Lotfi Karoui on Private Credit and the AI CapEx Supercycle

Please stay tuned after the conclusion of the podcast for additional important information.

Greg Hall: Hey, everybody. Welcome to another episode of Accrued Interest, PIMCO's podcast dedicated to serving financial advisors and the wealth management community. As always, I'm your host, Greg Hall. I lead the wealth management business for PIMCO here in the United States. And back by popular demand, Lotfi Karoui. Welcome, Lotfi. Nice to have you again.

Lotfi Karoui: Good to be back. Thanks for having me.

Greg Hall: Frequent listeners of the podcast, as I hope you all are, will remember that we had Lotfi on the pod back on March 17th, about six weeks into your tenure here at PIMCO.

Lotfi Karoui: Yep.

Greg Hall: Trial by fire.

Lotfi Karoui: Yep.

Greg Hall: And now you've been here about six months, so you're a veteran. I'm sure that you're not even remotely intimidated at the prospect of going through another podcast. Let me mention this upfront so I don't forget it later on.

In addition to doing fantastic work within the four walls of PIMCO and spending a lot of time out on the road with our clients, which I have to say, you've been phenomenal in your commitment to our clients over the last few months, and we really appreciate it. I know they do too. Lotfi has a weekly piece called the Credit Market Lens, which is published on the PIMCO website as well as through LinkedIn.

You can subscribe to it if you wanna navigate there during or after the pod and take a look at some of Lotfi's writing. It's really, in my biased opinion, quite excellent and will add some context and some flavor to some of the themes we'll talk about today. So today we're gonna be talking about credit markets.

We're gonna be talking about broader markets. And maybe we'll pick up sort of where we left off. When we last talked in early to mid-March, we were, I wouldn't say at the very outset of the stress in the direct lending universe, but we hadn't yet seen the gates go up and sort of the investor behavior follow through on some of the early warning signs and the news stories of late 2025. So before we get into more detailed topics, let's just talk about the overall picture for credit markets, what you like, what we ought to be focused on.

Lotfi Karoui: Sure. I mean, look, as you may remember, back then there were kind of like two competing views. One was the view that this was a little bit reminiscent of what we went through in the run-up to the global financial crisis, i.e., there was genuine concern that this stress in direct lending portfolios would gradually morph into something bigger and threaten financial stability.

And then there was a view, which we actually took back then, that this was merely a cyclical issue. Call it like a textbook example of how a credit cycle kind of unfolds a little bit. Direct lending as an asset class grew too quickly by several folds over a 10-year period. And every time you grow an asset class that fast, you're prone to imbalances and you start seeing some deterioration in underwriting standards and stress, etc.

Obviously, the heavy concentration in software doesn't help, and it kind of complicates the picture from a fundamental standpoint. Now, here we are three months later. I think the systemic debate has been more or less settled. I think there's a better understanding that you can always describe a state of the world in which this becomes a threat to financial stability, but it's a little bit of a stretch.

And I think the fundamental problems that direct lending as an asset class has to grapple with are also better understood today. But if I kind of have to take a little bit of a mark-to-market of where we are today, you're seeing continued outflows in semi-liquid vehicles. Automatically, that means financing conditions are actually tightening, at least on the margin. Still a lot of question marks with respect to software.

And I would say we have a clear, clean sort of price discovery process in the broadly syndicated loan market, and I would kind of use that as a proxy a little bit. But if you look at the performance of software loans on the public side, big decline in prices first quarter, no recovery since, and somewhat surprisingly, still very low dispersion.

Usually when you have a sector that gets sort of derated that way, the analyst community does the work a little bit, and then we start picking winners and losers. That hasn't happened, I think, in software today. And I personally find that quite interesting.

Because that tells you that, even if you're a seasoned analyst, you're still having a tough time answering the basic question with respect to these businesses, which is, what is the terminal value given the threat that AI actually poses?

Greg Hall: Okay.

Greg Hall: Right, right. And not differentiating amongst the different pieces of the software puzzle. I think the outflows — I think our listeners will know being alert people that watch the news — but the outflows in the wealth vehicles, in particular, have continued to run, by and large, 15% redemptions or plus or minus.

Lotfi Karoui: Yeah.

Greg Hall: 11, 12, 15, 16, 17. Most of the participants have chosen to prorate those redemptions and to hold them at their 5%, 6%, or 7% levels at this point. And so, you're sort of seeing this slow — maybe isn't quite the right word — but certainly a managed, measured bleed-out of this space.  You mentioned that that's tightened financial conditions.

