1
00:00:17,960 --> 00:00:20,840
Speaker 1: Hello, and welcome to this special edition of The Credit Edge.

2
00:00:20,880 --> 00:00:23,440
Speaker 1: My name is James Crombie. I'm a senior editor at Bloomberg.

3
00:00:23,640 --> 00:00:25,919
Speaker 1: We're here to discuss tech and AI, the story of

4
00:00:25,960 --> 00:00:28,600
Speaker 1: our time. We're joined by Rob Schiffman, who covers the

5
00:00:28,640 --> 00:00:30,960
Speaker 1: tech sector for Bloomberg Intelligence. How are you doing? Rob?

6
00:00:31,440 --> 00:00:31,960
Speaker 2: Never better?

7
00:00:32,080 --> 00:00:35,840
Speaker 1: James also delighted to have Annareg Renner, who covers tech

8
00:00:35,880 --> 00:00:37,760
Speaker 1: for BI as well. How's it going Anareg?

9
00:00:38,720 --> 00:00:40,480
Speaker 3: Fine? Thanks, I'd glad to be here.

10
00:00:41,280 --> 00:00:44,839
Speaker 1: A dynamic duo. Indeed, by all accounts, there is a

11
00:00:44,880 --> 00:00:47,519
Speaker 1: tech revolution going on. It's playing in surround sound and

12
00:00:47,560 --> 00:00:50,400
Speaker 1: impossible to ignore. The buildout is going to cost tech

13
00:00:50,440 --> 00:00:53,040
Speaker 1: and related companies trillions of dollars, a lot of it

14
00:00:53,080 --> 00:00:55,840
Speaker 1: to be raised in global capital markets. So Rob, I'm

15
00:00:55,840 --> 00:00:57,279
Speaker 1: going to start with you. Where are we in that

16
00:00:57,320 --> 00:00:59,000
Speaker 1: process and what have we learned so far?

17
00:00:59,320 --> 00:01:01,680
Speaker 2: Yeah, I'm like a little bit of a broken record,

18
00:01:01,880 --> 00:01:05,720
Speaker 2: like we are just beginning the amount of spending and

19
00:01:05,880 --> 00:01:09,959
Speaker 2: borrowing in order to get AI capacity as well as

20
00:01:10,000 --> 00:01:12,800
Speaker 2: products to where we need to get them. I don't

21
00:01:12,800 --> 00:01:17,000
Speaker 2: think we're anywhere close, So you know, we're moving up

22
00:01:17,000 --> 00:01:20,720
Speaker 2: our expectations for AI capex in the US over the

23
00:01:20,760 --> 00:01:24,160
Speaker 2: next five or six years from three trillion to four trillion.

24
00:01:24,480 --> 00:01:26,160
Speaker 2: I wouldn't be surprised by the end of the year

25
00:01:26,160 --> 00:01:28,480
Speaker 2: if we move that number up to five trillion. And

26
00:01:28,520 --> 00:01:30,559
Speaker 2: on the back end of that much spending, there's obviously

27
00:01:30,640 --> 00:01:33,560
Speaker 2: a ton of borrowing. So it's just really getting started.

28
00:01:33,600 --> 00:01:36,760
Speaker 2: And I think there's a lot of excess capacity in

29
00:01:36,800 --> 00:01:41,080
Speaker 2: the system, at least from the bond perspective, to take

30
00:01:41,120 --> 00:01:42,920
Speaker 2: all of that down to spend as much as they

31
00:01:42,959 --> 00:01:46,119
Speaker 2: need and not worry nearly as much as anog side

32
00:01:46,800 --> 00:01:48,680
Speaker 2: from the equity perspective, you.

33
00:01:48,640 --> 00:01:50,120
Speaker 1: Know, I'd like to be really specific when you bring

34
00:01:50,160 --> 00:01:52,200
Speaker 1: up numbers. Well, so it's going to be four trillion

35
00:01:52,640 --> 00:01:54,720
Speaker 1: in five or six, Yes, tell me which one, five

36
00:01:54,800 --> 00:01:55,320
Speaker 1: or six.

37
00:01:55,320 --> 00:01:58,320
Speaker 2: Yeah, it's it's four trillion plus through the end of

38
00:01:58,360 --> 00:02:01,800
Speaker 2: twenty thirty. Younumbers are moving so sort of quickly, like

39
00:02:01,840 --> 00:02:05,560
Speaker 2: we just saw through this last earning cycle, companies raising

40
00:02:05,600 --> 00:02:08,080
Speaker 2: numbers to levels that no one could have ever imagined,

41
00:02:08,600 --> 00:02:11,360
Speaker 2: except we're starting to imagine them, and we're actually seeing

42
00:02:11,400 --> 00:02:14,200
Speaker 2: it play out again through the amount of borrowing we've

43
00:02:14,200 --> 00:02:15,640
Speaker 2: seen just in the last couple of weeks.

44
00:02:15,800 --> 00:02:18,760
Speaker 3: And Eric, what do you say, you know, we see

45
00:02:18,800 --> 00:02:21,239
Speaker 3: the same exact thing, and you know, from our side,

46
00:02:22,360 --> 00:02:26,000
Speaker 3: almost every quarter matters because it's not just so the

47
00:02:26,040 --> 00:02:29,960
Speaker 3: even the you know, the number itself, but the tone

48
00:02:30,040 --> 00:02:33,520
Speaker 3: of management team is to which way are they leaning

49
00:02:33,560 --> 00:02:36,280
Speaker 3: towards in terms of the rate of growth that's going

50
00:02:36,320 --> 00:02:39,520
Speaker 3: to go up or art Because the important part is we,

51
00:02:39,720 --> 00:02:41,960
Speaker 3: as you know, you guys mentioned, we are in that

52
00:02:42,560 --> 00:02:45,680
Speaker 3: built phase right now. We also have to see the

53
00:02:45,720 --> 00:02:49,840
Speaker 3: diffusion phase or the use cases out there, because a

54
00:02:49,880 --> 00:02:55,200
Speaker 3: lot will depend on at what pace our industries embracing

55
00:02:55,240 --> 00:03:00,560
Speaker 3: these AI technologies and then how they are trying trans

56
00:03:00,720 --> 00:03:04,120
Speaker 3: voting or you could say transforming their own business to

57
00:03:04,200 --> 00:03:08,720
Speaker 3: embed motor AI capabilities into their application. So for us,

58
00:03:08,800 --> 00:03:11,720
Speaker 3: I think a lot matters is which way are the

59
00:03:11,880 --> 00:03:14,280
Speaker 3: leaning towards the rate of growth of CAPEX.

60
00:03:15,040 --> 00:03:19,080
Speaker 2: Yeah, it's very much a question of if they build it,

61
00:03:19,200 --> 00:03:23,400
Speaker 2: they will come versus they've already come. The demand is there,

62
00:03:23,760 --> 00:03:26,200
Speaker 2: so now they're building it, And I think that's what

63
00:03:26,280 --> 00:03:28,720
Speaker 2: people are questioning is whether or not that rate of

64
00:03:28,760 --> 00:03:32,560
Speaker 2: demand and the actual flow through of spending to cash

65
00:03:32,560 --> 00:03:34,800
Speaker 2: flow is ever going to be realized. But I think

66
00:03:34,840 --> 00:03:39,160
Speaker 2: from the largest Hyperscaler's perspective, all we're seeing is that

67
00:03:39,440 --> 00:03:43,120
Speaker 2: they can't build fast enough and they can't raise capital

68
00:03:43,200 --> 00:03:46,680
Speaker 2: fast enough in order to put their flag in the ground.

