WEBVTT - Markets Wrap: Trouble Ahead For The AI Money Machine?

0:00:02.759 --> 0:00:07.240
<v Speaker 1>Bloomberg Audio Studios, Podcasts, Radio News.

0:00:18.480 --> 0:00:20.959
<v Speaker 2>Welcome to the merriland Talks Money Market Wrap, where we

0:00:21.000 --> 0:00:23.280
<v Speaker 2>talk about the biggest moves and markets this week and

0:00:23.320 --> 0:00:26.560
<v Speaker 2>what's been driving them. I'm Join Stewick, senior reporter and

0:00:26.640 --> 0:00:29.240
<v Speaker 2>author of the Money Distilled newsletter, and then joining me

0:00:29.240 --> 0:00:32.440
<v Speaker 2>in the studio today is Bloomberg's Private Companies Managing Editor

0:00:32.520 --> 0:00:35.960
<v Speaker 2>Neil Callanan. Neil covers hedge funds, asset management, and real

0:00:36.080 --> 0:00:40.920
<v Speaker 2>estate in EMIA. So in today's show, I basically wanted

0:00:40.960 --> 0:00:44.280
<v Speaker 2>to talk about the AI industry. Wanted to give us

0:00:44.280 --> 0:00:46.840
<v Speaker 2>structure out of what it's actually about, lay it out

0:00:46.960 --> 0:00:50.680
<v Speaker 2>so that you can understand it better. They kind of

0:00:51.120 --> 0:00:54.720
<v Speaker 2>who's producing the chips and who's producing the models, and

0:00:54.760 --> 0:00:58.240
<v Speaker 2>then wanted to look at how that's funded and why.

0:00:58.320 --> 0:01:03.360
<v Speaker 2>There are some concerns about perhaps the circularity of some

0:01:03.440 --> 0:01:05.800
<v Speaker 2>of the deals in the industry, and Neil has written

0:01:05.800 --> 0:01:08.760
<v Speaker 2>a lot about that. Thanks very much for being with

0:01:08.920 --> 0:01:11.080
<v Speaker 2>us today, Neil. I was great to have you on

0:01:11.160 --> 0:01:16.800
<v Speaker 2>the show, having not at all so but recording this

0:01:17.000 --> 0:01:22.880
<v Speaker 2>on Wednesday. And probably the biggest story so far this

0:01:22.920 --> 0:01:26.000
<v Speaker 2>week has been the route in the South Korean Cosby

0:01:26.280 --> 0:01:30.200
<v Speaker 2>Index and Basically this is driven by the fact that

0:01:30.720 --> 0:01:35.600
<v Speaker 2>Cosby is half of it is comprised of two stocks

0:01:35.640 --> 0:01:38.479
<v Speaker 2>that are related to the AI boom, and that's skh

0:01:38.640 --> 0:01:43.560
<v Speaker 2>Nix and Samsung. And also there is an element of

0:01:44.200 --> 0:01:47.280
<v Speaker 2>retail investors getting in over their heads because they've been

0:01:47.319 --> 0:01:50.480
<v Speaker 2>able to buy all these leveraged ETFs and so the

0:01:50.520 --> 0:01:52.600
<v Speaker 2>costby went up a lot and then down a lot.

0:01:52.800 --> 0:01:54.640
<v Speaker 2>I mean, one of the stats that really jumped out

0:01:54.680 --> 0:01:57.200
<v Speaker 2>at me is that it's actually still up thirty five

0:01:57.200 --> 0:02:00.360
<v Speaker 2>percent on the year, but in the last month it's

0:02:00.400 --> 0:02:03.840
<v Speaker 2>fallen slightly formal set So yeah, okay, that's a proper

0:02:04.000 --> 0:02:08.400
<v Speaker 2>roll off course at the moment. But does this tale

0:02:08.480 --> 0:02:13.600
<v Speaker 2>is boy, does this suggest is about the overarching AI

0:02:14.120 --> 0:02:17.679
<v Speaker 2>story because it's not just about Korean leverage. There's other

0:02:17.800 --> 0:02:20.080
<v Speaker 2>stuff going on too, Isn't that absolutely?

0:02:20.120 --> 0:02:22.880
<v Speaker 1>And Korea in a way is a special case because

0:02:22.960 --> 0:02:27.000
<v Speaker 1>Korea has almost gamified the finance industry and the stock market.

0:02:27.080 --> 0:02:30.399
<v Speaker 1>And you mentioned the lever gts, but they've been big

0:02:30.440 --> 0:02:33.680
<v Speaker 1>on crypto and heather and things like that, stable coins

0:02:33.919 --> 0:02:37.600
<v Speaker 1>for a long time and and you know those a

0:02:37.600 --> 0:02:40.919
<v Speaker 1>lot of those investors in retail have grown up very

0:02:40.960 --> 0:02:43.280
<v Speaker 1>accepting of those levels of risk and in many cases

0:02:43.280 --> 0:02:45.480
<v Speaker 1>it has paid off. So, as you said, the COSPY

0:02:45.560 --> 0:02:48.440
<v Speaker 1>is still up this year despite the route in recent weeks.

0:02:48.440 --> 0:02:50.919
<v Speaker 1>But while it is driving a lot of the change

0:02:50.960 --> 0:02:53.280
<v Speaker 1>at the moment is the fear that China is catching up.