Are you seeing the pace of deployment in the direct lending and the private credit space react to that? Or has the institutional market stepped in and made—

Lotfi Karoui: Yeah. Well, unfortunately, the data is usually backward-looking, so I would sort of caution against overinterpreting the year-to-date flow or fund-raising activity data that we have. But if you take it at face value, it shouldn't come as a surprise that this is a mediocre year, at least by the standards of the last four to five years. Yeah. And so, there's definitely been a slowdown, but it's not a collapse on the institutional side.

Greg Hall: Right, right, right, right. And hence your point about this not necessarily posing a systemic risk. Have we learned anything in the last few months about the underlying credit quality of these portfolios? Or do you still see investors reacting primarily to software fears, if not realized yet? And maybe sometimes when some investors are frightened, other investors become frightened and the crowd takes over.

Lotfi Karoui: Yeah. I mean, look, we generally rely on BDC holdings as sort of a proxy for how the broader direct lending market would look. Sequentially, if you look at first-quarter data, I would say on a quarter-on-quarter basis, things have been roughly stable. On a year-over-year basis, there's still material deterioration.

And so, this feels to me like a slow-motion thing, you know, so far, now. Obviously, the bigger question mark is really with respect to the marks. And those are not only stale, at least that's the general perception among most investors, but they're also dispersed. And let me maybe take like 30 seconds to explain, what I mean.

Greg Hall: Okay.

Greg Hall: Yeah, yeah, yeah. Take all the time you want.

Lotfi Karoui: You know, if you look at loans that are held by two or three more BDCs, and you kind of calculate the price differential between the most optimistic manager and the most pessimistic one. On average, that's well over six, sometimes seven points, and so, if you're an investor and you see that, obviously you ask questions about where the marks are.

But the fact that they're so dispersed and the fact that there's so much disagreement across managers is probably a driver of that sort of cycle of redemption requests exceeding the caps. But that hasn't been resolved. And that kind of surprised me a little bit because I would've thought that by now, we would see more convergence in the marks. Doesn't matter what the level is, but I think there should have been more agreement across managers with respect to the marks, and we're clearly not there yet.

Greg Hall: Yeah. And part of that is, in these vehicles of forcing mechanism just works a lot more slowly.

Lotfi Karoui: Totally. I mean, on the public side, prices are the intersection of supply and demand. You can take any ISIN go on Bloomberg, or you can pull your favorite pricing engine and you get to see a price. By the way, I remember back in the days, before TRACE was introduced in the corporate bond market, it wasn't uncommon to see disagreement between index providers. I mean, you could take the same bond and it would be priced a little differently.

Greg Hall: Sure.

Lotfi Karoui: But that difference was like a point, a point and a half, not seven points. But I agree. I think part of the problem is you have a benchmark, and that's really a very healthy public credit market from a microstructure liquidity standpoint. And that benchmark sets the bar a little too high, I think, for direct lending.

Greg Hall: right?

Greg Hall: Yeah. Well, it's interesting because—and you've written on this too—there's an interesting debate or divide happening even amongst the dedicated private lenders out there. Hey, it’s the solution to this malaise that we're feeling right now introducing more transparency into marks or potentially more liquidity into the loans themselves? And you've got strong feelings on this. Maybe we talk about it.

Lotfi Karoui: Yeah, yeah. I mean, look, any steps that are taken to improve transparency are obviously welcome. The less disagreement we can have among managers, the better the outcome for investors. There's absolutely no question about that. Now, there's a variety of ways you can achieve that. And there's certainly been a lot of ideas floating around. One of them was to provide daily pricing and I think you are referring to —

Greg Hall: Back in May, I think. Yeah.

Lotfi Karoui: Back in May, where we said daily pricing is not daily liquidity. And those are two different things. I mean, I can price anything intraday if you want. You can give me a model, I can put an input in it, and then get—

Greg Hall: Yeah

Greg Hall: It's effectively what we're doing now. We take a model—

Lotfi Karoui: Which is sort of what we're doing.

Greg Hall: When I say "we," I mean the direct lending and private credit industry, broadly speaking, is taking a model, interpreting a price. It's just not a price that's being transacted on directly or being set by transactions.

Lotfi Karoui: Yeah. And so, to get real liquidity, you basically need two-way flows. And there's lots of obstacles, I think, that would prevent that from happening overnight. And in my view, the most obvious one is constraints on ownership transferability. I mean, these are privately negotiated agreements.

If you're a new investor and you don't have a pre-existing relationship with the borrower, it's actually hard to settle. Settlement is another constraint. Quality of the documentation. On the public side, every security has an ISIN. It's well documented. You kind of need to create that plumbing.

And then, I think, bigger picture, I would say, I'm not sure incentives are aligned, actually, to move to a liquid direct lending market. Because the whole point about buying a direct loan, and more generally any private asset, it could be private equity or private real estate, is that you're harvesting some illiquidity premium. That's sort of the trade-off.