69
00:03:47,120 --> 00:03:49,160
Speaker 1: But as you know, we've talked about the volume of

70
00:03:49,160 --> 00:03:52,080
Speaker 1: cash that they need to build this stuff, we pretty

71
00:03:52,120 --> 00:03:55,800
Speaker 1: much don't know when or how they'll get the money

72
00:03:55,800 --> 00:03:57,360
Speaker 1: back for it. You say that there is demand, but

73
00:03:57,720 --> 00:03:59,920
Speaker 1: the people actually buying the products already use seeing that

74
00:04:00,160 --> 00:04:03,320
Speaker 1: real cash flow coming back to the companies.

75
00:04:03,440 --> 00:04:05,320
Speaker 2: Well, it's a little bit of yes and no. So

76
00:04:06,120 --> 00:04:08,280
Speaker 2: where are they seeing cash flow today. They're seeing cash

77
00:04:08,280 --> 00:04:15,040
Speaker 2: flow today from more of their legacy businesses, so traditional cloud, advertising,

78
00:04:15,760 --> 00:04:21,280
Speaker 2: hardware services, and that's what's funding this build. We're starting

79
00:04:21,320 --> 00:04:23,920
Speaker 2: to see AI revenues an a ROD could be more

80
00:04:23,920 --> 00:04:26,840
Speaker 2: specific on who's generating wants. So we're seeing tens of

81
00:04:26,839 --> 00:04:31,280
Speaker 2: billions of AI revenues, but you know, very early stages

82
00:04:31,320 --> 00:04:34,160
Speaker 2: of the game when it comes to the AI collection

83
00:04:34,360 --> 00:04:38,080
Speaker 2: versus the spend. That being said, you know, we've said

84
00:04:38,080 --> 00:04:40,880
Speaker 2: all along the starting point for us is the mount

85
00:04:41,000 --> 00:04:45,360
Speaker 2: rushmore of credits. There's so much capacity, there's so much

86
00:04:45,440 --> 00:04:49,440
Speaker 2: room on these balance sheets to add leverage to fund

87
00:04:49,680 --> 00:04:52,919
Speaker 2: long term growth. Most of these companies, they're just not

88
00:04:53,120 --> 00:04:59,159
Speaker 2: worried right now about concern about equity prices. They are

89
00:04:59,200 --> 00:05:02,560
Speaker 2: building long term businesses, so they're willing to take short

90
00:05:02,680 --> 00:05:05,359
Speaker 2: term hits with the confidence that the cash flow is

91
00:05:05,400 --> 00:05:09,360
Speaker 2: going to be there over that four or five year period.

92
00:05:09,120 --> 00:05:11,320
Speaker 1: That I mentioned and Directorge to weigh in on that.

93
00:05:11,800 --> 00:05:13,719
Speaker 3: Yeah, James, So you know, for us, we divide the

94
00:05:13,720 --> 00:05:16,159
Speaker 3: whole world in two buckets when it look at the

95
00:05:16,320 --> 00:05:20,120
Speaker 3: realization of revenue. One is the training revenue that they're

96
00:05:20,120 --> 00:05:23,359
Speaker 3: getting for these large language models, and the other is

97
00:05:23,360 --> 00:05:28,159
Speaker 3: the inference revenue. As of right now, a large portion

98
00:05:28,320 --> 00:05:33,000
Speaker 3: of what we are seeing that these people are investing

99
00:05:33,040 --> 00:05:35,800
Speaker 3: in is to train these large language models. I mean

100
00:05:35,880 --> 00:05:39,120
Speaker 3: you see you know Core viaves, Backlog or Oracle for

101
00:05:39,160 --> 00:05:43,160
Speaker 3: that matter, and so forth. That's a large portion of

102
00:05:43,160 --> 00:05:49,440
Speaker 3: that is the training revenue for some of these big companies.

103
00:05:49,760 --> 00:05:52,400
Speaker 3: So as long as the funding environment is great and

104
00:05:52,480 --> 00:05:55,159
Speaker 3: these companies can go out and raise billions of dollars

105
00:05:55,240 --> 00:05:57,840
Speaker 3: and then they move on to the cloud providers and say, well,

106
00:05:57,880 --> 00:05:59,960
Speaker 3: I'm going to borrow your infrastructure to train these more.

107
00:06:00,600 --> 00:06:04,440
Speaker 3: You know, that's one sided the equation for us, It's okay,

108
00:06:04,480 --> 00:06:07,560
Speaker 3: there's nothing wrong in that dollar coming in. But for us,

109
00:06:07,600 --> 00:06:11,720
Speaker 3: that's not an infinite revenue basis, or it's not the

110
00:06:11,760 --> 00:06:15,279
Speaker 3: sticky revenue base that we are looking for. We want

111
00:06:15,279 --> 00:06:18,960
Speaker 3: to see inference revenue that's coming from the end markets

112
00:06:18,960 --> 00:06:21,480
Speaker 3: of companies. So let's say, for the sake of discussion,

113
00:06:21,520 --> 00:06:24,880
Speaker 3: that a bank decides to embed an AI more AI

114
00:06:24,960 --> 00:06:27,880
Speaker 3: into their products. They are going to create a chat

115
00:06:27,920 --> 00:06:32,320
Speaker 3: pot which is going to be used to buy internally

116
00:06:32,520 --> 00:06:35,760
Speaker 3: or by externally their clients to pull in all the

117
00:06:35,800 --> 00:06:40,120
Speaker 3: information for better customer service, better product development, et cetera,

118
00:06:40,160 --> 00:06:43,640
Speaker 3: et cetera. It's those revenues that we want to see

119
00:06:43,680 --> 00:06:47,080
Speaker 3: because that's where the true value is for us. When

120
00:06:47,080 --> 00:06:49,400
Speaker 3: you look at a company or a big banks that

121
00:06:50,160 --> 00:06:53,359
Speaker 3: starts an application to build an application, let's say on

122
00:06:53,400 --> 00:06:59,159
Speaker 3: Amazon's platform using anthropics model, that particular app is going

123
00:06:59,200 --> 00:07:02,360
Speaker 3: to be the one that gives me recurring revenue down

124
00:07:02,440 --> 00:07:05,760
Speaker 3: the road. Think of this as something similar that happened

125
00:07:05,800 --> 00:07:09,480
Speaker 3: in the past, the previous big boom of cloud computing,

126
00:07:09,520 --> 00:07:12,480
Speaker 3: where you know, you created the uber app or a

127
00:07:12,680 --> 00:07:15,880
Speaker 3: liftop on some of these cloud providers, and then the

128
00:07:15,920 --> 00:07:19,000
Speaker 3: app will grow and you're going to make those revenue

129
00:07:19,000 --> 00:07:22,000
Speaker 3: along with it. So that's the business we think is

130
00:07:22,000 --> 00:07:24,800
Speaker 3: the more important one, but that's going to take time

131
00:07:24,840 --> 00:07:27,720
Speaker 3: because we are just in the initial phase of those

132
00:07:28,280 --> 00:07:29,160
Speaker 3: revenues coming in.

133
00:07:29,800 --> 00:07:33,040
Speaker 2: Yeah, so we're going to shift to negative net free

134
00:07:33,040 --> 00:07:35,040
Speaker 2: cash flow for a lot of these companies. So that's

135
00:07:35,360 --> 00:07:39,120
Speaker 2: operating cash flow, less capex, less shareholder returns, and that's

136
00:07:39,120 --> 00:07:41,560
Speaker 2: why you're going to see some more funding. So you know,

137
00:07:41,640 --> 00:07:44,480
Speaker 2: it's a lot of skiing uphill for the next few years.

138
00:07:44,480 --> 00:07:46,880
Speaker 2: But the thing is, once you can reach that top

139
00:07:46,920 --> 00:07:49,880
Speaker 2: of the mountain, that inflection point, Wow, just imagine the

140
00:07:49,880 --> 00:07:52,520
Speaker 2: type of cash flows that you can generate once the

141
00:07:52,600 --> 00:07:56,880
Speaker 2: capital is already spent and you can really start monetizing.