0:02:53.960 --> 0:02:58.120
<v Speaker 1>And China's emergence in AI, which was always going to

0:02:58.160 --> 0:03:02.160
<v Speaker 1>happen but perhaps faster than people, is really concentrating minds

0:03:02.160 --> 0:03:03.880
<v Speaker 1>at the moment. And we have seen in the past

0:03:03.919 --> 0:03:06.040
<v Speaker 1>when Deep Sea came out of nowhere and released its

0:03:06.080 --> 0:03:09.280
<v Speaker 1>models at the market disruption that that cause. And now

0:03:09.320 --> 0:03:11.240
<v Speaker 1>we're going to a kind of second bout of that

0:03:11.720 --> 0:03:14.639
<v Speaker 1>where in the space of just over a week you've

0:03:14.639 --> 0:03:17.440
<v Speaker 1>seen massive progress in terms of memory chips, in terms

0:03:17.440 --> 0:03:20.520
<v Speaker 1>of language models, and so China is definitely emerging. And

0:03:21.200 --> 0:03:23.359
<v Speaker 1>one of the things around all is spending that we're

0:03:23.360 --> 0:03:25.760
<v Speaker 1>seeing in AI is the assumption is that people are

0:03:25.760 --> 0:03:28.360
<v Speaker 1>willing to pay for it and won't pay a lot

0:03:28.400 --> 0:03:30.480
<v Speaker 1>for it going forward. And then suddenly if you have

0:03:30.520 --> 0:03:33.560
<v Speaker 1>these cheaper models coming out of China and people start

0:03:33.600 --> 0:03:35.920
<v Speaker 1>switching to that, then what happens with the names we're

0:03:35.920 --> 0:03:41.520
<v Speaker 1>all familiar like with like open AI and Anthropic and

0:03:41.600 --> 0:03:45.680
<v Speaker 1>so you know, there's definitely fears in the market about

0:03:45.680 --> 0:03:49.160
<v Speaker 1>that at the moment, and I think people are definitely

0:03:49.240 --> 0:03:52.880
<v Speaker 1>taking pots globally as a result of that. At the

0:03:52.880 --> 0:03:54.800
<v Speaker 1>same time, a lot of this money is going to

0:03:54.840 --> 0:03:59.440
<v Speaker 1>be spent. There is an investment case for it. It's

0:03:59.480 --> 0:04:01.760
<v Speaker 1>whether it's getting ahead of itself at the moment. In

0:04:01.760 --> 0:04:03.720
<v Speaker 1>certainly in terms of graduations, it did seem to be

0:04:03.760 --> 0:04:05.320
<v Speaker 1>getting steamy at one point this year.

0:04:05.520 --> 0:04:07.960
<v Speaker 2>Yeah, I mean, I thought we'd just kind of lay

0:04:08.000 --> 0:04:12.680
<v Speaker 2>out roughly the AI industry and the business mode because

0:04:12.720 --> 0:04:14.280
<v Speaker 2>I think a lot of the time, you know, we

0:04:14.360 --> 0:04:18.160
<v Speaker 2>sort about AI and it's quite an amorphous kind of blob,

0:04:18.520 --> 0:04:20.760
<v Speaker 2>and a lot of time we're thinking about chat GPT.

0:04:21.360 --> 0:04:23.920
<v Speaker 2>I mean, the video is in there somewhere and all

0:04:23.920 --> 0:04:27.160
<v Speaker 2>the rest of these things. But so always kind of

0:04:27.200 --> 0:04:29.400
<v Speaker 2>looking at it. And that's what I was talking to

0:04:29.440 --> 0:04:32.320
<v Speaker 2>you to AI about it and getting a sense of

0:04:32.360 --> 0:04:34.600
<v Speaker 2>what the kind of value chain it. So you've got

0:04:34.680 --> 0:04:38.000
<v Speaker 2>your you've got the machines that make them, So you've

0:04:38.000 --> 0:04:40.279
<v Speaker 2>got the companies that make the machines that make the chips,

0:04:40.600 --> 0:04:44.120
<v Speaker 2>and that's like your SML that makes the kind of

0:04:44.120 --> 0:04:47.320
<v Speaker 2>lift up lithography machines and then you get the chip

0:04:47.360 --> 0:04:51.640
<v Speaker 2>foundries themselves, and that's like the Taiwanese company t s MC.

0:04:53.000 --> 0:04:55.960
<v Speaker 2>You've got chip designers and that's the video. So they're

0:04:56.000 --> 0:04:58.440
<v Speaker 2>they're the ones that basically do the kind of value

0:04:58.480 --> 0:05:01.480
<v Speaker 2>add element, isn't it. It's like they are they're making

0:05:01.480 --> 0:05:05.000
<v Speaker 2>the brains for these things. And then all of these

0:05:05.080 --> 0:05:09.719
<v Speaker 2>chips go into big AI hotels, the kind of data centers,

0:05:09.800 --> 0:05:11.920
<v Speaker 2>and they're run by the same people that run what

0:05:11.960 --> 0:05:14.760
<v Speaker 2>we used to call Internet hotels, the kind of data servers,

0:05:14.760 --> 0:05:17.160
<v Speaker 2>like the Amazons and all the rest of this world.

0:05:17.680 --> 0:05:19.599
<v Speaker 2>And then you've got the people who actually make the

0:05:19.680 --> 0:05:23.719
<v Speaker 2>AI models and the AI models the brains of the

0:05:23.760 --> 0:05:27.240
<v Speaker 2>AI live on the data centers, and that's like your

0:05:27.240 --> 0:05:31.320
<v Speaker 2>open AI. And then they're selling them companies either big

0:05:31.480 --> 0:05:35.480
<v Speaker 2>enterprise providers who kind of like garnish their existing offerings

0:05:35.480 --> 0:05:40.920
<v Speaker 2>with AI, orange users like me and you, are smaller companies.