Lotfi Karoui: You're committing capital for a certain number of years. And you're trying to say, am I being compensated for that illiquidity premium? If I take away that illiquidity premium from you, well then what is the point of actually deploying capital in direct lending to begin with? And so, to me, the flaw is not in the illiquidity of the asset class.

I mean, that's sort of an attribute. The flaw is that that compensation for illiquidity has gotten way too thin, and it needs to reset to more attractive levels. And we sort of need time, I think, to achieve that.

Greg Hall: Yeah. Those are really, really important points. I'd say probably the other barrier to adoption —and this is where I think the debate gets really interesting, because on the one hand, you've had some really prominent dedicated private lenders attempt to build trading platforms for the underlying loans or to promote this transparency and liquidity. And like you said, I think, generally we as market practitioners, view that as laudable.

Lotfi Karoui: Yeah.

Greg Hall: On the other hand, I think a lot of advisors listening to the podcast right now would probably admit it to themselves and to us, that the allure of some of these products has been that they don't get marked that frequently. And so, the journey for the client, assuming you're going to a good place, which I think everybody has believed has been the case until very recently, is smoother and it's easier to be on.

And so, if you introduce volatility, then the value proposition of owning private versus public securities, right, it loses a lot of its appeal. And so, it'll be interesting to see how that resolves itself over time.

I think my view for the wealth market is that, there hasn't really been a time in our lives' investing history where the typical two-legged taxable US investor has been desperate to own a lot of floating-rate credit in highly levered companies, other than immediately following the financial crisis, when, to your point, spreads were really wide and there was a lot of excess profit to be reaped.

And then this volatility aspect of the investment, of the asset class, has been really appealing, but that's eroded with the headlines and the gates going up. So, for me, it all sort of speaks to not a collapse of the asset class, but probably a long-term normalization to a much smaller level over time.

Lotfi Karoui: Totally.

Greg Hall: Yeah.

Lotfi Karoui: I hundred percent agree. And I think the other normalization that will happen is that we'll probably go back to a structure of direct lending that looks a little bit like direct lending 1.0.

Greg Hall: Mm.

Lotfi Karoui: More focused on the middle market.

Greg Hall: Yeah. It could be. And also, maybe we see a re-evaluation of the asset-liability mismatch and appropriateness for which vehicles actually hold some of this, right?

Lotfi Karoui: Yeah.

GREG HALL: While we're, on the topic of private lending, generally, one of the things I've noticed is, especially with some of these bigger transactions going on in the marketplace right now, there seems to be a little bit of a proliferation of credit types. We see true private types of bespoke financings. We see stuff that's regular way in the listed market or arranged by banks.

And there's a few things, you know, we've been involved in them too, that seem to occupy maybe a hybrid zone in the 144A market, or, I mean, actually I'm gonna ask you to maybe explain some of these different types, but there's a lot of letters and numbers involved, and so I thought maybe for advisors listening, you could just sort of walk through the different ways that people are raising debt financing right now.

LOTFI KAROUI: Yeah. I mean, look, the argument that you hear more and more frequently is along the lines of like, private credit is already more liquid than commonly assumed. And that argument typically points to the 144A market as evidence of that liquidity.

GREG HALL: Right.

LOTFI KARAUI: Now, the idea that 144A securities are privately placed, yet often trade with liquidity comparable to corporate bond markets. But I think that comparison is actually somewhat misleading. And it essentially conflates two very different types of instrument. So let's start with a 144A bond.

I mean, that's technically sold through a private placement exemption, which essentially means that the issuer can sell the bond directly to qualified institutional investors without going through the full SEC registration process. That is typically required for a public offering, but that exemption actually affects how the bond is initially sold, not how it's structured or owned or traded in almost in a very practical sense, I would say 144A securities behave pretty much like public bonds.

They're syndicated, they carry CUSIP or an ISIN and so documentation is actually fairly standardized. They rated the vast majority of the time. And then I would say perhaps more importantly they're reported on trace. Whenever a secondary market transaction occurs, you know, it is recorded almost real time.

And it's traded by a very diverse investor base, including insurance companies, pension plans, and of course asset managers. And so the price discovery process that is behind a 144A security, is nearly identical to a public bond. There's actually virtually no difference. And in fact, if you look at the high yield bond market, the majority of the high yield bond market today is 144A securities.

Now, why am I bringing that up? Because the liquidity of 144A bonds actually comes from their standardization, the fact that they're very widely distributed across a diverse set of investors, not because they were issued under a private placement.

And so this is fundamentally different from a corporate direct loan and NAV loan, CRE loan or even like an asset based finance type of structures. Those instruments are privately negotiated between a borrower and a small group of lenders, sometimes actually one lender. And each deal has its own documentation, collateral structure, and that just means a higher bar for an ownership, transfer. Now, just to be very clear none of this implies that illiquidity is a flaw, you know?