142
00:07:57,000 --> 00:08:01,000
Speaker 2: These companies become monstrous free cash flow. Ma, I'm as

143
00:08:01,000 --> 00:08:03,160
Speaker 2: bullish as I've ever been. I just think in a

144
00:08:03,200 --> 00:08:05,680
Speaker 2: handful of years we're going to go back to this

145
00:08:05,760 --> 00:08:07,640
Speaker 2: question is what are these companies going to be doing

146
00:08:07,680 --> 00:08:10,280
Speaker 2: with all their cash. They're going to have hundreds of

147
00:08:10,280 --> 00:08:12,840
Speaker 2: billions of dollars of excess cash, and now we're going

148
00:08:12,880 --> 00:08:15,000
Speaker 2: to be questioning if they invested enough or should they

149
00:08:15,040 --> 00:08:16,000
Speaker 2: have invested more.

150
00:08:16,640 --> 00:08:19,720
Speaker 1: Again, you assume that they get paid back and the

151
00:08:19,760 --> 00:08:22,640
Speaker 1: technology is not obsolete or replaced with something much cheaper

152
00:08:22,640 --> 00:08:25,680
Speaker 1: from China, or you know, something else displaces them. Things

153
00:08:25,720 --> 00:08:27,520
Speaker 1: are changing so fast. I don't know how we can

154
00:08:27,840 --> 00:08:31,880
Speaker 1: move into the future with such certainty. And then, you know,

155
00:08:31,880 --> 00:08:33,959
Speaker 1: it seems on the one handred it's a good equity story.

156
00:08:33,960 --> 00:08:36,040
Speaker 1: But on the debt side, why am I buying one

157
00:08:36,120 --> 00:08:38,240
Speaker 1: hundred g of paper from Google?

158
00:08:39,440 --> 00:08:41,480
Speaker 2: Yeah? Well, listen, here's a there's a there's a little

159
00:08:41,480 --> 00:08:43,720
Speaker 2: bit to break down there. First, Let's just think about

160
00:08:43,760 --> 00:08:47,840
Speaker 2: the type of capacity a Google has SMP when they

161
00:08:47,840 --> 00:08:50,560
Speaker 2: affirm their double A plus rating this morning said that

162
00:08:50,679 --> 00:08:53,160
Speaker 2: they have one hundred They can add one hundred and

163
00:08:53,240 --> 00:08:58,199
Speaker 2: eighty billion of incremental net debt before triggering a downgrade.

164
00:08:58,320 --> 00:09:01,240
Speaker 2: That's a downgrade for double A plus, So what's the

165
00:09:01,320 --> 00:09:03,680
Speaker 2: number to go to? Triple B is at a trillion dollars.

166
00:09:03,920 --> 00:09:07,040
Speaker 2: So the amount of capacity that these companies have to

167
00:09:07,280 --> 00:09:10,800
Speaker 2: spend and not necessarily make everything back right away is

168
00:09:11,120 --> 00:09:13,720
Speaker 2: very high. Why is Google barring a hundred year bond.

169
00:09:13,840 --> 00:09:15,199
Speaker 2: None of us are going to be here in a

170
00:09:15,280 --> 00:09:17,360
Speaker 2: hundred years, so we certainly don't know if we're going

171
00:09:17,440 --> 00:09:19,480
Speaker 2: to get paid off or not. But the reality is

172
00:09:19,520 --> 00:09:22,199
Speaker 2: from the bond market perspective. If you think about the

173
00:09:22,320 --> 00:09:25,439
Speaker 2: duration of one hundred year versus a more traditional long

174
00:09:25,480 --> 00:09:28,000
Speaker 2: bomb like a thirty year, it's effectively the same. It

175
00:09:28,240 --> 00:09:31,680
Speaker 2: sounds really cool, it's like, oh my gosh, like who's

176
00:09:31,679 --> 00:09:33,560
Speaker 2: ever going to be around for one hundred years? Like

177
00:09:33,600 --> 00:09:37,280
Speaker 2: how many legacy tech companies have disappeared? But when you

178
00:09:37,320 --> 00:09:40,280
Speaker 2: actually go down to the simple math, there's not really

179
00:09:40,360 --> 00:09:42,600
Speaker 2: much of a difference between one hundred year and a

180
00:09:42,640 --> 00:09:46,360
Speaker 2: thirty year my senses, there is probably reverse incurry. Somebody

181
00:09:46,400 --> 00:09:48,600
Speaker 2: came in and said, you know what, we'd love to

182
00:09:48,640 --> 00:09:52,000
Speaker 2: take down this long dated asset. We've got long dated

183
00:09:52,280 --> 00:09:54,240
Speaker 2: liabilities and we'll just match it up and they're going

184
00:09:54,280 --> 00:09:56,080
Speaker 2: to put it away and it's going to go away.

185
00:09:56,720 --> 00:09:59,079
Speaker 2: So that doesn't add concern to me whatsoever.

186
00:10:00,240 --> 00:10:02,960
Speaker 1: Just you know, shorter than that, I mean thirty years,

187
00:10:02,960 --> 00:10:05,480
Speaker 1: forty years? Does that make sense in the same way

188
00:10:05,600 --> 00:10:07,160
Speaker 1: I mean using your bone.

189
00:10:06,880 --> 00:10:10,160
Speaker 2: Mouth, Yeah, of course it does. I mean, look at again,

190
00:10:10,559 --> 00:10:13,840
Speaker 2: we're looking at credit quality. Credit quality of Microsoft is

191
00:10:13,920 --> 00:10:17,319
Speaker 2: higher than the US government. Why are you buying treasuries

192
00:10:17,360 --> 00:10:19,560
Speaker 2: if you think that there is a US treasury going

193
00:10:19,640 --> 00:10:22,400
Speaker 2: to be around in thirty years maybe probably. I guess

194
00:10:22,440 --> 00:10:24,839
Speaker 2: they can print money, but it's the same sort of question.

195
00:10:24,880 --> 00:10:27,040
Speaker 2: I just think when it boils down to it, it's

196
00:10:27,080 --> 00:10:30,240
Speaker 2: all about cashflow, and that the probability of that the

197
00:10:30,280 --> 00:10:31,960
Speaker 2: cash flows are going to be there at that time

198
00:10:32,559 --> 00:10:36,040
Speaker 2: are super super high, even if they're making massive mistakes.

199
00:10:36,080 --> 00:10:38,679
Speaker 2: Like to me, you know, it's much different. There's a

200
00:10:39,080 --> 00:10:41,360
Speaker 2: bifurcated way of looking at this, and an a Rock's

201
00:10:41,400 --> 00:10:44,160
Speaker 2: perspective is going to be meaningfully different than mine. That

202
00:10:44,640 --> 00:10:47,560
Speaker 2: if the returns on investment are much lower than what

203
00:10:47,600 --> 00:10:51,960
Speaker 2: the market is anticipating now, equity prices could go down dramatically.

204
00:10:52,040 --> 00:10:56,120
Speaker 2: You're actually seeing that every quarter. If a Microsoft grows

205
00:10:56,160 --> 00:10:58,520
Speaker 2: at forty percent, but the market thought they were supposed

206
00:10:58,520 --> 00:11:01,880
Speaker 2: to grow their Azure business at forty two, they go,

207
00:11:02,080 --> 00:11:04,840
Speaker 2: they lose five hundred billion dollars in market cap. Bonds

208
00:11:04,880 --> 00:11:07,960
Speaker 2: don't move, you know, the corporate bonds in general, James,

209
00:11:07,960 --> 00:11:10,440
Speaker 2: you know this better than everyone. We're near historical tights,

210
00:11:10,640 --> 00:11:12,680
Speaker 2: and even though tech has wind out over the last

211
00:11:13,160 --> 00:11:16,120
Speaker 2: three to six months, it's still near the tightest trading

212
00:11:16,200 --> 00:11:19,240
Speaker 2: sector of the corporate bond market at historical all time tight.

