00:00Now concerns over the size and increasingly circular nature of AI financing deals have been
00:06given credit markets pause. Now NVIDIA is moving to bring more outside capital into the AI build
00:10out and potentially give investors some reassurance. Earlier this week the chipmaker said it was
00:16enlisting six financial heavyweights in an effort to line up as much as $500 billion in financing.
00:23Now let's discuss all of this and other private market developments with Neil Kalinan who leads
00:27global coverage of private companies for Bloomberg. Neil as always thank you so much for
00:32joining us. I mean this is a big job because you're also trying to find I mean this kind of
00:35like
00:35connects everything all at once because financing in general is changing. What can you tell us about
00:40this AI circular economy? Yeah it's a huge obviously part of credit now. You know even two years ago
00:47there was very little hyperscale debt in the market and now it's coming up to 20-30% of indexes
00:54in terms of issuance this year. So it's a massive change from before and there are a lot of dangers
00:59inherent in this and NVIDIA decided this week to step in and kind of calm credit markets by providing
01:05this computing power as a form of securitization which will ultimately they think help finance other
01:12companies and allow them to invest in AI. The issue for people is that that will largely be done through
01:17residual value guarantees which is basically a promise that they will cover shortfalls if that
01:23was to happen but effectively and this is the big question investors now face is that creates a big
01:29unknown going forward and you have to take a view on what the value of these assets will be worth
01:36in a
01:36few years and the real danger is that somebody like China comes along creates a significantly cheaper
01:42compute and then these residual values that you thought would be there are gone or certainly written
01:48down and then suddenly a massive liability is appearing on a balance sheet that can create problems
01:52obviously for companies going forward. But so the risk I mean could it be systemic? Is that the ultimate
01:57and that's obviously the ultimate risk? Yeah at this stage it wouldn't be but the danger is it could
02:03be could become so and there's two main reasons that are two main ways that would happen. One is
02:09on the banking side so if banking gets overly exposed and doesn't fully realize the these residual value
02:15guarantees can't fully monitor them and then suddenly the companies they back end up with a bunch of
02:20liabilities that weren't expected and the companies run into trouble. The other side is if these residual
02:26value guarantees are being insured and those insurance wrappers are being then sold off to many of the
02:34companies who are financing the companies in the first place because of the like circular nature of
02:40private markets as well. They're financing things up front but they're also insuring things at the
02:45back end and whether we could get into a circular space there that could cause trouble going forward.
02:49Overall how is the private credit space? Private credit you know it's been a roller coaster year for them
02:57to put it by. So they came into the air expecting this massive inflows of pension cash and 401k cash
03:03and
03:04they were looking for ways to set up products to put it to work. Then the Saskpocalypse came along in
03:09March
03:10the sentiment completely shifted away from that and actually people started pulling money particularly from BDCs
03:17and so now things are starting to turn around again for private credit. Certainly they feel like the
03:23institutional money that stuck with it is seeing much better opportunities. They're getting better spreads.
03:30Returns are good. So far earnings have held up but there are some potential problems coming down the line there
03:35particularly where it relates to software problems and software issues going forward.
03:39Yeah which is what we call the cesspocalypse which I don't know who found that term but it's absolutely genius.
03:44Where I mean could we see another you know this was uh basically because of a report I mean could
03:50we
03:50see it flare up again or does it normalize because they're structured differently? I think it's going
03:56to come back onto the radar in a big way in the next few months and the reason for that
03:59is there is this
04:00massive maturity wall of software debt coming due in 2028. Usually these things started getting refinanced
04:07about 18 months out so it's around now that people are starting to look around and deciding what to do.
04:12The issue for software companies is they have been growing so the lenders have been happy with that.
04:18The problem is they may not grow fast enough to grow into the debt that they've taken on and the
04:24problem with that is many of the loans were taken out in the COVID era very cheap debt. Where companies
04:30ran into trouble they then took on these what's called pick debt so it's very expensive debt but you
04:36have to pay until the loan comes due so now you're facing the original loans you had are going to
04:41be
04:41much more expensive the pick loans which have grown in size are very expensive are going to come due
04:46and so your blended average of debt which was okay now looks very expensive and that can limit your growth
04:51going forward. Who should we worry about again is it you know you have small companies you have
04:57large companies you have them with exposure to private equity like what would you watch out for?
05:02So any company that is built on a database of publicly accessible data that's going to face real
05:10trouble from AI so if you if you don't own your own data if you don't create your own data
05:15then you're
05:16very very vulnerable to AI and so a lot of enterprise-based software companies maybe we're looking at going
05:22forward and there are going to be problems in that sector definitely. Overall I mean there's such a you
05:28know we had SpaceX we were talking about Anthropik I mean certain media reports have a valuation of
05:33like one to two trillion I mean the sizes of these companies that were private coming in the public
05:38markets does it completely change the nature of how we cover markets? Yes it does because I mean
05:45there is only really one story in town right I mean it is AI and that is all that anybody
05:49wants to know
05:50and certainly the public markets are a good reflection of that and you know SpaceX IPO
05:55uh started great turned bad as sentiment changed um at one point the short interest in the free
06:02flow to SpaceX was above 40 percent much lower now as things have normalized and think volatility has
06:09reduced but where you're going to see that is in things like the IPO market so if you have an
06:15AI or
06:15you're an AI adjacent story that's great you're the demand is going to be there people love it and they
06:21definitely want to be part of that if you're a consumer company and you're trying to IPO it's much
06:27much more difficult to get backing from public investors for that because they want to be in
06:32this AI fast-growing story you're also concentrating risk much more absolutely and and for now it's still
06:39in the scheme of things quite small but it's as it grows and adds to spending which is massive uh
06:45goes
06:45forward we we added up the hyperscalers future spending commitments it's nearly two and a half trillion
06:51it's basically um almost doubled in this in three months so the the commitments they're making
06:57to spending are huge and people are very much paying attention to that um what we want to see
07:02now is whether the revenue can grow as quickly as the spending is growing core we have a couple of
07:07days ago seemed to be a positive in that side but obviously um we need to see on a broader
07:12scale
07:13what's happening in that respect
Comments