1
00:00:00,080 --> 00:00:04,080
Speaker 1: Hello, Odd Lots listeners. I'm Joe Wisenthal and I'm Tracy Alloway. Tracy,

2
00:00:04,120 --> 00:00:06,280
Speaker 1: we're doing another live show and it's right here in

3
00:00:06,280 --> 00:00:06,880
Speaker 1: New York City.

4
00:00:07,080 --> 00:00:10,080
Speaker 2: Yeah, this one should be our biggest yet, and we're

5
00:00:10,080 --> 00:00:12,480
Speaker 2: going to have a bunch of Odd Lots favorites and

6
00:00:13,160 --> 00:00:15,880
Speaker 2: do something maybe a little different to some of our

7
00:00:15,960 --> 00:00:18,200
Speaker 2: previous live podcast recordings.

8
00:00:18,280 --> 00:00:21,079
Speaker 1: When the guests are revealed, the show is going to

9
00:00:21,120 --> 00:00:22,840
Speaker 1: sell out right away, so you should really just go

10
00:00:22,840 --> 00:00:25,960
Speaker 1: get your ticket right now. It's June twenty sixth. It's

11
00:00:25,960 --> 00:00:28,560
Speaker 1: at Record NYC, and you can find a ticket link

12
00:00:28,600 --> 00:00:31,800
Speaker 1: at Bloomberg dot com slash odd Lots or Bloomberg Events

13
00:00:31,840 --> 00:00:34,440
Speaker 1: dot com slash odd Lots Live and why.

14
00:00:34,360 --> 00:00:35,520
Speaker 2: We hope to see you there.

15
00:00:38,479 --> 00:00:52,400
Speaker 3: Bloomberg Audio Studios, Podcasts, Radio News.

16
00:00:54,040 --> 00:00:57,160
Speaker 2: Hello and welcome to another episode of the Odd Loots podcast.

17
00:00:57,280 --> 00:00:58,720
Speaker 2: I'm Tracy Alloway.

18
00:00:58,400 --> 00:00:59,600
Speaker 1: And I'm Joe Wisenthal.

19
00:01:00,000 --> 00:01:02,960
Speaker 2: You know, does it feel to you like every day

20
00:01:03,080 --> 00:01:05,360
Speaker 2: when you wake up in markets there is a new

21
00:01:05,520 --> 00:01:09,640
Speaker 2: sort of dominant narrative, and often it is the exact

22
00:01:09,840 --> 00:01:12,880
Speaker 2: opposite of the narrative that was dominant a day or

23
00:01:12,920 --> 00:01:13,360
Speaker 2: two ago.

24
00:01:15,160 --> 00:01:17,000
Speaker 1: Go on, say more what you mean, where are you

25
00:01:17,000 --> 00:01:18,840
Speaker 1: going with this? I mean, I agree in some sense,

26
00:01:18,880 --> 00:01:20,760
Speaker 1: but I think you have something specific in mind.

27
00:01:21,040 --> 00:01:23,000
Speaker 2: Well, no, I actually have a bunch of things in mind.

28
00:01:23,080 --> 00:01:26,080
Speaker 2: So for instance, you know, recently there was a lot

29
00:01:26,080 --> 00:01:28,800
Speaker 2: of talk about the bond vigilantes. Yeah, back and people

30
00:01:28,840 --> 00:01:30,319
Speaker 2: worried about the fiscal deficit.

31
00:01:30,480 --> 00:01:31,160
Speaker 1: Yeah, and then this.

32
00:01:31,200 --> 00:01:34,480
Speaker 2: Week that seems to have suddenly gone away, or at

33
00:01:34,560 --> 00:01:38,400
Speaker 2: least it's not the only thing. People are talking about tariffs.

34
00:01:39,040 --> 00:01:39,200
Speaker 4: You know.

35
00:01:39,319 --> 00:01:42,600
Speaker 2: Obviously we had the big market freak out in April,

36
00:01:42,720 --> 00:01:46,880
Speaker 2: and then we had subsequent delays and extensions of the deadlines.

37
00:01:47,200 --> 00:01:50,400
Speaker 2: I don't hear people talking about tariffs as much as

38
00:01:50,400 --> 00:01:51,600
Speaker 2: they were in April.

39
00:01:51,800 --> 00:01:55,639
Speaker 1: So then through it all is you know, on days

40
00:01:55,680 --> 00:01:58,360
Speaker 1: when we don't have to talk about tariffs or bond vigilantes,

41
00:01:58,480 --> 00:02:00,880
Speaker 1: then we immediately turned to AI. Well this is what

42
00:02:00,920 --> 00:02:03,000
Speaker 1: I was saying, you know, So this is the thing.