213
00:11:19,320 --> 00:11:22,040
Speaker 2: So the bond market is telling us something different. We're

214
00:11:22,080 --> 00:11:25,160
Speaker 2: not worried, and there's ways of seeing that. One is

215
00:11:25,600 --> 00:11:28,280
Speaker 2: where bonds are trading today. Two is the type of

216
00:11:28,320 --> 00:11:31,040
Speaker 2: demand you're seeing. So I'm not shocked that a double

217
00:11:31,040 --> 00:11:33,560
Speaker 2: A rated name is trying to raise fifteen billion dollars

218
00:11:33,559 --> 00:11:35,720
Speaker 2: and has one hundred billion dollars in demand. But if

219
00:11:35,760 --> 00:11:38,280
Speaker 2: you really thought that there was concern about, you know,

220
00:11:38,320 --> 00:11:40,560
Speaker 2: can these companies raise money? Look at Oracle. They were

221
00:11:40,559 --> 00:11:43,360
Speaker 2: the poster child for what everybody thought was going wrong

222
00:11:44,040 --> 00:11:46,560
Speaker 2: in the AI tech space, spending so much more money

223
00:11:46,600 --> 00:11:50,520
Speaker 2: that they have, leveraging up, no certainty that Open AI

224
00:11:50,679 --> 00:11:53,800
Speaker 2: is ever going to pay them, the potential of the

225
00:11:53,960 --> 00:11:58,400
Speaker 2: largest company in ig falling to junk. And what happens

226
00:11:58,440 --> 00:12:00,319
Speaker 2: They go out and do a twenty five billion dollar

227
00:12:00,360 --> 00:12:02,800
Speaker 2: bond deal over one hundred and thirty billion dollars demand,

228
00:12:03,280 --> 00:12:06,200
Speaker 2: issue twenty five billion dollars of equity, and we have

229
00:12:06,240 --> 00:12:09,120
Speaker 2: an equity price it's up ten percent today so and

230
00:12:09,200 --> 00:12:11,880
Speaker 2: bonds are tighter. So uh, I just think that if

231
00:12:11,880 --> 00:12:13,920
Speaker 2: you're looking at it through an equity lens, there's so

232
00:12:14,040 --> 00:12:16,400
Speaker 2: much more that can go wrong. Then I think what

233
00:12:16,480 --> 00:12:19,000
Speaker 2: can go happen from a bond holder's Lenes.

234
00:12:19,040 --> 00:12:20,040
Speaker 1: Are you worried down to EG?

235
00:12:21,160 --> 00:12:24,640
Speaker 3: So here's the thing. You're right. As an equity analyst,

236
00:12:24,679 --> 00:12:27,200
Speaker 3: I do worry about a handful of things, and I

237
00:12:27,200 --> 00:12:29,599
Speaker 3: think the biggest thing for us is we need to

238
00:12:29,640 --> 00:12:34,040
Speaker 3: see the cost curve of token generation go down a lot.

239
00:12:34,880 --> 00:12:37,520
Speaker 3: To me, that's the biggest risk we have right now,

240
00:12:37,520 --> 00:12:41,240
Speaker 3: because I'm not concerned that Microsoft is spending you know,

241
00:12:41,240 --> 00:12:44,360
Speaker 3: one hundred plus billion dollars, but my I'm concerned that,

242
00:12:44,640 --> 00:12:47,560
Speaker 3: you know, fifty sixty percent of that is going towards

243
00:12:47,600 --> 00:12:50,960
Speaker 3: buying in vidio chips that become obsolete in six years.

244
00:12:51,200 --> 00:12:54,280
Speaker 3: So that that, to me, is an issue because that

245
00:12:54,440 --> 00:12:57,559
Speaker 3: means you can only run some of these transactions on

246
00:12:57,760 --> 00:13:02,000
Speaker 3: very expensive chips. Now, I'm not saying they need to

247
00:13:02,000 --> 00:13:05,479
Speaker 3: buy it lower chips. They would buy what the system demands.

248
00:13:05,920 --> 00:13:08,680
Speaker 3: But something needs to happen on a from a tech

249
00:13:08,720 --> 00:13:11,640
Speaker 3: point of view, so that this cost curve goes down.

250
00:13:12,080 --> 00:13:15,280
Speaker 3: Whether that happens on the chip level, software level, a

251
00:13:15,320 --> 00:13:18,400
Speaker 3: combination of two, all of those things need to happen

252
00:13:18,480 --> 00:13:21,440
Speaker 3: because in order for us to see mass adoption of

253
00:13:21,559 --> 00:13:24,640
Speaker 3: some of these technologies, that token cost has to go down.

254
00:13:24,679 --> 00:13:26,960
Speaker 3: So that is the only concern we have I would

255
00:13:26,960 --> 00:13:30,640
Speaker 3: say in the short term. Longer term, we have seen

256
00:13:30,679 --> 00:13:34,040
Speaker 3: this movie play very well in the cloud era if

257
00:13:34,080 --> 00:13:36,959
Speaker 3: you go back and see how AWS persformed, and in fact,

258
00:13:37,640 --> 00:13:41,240
Speaker 3: for most of us, we didn't see the negative you

259
00:13:41,280 --> 00:13:43,600
Speaker 3: could say, free cash flow of that or the high

260
00:13:43,600 --> 00:13:46,720
Speaker 3: capex that goes into it because it was under the

261
00:13:46,920 --> 00:13:50,000
Speaker 3: Amazon umbrella. Then boom one year they come out and

262
00:13:50,040 --> 00:13:53,560
Speaker 3: show how great this business is. And now AWS before

263
00:13:53,600 --> 00:13:56,880
Speaker 3: the investment started, you know, I think it the peak

264
00:13:57,000 --> 00:14:00,800
Speaker 3: margins of AWS was somewhere north of thirty five. So

265
00:14:01,200 --> 00:14:04,360
Speaker 3: these are profitable businesses once you scale them up. But

266
00:14:04,480 --> 00:14:06,600
Speaker 3: at the same time, the concern for us is that

267
00:14:06,880 --> 00:14:08,360
Speaker 3: the cost CUD needs to come down.

268
00:14:09,360 --> 00:14:11,120
Speaker 1: I'm glad you mentioned the chips because there seems very

269
00:14:11,160 --> 00:14:14,000
Speaker 1: global chip war going on. There's a real power struggle,

270
00:14:14,640 --> 00:14:17,600
Speaker 1: which is you know, geopolitical in some ways. There is

271
00:14:17,760 --> 00:14:21,600
Speaker 1: potential for a lot of political meddling in and all

272
00:14:21,640 --> 00:14:24,080
Speaker 1: this stuff. What's the political risk for the for the market?

273
00:14:25,600 --> 00:14:27,920
Speaker 3: I mean it is high. I get the chip part

274
00:14:27,920 --> 00:14:30,840
Speaker 3: of it, and again it's it's going to be a

275
00:14:31,240 --> 00:14:34,280
Speaker 3: you know, shortage problem in whether it's memory in some

276
00:14:34,400 --> 00:14:38,360
Speaker 3: other areas. I I don't have any strong opinion on

277
00:14:38,960 --> 00:14:41,040
Speaker 3: you know, the fight between US and China on it.

278
00:14:41,160 --> 00:14:44,600
Speaker 3: I think, you know, from our side, I don't see

279
00:14:44,800 --> 00:14:48,520
Speaker 3: US companies using a Chinese model to build their application,

280
00:14:48,960 --> 00:14:51,800
Speaker 3: and I don't expect the Chinese companies to use a

281
00:14:51,920 --> 00:14:54,080
Speaker 3: US model. I mean it's a and again, you know,

282
00:14:54,120 --> 00:14:57,160
Speaker 3: there may be advantages of one AI over the others,

283
00:14:57,160 --> 00:14:59,640
Speaker 3: but I think they're both going to be in you know,

284
00:14:59,680 --> 00:15:02,720
Speaker 3: playing in a or swimming in different ponds. So I

285
00:15:03,000 --> 00:15:05,320
Speaker 3: don't spend that much time, to be honest, thinking about it.

