00:00Do you see the potential losses in private credit as more significant than people expect,
00:05even without some sort of true systemic risk? No, I don't. But I think that the systemic risk
00:14point is perhaps most significant here. In terms of the misperceptions that exist,
00:22these are structures that essentially grew after the global financial crisis precisely because
00:29the liability structures are so durable, that there is great comfort taken in regulators,
00:35moving risk out of a financial system, banking that largely has runnable liabilities, depending
00:43on deposits, depending on wholesale finance, depending on repos, money markets, and moving
00:50it to balance sheets where you have much more paid-in capital. There's, of course, three to five
00:56times as much equity capital backing loans. And then also, of course, longer-term liabilities.
01:02The leverage that's taken on by private credit funds, direct lenders, again, relatively modest,
01:10but also it has maturities that generally extend beyond the maturities of the loans that are held
01:16on the asset side of the balance sheet. So again, very durable structures. I think that right now you
01:22have default rates that are running in about two and a quarter percent range, maybe two and a half
01:27percent. So again, this is an area where they ticked up very modestly. But I think it's nowhere near
01:35the discussion. The amount of attention that it's getting just feels disproportionate to me at this
01:41point.
01:41Jason, has that become actionable in the sense that have you seen assets in the private credit
01:46space that are being sold at a discount? Or is this just histrionic rhetoric that isn't really
01:51transpiring through a market revaluation?
01:55I think it's important, again, to think about we're talking about whole loans. These are non-traded
02:02assets. So really, what this is comparable to is, of course, whole loans on bank balance sheets.
02:10For JP Morgan, for instance, has about $350 billion of CNI loans, commercial industrial loans.
02:19No one would think that JP Morgan can liquidate that portfolio or a large portion of it in a
02:26relatively short period of time without discounts to the carrying value of those assets. So again,
02:33when you have non-traded assets, it generally takes time to underwrite. There's an illiquidity,
02:39of course, there needs to be an illiquidity premium for investors with exposure to these
02:45assets. But when you talk about discounts, of course, in non-traded illiquid assets,
02:52there's going to be some discount if they have to be liquidated quickly. And that's precisely why
02:56you have these durable liability structures that mean that these assets do not need to be
03:04liquidated. They can be held to maturity. Jason, the broader credit markets, though,
03:09it was a painful February. That's been reflected in some of the BDC earnings. But leveraged loan markets
03:14had a very painful month, too. And a lot of that has been attributed to what happened with software
03:20companies. I wonder if you can make the parallel, if you do, as some people do, to 2014 with energy
03:26companies. As you started to see defaults pile up from them, could you see a repeat with a different
03:32flavor of something coming from software companies that maybe not fully forces a credit cycle, but at
03:38least does introduce more pain into the system?
03:41I think so. It's interesting. We've, of course, looked through our portfolio. We have about 293
03:49companies and not much software exposure. It's very minimal. But to talk to those companies,
03:55CTOs, to the CIOs, to look at how they're spending on information services, on software, to look at how
04:03they're embracing AI. Of course, AI strategies, moving out to the front of the technological frontier
04:10is very important. And to understand, you know, how they're spending and how software may be
04:16disintermediated. So this has been an area of focus. And I do think at this stage, some of the AI
04:23disintermediation is overstated. I do think at greater risk right now is consultants. And I think that when
04:31you look, when you see how they're trying to spend on AI and integrate it into workflows, it could be
04:38that software becomes a key distribution channel for some of the frontier AI models. I think pricing
04:45is going to be under pressure, certainly. I think that the economics of software may change because
04:52now all of a sudden you have to deal with compute and inference costs. Of course, many software platforms
04:58have been predicated on the idea of zero marginal costs. So there could be some material changes
05:05there. And then I think also when you look back to how richly valued the software sector was, people
05:14talk very often about terminal values in software. Well, why is that? It's because there's so little
05:20cash generation for many of these businesses today. So it's really the free cash flow from these businesses
05:28is arriving further into the future, which of course makes it more sensitized to, again, terminal values
05:35or expectations five years, 10 years forward. So again, I don't think that one can just say that
05:43there's nothing to see here. But at this stage, I do think that some of the talk about total
05:49disintermediation is really overstated and it's not consistent with the way that CIOs, CTOs
05:56are actually integrating AI into their businesses at the moment.
Comments