43
00:02:03,040 --> 00:02:05,800
Speaker 1: There's like this default thing that's in the background, which

44
00:02:05,840 --> 00:02:08,920
Speaker 1: is AI. And then that of course leads to things

45
00:02:09,000 --> 00:02:12,560
Speaker 1: like power and electricity and so forth. So yes, I

46
00:02:12,639 --> 00:02:14,680
Speaker 1: now see what you're saying, which is that on any

47
00:02:14,720 --> 00:02:17,520
Speaker 1: given day, it's like you spin the roulette wheel or

48
00:02:17,520 --> 00:02:20,160
Speaker 1: the wheel of fortune and then it's like, Okay, it's

49
00:02:20,200 --> 00:02:23,280
Speaker 1: an AI day. Actually we're recording this May twenty eighth.

50
00:02:23,440 --> 00:02:25,239
Speaker 1: By the time people are listening to this and video

51
00:02:25,280 --> 00:02:27,040
Speaker 1: earnings are going to be out. Yeah, and so today

52
00:02:27,080 --> 00:02:28,359
Speaker 1: is an AI day.

53
00:02:28,360 --> 00:02:29,240
Speaker 4: It might be an AI day.

54
00:02:29,480 --> 00:02:30,880
Speaker 2: What I was going to say is one of our

55
00:02:30,960 --> 00:02:34,680
Speaker 2: previous guests, Victor Schwetz over at McCrory. He framed it

56
00:02:34,760 --> 00:02:37,360
Speaker 2: really well where he was basically saying, if you know,

57
00:02:37,400 --> 00:02:40,720
Speaker 2: if you can't gauge what the dominant narrative is going

58
00:02:40,760 --> 00:02:43,720
Speaker 2: to be every day, or if everything that's happening is

59
00:02:43,800 --> 00:02:46,679
Speaker 2: so volatile and so dramatic and we're talking about really

60
00:02:46,720 --> 00:02:48,919
Speaker 2: big changes, then you kind of have to go back

61
00:02:48,960 --> 00:02:53,040
Speaker 2: to doing the normal thing. And the normal thing before

62
00:02:53,360 --> 00:02:55,040
Speaker 2: all of this kind of started in the New Year,

63
00:02:55,280 --> 00:02:58,560
Speaker 2: was worrying about AI, right, and thinking about AI and

64
00:02:58,960 --> 00:03:01,920
Speaker 2: whether or not you know, all of that AI spending

65
00:03:01,960 --> 00:03:05,320
Speaker 2: is going to be backed by revenue. So on that note,

66
00:03:05,480 --> 00:03:07,200
Speaker 2: that's what we're going to do today. We're going to

67
00:03:07,240 --> 00:03:09,440
Speaker 2: go back to normality, although we are going to talk

68
00:03:09,440 --> 00:03:11,680
Speaker 2: about some of the current stuff as well. And I

69
00:03:11,720 --> 00:03:13,760
Speaker 2: am very pleased to say we have someone who I

70
00:03:13,800 --> 00:03:16,680
Speaker 2: have wanted to get on the podcast for a really

71
00:03:16,880 --> 00:03:20,200
Speaker 2: long time, and he's finally sort of been let out

72
00:03:20,280 --> 00:03:22,960
Speaker 2: into the wild. So we are very excited we're going

73
00:03:23,040 --> 00:03:25,560
Speaker 2: to be speaking with Michael semblest He is, of course,

74
00:03:25,600 --> 00:03:29,280
Speaker 2: the chairman of market and investment Strategy for JP Morgan

75
00:03:29,400 --> 00:03:32,959
Speaker 2: Asset Management. He writes a great, great piece called I

76
00:03:33,160 --> 00:03:35,280
Speaker 2: on the Market, which he's been doing for which has a.

77
00:03:35,200 --> 00:03:36,600
Speaker 4: Cult following, cult following.

78
00:03:36,720 --> 00:03:39,720
Speaker 1: Every time a new symbolist piece comes out, everyone talks

79
00:03:39,720 --> 00:03:40,560
Speaker 1: about it and reads it.

80
00:03:40,720 --> 00:03:42,640
Speaker 2: Yeah, and I think it's been going on for something

81
00:03:42,680 --> 00:03:45,280
Speaker 2: like twenty years now. And he himself has been at

82
00:03:45,360 --> 00:03:48,320
Speaker 2: JP Morgan since nineteen eighty seven, so, you know, a

83
00:03:48,360 --> 00:03:52,320
Speaker 2: real witness to financial market history. So I'm very excited. Michael.