0:05:41.279 --> 0:05:43.240
<v Speaker 2>Is that about the size of it?

0:05:43.279 --> 0:05:43.440
<v Speaker 1>Is that?

0:05:43.520 --> 0:05:46.560
<v Speaker 2>Does that kind of sound about right in terms of

0:05:47.160 --> 0:05:49.919
<v Speaker 2>what the EI value chain is and who all the

0:05:49.920 --> 0:05:51.240
<v Speaker 2>companies are involved in this?

0:05:52.400 --> 0:05:56.520
<v Speaker 1>But like this is so massive now that encompasses everything

0:05:56.560 --> 0:05:59.599
<v Speaker 1>else as well. So they need energy and so we're

0:05:59.600 --> 0:06:02.640
<v Speaker 1>saying boom and energy, and people are talking about leading

0:06:02.720 --> 0:06:06.040
<v Speaker 1>up to three hundred gigawatts of additional energy power by

0:06:06.080 --> 0:06:09.360
<v Speaker 1>twenty thirty. That's enough to power two hundred and twenty

0:06:09.400 --> 0:06:13.840
<v Speaker 1>five million homes for a year just for data centers. Now,

0:06:13.839 --> 0:06:15.599
<v Speaker 1>those things have to be built as well. You have

0:06:15.640 --> 0:06:17.799
<v Speaker 1>to get people to build the data centers and build

0:06:17.800 --> 0:06:21.120
<v Speaker 1>the energy infrastructure, et cetera. So construction firms are taking

0:06:21.120 --> 0:06:23.720
<v Speaker 1>off as well. And so you know, in the US

0:06:23.760 --> 0:06:26.480
<v Speaker 1>in particular, it's driving a lot of the economic growth,

0:06:26.520 --> 0:06:28.839
<v Speaker 1>and any slowdown in this kind of spending will be

0:06:28.880 --> 0:06:32.680
<v Speaker 1>negative for growth, not necessarily turn recessionary anything, but it

0:06:32.680 --> 0:06:36.720
<v Speaker 1>would be negative for growth in the US. But this

0:06:36.839 --> 0:06:38.800
<v Speaker 1>goes back your point about in the end. It comes

0:06:38.880 --> 0:06:40.920
<v Speaker 1>back to the end users. Goes back to my point

0:06:40.960 --> 0:06:43.039
<v Speaker 1>about China in a way, which is that you know,

0:06:43.120 --> 0:06:44.800
<v Speaker 1>we need people at the end to be paying for

0:06:44.839 --> 0:06:48.120
<v Speaker 1>these services in order to justify these investments. And when

0:06:48.160 --> 0:06:50.200
<v Speaker 1>you look at the amount of money that are being

0:06:50.200 --> 0:06:52.960
<v Speaker 1>spent by some of the hyperscalers in particular, which are

0:06:53.360 --> 0:06:57.160
<v Speaker 1>these Internet hotels originally as you describe them, you know,

0:06:57.240 --> 0:07:00.880
<v Speaker 1>if you look at alphabet th recent filing, it's forward

0:07:01.080 --> 0:07:06.600
<v Speaker 1>spending commitments rows five hundred billion dollars essentially in three

0:07:06.680 --> 0:07:11.080
<v Speaker 1>months now. Revenue, the future revenue growth grew but not

0:07:11.440 --> 0:07:14.000
<v Speaker 1>by anything close to that, and that goes back to

0:07:14.040 --> 0:07:16.720
<v Speaker 1>the fears people have about this entire ecosystem that is

0:07:16.800 --> 0:07:19.880
<v Speaker 1>sucking up so much money and so much capital at

0:07:19.880 --> 0:07:22.400
<v Speaker 1>the moment that people are kind of going, well, well,

0:07:22.640 --> 0:07:24.920
<v Speaker 1>I don't know whether I want to have as much

0:07:24.960 --> 0:07:27.200
<v Speaker 1>exposure as I could have to this. And you're starting

0:07:27.200 --> 0:07:29.240
<v Speaker 1>to see that in the credit markets as well, where

0:07:29.240 --> 0:07:32.080
<v Speaker 1>people are becoming much more discriminating about the deals they

0:07:32.080 --> 0:07:34.600
<v Speaker 1>invest in and the price they're wanting to pay. And

0:07:34.640 --> 0:07:37.680
<v Speaker 1>they're also hedging a bit more so this week Core

0:07:37.720 --> 0:07:42.200
<v Speaker 1>Weaves CDs, which is the form of hedging against the

0:07:42.280 --> 0:07:45.600
<v Speaker 1>fault risk that's risen't almost a record. The core Weave

0:07:45.680 --> 0:07:47.920
<v Speaker 1>is one of these companies that is like a wee

0:07:48.000 --> 0:07:51.720
<v Speaker 1>work of the GPU world, which is the chips, and

0:07:51.800 --> 0:07:55.400
<v Speaker 1>they basically rent out other people what they're doing. And

0:07:55.960 --> 0:07:58.920
<v Speaker 1>you know, people have become much more cynical in the

0:07:58.960 --> 0:08:01.040
<v Speaker 1>last few weeks about deals, partly because they know so

0:08:01.120 --> 0:08:03.480
<v Speaker 1>much of this stuff is coming to the credit markets

0:08:03.480 --> 0:08:06.200
<v Speaker 1>that they don't have to buy everything and The question

0:08:06.320 --> 0:08:11.160
<v Speaker 1>is whether people in the AI industry as a wider

0:08:11.240 --> 0:08:14.080
<v Speaker 1>thing have become too complaisent about the idea that the

0:08:14.120 --> 0:08:16.520
<v Speaker 1>credit markets will be there and be supportive of them,

0:08:17.040 --> 0:08:19.280
<v Speaker 1>and that's not necessarily always going to be the case.