GREG HALL: Yeah.

LOTFI KAROUI: Illiquid assets actually play a very important role in a multi-asset diversified portfolio, but to me, the question is not whether, you know, an asset is illiquid or not, is whether you're being compensated for that illiquidity. And so that's I think, the biggest question that investors, advisors alike need to ask themselves, which is, you know, are you being sort of appropriately rewarded for essentially giving up on liquidity?

And to be quite frank, I think assessing that trade off, you know, requires an apples to apples comparison. And if you conflate 144A bonds with generally private debt type of assets, you're essentially overstating that, you know, that liquidity of private debt and you're underestimated how much compensation you're getting.

GREG HALL: I think it is a really tricky time for advisors trying to unpack everything going on in the private credit landscape because of course liquidity's risen to the fore because folks are redeeming from some of the dedicated products and they're being told to wait for a period of time until the redemption prorations open up.

And that can be frustrating for an advisor. And so you see a lot of rhetoric about the liquidity of the underlying loans and or marking portfolios on a daily basis, which, you know, can add transparency. And I think, you know, you and I have talked about this we've talked about it in this session that you know, we applaud, you know, moves for more transparency and more information for investors, but not everything is exactly the same.

LOTFI KAROUI: Totally.

Greg Hall: Let me return for a second to the systemic discussion around direct lending, because there were two things in our most recent secular outlook, which listeners of the pod will recognize as a piece that we publish once a year that tries to look out over five or ten years, as opposed to our quarterly pieces that look out maybe over more of like a six-month cyclical timeframe.

And there were two things I picked up in that report that I hadn't heard us say before, and I thought they were pretty interesting, and you're in a great position to offer some context.

The first is, I think we stated quite clearly, and Dan has come back to this topic a few times since, that we feel like the default cycle in credit has begun, or I shouldn't say the default cycle, but a default cycle.

And the second thing we noted was that we are concerned—not running for the hills—about some elements of financial engineering that have begun to take hold in markets. So maybe we could talk about those two topics for a little bit.

Lotfi Karoui: Well, let me start with defaults. Look, I mean, typically defaults are a highly cyclical variable. You know, I mean, the example that I always give, take a time series of 12-month trailing default rate in the high-yield bond market going back 40 years.

And what you see is four humps, basically, and those were the four last recessions. And so, typically it does take an economic downturn to actually fuel a rise in financial distress in corporate credit.

This time around, I look back at the last six months. We are seeing signs in some corners of leveraged finance, most notably direct lending and to a lesser extent the broadly syndicated loan market, where I think that default cycle is kind of starting to decouple a little bit because the economy is still in great shape, at least in aggregate. I mean, there are some pockets of the economy that are doing better than others, but in aggregate—

Greg Hall: Yeah. The data moves around.

Lotfi Karoui: But this is still an economy that is growing. And so is there a risk actually that this time might feel a little different? I think that risk is high, yes. And I do think that for direct lending, and to a lesser extent the broadly syndicated loan market, you have a few headwinds that I think will likely fuel a gradual increase in financial distress.

But one is sort of an obvious point. Most of the liabilities are actually floating in nature. Well, policy rates are still high, so there's a little bit of a lagged effect from—

Greg Hall: And they've been high persistently for quite some time.

Lotfi Karoui: So that 500 basis points of cumulative hikes that the Fed delivered between '22 and '23 is actually still putting pressure on those capital structures. Software doesn't help. That's a story that will take possibly years to get resolved.

And by the way, there is a view out there that because refinancing needs are not that high in the near term, it should be okay. If you look at the vast majority of defaults in the United States, they tend to be strategic in nature.

You don't wait until you have to refinance in order to restructure the balance sheet. You generally do it before, and you do it either out of court or in court. It doesn't really matter how you do it, but I think you have enough flexibility to do it before actually the time comes to refinance.

And then I think what we're discovering too is that, this is true for direct lending because the asset class has grown so fast. At some point there's a limit to how much capital you can deploy. And so, you start loosening your underwriting standards a little bit and you start to write bigger checks and things like that.

And so, there's an element of loosening underwriting standards, I think, that you definitely see in direct lending that you don't see in markets like the high-yield bond market, for example.

I mean, this will come as a surprise to many listeners, but the size of the high-yield market has been roughly flat in the last 10 years. I mean, to me, that's evidence that you're not misallocating capital. If anything, it's the opposite. It's a market that's been remarkably disciplined.

You look at the sector composition, you see a lot of the old economy, energy, chemicals, asset-heavy type of companies. It's a little less than 40% secured. It's 55% BB-rated. And so, it's a much higher-quality market relative to its own history. Most of the stress, I think, and the excesses have built in some of the adjacent markets, and direct lending is one of them.