286
00:15:05,960 --> 00:15:07,960
Speaker 2: Yeah, this to me is actually the biggest risk that's

287
00:15:08,000 --> 00:15:10,200
Speaker 2: out there. I'm not worried about sort of day to

288
00:15:10,280 --> 00:15:14,360
Speaker 2: day fundamentals and whether or not AWS or AZURE can

289
00:15:14,400 --> 00:15:17,240
Speaker 2: grow at you know, twenty five percent for all eternity

290
00:15:17,360 --> 00:15:20,680
Speaker 2: or not, or whether it's twenty eight percent, but it's

291
00:15:20,680 --> 00:15:23,560
Speaker 2: it's it's definitely much more tail risk and macro risk

292
00:15:23,680 --> 00:15:27,680
Speaker 2: that if there is some sort of real geopolitical issues.

293
00:15:28,160 --> 00:15:32,720
Speaker 2: It's hard to make chips when it's being done primarily overseas.

294
00:15:32,760 --> 00:15:35,440
Speaker 2: I mean, the Chips Act was intended to get more

295
00:15:35,480 --> 00:15:39,280
Speaker 2: manufacturing in the US, but it's you know, it's just

296
00:15:39,320 --> 00:15:42,560
Speaker 2: a couple of drips. You know, into an ocean. So

297
00:15:43,440 --> 00:15:47,200
Speaker 2: that's where I really worry if there's much bigger geopolitical tensions,

298
00:15:47,280 --> 00:15:50,960
Speaker 2: if China, you know, listen, everyone really sort of needs

299
00:15:51,000 --> 00:15:53,040
Speaker 2: to work together here, it doesn't, it doesn't help to

300
00:15:53,360 --> 00:15:55,760
Speaker 2: blow each other up. So there's always going to be

301
00:15:55,800 --> 00:15:57,840
Speaker 2: a lot of rhetoric. But that's where it does get

302
00:15:57,840 --> 00:15:59,440
Speaker 2: a little bit scary. You actually do you have to

303
00:15:59,480 --> 00:16:02,480
Speaker 2: make stuff, and you know, if you're making the vast

304
00:16:02,520 --> 00:16:08,960
Speaker 2: majority of the most expensive needed components overseas uh, particularly

305
00:16:08,960 --> 00:16:13,480
Speaker 2: particularly in UH in Taiwan, those sort of risks are high,

306
00:16:13,600 --> 00:16:16,680
Speaker 2: and that's something that you can never really get away from.

307
00:16:17,200 --> 00:16:18,880
Speaker 1: I mean, just it. Also in these sort of broader

308
00:16:18,920 --> 00:16:22,040
Speaker 1: market impact to these companies. They obviously dominate the equity

309
00:16:22,160 --> 00:16:24,520
Speaker 1: indexes and the you know, the mag seven moves the

310
00:16:24,520 --> 00:16:27,000
Speaker 1: whole market every day, but on the credit side, they're

311
00:16:27,040 --> 00:16:30,000
Speaker 1: relatively small issues right now, they are expected to get

312
00:16:30,000 --> 00:16:31,600
Speaker 1: a lot bigger if you know, the numbers that you're

313
00:16:31,640 --> 00:16:35,400
Speaker 1: giving us the top rubber actually true, in which case

314
00:16:35,440 --> 00:16:38,160
Speaker 1: does that reshape the entire investment grade credit market?

315
00:16:38,560 --> 00:16:42,400
Speaker 2: No, not really. I mean technology is some low double

316
00:16:42,440 --> 00:16:45,760
Speaker 2: digit percentage of the high grade index, and you know,

317
00:16:45,760 --> 00:16:48,360
Speaker 2: the high grade index is really big. So if you

318
00:16:48,440 --> 00:16:51,360
Speaker 2: add you know, two or three hundred odd plus a

319
00:16:51,440 --> 00:16:55,320
Speaker 2: year in IG bonds, you know, it also doesn't move

320
00:16:55,360 --> 00:16:58,239
Speaker 2: the needle that much. We're going to be well behind

321
00:16:58,640 --> 00:17:03,840
Speaker 2: UH financials. And you know, as long as again, as

322
00:17:03,880 --> 00:17:06,800
Speaker 2: long as it's dominated by the highest rated credits, I

323
00:17:06,800 --> 00:17:08,960
Speaker 2: don't think there's that much more of a concern. So

324
00:17:09,680 --> 00:17:11,760
Speaker 2: the real issue is not so much as part of

325
00:17:11,800 --> 00:17:14,400
Speaker 2: the index, it's just do people get full on these names?

326
00:17:14,440 --> 00:17:16,520
Speaker 2: How much exposure do you want to have to any

327
00:17:16,560 --> 00:17:19,920
Speaker 2: individual name? Like if you were already long Oracle going

328
00:17:19,960 --> 00:17:21,199
Speaker 2: into the last bond deal, you know they had one

329
00:17:21,240 --> 00:17:23,439
Speaker 2: hundred million dollars in bonds outstanding, Like how do they

330
00:17:23,440 --> 00:17:25,879
Speaker 2: price another twenty five billion? And I think this is

331
00:17:25,880 --> 00:17:28,640
Speaker 2: actually very simple because people always say, isn't the market

332
00:17:28,760 --> 00:17:30,359
Speaker 2: just going to say we're full and we can't be

333
00:17:30,400 --> 00:17:33,280
Speaker 2: overexposed to any of these individual names. And this is

334
00:17:33,359 --> 00:17:38,000
Speaker 2: where price matters. If you price something cheap enough, they

335
00:17:38,040 --> 00:17:41,760
Speaker 2: will come. So that's where we certainly learn that throughout history.

336
00:17:41,800 --> 00:17:43,560
Speaker 2: And I just think, you know, if you try to

337
00:17:43,600 --> 00:17:48,000
Speaker 2: price a forty year alphabet bond a couple of years ago,

338
00:17:48,480 --> 00:17:50,359
Speaker 2: you know what might cost might have cost half as

339
00:17:50,400 --> 00:17:52,800
Speaker 2: much as it costs today. But within in the big

340
00:17:52,840 --> 00:17:57,119
Speaker 2: picture of absolute cost, total coupons, total yields, it's really

341
00:17:57,160 --> 00:17:59,480
Speaker 2: not that much. And in order to get more people

342
00:17:59,480 --> 00:18:02,359
Speaker 2: to buy, you make it ten or fifteen basis points cheaper,

343
00:18:02,840 --> 00:18:05,000
Speaker 2: and you're going to get you're going to get that interest.

344
00:18:05,000 --> 00:18:07,600
Speaker 2: So I just don't I don't think that argument really

345
00:18:07,680 --> 00:18:10,800
Speaker 2: works or holds up. And again, we're starting from a

346
00:18:10,800 --> 00:18:14,000
Speaker 2: different base. We're starting from the highest quality basis of

347
00:18:14,040 --> 00:18:16,879
Speaker 2: any corporate credits that exist. And if you're ever going

348
00:18:16,920 --> 00:18:19,879
Speaker 2: to take risks in any of these names, this is

349
00:18:20,000 --> 00:18:20,719
Speaker 2: space to do it.

350
00:18:21,080 --> 00:18:24,120
Speaker 1: Do you expect the wholesale winding out of the tech

351
00:18:24,160 --> 00:18:26,720
Speaker 1: sector because we've seen it trade wide to the IG

352
00:18:26,840 --> 00:18:29,919
Speaker 1: index it generally trades tie to. Is that the effect

353
00:18:30,000 --> 00:18:30,639
Speaker 1: of all the supply?