0:08:19.600 --> 0:08:21.920
<v Speaker 2>Well, I mean, I think that this is it is

0:08:21.960 --> 0:08:24.880
<v Speaker 2>fascinating from that point of view, and that there really

0:08:24.920 --> 0:08:27.920
<v Speaker 2>has been a wug bond issue and hasn't. And I

0:08:28.000 --> 0:08:33.160
<v Speaker 2>mean we're talking about by some measures now the hyperscalers

0:08:33.200 --> 0:08:37.160
<v Speaker 2>are the biggest issues of corporate investment grade debt, and

0:08:37.760 --> 0:08:40.520
<v Speaker 2>you're used to being the banks and the other point

0:08:40.600 --> 0:08:44.360
<v Speaker 2>but hyperscalers is that before all this they were they

0:08:44.360 --> 0:08:48.000
<v Speaker 2>were basically running on their own cash generation. They were

0:08:48.160 --> 0:08:53.240
<v Speaker 2>deemed as being essentially impregnable balance sheets. And while it's

0:08:53.240 --> 0:08:56.840
<v Speaker 2>not negative that they've started, you know, recent or not started,

0:08:56.880 --> 0:08:59.760
<v Speaker 2>they've now got a lot of debt. And for example,

0:08:59.840 --> 0:09:02.800
<v Speaker 2>our for bet It's latest quarter was its first negative

0:09:02.800 --> 0:09:07.840
<v Speaker 2>free cash flow quarter ever since it lest date. It's

0:09:07.920 --> 0:09:10.080
<v Speaker 2>just a sort of sign of how much is out

0:09:10.120 --> 0:09:14.000
<v Speaker 2>there and how these these businesses are kind of fundamentally

0:09:14.000 --> 0:09:18.000
<v Speaker 2>being changed by this. And then you throw in China

0:09:18.080 --> 0:09:20.800
<v Speaker 2>maybe turning around and doing it all much much cheaper,

0:09:21.400 --> 0:09:23.920
<v Speaker 2>and suddenly you're kind of oh, your onions are going

0:09:23.960 --> 0:09:27.680
<v Speaker 2>to be crushed, or so the Indians that we're hoping

0:09:27.800 --> 0:09:30.760
<v Speaker 2>that you get might get crushed by this. And the

0:09:30.880 --> 0:09:32.720
<v Speaker 2>one thing I thought was really interesting, going back to

0:09:32.760 --> 0:09:36.680
<v Speaker 2>the debt point, as you involved in putting together this

0:09:36.880 --> 0:09:41.280
<v Speaker 2>very complicated and now famous in a sort of local

0:09:41.360 --> 0:09:46.160
<v Speaker 2>way chart about how this is all being financed. And

0:09:46.240 --> 0:09:49.640
<v Speaker 2>one of the big things that's I think slightly worrying

0:09:49.679 --> 0:09:52.120
<v Speaker 2>people as well is that a lot of the money

0:09:52.559 --> 0:09:56.920
<v Speaker 2>seems to be coming from the people who raise the

0:09:56.960 --> 0:10:00.520
<v Speaker 2>money in the first place, so like almost like a

0:10:00.559 --> 0:10:04.400
<v Speaker 2>form of vendor financing. You've got companies that kind of

0:10:04.400 --> 0:10:07.400
<v Speaker 2>make the chips, paying the people who buy the chips

0:10:07.520 --> 0:10:09.360
<v Speaker 2>or who are going to rent the data centers that

0:10:09.520 --> 0:10:11.560
<v Speaker 2>use the chips. Can you talk to us a bit

0:10:11.559 --> 0:10:16.040
<v Speaker 2>more about about that and how that's kind of panning out.

0:10:16.679 --> 0:10:19.959
<v Speaker 1>Yeah, what happened late last year was I was sitting

0:10:20.000 --> 0:10:24.559
<v Speaker 1>there reading story after story about these deals where companies

0:10:24.600 --> 0:10:27.640
<v Speaker 1>were doing deals with other companies in the universe, and

0:10:27.800 --> 0:10:30.440
<v Speaker 1>video has always been at the center of this, but

0:10:30.480 --> 0:10:32.959
<v Speaker 1>the likes of Google, Ananthropic and open Ai were doing

0:10:32.960 --> 0:10:35.040
<v Speaker 1>all these as well. And I literally couldn't keep up,

0:10:35.559 --> 0:10:37.920
<v Speaker 1>and so we just had the idea and maybe we

0:10:38.000 --> 0:10:40.680
<v Speaker 1>need to do a visual here and kind of show

0:10:40.960 --> 0:10:45.080
<v Speaker 1>the levels of circular deals that are happening. And you know,

0:10:45.400 --> 0:10:48.560
<v Speaker 1>this became something of a bad word in the nineteen

0:10:48.600 --> 0:10:50.800
<v Speaker 1>nineties with fiber optic and when there was a big