Greg Hall: Yeah. I mean, it's an incredibly interesting point. In this massive post-crisis expansion in credit, or recovery in credit deployment, it hasn't been nearly as much a participant—

Lotfi Karoui: On the public side.

Greg Hall: Yeah. On the high-yield side.

Lotfi Karoui: If anything, it's been the opposite.

On the return of financial engineers, look, we're late cycle. Obviously, you have really strong levels of yield support in fixed income. But if you look at the spread component, it's very thin. And it doesn't matter where you look. You can look at IG, high yield, the United States, outside of the United States. The level of risk premium is very thin.

Greg Hall: All the ninety plus percentile….

Lotfi Karoui: Totally. Like, at all-time tights or close to all-time tights. And so, every time that happens, people start to get a little creative about ways to generate better returns, using forms of financial engineering.

Greg Hall: How do I make a thin spread look like a fat spread?

Lotfi Karoui: Right. And so, what's the typical ingredient? Leverage. You know, just look at the growth of leveraged ETF AUM, for example. You know, that one has quadrupled over the last couple of years. The return of structures that have a little bit of leverage-on-leverage type of thing.

Now, I just want to be clear. All these corners of the market are small in and of themselves. They kind of look unrelated a little bit. So, unlike the run-up to the global financial crisis, where you had clear evidence that the housing market was caught in a bubble, and at the same time investors started to grow mismatches between assets and liabilities, leverage rose way faster than today. I think this is not it. But you're seeing signs of—

Greg Hall: What are some of the signs? What are some of the structures that we're seeing? Again, I don't think you are forecasting doom. It's just, but there's stuff that, on the margin, we —

Lotfi Karoui: I mean, the growth of leveraged ETFs is one example.

Greg Hall: In the equity space.

Lotfi Karoui: These vehicles work great until they don't. They work great in a bull market, but once you unwind that, you have sort of a reverse effect. I think we've seen growth of fixed income CLOs that reference illiquid underlying assets. Again, that works okay until you start having outflows, and you'll test it.

By the way, corporate bond ETFs have proven their resilience the last 15 years, but only because the underlying liquidity in the corporate bond market has supported that. Unclear to me whether an ETF on a structured product can actually benefit from the same resilience. But that's something that will be tested.

And certainly, the AUM has grown a lot. It's still small when you put it in the context of the broader credit market, but it's grown fast enough that I think it's something that will get tested at some point.

You know, there's an element of interconnectedness between financial institutions. There's a lot of debt capital that is recycled in the same ecosystem. That structure, in my view, is still a lot better than what we had pre-GFC, but it's not perfect. And it will be tested.

Greg Hall: Right. You've got private equity sponsors that have lending arms that lend to their own deals. They may be constrained to a certain percentage, or somebody else sets the price, but you do have an interrelation between the two arms of business.

Lotfi Karoui: Basically, institutions sitting on various levels of that financing supply chain end up lending to each other. And that increases the interconnectedness. And it's something that will be tested.

Greg Hall: Yeah. And I think the other thing that we cited, and Dan mentioned a few times, is also a little bit of a reliance on the rating agencies to preserve us from taking too much risk. But of course, there are limitations there too.

They don't always get it right. You can try really hard to rate credits, but as the structures become more creative, it's harder and harder to keep up with the structures themselves. That presents some issues as well.

So, we spoke about software, and we touched on AI briefly. But as we turn to kind of the view forward, of course, one of the major themes that's come to prominence, since the last time you and I spoke, or at least its prominence has been reinforced, has been this AI CapEx Supercycle.

I'm curious how you're thinking about that these days. Are we talking more hype, more reality? How do you parse some of the huge numbers that are coming out of the big research houses or the consultancies at this point?

Lotfi Karoui: Yeah. It's definitely reality. I mean, if you take consensus estimates at face value, we're expecting a little over $1.6 trillion, maybe even more. And these numbers actually change on a daily basis almost. But let's take $1.6 trillion worth of CapEx just for '26 and '27, and a big chunk of that will be funded in debt markets.

And so that's really been the shift, is that the AI CapEx cycle went from a cycle that was almost entirely relying on operating cash flows to a cycle that is now very much debt-funded. And that changes the calculus completely, I think, for credit investors and fixed income investors more broadly.

And so, as it stands right now, I would say the releveraging impulse among the hyperscalers is certainly a reality. That is hard to ignore. I mean, a lot of these hyperscalers had negative net debt three years ago.

Greg Hall: Yeah.

Lotfi Karoui: You know, positive net debt. And so, there's definitely been more accumulation of—

Greg Hall: Just so I understand, the major tech companies, the seven or eight biggest, shifting from funding all this massive investment from the cash flow of their businesses to debt.