354
00:18:30,800 --> 00:18:32,879
Speaker 2: I think if you just listen to the talking heads

355
00:18:32,960 --> 00:18:36,560
Speaker 2: on whether it's Bloomberg or other financial stations, you would

356
00:18:36,560 --> 00:18:39,560
Speaker 2: have thought we'd be distressed by now. I think it's

357
00:18:40,200 --> 00:18:43,719
Speaker 2: easy to cry wolf, but the reality is not the

358
00:18:43,760 --> 00:18:47,520
Speaker 2: same as the perception here. And there's this, you know,

359
00:18:47,640 --> 00:18:50,679
Speaker 2: I think we finally stopped talking about an AI bubble.

360
00:18:50,800 --> 00:18:53,359
Speaker 2: You know, that was the you know, I knew that

361
00:18:53,440 --> 00:18:57,320
Speaker 2: was overdone when my mother asked me in December at

362
00:18:57,359 --> 00:19:00,320
Speaker 2: Knica whether or not we're in an AI bubble, and

363
00:19:00,520 --> 00:19:02,480
Speaker 2: I told her she doesn't even know how to spell AI, Like,

364
00:19:02,520 --> 00:19:04,680
Speaker 2: what are you worried about? But that's that's the that's

365
00:19:04,680 --> 00:19:06,400
Speaker 2: the sort of talking point I think it's it's it's

366
00:19:06,520 --> 00:19:09,560
Speaker 2: very much overdone again. I understand some of the equity

367
00:19:09,640 --> 00:19:13,880
Speaker 2: valuation type concerns. My view is like, if everything repriced

368
00:19:13,880 --> 00:19:16,440
Speaker 2: at twenty five percent, you know the world would still

369
00:19:16,440 --> 00:19:18,160
Speaker 2: go on, and then we'd make up that twenty five

370
00:19:18,160 --> 00:19:21,359
Speaker 2: percent over the next next few years. I just know

371
00:19:21,520 --> 00:19:24,359
Speaker 2: this is not two thousand. This is not the dot

372
00:19:24,400 --> 00:19:26,959
Speaker 2: com bubble. Somebody asked me today about it. The last time,

373
00:19:27,119 --> 00:19:29,240
Speaker 2: maybe you said it earlier. Last time one hundred year

374
00:19:29,240 --> 00:19:31,440
Speaker 2: bond came was during the dot com era. Well, dot

375
00:19:31,440 --> 00:19:35,720
Speaker 2: com price things on eyeballs and clicks, not on revenues.

376
00:19:36,680 --> 00:19:39,160
Speaker 2: You know what we're doing is we actually have companies

377
00:19:39,200 --> 00:19:43,000
Speaker 2: that are generating hundreds of billions. I put out a

378
00:19:43,000 --> 00:19:45,840
Speaker 2: note the other day about Amazon's path to a trillion

379
00:19:45,880 --> 00:19:49,760
Speaker 2: dollars in annual revenues, hundreds of billion dollars dollars of

380
00:19:50,119 --> 00:19:53,360
Speaker 2: cash flow so's it's it's not the same. Then it's

381
00:19:53,359 --> 00:19:58,000
Speaker 2: also not the financial crisis where everything was dramatically over levered.

382
00:19:58,000 --> 00:20:00,800
Speaker 2: We had no clue. There were black boxes of over

383
00:20:00,920 --> 00:20:04,080
Speaker 2: leverage in the system, and that's just not the case

384
00:20:04,119 --> 00:20:07,160
Speaker 2: right now. These companies again are under levered. So are

385
00:20:07,160 --> 00:20:08,400
Speaker 2: we are we getting more debt?

386
00:20:08,600 --> 00:20:08,879
Speaker 3: Yes?

387
00:20:09,160 --> 00:20:11,280
Speaker 2: Or is leverage going up? To me? I'll tell you

388
00:20:11,320 --> 00:20:14,440
Speaker 2: a lot of these companies have negative net debt, i e.

389
00:20:14,600 --> 00:20:17,680
Speaker 2: More cash than debt. So does it really matter if

390
00:20:17,680 --> 00:20:23,600
Speaker 2: a company is negative leverage or net debt neutral? The

391
00:20:23,640 --> 00:20:26,320
Speaker 2: answer is no. So is there more leverage in the system?

392
00:20:26,400 --> 00:20:26,640
Speaker 3: Yes?

393
00:20:26,760 --> 00:20:29,360
Speaker 2: Is there inherently much more risk in the system right now?

394
00:20:29,440 --> 00:20:29,560
Speaker 1: Now?

395
00:20:30,040 --> 00:20:34,359
Speaker 2: If this path continues for a decade, maybe, but today now.

396
00:20:34,720 --> 00:20:36,960
Speaker 1: I think they're to defend on my fellow journalists whenever

397
00:20:36,960 --> 00:20:38,679
Speaker 1: we see this amount of cash being raised, or this

398
00:20:38,720 --> 00:20:42,159
Speaker 1: amount of money money flowing into anything, and then you

399
00:20:42,200 --> 00:20:44,399
Speaker 1: know a lot of it is into the unknown in

400
00:20:44,480 --> 00:20:46,280
Speaker 1: terms of this journey we're on, in terms of AI,

401
00:20:46,359 --> 00:20:48,480
Speaker 1: I think I think we immediately scratch our heads and

402
00:20:48,480 --> 00:20:51,600
Speaker 1: wonder whether there is maybe a bit of for us there.

403
00:20:51,680 --> 00:20:52,760
Speaker 1: But what about you Andack?

404
00:20:52,800 --> 00:20:54,960
Speaker 3: Do you think there's about I was I was about

405
00:20:54,960 --> 00:20:57,320
Speaker 3: to jump in because I was eager to give you

406
00:20:57,400 --> 00:21:00,160
Speaker 3: my two cents on this thing, because you see, where

407
00:21:00,200 --> 00:21:04,080
Speaker 3: is the thing? These are very responsible CEOs of very

408
00:21:04,119 --> 00:21:08,919
Speaker 3: responsible companies. And we saw that in the COVID crisis.

409
00:21:09,040 --> 00:21:14,040
Speaker 3: When Amazon needed more distribution capacity, they went and raised

410
00:21:14,080 --> 00:21:16,920
Speaker 3: capital and might actually don't remember if they went out

411
00:21:16,960 --> 00:21:20,880
Speaker 3: and issued bonds, but they went. They spent a crazy

412
00:21:20,880 --> 00:21:23,680
Speaker 3: amount of money to I think double up their logistics

413
00:21:23,720 --> 00:21:25,960
Speaker 3: network more than double it up for some time and

414
00:21:26,200 --> 00:21:28,280
Speaker 3: because they needed it. And guess what they did after that,

415
00:21:28,560 --> 00:21:33,480
Speaker 3: they didn't spend as much on it. So if Microsoft

416
00:21:33,600 --> 00:21:37,359
Speaker 3: CEO is not seeing the demand coming through from the

417
00:21:37,400 --> 00:21:42,040
Speaker 3: pipeline from the clients, I'm very confident that big number

418
00:21:42,080 --> 00:21:44,280
Speaker 3: that you have that you are looking through, you know

419
00:21:44,320 --> 00:21:47,800
Speaker 3: at right now is going to go down dramatically because

420
00:21:48,040 --> 00:21:50,960
Speaker 3: they are the buyers of Nvidia chips. I mean, can

421
00:21:51,160 --> 00:21:53,679
Speaker 3: I understand when you know, Jensen gets on the stage

422
00:21:53,720 --> 00:21:56,119
Speaker 3: and talks about that is in finite demand and all

423
00:21:56,160 --> 00:21:58,720
Speaker 3: that because he's the seller of that product. But you know,

424
00:21:58,800 --> 00:22:00,679
Speaker 3: you got to go and listen to the iyres of

425
00:22:00,800 --> 00:22:04,040
Speaker 3: this thing, and you know, for for Google to go

426
00:22:04,119 --> 00:22:09,600
Speaker 3: out and basically you know, talk about this much capex. Remember,

427
00:22:09,800 --> 00:22:14,280
Speaker 3: these guys have very strong businesses that are non dependent

428
00:22:14,320 --> 00:22:17,280
Speaker 3: on that AI capex. I mean they control or they

429
00:22:17,320 --> 00:22:20,800
Speaker 3: basically have one of the most profitable business models of

430
00:22:20,840 --> 00:22:23,240
Speaker 3: all times. When you look at Google with search, and

431
00:22:23,359 --> 00:22:25,359
Speaker 3: you know Amazon with both the e commerce and the

432
00:22:25,480 --> 00:22:28,520
Speaker 3: digital advertising and the AWS and so forth, and you

433
00:22:28,560 --> 00:22:32,760
Speaker 3: can you know, dissectively company, they are very responsible in

434
00:22:32,840 --> 00:22:36,120
Speaker 3: terms of if they are not seeing the demand, it's

435
00:22:36,119 --> 00:22:38,200
Speaker 3: going to go down. And that's the comfort I have

436
00:22:38,640 --> 00:22:41,399
Speaker 3: when when we look at it from the equity side.