0:10:50.880 --> 0:10:53.599
<v Speaker 1>rollout of that, lots of spending and there was vendor financing,

0:10:53.679 --> 0:10:57.040
<v Speaker 1>and there was capacity sharing and various other things, and

0:10:57.080 --> 0:10:59.040
<v Speaker 1>then a lot of those companies went bus when the

0:10:59.400 --> 0:11:01.600
<v Speaker 1>demand was in there at the end having invested in

0:11:01.600 --> 0:11:03.839
<v Speaker 1>all that money. And of course the irony, of course

0:11:03.920 --> 0:11:07.920
<v Speaker 1>is that long term the economy benefit and massively from

0:11:08.360 --> 0:11:11.000
<v Speaker 1>that level of spending. It's just the companies involved. But

0:11:11.200 --> 0:11:13.120
<v Speaker 1>noways so many of them ended up going bankrupt.

0:11:13.320 --> 0:11:17.439
<v Speaker 2>Well, that's that's the infrastructure story, isn't it. The always

0:11:17.440 --> 0:11:19.439
<v Speaker 2>are still hearing we still have trained a lot of

0:11:19.440 --> 0:11:22.040
<v Speaker 2>the companies about them went bused and seeing me dot

0:11:22.080 --> 0:11:22.840
<v Speaker 2>com and.

0:11:23.160 --> 0:11:25.880
<v Speaker 1>That happens with most new technologies. And again that goes

0:11:25.880 --> 0:11:29.000
<v Speaker 1>back to why investors are being somewhat skeptical at the

0:11:29.040 --> 0:11:31.800
<v Speaker 1>moment about that. But these circlear deals, they can create

0:11:31.880 --> 0:11:37.160
<v Speaker 1>misaligned incentives around things like are you making the decision

0:11:37.240 --> 0:11:39.760
<v Speaker 1>for the in the best interests of the company. Are

0:11:39.800 --> 0:11:41.800
<v Speaker 1>you making it in the best interest of the company

0:11:41.840 --> 0:11:45.439
<v Speaker 1>that invested in you and is one of your biggest customers, etcetera.

0:11:45.480 --> 0:11:48.120
<v Speaker 1>And does that mean you're not the voting resources to

0:11:48.160 --> 0:11:51.200
<v Speaker 1>something else when perhaps you should be. It also raises

0:11:51.320 --> 0:11:54.880
<v Speaker 1>the question of whether there might be misaligned values. So, yeah,

0:11:54.920 --> 0:11:57.040
<v Speaker 1>the deal has been done out let's just say seven

0:11:57.200 --> 0:12:01.200
<v Speaker 1>billion valuation, but if that is with somebody else might

0:12:01.280 --> 0:12:05.040
<v Speaker 1>have been a four billion. Yeah, you're also creating customers

0:12:05.040 --> 0:12:09.120
<v Speaker 1>for you who are beholding by those contracts. And then

0:12:09.520 --> 0:12:12.960
<v Speaker 1>that like essentially creates regulatory risk as well. But regulators

0:12:12.960 --> 0:12:15.760
<v Speaker 1>may come alongo, actually we don't like that, and we're

0:12:15.800 --> 0:12:18.439
<v Speaker 1>not sure whether that's the best deal for the consumer.

0:12:18.920 --> 0:12:21.400
<v Speaker 1>And so, you know, all these risks are immerging up

0:12:21.400 --> 0:12:24.920
<v Speaker 1>perhaps we didn't have before. But the biggest one, probably

0:12:24.960 --> 0:12:27.720
<v Speaker 1>of all, is that these circular deals can create a

0:12:27.720 --> 0:12:31.040
<v Speaker 1>false impression of demand. And you may think that all

0:12:31.120 --> 0:12:33.679
<v Speaker 1>these companies are generating massive revenues, but if it's all

0:12:33.720 --> 0:12:36.640
<v Speaker 1>just moving around and slashing around, then if something falls

0:12:36.679 --> 0:12:39.080
<v Speaker 1>out of bed, then there could be wider implications and

0:12:39.120 --> 0:12:41.120
<v Speaker 1>that could become systemic.

0:12:52.840 --> 0:12:54.800
<v Speaker 2>The other thing I thought was interested in this week

0:12:54.920 --> 0:12:58.840
<v Speaker 2>is that athough cost we's fallen out of bed the

0:12:58.960 --> 0:13:01.840
<v Speaker 2>nasdyke to cup a bump, but not a big one.

0:13:02.840 --> 0:13:06.760
<v Speaker 2>It so far seems to be restricted largely to the

0:13:06.800 --> 0:13:10.040
<v Speaker 2>AI and the tech sector. And actually plenty of other

0:13:10.280 --> 0:13:14.040
<v Speaker 2>stocks are doing fine. They equal weighted S and P

0:13:14.240 --> 0:13:16.040
<v Speaker 2>is doing fine as there's in that's the S and

0:13:16.080 --> 0:13:20.520
<v Speaker 2>P that's not wholly invested in the tech sector. And

0:13:20.559 --> 0:13:23.120
<v Speaker 2>also the FOOTS one hundred is almost back at a

0:13:23.160 --> 0:13:28.040
<v Speaker 2>record high, the kind of laggard of the global stock markets.

0:13:28.520 --> 0:13:32.680
<v Speaker 2>But is this issue of how much does this spread

0:13:32.760 --> 0:13:36.000
<v Speaker 2>or could it spread beyond the tech sector? So who

0:13:36.040 --> 0:13:40.600
<v Speaker 2>else is involved in lending to these companies and if

0:13:40.640 --> 0:13:43.719
<v Speaker 2>something did break down there sort of where are the

0:13:43.880 --> 0:13:47.080
<v Speaker 2>contagion sort of vectors.