Lotfi Karoui: to debt. And that's partly because CapEx is now absorbing the entirety of your cash flow from operations. And so, it's depleting all of that. And so, you have to sort of diversify your sources of funding.

Greg Hall: Right.

Greg Hall: Gotta go find money elsewhere.

Lotfi Karoui: And the most natural way to do that is debt capital markets because equity capital is obviously more expensive. Yeah. And so that's been shift number one. I think shift number two that I kind of found actually interesting is the currency mix has also shifted. Usually most of these companies rely on the dollar market. We've seen them go into other   to investors in other jurisdictions, Euro, Sterling, Japanese Yen, Swiss Franc, Canadian Dollar, and so it’s opened up other markets and gave up opportunities to investors in other jurisdictions to sort of get a bite at that apple a little.

Greg Hall: Are they just managing curves and FX, or is there some deeper reason for it?

Lotfi Karoui: There's some of it. I think, the other reason is that you don't want to hit your concentration limits in dollar indices. And so, you look for ways to diversify the currency mix a little bit. And there's demand outside the West.

Greg Hall: Sure. Yeah. I would imagine so.

Lotfi Karoui: Now really the key for all this buildout to end well is the pace and the timing of monetization. If you create a virtuous circle where you're able to monetize all this CapEx at the same time that you continue to build great.

Greg Hall: Right.

Lotfi Karoui: If not, well, there's a risk that we kind of repeat the same mistakes that we made in the late 1990s, which is, you go a little bit too fast, you overbuild, and eventually demand is unable to keep up with supply. But for now, I'm not seeing a lot of evidence that there's some overcapacity that we're building up in the system.

Greg Hall: Right. The metrics still appear a lot more disciplined than what we saw in—

Lotfi Karoui: And I think that's a really excellent point that you're making, which is discipline has been way, way more present among investors this time around. You know, look no further than the equity market. Look at the performance of the Mag Seven relative to the broader index. It hasn't been particularly stellar.

You know, over the last two years, dispersed winners and losers. But as a group, I think they've stopped outperforming the broader market the same way they had prior to the onset of the cycle. And then on the credit side, the same thing is happening. Actually, some of the way the debt is being absorbed is also happening via wider spreads, particularly if you look at long-dated bonds — so, 30-year versus 10-year bonds.

That extra premium that you have to pay to issue a 30-year bond, that's been rising for that group. So, to me, that's evidence that the market is actually paying attention. And it feels to me—I mean, I have lived through that late 1990s period, that feels to me that there's more discipline this time around relative to the late 1990s.

Greg Hall: Yeah. Our CIO, Dan Ivascyn, who was with us a couple weeks ago. He does like to make the point, especially in talking about AI, that as the bond guys, we don't capture the upside if it works, only the downside if it doesn't. And so, you can imagine a world in which we and our brethren out in the markets are a little bit more focused on protecting against risk out 10, 20, 30 years.

On top of the AI CapEx cycle, of course, we had Greg Chernow on the pod talking about the commodity complex a couple of weeks ago, and we talked a lot about CapEx in global energy security and the reshoring of energy production and transportation. Maybe the need to invest in different ways of getting oil around the world and the Strait of Hormuz.

We also, you know, touched on —but didn't go into— he mentioned the impact of defence spending as countries around the world maybe come to the conclusion that the US is not gonna play the same role that it has in the last 50 years as the global protector. And so, they need to rearm.

And he mentioned that in the context of different materials and commodities that that creates demand for. I guess my question for you is, if all these things happen at the same time, what do you think about aggregate demand for financing, and then where does this money come from? Where does this come from? Especially in a world where banks are still pulling back.

Lotfi Karoui: Totally, I mean look, leaving aside the source of the financing, I think you gotta go back to the basics, which is supply and demand.

If everyone is spending at the same time, you know, defence, energy, AI, well, that's a positive aggregate demand shock essentially. And there's only a finite amount of savings in the system. And so, how does the adjustment happen? It happens via probably higher real yields. And it means that yields are gonna stay a little higher for a little longer here, right? But to me, that's the most obvious way you address this.

Now remember, I talked earlier about the timing. You know, there's the AI buildout, and then there's AI adoption, right? If you get adoption to come quicker than now, I have a lot of sympathy with the view that AI adoption is ultimately disinflationary because the marginal cost of producing a digital good is very low.

And so, I tend to view the buildout compounded with defence spending, energy CapEx, etc, as a demand shock. I think the adoption is a supply shock, and kind of the intersection of those two things will determine where yields and ultimately how the inflation outlook will look. But the sooner you get adoption, I think the higher the odds that you set in motion of a virtuous cycle, in my view, where adoption is disinflationary. It absorbs some of the inflationary impulse from the forces that you discussed.

Greg Hall: Right.