84
00:03:52,480 --> 00:03:54,640
Speaker 2: Thank you so much for finally coming on the show.

85
00:03:54,920 --> 00:03:56,520
Speaker 4: Thank you very much. Good to be here.

86
00:03:57,000 --> 00:03:59,280
Speaker 2: I guess my first question is can you explain what

87
00:03:59,320 --> 00:04:02,600
Speaker 2: you do at JP Morgan Asset Management. I on the

88
00:04:02,600 --> 00:04:05,360
Speaker 2: Market is a phenomenal publication. I've been a fan for

89
00:04:05,400 --> 00:04:07,600
Speaker 2: a long time, but I always wonder like, is there

90
00:04:07,600 --> 00:04:12,240
Speaker 2: a difference between writing investment outlooks for asset management clients

91
00:04:12,320 --> 00:04:15,240
Speaker 2: versus writing them if you're on the cell side.

92
00:04:15,720 --> 00:04:20,120
Speaker 4: Well, yeah, there's a huge difference. First of all, you know,

93
00:04:20,160 --> 00:04:22,720
Speaker 4: thank god for the Chinese wall, Arisa rules and things

94
00:04:22,760 --> 00:04:25,960
Speaker 4: like that, because I have the independence, and what our

95
00:04:26,000 --> 00:04:28,560
Speaker 4: clients want is for me to share what I'm thinking.

96
00:04:29,040 --> 00:04:31,680
Speaker 4: The best example of that is in February twenty twenty one,

97
00:04:32,120 --> 00:04:35,160
Speaker 4: the investment bank was making money handover all investment banks

98
00:04:35,200 --> 00:04:37,839
Speaker 4: for making money hand over fists in the back market,

99
00:04:38,200 --> 00:04:41,080
Speaker 4: and I published forgot about it. I published, you know,

100
00:04:41,160 --> 00:04:45,039
Speaker 4: two really scathing pieces on SPACs. The first one was

101
00:04:45,040 --> 00:04:47,600
Speaker 4: called Hydraulics Backing, and the next one, because it was

102
00:04:47,680 --> 00:04:50,760
Speaker 4: during COVID, was called spacscene hesitancy, where I kind of

103
00:04:50,760 --> 00:04:53,160
Speaker 4: pointed out that the sponsors were the only ones making

104
00:04:53,200 --> 00:04:56,640
Speaker 4: money and that their break even threshold was something like

105
00:04:56,680 --> 00:04:59,760
Speaker 4: an eighty percent decline in the merged stock price. Was

106
00:04:59,760 --> 00:05:03,680
Speaker 4: there so anything better than that? They would make money

107
00:05:03,920 --> 00:05:06,520
Speaker 4: and everybody else was getting slaughtered. So, you know, being

108
00:05:06,760 --> 00:05:10,160
Speaker 4: behind the fiduciary wall allows me to kind of call

109
00:05:10,200 --> 00:05:12,200
Speaker 4: it like I see it. And the eye of the

110
00:05:12,200 --> 00:05:15,760
Speaker 4: market is the external presentation of our thoughts and views,

111
00:05:16,120 --> 00:05:18,920
Speaker 4: and you know, on an internal basis, I'm participating our

112
00:05:18,920 --> 00:05:23,040
Speaker 4: investment committees dealing with asset allocation and security selection and

113
00:05:23,800 --> 00:05:27,600
Speaker 4: everything else that goes along with managing money for Dowmence foundations,

114
00:05:27,600 --> 00:05:30,880
Speaker 4: pension plans, insurance companies, sovereign wealth funds, and individuals.

115
00:05:31,080 --> 00:05:34,160
Speaker 1: It does feel like on the media side we are

116
00:05:34,400 --> 00:05:38,080
Speaker 1: at the whims of narratives and all the things that

117
00:05:38,120 --> 00:05:40,960
Speaker 1: Tracy talked about at the beginning, Whereas if you talk

118
00:05:41,120 --> 00:05:44,880
Speaker 1: to people who are engaged in the art of security selection,

119
00:05:45,760 --> 00:05:49,600
Speaker 1: it feels that actually might be liberating because on some level,

120
00:05:50,400 --> 00:05:52,720
Speaker 1: whatever is going on, you could still look at an

121
00:05:52,720 --> 00:05:55,440
Speaker 1: instrument and say is this a well priced instrument or not,

122
00:05:55,760 --> 00:05:58,359
Speaker 1: and not having to get caught up so much in

123
00:05:58,680 --> 00:06:01,960
Speaker 1: just you know, you know, whatever the headline of the.