0:13:49.040 --> 0:13:52.040
<v Speaker 1>The demand means that the AI industry has had to

0:13:52.040 --> 0:13:54.360
<v Speaker 1>go to pretty much every corner of the credit market

0:13:54.400 --> 0:13:56.600
<v Speaker 1>and hand out the ball and kind of say, give

0:13:56.679 --> 0:14:00.600
<v Speaker 1>us a few quid what we need going for. And

0:14:00.960 --> 0:14:04.200
<v Speaker 1>so you're seeing everything from direct lending, which is a

0:14:04.200 --> 0:14:07.120
<v Speaker 1>private credit usually for infrastructure, so like just to build

0:14:07.160 --> 0:14:10.920
<v Speaker 1>a data center, and you're seeing investment grade. You're also

0:14:10.960 --> 0:14:14.040
<v Speaker 1>seeing these neo clouds like core Weave who I mentioned to,

0:14:14.160 --> 0:14:17.720
<v Speaker 1>are sub investment grade. Sometimes not all of them, but

0:14:17.760 --> 0:14:19.440
<v Speaker 1>some of them are sub investment grades. So that's the

0:14:19.520 --> 0:14:22.320
<v Speaker 1>high year market. The junk bonds that people might know

0:14:23.200 --> 0:14:27.880
<v Speaker 1>they're in the structured credit markets that are you know,

0:14:27.960 --> 0:14:31.120
<v Speaker 1>these are basically taking bond payments that are gue and

0:14:31.160 --> 0:14:33.480
<v Speaker 1>slicing them up by risk and selling them off to people.

0:14:36.320 --> 0:14:39.560
<v Speaker 2>It's didn't something like that hal in the mortgage market,

0:14:40.200 --> 0:14:41.280
<v Speaker 2>was it twenty years ago?

0:14:42.320 --> 0:14:44.720
<v Speaker 1>We haven't got the kind of the CDO level or

0:14:44.800 --> 0:14:52.440
<v Speaker 1>CDO squared level. But like where I would be concerned

0:14:52.480 --> 0:14:55.200
<v Speaker 1>around that. Lending in particular is around what in real

0:14:55.320 --> 0:14:58.440
<v Speaker 1>estate is called speculative lending, which means something else too

0:14:58.480 --> 0:15:01.760
<v Speaker 1>many people. But in real estate it's basically building the

0:15:01.760 --> 0:15:04.120
<v Speaker 1>stuff before you have a tenant, and if you can

0:15:04.160 --> 0:15:07.760
<v Speaker 1>get finance, and that from a bank or from a

0:15:07.760 --> 0:15:10.840
<v Speaker 1>private credit lender, that says something about bubble territory because

0:15:10.960 --> 0:15:13.920
<v Speaker 1>they are taking a complete risk that you are building

0:15:13.920 --> 0:15:15.440
<v Speaker 1>this thing over the course of five years and you

0:15:15.480 --> 0:15:18.560
<v Speaker 1>will find somebody to occupy it. If not, or if

0:15:18.600 --> 0:15:20.720
<v Speaker 1>the industry moves on and technology moves on, you can

0:15:20.800 --> 0:15:24.480
<v Speaker 1>end up with a very expensive elephant at the end,

0:15:24.520 --> 0:15:26.920
<v Speaker 1>which may not have much residual value for people.

0:15:27.200 --> 0:15:29.800
<v Speaker 2>If you build it, they will come only walks in

0:15:29.840 --> 0:15:31.080
<v Speaker 2>the field of dreams done that.

0:15:32.800 --> 0:15:34.560
<v Speaker 1>I mean, there can be a first moving thing where

0:15:34.600 --> 0:15:36.320
<v Speaker 1>you can get away with it. But if you're spending

0:15:36.320 --> 0:15:40.000
<v Speaker 1>five billion credits building a data center, I would want

0:15:40.320 --> 0:15:42.440
<v Speaker 1>to be more certain of that. But we are beginning

0:15:42.480 --> 0:15:45.400
<v Speaker 1>to see elements of that, and we're also beginning to

0:15:45.400 --> 0:15:53.640
<v Speaker 1>see terms being pushed within those deals for lending that

0:15:54.520 --> 0:15:57.800
<v Speaker 1>maybe are too generous to the hyperscouts. That is the

0:15:57.840 --> 0:16:00.600
<v Speaker 1>point of view of the credit markets. So you're starting

0:16:00.600 --> 0:16:03.000
<v Speaker 1>to hear things like after a certain amount of time,

0:16:03.160 --> 0:16:05.600
<v Speaker 1>if the project is delayed, then the hyperscat to come

0:16:05.680 --> 0:16:09.280
<v Speaker 1>back out. Now, if you've invested five years and spent

0:16:09.440 --> 0:16:11.400
<v Speaker 1>as I say, five billion dollars in billiing a data

0:16:11.400 --> 0:16:13.960
<v Speaker 1>center and it gets a bit delayed because let's just

0:16:13.960 --> 0:16:16.080
<v Speaker 1>say something got stuck in the straits of her moves

0:16:16.120 --> 0:16:19.280
<v Speaker 1>at the moment, that's a very big risk. And so

0:16:19.600 --> 0:16:21.760
<v Speaker 1>again that's part of what's happening with the credit markets

0:16:21.760 --> 0:16:25.440
<v Speaker 1>and the pullback. They're kind of reconsidering some of the

0:16:25.520 --> 0:16:27.640
<v Speaker 1>levels of risk that they're accepting at the moment.