Lotfi Karoui: So, there's a lot of moving parts.

Greg Hall: I think, there's a lot of nuances in this conversation.

Lotfi Karoui: And sometimes you hear people overly focused on the demand aspect of this. Absolutely true. Look, you know, you're spending more money. There's only a finite amount of capital in the system. And so, the cost of capital has to adjust higher. That's the most obvious way you can clear supply and demand. But then you have adoption, and that, I think, frees up a lot of excess capacity in the system.

Greg Hall: Right. That's disinflationary and potentially has an offsetting pull. And then there's what goes on in the broader economy at the same time.

Lotfi Karoui: And it's also structural. It's sticky.

Greg Hall: Yeah, you wrote a nice piece on emerging markets, which, to be honest, I think for the typical US financial advisor, that's not a huge allocation. It is something that we've had more and more people ask us about, especially as we've had our own views about finding value internationally. So maybe we can talk about that for a little bit, but also don't let me presuppose the answer.

Lotfi Karoui: Yeah.

Greg Hall: So, you tell me where you're seeing some interesting—

Lotfi Karoui: Yeah. I think we're starting with a broader context here, which is we're late cycle, right? And that doesn't mean the cycle is gonna inflect tomorrow, but we are late cycle.

Greg Hall: What do you say? Cycles don't die of old age.

Lotfi Karoui: They don't really. I mean, we're late cycle based on a bunch of metrics. You have an economy that's at full employment. You have valuation metrics that are at extreme levels. And that's true in fixed income. It's true in the equity market. It's basically true in the broader risk asset complex. And so, the key ingredient to portfolio construction, in my view, is diversification and trying to be ready for that inflection point.

Now, the good news in fixed income is that you're still getting the best level of yield support you've had in 20 years. And that's really important because it means that bonds have more convexity embedded in them. And it means that the ability of the risk-free component to protect you in a bad state of the world is still remarkably strong.

Greg Hall: Pause on that for a second and make sure I understand what you mean by bonds have more convexity embedded.

Lotfi Karoui: So, let's go to 2022. That's a good example which was traumatizing for a lot of investors.

Greg Hall: I was gonna say, can we please not?

Lotfi Karoui: But it's a useful example because back then, most multi-asset portfolios were hit by a double whammy of higher yields, so lower bond prices, and then lower risk assets. Why did that happen?

It happened because the primary driver of higher yields was a massive inflationary shock that led the Fed to tighten monetary policy very aggressively over a very short period of time. But the other reason is the starting level of yields was very low. I mean, policy rates were basically at zero. Ten-year yields were less than 2%.

Today you've reset basically the level of bond yield premium at much higher levels. Take 10 years as an example. You're at 4.5%. If tomorrow you have a recession, basically the Fed is sitting on 300 basis points of cuts that they can deliver right away. And so, the whole curve is going to move downward, and you're gonna get a nice embedded stability from owning bonds.

And I think that point—it's interesting you said, I've been talking to a lot of clients but I've been on the road the last two months or so.

Greg Hall: Yeah.

Lotfi Karoui: And it was striking to me to see that there's a lot of views that are still anchored on either some of the assumptions that prevailed in that '22 period or even sometimes the post-GFC, pre-COVID type of period. This is different.

Greg Hall: You couldn't be more right. When we talk to advisors, they're still—I shouldn't say still—but they've got every reason to hold this belief, but they're very distrustful of bonds playing that traditional role in their portfolio as the non-correlated diversifier. And based on the experience in '22, I think that's driven a lot of the interest in floating-rate credit markets.

Lotfi Karoui: Yeah. I totally agree. Now, today, you can pick your capital market assumption for equities, how much you think we're gonna have for the next five years.

Greg Hall: Your return projection.

Lotfi Karoui: Seven, 8%. Even if you make super bullish assumptions about that, the beauty of bonds is that the starting level of yields pretty much determines your future return for the next four or five years. And so, if you take the Bloomberg Agg as an example, the yield, as we're recording this is around 4.75%. That's gonna be your return for the next five years. But if you have an accident.

Greg Hall: Potentially. Yeah, yeah.

Lotfi Karoui: Then you get paid. Then that convexity in bonds kind of kicks in. And so, in general, I would advocate for a return to a more balanced multi-asset portfolio. It could be 60/40, whatever that mix is. But I think it makes sense to tilt things a bit more towards bonds because of that convexity in the event the economy inflects and the cycle dies.

Along the same lines of diversification is emerging markets. That was an asset class that was sort of unloved for almost 15 years. I think now there's realization among a lot of investors that, this is not your older brother's EM market. There's more credibility from the central banks. A lot of the external vulnerabilities that have typically fuelled crises in EMs are just not there.