437
00:22:41,440 --> 00:22:44,760
Speaker 2: And these are true visionaries, like when we're worth you know,

438
00:22:44,960 --> 00:22:47,400
Speaker 2: I know, I know a lot of these well known

439
00:22:47,640 --> 00:22:51,760
Speaker 2: tech leaders are sort of seen as as maybe a

440
00:22:51,760 --> 00:22:55,560
Speaker 2: bit crazy, but they are true visionaries. They are seeing

441
00:22:55,600 --> 00:22:59,520
Speaker 2: things we do not see. They are seeing products and

442
00:22:59,680 --> 00:23:04,000
Speaker 2: serve us and contracts for those products and services that

443
00:23:04,119 --> 00:23:07,760
Speaker 2: we cannot see, and that's how they built their businesses.

444
00:23:07,840 --> 00:23:10,640
Speaker 2: So I think at some point you have to put

445
00:23:10,680 --> 00:23:13,760
Speaker 2: trust in the smartest guys in the room. So I listen,

446
00:23:13,800 --> 00:23:16,679
Speaker 2: I don't want to pooh pooh. There's obviously reasons to

447
00:23:16,720 --> 00:23:19,720
Speaker 2: be have some sort of concern and as part of

448
00:23:19,760 --> 00:23:22,120
Speaker 2: the system, obviously lower down the credit scale. I mean,

449
00:23:22,280 --> 00:23:23,959
Speaker 2: we're seeing a lot of strain when it comes to

450
00:23:24,560 --> 00:23:27,640
Speaker 2: levered loans and the software space. And you know this

451
00:23:27,720 --> 00:23:30,000
Speaker 2: is not one big party where it's just all flowers

452
00:23:30,000 --> 00:23:33,320
Speaker 2: and candy. I'm just talking about from the largest company,

453
00:23:33,520 --> 00:23:38,240
Speaker 2: largest debtish or highest rated, high grade perspective. A lot

454
00:23:38,240 --> 00:23:41,199
Speaker 2: of these risks are overblown. I do think there's going

455
00:23:41,240 --> 00:23:44,640
Speaker 2: to be major fallouts though, across tech and across other

456
00:23:44,680 --> 00:23:48,160
Speaker 2: industries as this stuff plays out. It's just who has

457
00:23:48,280 --> 00:23:53,000
Speaker 2: the strongest defensive gear on and that's the companies that

458
00:23:53,040 --> 00:23:53,520
Speaker 2: I follow.

459
00:23:56,119 --> 00:23:57,800
Speaker 3: I do have a follow up on that, and then

460
00:23:57,920 --> 00:24:01,000
Speaker 3: this is what I tell you know. Westers are asking

461
00:24:01,080 --> 00:24:04,520
Speaker 3: us about the comparison between the dot com bubble and

462
00:24:04,680 --> 00:24:08,320
Speaker 3: you know, the AI bubble or AI investments. When you

463
00:24:08,440 --> 00:24:11,240
Speaker 3: looked at the dot com side of things, It's true

464
00:24:11,280 --> 00:24:14,000
Speaker 3: they built up a whole lot of capacity with the

465
00:24:14,040 --> 00:24:18,240
Speaker 3: hope business is going to come. At this point, business

466
00:24:18,320 --> 00:24:21,439
Speaker 3: is already there. They can't get enough boxes up so

467
00:24:21,480 --> 00:24:25,879
Speaker 3: that they can realize that business into revenue. It is

468
00:24:26,080 --> 00:24:29,560
Speaker 3: very different when you're leading off with a very very

469
00:24:29,600 --> 00:24:34,159
Speaker 3: successful consumer application, because the business leaders see it. It

470
00:24:34,440 --> 00:24:38,200
Speaker 3: wasn't you know when everybody understood the capabilities of chat,

471
00:24:38,240 --> 00:24:42,400
Speaker 3: GPT executives or the leaderships across the world, they're like, oh,

472
00:24:42,440 --> 00:24:46,040
Speaker 3: we want this into our own product. E Commerce or

473
00:24:46,080 --> 00:24:49,120
Speaker 3: the digital commerce at that time was an unknown concept

474
00:24:49,200 --> 00:24:52,399
Speaker 3: and nobody understood what it could do their do their business.

475
00:24:52,400 --> 00:24:56,439
Speaker 3: So it's there this stark difference of the business already

476
00:24:56,480 --> 00:24:59,400
Speaker 3: in your hands, and then you're lighting up the database

477
00:24:59,480 --> 00:25:02,960
Speaker 3: the data centers is very different than what we saw before.

478
00:25:03,680 --> 00:25:05,400
Speaker 2: Yeah, and James, this is the tip of the iceberg.

479
00:25:05,680 --> 00:25:11,520
Speaker 2: You know we saw earlier this year, Broadcomcom, IBM, obviously,

480
00:25:11,560 --> 00:25:15,160
Speaker 2: Oracle now Alphabet. I think there's more in the way.

481
00:25:15,359 --> 00:25:20,359
Speaker 2: There's Meta, there's Microsoft, there's Apple, So we're going to

482
00:25:20,400 --> 00:25:22,960
Speaker 2: see a couple more big deals. I do think some

483
00:25:23,040 --> 00:25:26,240
Speaker 2: of the way that they get exposure into different hands

484
00:25:26,280 --> 00:25:29,680
Speaker 2: and improve technicals is that there's buyers from in other

485
00:25:29,720 --> 00:25:34,200
Speaker 2: currencies Swiss francs, pounds, maybe euros, yen, and there's also

486
00:25:34,520 --> 00:25:37,120
Speaker 2: a variety of creative financing done through these asset back

487
00:25:37,160 --> 00:25:40,919
Speaker 2: private deals. So this is not going to be our

488
00:25:41,000 --> 00:25:43,639
Speaker 2: last conversation on this, and there's obviously not going to

489
00:25:43,680 --> 00:25:46,560
Speaker 2: be any answers for a while. We're still going to

490
00:25:46,600 --> 00:25:49,520
Speaker 2: be asking these questions because there's going to be more

491
00:25:49,560 --> 00:25:52,480
Speaker 2: spending and more borrowing, and I think it creates a

492
00:25:52,560 --> 00:25:54,600
Speaker 2: good debate and it's like we're gonna have to wait

493
00:25:54,600 --> 00:25:56,640
Speaker 2: and see a little bit. In the meantime, I think

494
00:25:56,680 --> 00:25:58,399
Speaker 2: if we just go back and look at last quarters

495
00:25:58,400 --> 00:26:01,320
Speaker 2: results across the board, all of these names are doing

496
00:26:01,440 --> 00:26:04,840
Speaker 2: ten percent plus top line revenues and whether or not

497
00:26:04,840 --> 00:26:07,680
Speaker 2: that's not that's lower than expectations or not. It's hard

498
00:26:07,680 --> 00:26:12,399
Speaker 2: to argue just from a an absolute perspective or in

499
00:26:12,480 --> 00:26:13,639
Speaker 2: any sort of sense of trouble.