0:16:28.400 --> 0:16:31.720
<v Speaker 2>I mean from that point of view, this arguably, as

0:16:31.720 --> 0:16:33.840
<v Speaker 2>long as it's not already going too far, maybe a

0:16:33.880 --> 0:16:36.880
<v Speaker 2>good thing, and then maybe it gets a arein end

0:16:36.920 --> 0:16:41.520
<v Speaker 2>of horns a bit before it does go properly payer shaped.

0:16:42.080 --> 0:16:44.040
<v Speaker 1>Yeah, and I think it was always going to happen.

0:16:45.280 --> 0:16:46.800
<v Speaker 1>And to be fair, like you look at something like

0:16:46.800 --> 0:16:49.120
<v Speaker 1>the Cosby and it's still up. You know, many of

0:16:49.160 --> 0:16:51.200
<v Speaker 1>the companies in are still up over one hundred percent

0:16:51.320 --> 0:16:55.680
<v Speaker 1>here today in some cases with AIA adjacent companies are

0:16:55.760 --> 0:16:57.880
<v Speaker 1>up three hundred percent for the year, so that you know,

0:16:58.960 --> 0:17:01.520
<v Speaker 1>and there was an act moment for a pause anyway.

0:17:02.160 --> 0:17:07.400
<v Speaker 1>And while we are seeing in terms of the Chinese evolution,

0:17:08.080 --> 0:17:10.639
<v Speaker 1>is that that was always going to happen as well.

0:17:10.920 --> 0:17:14.639
<v Speaker 1>And China's big advantages. It has cheap electricity, so the

0:17:14.680 --> 0:17:18.200
<v Speaker 1>tokenization is cheaper. So if you're a company that's also

0:17:18.400 --> 0:17:21.159
<v Speaker 1>been spending loads of money and AI and your staff,

0:17:21.160 --> 0:17:24.119
<v Speaker 1>it turns out are using it to convert excel files

0:17:24.119 --> 0:17:27.440
<v Speaker 1>into PDFs rather than actually, you know, using a much

0:17:27.480 --> 0:17:29.680
<v Speaker 1>cheaper technology for that, you're going to be scatting around

0:17:29.720 --> 0:17:33.200
<v Speaker 1>for cheaper prices. And so you know, China and the

0:17:33.280 --> 0:17:37.280
<v Speaker 1>Chinese LM's become a natural kind of success story from that,

0:17:37.880 --> 0:17:43.920
<v Speaker 1>and you know that was all to be expected, I

0:17:43.960 --> 0:17:47.080
<v Speaker 1>think when a certain reason, and it's just all happened

0:17:47.160 --> 0:17:48.880
<v Speaker 1>very quickly and all at once, as it does these

0:17:48.920 --> 0:17:51.439
<v Speaker 1>days and markets, and there's a bit of panic in

0:17:51.480 --> 0:17:54.520
<v Speaker 1>certain areas, particularly when it comes to retail money. You

0:17:54.600 --> 0:17:57.400
<v Speaker 1>and I have spent years writing about retail money being

0:17:57.600 --> 0:18:00.000
<v Speaker 1>hot money and how people can panic, and to be fair,

0:18:00.040 --> 0:18:02.479
<v Speaker 1>if I had leveraged several times, I've been looking at

0:18:02.520 --> 0:18:04.320
<v Speaker 1>that and kind of gone, oh my god. So it's

0:18:04.400 --> 0:18:06.720
<v Speaker 1>understandable in a way, but you know, if people are

0:18:06.760 --> 0:18:10.119
<v Speaker 1>sensible about how the investors still major opportunities there.

0:18:09.960 --> 0:18:12.160
<v Speaker 2>Yeah, I mean yeah. My heart goes out to the

0:18:12.280 --> 0:18:16.000
<v Speaker 2>various kind of rookie retail investors in South Korea who

0:18:16.000 --> 0:18:19.440
<v Speaker 2>are now looking at, you know, some really nasty losses.

0:18:20.040 --> 0:18:22.159
<v Speaker 2>I'm hoping that they were all young enough to bounce

0:18:22.240 --> 0:18:25.240
<v Speaker 2>back from it. The only thanks very much for this

0:18:25.600 --> 0:18:28.040
<v Speaker 2>was really helpful. The one other thing I was asking

0:18:28.080 --> 0:18:32.720
<v Speaker 2>about is on the debt side to flood, the kind

0:18:32.760 --> 0:18:35.199
<v Speaker 2>of the debt market, if you like, And we have

0:18:35.280 --> 0:18:38.959
<v Speaker 2>also seen a bit of equity issuings, and I suppose

0:18:39.000 --> 0:18:42.400
<v Speaker 2>the other thing I'm wondering about, is how much can

0:18:42.600 --> 0:18:48.200
<v Speaker 2>the market take whenever We've had decades of de equitization,

0:18:48.480 --> 0:18:52.400
<v Speaker 2>so companies buying back order shares or getting bought off

0:18:52.400 --> 0:18:57.159
<v Speaker 2>the market, and now we're going to get quite unusually

0:18:57.320 --> 0:19:01.440
<v Speaker 2>kind of net equity issues quite possibly year. I mean,

0:19:01.440 --> 0:19:05.000
<v Speaker 2>my design city about what might happen regarding the tightness

0:19:05.040 --> 0:19:09.320
<v Speaker 2>of my overall as They're just not going to be

0:19:10.000 --> 0:19:12.040
<v Speaker 2>enough to go around for all the hungry mouths.