That doesn't mean that there are no issues. There are a lot of countries still running inflation that is a little too high. There are some fiscal issues here and there. And so again, country selection is key. But, by and large, I think there's definitely premium in EM. And I think the ability to give you diversification in a bad state of the world is also a lot better.

The final thing that I would add is that you mentioned 2022, which I agree was a very traumatizing experience for a lot of multi-asset investors. But here's another difference between today and 2022. Economies globally, even if you look at the developed world, are much less synchronized today.

Back in 2022, if you looked at the cumulative change in 10-year yields, you could pick Japan, Germany, Australia, Canada, US, UK—everything moved together. Everything moved in the same direction. That's not the case today. And in fact, it will probably come as a surprise to a lot of listeners, but 10-year yields in the US have been kind of flat since January 2025. It doesn't feel that way.

Greg Hall: Yeah.

Lotfi Karoui: Because you have a lot of vol and every now and then some geopolitical Shock.

Greg Hall: It feels like we've been on a roller coaster, but we've barely moved.

Lotfi Karoui: Yeah, but that's where you've been the last 18 months. And so, what happens in the meantime is that you're just accumulating income. Meanwhile, you have other jurisdictions like Japan and Germany where yields are still drifting higher. And so, your ability to generate alpha by just moving from one jurisdiction to another is remarkably better today than it was back in '22 when everything was correlated and moving in a very synchronized way.

Greg Hall: You know, it's interesting. We have been doing a lot of research on where the average advisor and average client portfolios are situated in the US today. And, you know, 60/40, we’re so far from 60/40, we're not even close anywhere near it. I mean, 90/10 is probably a better estimate.

Lotfi Karoui: By the way, that was the right call.

Greg Hall: Oh. Absolutely.

Lotfi Karoui: Ninety-ten was the right call, with the benefit of hindsight.

Greg Hall: Yeah.

Lotfi Karoui: But this is why I said earlier, I think we shouldn't sort of anchor portfolio construction today to the same assumptions that prevailed the last 15 years.

Greg Hall: And advisors out there, you know, they've earned their way into 90/10, right?

Lotfi Karoui: Yeah.

Greg Hall: They've compounded their way there in the equity markets. I think one of the things that matters a lot in the US—and it's different around the world—but in the US, if you've got the privilege of being in that place of having generated such great returns from the equity markets over the last 5, 10, 15 years, you're sitting on a lot of low-basis stock that's hard to sell if you want to rebalance. If you've taken advantage of some of the great innovations in tax-loss harvesting, that problem is even more pronounced.

And one of the things that we see in industry flows, in our own data, is that advisors are acting on the marginal dollar. So, they're following your prescription on the marginal dollar. It's just gonna take a while, I think, for us to get ourselves back into a more traditional state.

Lotfi Karoui: 60/40. Yeah.

Greg Hall: If we ever get to 60/40 again. But certainly, we could move quite a lot from where we are today back towards that.

Lotfi Karoui: And the point shouldn't be about bonds beating stocks or anything like that. It's just that you're getting better yield support. And in a state of the world where the cycle inflects, it doesn't matter what the shock is. It could be the end of the AI CapEx cycle. It could be something else. But, you know, at some point the cycle will end. You have an extra embedded protection in bonds that I think stocks don't give you.

Greg Hall: Right. And you made the point earlier about companies that are coming up on a refinancing wall, and it's best to be prepared going in and avoid the cliff or the calamity when it comes, which I think we'd all subscribe to. So, you know, not exactly breaking news here at PIMCO that we like the idea of adding bonds to the portfolio at this point.

But I think we've been pretty humble out there about the relatively unattractive state of bonds, you know, going back to 2021 where the world was pretty much at the zero bound, like you said. Compare that to today, and you're just in a very different neighbourhood.

Lotfi Karoui: Yep. Yep.

Greg Hall: Alright. Well, Lotfi, as always, this is an absolute pleasure to have you on today. Thanks for joining us. Thanks again for all the good hard work you're doing, not just analytically, but spending a lot of real time out there with clients, helping them interpret and understand what's going on in markets.

If you liked what you heard today or you're interested in going a little bit deeper on any of these topics, we'd really encourage you to hit the like button, subscribe to the podcast. Knowing you're out there helps us bring you more content and refine it to your tastes.

If you would like to read any of Lotfi's research, if you'd want to go into any other topics that PIMCO's published on, visit us at pimco.com or the website in your home country. If you identify yourself as a financial advisor, you'll be taken to Advisor Forum. Advisor Forum is our one-stop destination for you to find what you need from PIMCO quickly, pragmatically, and then move on with your day, making sure that you can deliver the best analysis and content to your clients.

We'll be back with you in a couple of weeks with our next episode. Until then, hope you're having a great summer, hope you enjoyed your Fourth of July, and we'll see you soon.