124
00:06:01,960 --> 00:06:04,400
Speaker 4: Day is right, And actually, you know, if you're managing

125
00:06:04,440 --> 00:06:08,839
Speaker 4: money other than kind of a fast money macro hedge fund,

126
00:06:09,400 --> 00:06:12,400
Speaker 4: you don't have the luxury of changing your asset allocation

127
00:06:12,480 --> 00:06:15,000
Speaker 4: every time a new narrative comes along, so you to

128
00:06:15,000 --> 00:06:17,680
Speaker 4: some extent you have to kind of choose your poison,

129
00:06:17,800 --> 00:06:21,760
Speaker 4: make your asset allocation decisions, and then be either rewarded

130
00:06:21,960 --> 00:06:23,599
Speaker 4: or punished based on the outcome.

131
00:06:24,320 --> 00:06:26,000
Speaker 2: So one of the reasons we wanted to get you

132
00:06:26,040 --> 00:06:28,520
Speaker 2: on the show was because you've been writing in great

133
00:06:28,600 --> 00:06:32,680
Speaker 2: detail about AI. So let me just ask the basic question.

134
00:06:32,720 --> 00:06:35,320
Speaker 2: If we're not gonna worry about you know, Trump and

135
00:06:35,440 --> 00:06:38,719
Speaker 2: tariffs and bond vigilantes and all the sort of headliney stuff,

136
00:06:38,920 --> 00:06:41,440
Speaker 2: if we're going back to considering, I guess the macro

137
00:06:41,560 --> 00:06:45,720
Speaker 2: and market importance of AI at the moment. How dominant

138
00:06:45,880 --> 00:06:49,240
Speaker 2: is AI when it comes to market valuations at the moment.

139
00:06:49,240 --> 00:06:52,120
Speaker 2: I guess how fundamental is AI when it comes to

140
00:06:52,320 --> 00:06:53,360
Speaker 2: the wider stock market.

141
00:06:53,920 --> 00:06:59,120
Speaker 4: I don't think you can overstate the importance of this stuff. Wow,

142
00:06:59,200 --> 00:07:03,160
Speaker 4: when you dec impose the evolution of S and P

143
00:07:03,160 --> 00:07:07,719
Speaker 4: profit margins and you strip off the MAG seven versus

144
00:07:07,800 --> 00:07:10,080
Speaker 4: the rest of the market, it's like East and West

145
00:07:10,080 --> 00:07:14,320
Speaker 4: Berlin right now, I'm dating myself, right, But that's how different,

146
00:07:14,400 --> 00:07:18,119
Speaker 4: that's how divergent those are. If you look at Earning's growth,

147
00:07:18,560 --> 00:07:21,280
Speaker 4: I think Earning's growth for the MAG seven over the

148
00:07:21,360 --> 00:07:24,240
Speaker 4: last ten years is close to twenty percent. It's six

149
00:07:24,280 --> 00:07:26,240
Speaker 4: percent for the S and P four ninety three. So

150
00:07:26,280 --> 00:07:30,160
Speaker 4: it's almost like two completely different universes of companies. And

151
00:07:30,480 --> 00:07:35,680
Speaker 4: so they are extremely profitable. They are spending somewhere between,

152
00:07:35,720 --> 00:07:38,640
Speaker 4: depending on a quarter, twenty five to forty percent of

153
00:07:38,680 --> 00:07:40,720
Speaker 4: their revenues on capital spending in R and D. Those

154
00:07:40,760 --> 00:07:43,880
Speaker 4: numbers are unprecedented and in the history of the markets,

155
00:07:44,040 --> 00:07:46,440
Speaker 4: even going back to the nineteen sixties main frame heer

156
00:07:46,440 --> 00:07:49,080
Speaker 4: and things like that, we've never seen anything quite like this.

157
00:07:49,600 --> 00:07:52,480
Speaker 4: And what the hyperscalers are doing I would describe. I know,

158
00:07:52,480 --> 00:07:54,600
Speaker 4: we're only twenty five years into the century, but this

159
00:07:54,640 --> 00:07:57,000
Speaker 4: is the bed of the century that you can spend

160
00:07:57,040 --> 00:08:00,520
Speaker 4: this much on AI infrastructure, you know, us your free

161
00:08:00,520 --> 00:08:02,800
Speaker 4: cash flow margins for some period of time and wait

162
00:08:02,840 --> 00:08:03,840
Speaker 4: for the ultimate payoff.