0:19:12.080 --> 0:19:14.960
<v Speaker 1>So I think investors would be discerning. I mean, you

0:19:15.000 --> 0:19:18.920
<v Speaker 1>can tell, you can tell an equity story that people

0:19:18.960 --> 0:19:22.560
<v Speaker 1>will follow. I mean, just thinking within AI, how many

0:19:22.600 --> 0:19:24.400
<v Speaker 1>of these companies we had heard of three years ago?

0:19:24.720 --> 0:19:27.760
<v Speaker 1>Very few of Annie and yeah you know that, and

0:19:27.760 --> 0:19:30.400
<v Speaker 1>they may not have very much revenue, but people are

0:19:30.400 --> 0:19:34.320
<v Speaker 1>wenting to bet on a deal. But there's also a

0:19:34.359 --> 0:19:37.120
<v Speaker 1>thing of like it can go wrong quite quickly, and

0:19:37.320 --> 0:19:39.520
<v Speaker 1>a SpaceX is probably a good example of that, and

0:19:39.640 --> 0:19:42.320
<v Speaker 1>unless say, went wrong, but there was so much hype

0:19:42.359 --> 0:19:45.200
<v Speaker 1>and now it's obviously down since its IPO price, which

0:19:45.359 --> 0:19:49.359
<v Speaker 1>was ambitious in the first place. I looked recently and

0:19:49.400 --> 0:19:52.399
<v Speaker 1>the short interests on the stock was at near forty percent,

0:19:52.840 --> 0:19:55.240
<v Speaker 1>So people are being very bearish on the future for it,

0:19:55.520 --> 0:19:59.000
<v Speaker 1>and that's probably the big question for investors at the moment,

0:19:59.040 --> 0:20:02.280
<v Speaker 1>SpaceX being partially the NAI story with data centers in space,

0:20:02.320 --> 0:20:07.760
<v Speaker 1>et cetera. Is has that and what happened to that

0:20:07.920 --> 0:20:10.600
<v Speaker 1>stock closed the I P O window for a while,

0:20:11.480 --> 0:20:15.440
<v Speaker 1>and in that sense, then the focus would shift back

0:20:15.480 --> 0:20:18.879
<v Speaker 1>to the equity markets. I started to the credit markets

0:20:19.200 --> 0:20:22.439
<v Speaker 1>and the credit markets, and Sculptor I think it was,

0:20:22.560 --> 0:20:25.760
<v Speaker 1>put an investor letter out recently and they said they

0:20:25.840 --> 0:20:28.480
<v Speaker 1>talked about how credit capacity is needed. It's great, it's

0:20:28.480 --> 0:20:32.200
<v Speaker 1>the extent just as people are pulling back, and that's

0:20:32.200 --> 0:20:33.959
<v Speaker 1>something you have to be a bit fearful of. I

0:20:33.960 --> 0:20:37.760
<v Speaker 1>think now often we have taught these things in two

0:20:37.800 --> 0:20:40.679
<v Speaker 1>weeks later, it's the market is that over sold and

0:20:40.680 --> 0:20:44.240
<v Speaker 1>it's all right back up and often exceeding the previous

0:20:44.280 --> 0:20:47.760
<v Speaker 1>size obviously, but at the moment, it's definitely a moment.

0:20:49.119 --> 0:20:52.160
<v Speaker 1>It's just the length that that moment takes.

0:20:52.520 --> 0:20:55.280
<v Speaker 2>Oay, well, look thanks a lot, Neil. I think that

0:20:55.400 --> 0:20:56.320
<v Speaker 2>was really helpful.

0:20:56.440 --> 0:20:57.200
<v Speaker 1>I hope.

0:20:58.400 --> 0:21:00.439
<v Speaker 2>That the main point it was to try and explain

0:21:00.480 --> 0:21:02.480
<v Speaker 2>to people what's going on, and I think you've done

0:21:02.520 --> 0:21:06.879
<v Speaker 2>that excellently. And obviously the market can remain excitable for

0:21:06.960 --> 0:21:10.080
<v Speaker 2>longer than anyone can remain solvent, especially if you're invested

0:21:10.119 --> 0:21:13.560
<v Speaker 2>in a leverage VTF, so just be careful out there.

0:21:13.880 --> 0:21:23.200
<v Speaker 2>Thanks again, Neil, Thank you, thanks for listening this week's

0:21:23.200 --> 0:21:25.680
<v Speaker 2>Merton Talks Money Markets RAP. If you like a show,

0:21:25.800 --> 0:21:28.520
<v Speaker 2>rate review and subscribe whever you listen to podcasts, and

0:21:28.600 --> 0:21:30.840
<v Speaker 2>also be sure to follow me on x or Twitter

0:21:30.960 --> 0:21:34.600
<v Speaker 2>at Joint Underscore Stepic. This episode was produced by some

0:21:34.760 --> 0:21:38.399
<v Speaker 2>of Sadi and Moses and am. Questions and comments on

0:21:38.440 --> 0:21:41.159
<v Speaker 2>the show are always welcome. Our show email is Merrion

0:21:41.200 --> 0:21:45.359
<v Speaker 2>Money at Bloomberg dot Net. Special thanks to Neil Callen.