500
00:26:14,200 --> 00:26:16,359
Speaker 1: Just the deeply cynical journalists to me that and I

501
00:26:16,400 --> 00:26:19,320
Speaker 1: have more vision and related to corporate leader, I always

502
00:26:19,359 --> 00:26:19,840
Speaker 1: run for the door.

503
00:26:19,920 --> 00:26:22,439
Speaker 3: Sorry, well, jims, let me let me give you something

504
00:26:22,520 --> 00:26:24,639
Speaker 3: on the other side of the equation. But you have

505
00:26:24,680 --> 00:26:27,760
Speaker 3: to worry about, you know, the neo clouds or I

506
00:26:27,800 --> 00:26:29,600
Speaker 3: shouldn't say you shoud to worry about it, but you

507
00:26:29,680 --> 00:26:31,879
Speaker 3: have to be very cautious and see that you know,

508
00:26:31,880 --> 00:26:34,080
Speaker 3: how is code we're going to take care of that stuff?

509
00:26:34,240 --> 00:26:36,320
Speaker 3: How is Neby is going to take care of that stuff?

510
00:26:36,480 --> 00:26:40,320
Speaker 3: Will Oracle have enough? You know, revenue coming through? There

511
00:26:40,359 --> 00:26:42,840
Speaker 3: are a handful of places where there could be certain.

512
00:26:43,560 --> 00:26:46,320
Speaker 3: You know, you could say excesses if there is a

513
00:26:46,359 --> 00:26:49,320
Speaker 3: demand fallout, but I won't say that that You can

514
00:26:49,440 --> 00:26:53,680
Speaker 3: put the hyperskill stut providers into that bucket. There are

515
00:26:53,760 --> 00:26:56,800
Speaker 3: there are going to be blowouts that there. I'm not

516
00:26:56,840 --> 00:26:59,280
Speaker 3: saying the names I mentioned, but there will be smaller

517
00:26:59,280 --> 00:27:03,160
Speaker 3: companies or all our startups that country handle if there

518
00:27:03,240 --> 00:27:05,560
Speaker 3: is a pullback, even a minor pulled back and spending.

519
00:27:06,320 --> 00:27:07,840
Speaker 1: Let me ask you both, as the closer, what's the

520
00:27:07,920 --> 00:27:10,080
Speaker 1: one thing you're concerned about for this year that might

521
00:27:10,160 --> 00:27:12,760
Speaker 1: make you revisit your bullish thesis on the tech set

522
00:27:12,840 --> 00:27:14,280
Speaker 1: to Starting with rub.

523
00:27:15,359 --> 00:27:18,199
Speaker 2: Well, it's hard to get me out of this this

524
00:27:18,280 --> 00:27:23,480
Speaker 2: bullish position. But there's two things. One is where I

525
00:27:23,480 --> 00:27:26,680
Speaker 2: don't think we're going is that capex actually gets reduced

526
00:27:26,880 --> 00:27:29,000
Speaker 2: through the end of the year, and that would mean

527
00:27:29,240 --> 00:27:32,000
Speaker 2: longer term demand is starting to decline. So I actually

528
00:27:32,160 --> 00:27:35,200
Speaker 2: like that we're seeing consistent increases in CAPEX. But two

529
00:27:35,280 --> 00:27:37,040
Speaker 2: is I actually think it's a yield issue, and that's

530
00:27:37,040 --> 00:27:40,840
Speaker 2: everything that drives the bomb market, which is if yield dries,

531
00:27:41,560 --> 00:27:43,119
Speaker 2: we have a new fed share that's going to be

532
00:27:43,160 --> 00:27:45,960
Speaker 2: coming in, and you know, if we start to get

533
00:27:45,960 --> 00:27:53,680
Speaker 2: to uneconomic borrowing levels. That has an enormous impact on spreads,

534
00:27:53,359 --> 00:27:57,000
Speaker 2: on the concept of future cash flows. So I actually

535
00:27:57,040 --> 00:28:00,240
Speaker 2: think it's more of a macro concern than a micro conscer.

536
00:28:01,040 --> 00:28:05,520
Speaker 3: And Eric, yeah, see, for me, I think the I'm

537
00:28:05,640 --> 00:28:09,120
Speaker 3: very concerned or I'm very much looking forward to anything

538
00:28:09,160 --> 00:28:14,840
Speaker 3: that can give us an idea how the cost of

539
00:28:15,240 --> 00:28:17,720
Speaker 3: this token generation is going to go down, whether it

540
00:28:17,800 --> 00:28:21,200
Speaker 3: is on the software side or because we want. We

541
00:28:21,520 --> 00:28:25,280
Speaker 3: know people are experimenting with AI applications within their company,

542
00:28:25,400 --> 00:28:28,600
Speaker 3: but the mass adoption of that product would only come

543
00:28:28,600 --> 00:28:31,600
Speaker 3: out when the cost of running that query is not

544
00:28:31,720 --> 00:28:35,680
Speaker 3: going to be high. So you know, if I know,

545
00:28:35,720 --> 00:28:38,680
Speaker 3: if I'm doing a query in my little chat pot

546
00:28:38,800 --> 00:28:42,320
Speaker 3: and if it's costing the company several tens of dollars

547
00:28:42,360 --> 00:28:45,720
Speaker 3: in per quarry to run the long term, that business

548
00:28:45,840 --> 00:28:48,160
Speaker 3: is not to be sustainable. And to me that the

549
00:28:48,320 --> 00:28:51,960
Speaker 3: unit economics is extremely important because that really dictates then

550
00:28:52,000 --> 00:28:56,320
Speaker 3: the wider adoption of that particular technology, which in turn

551
00:28:56,400 --> 00:28:58,680
Speaker 3: means CAPEX and all the other things that go into it.

552
00:29:00,600 --> 00:29:04,040
Speaker 1: Great stuff, Rob Schiffman and Anareg Runner from Bloomberg Intelligence.

553
00:29:04,080 --> 00:29:05,680
Speaker 1: It's been a pleasure having you on the credit edge money.

554
00:29:05,720 --> 00:29:05,960
Speaker 3: Thanks.

555
00:29:06,040 --> 00:29:10,000
Speaker 1: Thanks James for even more analysis, Read all of Anareg

556
00:29:10,120 --> 00:29:13,280
Speaker 1: and Rob's great work on the Bloomberg Terminal, or just

557
00:29:13,320 --> 00:29:16,000
Speaker 1: call them up. They're great value. They will answer your questions.

558
00:29:16,120 --> 00:29:18,800
Speaker 1: Bloomberg Intelligence is part of our research department, with five

559
00:29:18,880 --> 00:29:21,719
Speaker 1: hundred analysts and stretchers working across all the markets. Coverage

560
00:29:21,720 --> 00:29:24,320
Speaker 1: includes over two thousand equities and credits and outlooks on

561
00:29:24,320 --> 00:29:27,800
Speaker 1: more than ninety industries and one hundred market indices, currencies

562
00:29:27,840 --> 00:29:31,160
Speaker 1: and commodities. Please do subscribe to the Credit Edge wherever

563
00:29:31,200 --> 00:29:33,960
Speaker 1: you get your podcasts. We're on Apple, Spotify, and all

564
00:29:34,000 --> 00:29:37,680
Speaker 1: other good podcast providers, including the Bloomberg Terminal at bpod Go.

565
00:29:38,240 --> 00:29:40,640
Speaker 1: Give us a review, tell your friends, or email me

566
00:29:40,720 --> 00:29:44,800
Speaker 1: directly at jcrumby eight at Bloomberg dot net. I'm James Crombie.

567
00:29:44,840 --> 00:29:47,040
Speaker 1: It's been a pleasure having you join us again, next

568
00:29:47,040 --> 00:30:07,040
Speaker 1: time on the Credit Edge