on your Q4 dance card?

John: The midterms are going to be really interesting. There's an assumption in markets that AI growth is a bipartisan issue—we have to beat China, everyone agrees. But if you talk to anyone outside of Wall Street, there's a lot of concern. That could become a significant issue in the midterms and, given how correlated these risks are to markets, could create real volatility. There are also big IPOs coming, a significant amount of debt issuance expected in September and October. I don't think it'll be smooth sailing between now and year end.

Brian: Convergence is happening in real time. The dispersion theme is playing out in real time. And I think the market structure change we're living through is actually creating opportunity—not just forcing change. People now have the chance to say: I can choose my fixed income factors better than ever before. The menu is as big as it's ever been. I don't have to own the Ag, I don't have to own just money market funds or just the big banks. For equity investors, there have been uncapped upside trades for years. But in fixed income, there's now a clearing level where you can lock in a great yield, own great companies, and sleep well at night. It's been a very long time since investors had that opportunity.

John: The banks are 20 to 25% of the IG market partly because they're viewed as quasi-sovereign—they have an effective government backstop. The interesting question to think about over the next six to twelve months is whether the largest five to ten public companies in the world, as they become the largest issuers of debt in the world and everyone owns exposure to them everywhere, start to invite a similar level of government oversight, regulation, or even some form of ownership relationship. That could create some fireworks.

Note: This transcript has been edited for clarity. It does not represent a verbatim transcript of the podcast. Podcast recorded on August 17, 2026.