163
00:08:04,480 --> 00:08:06,840
Speaker 1: Even prior to the AI you know, you mentioned this

164
00:08:06,960 --> 00:08:10,840
Speaker 1: extraordinary gap. I mean, there are two things about hearing

165
00:08:10,880 --> 00:08:13,280
Speaker 1: you describe the earnings picture within the S and P

166
00:08:13,400 --> 00:08:16,360
Speaker 1: five hundred that strike me. One is the absolute gap.

167
00:08:16,760 --> 00:08:20,000
Speaker 1: But then there's also the fact that it's the biggest

168
00:08:20,040 --> 00:08:22,720
Speaker 1: companies growing really fast. And I in my mind, when

169
00:08:22,720 --> 00:08:25,080
Speaker 1: I was younger, assumed that as big companies got bigger,

170
00:08:25,080 --> 00:08:29,480
Speaker 1: they grew slower and smaller companies grew faster. When you

171
00:08:29,520 --> 00:08:33,280
Speaker 1: look at stock market history, how rare is the type

172
00:08:33,280 --> 00:08:35,960
Speaker 1: of earnings growth the year on year, the huge numbers

173
00:08:35,960 --> 00:08:39,000
Speaker 1: that they put up earnings growth wise, How unusual is

174
00:08:39,040 --> 00:08:42,160
Speaker 1: it for stock market history for the biggest companies to

175
00:08:42,280 --> 00:08:45,120
Speaker 1: also just be galloping past everyone else like this, you.

176
00:08:45,080 --> 00:08:47,560
Speaker 4: Can't find it. You can't find it. I mean stock

177
00:08:47,600 --> 00:08:49,760
Speaker 4: market history on a security level basis goes back to

178
00:08:49,800 --> 00:08:52,880
Speaker 4: around nineteen eighty. There's a variety of databases through facts

179
00:08:52,880 --> 00:08:55,760
Speaker 4: that through the University of Chicago. That's about as far back.

180
00:08:55,800 --> 00:08:56,720
Speaker 1: Really, you can't go much.

181
00:08:57,160 --> 00:08:59,080
Speaker 4: It's hard to go back to users in that there's

182
00:08:59,120 --> 00:09:01,920
Speaker 4: some Wilsher data, but it's it's difficult. It's very spody.

183
00:09:02,360 --> 00:09:04,240
Speaker 4: So if you're looking for things like earnings growth, if

184
00:09:04,280 --> 00:09:07,240
Speaker 4: he just wants stock price growth, but if you're looking

185
00:09:07,320 --> 00:09:10,600
Speaker 4: for earnings and margins and things like that's hard. So

186
00:09:10,640 --> 00:09:13,920
Speaker 4: we can't find We can't find any comparable period where

187
00:09:13,920 --> 00:09:16,600
Speaker 4: the bigger companies are growing faster. And even within the

188
00:09:16,600 --> 00:09:20,680
Speaker 4: private equity universe, your theory holds true. The generally smaller

189
00:09:20,720 --> 00:09:23,480
Speaker 4: and midtized funds of usually you're going to outperform over

190
00:09:23,520 --> 00:09:25,560
Speaker 4: the long periods of time the larger funds. This is

191
00:09:25,600 --> 00:09:29,480
Speaker 4: negative size bias. So this is pretty unique. And you know,

192
00:09:29,520 --> 00:09:31,760
Speaker 4: over the years, I've always thought that the biggest risk

193
00:09:31,840 --> 00:09:35,000
Speaker 4: to all of this was going to be some reinterpretation

194
00:09:35,080 --> 00:09:37,440
Speaker 4: of the Sherman Act, and you know, anti trust And

195
00:09:37,600 --> 00:09:41,040
Speaker 4: while Epic Games has won a lawsuit, here are there

196
00:09:41,320 --> 00:09:44,560
Speaker 4: and the Department of Justice is picking away at Amazon,

197
00:09:44,559 --> 00:09:47,360
Speaker 4: Apple and Google for different reasons, there really hasn't been

198
00:09:47,360 --> 00:09:50,560
Speaker 4: a fundamental rethinking of anti trust law that would get

199
00:09:50,559 --> 00:09:51,040
Speaker 4: in the way of this.

200
00:09:51,480 --> 00:09:53,040
Speaker 2: Can you talk a little bit more about that. What

201
00:09:53,080 --> 00:09:56,040
Speaker 2: are the drivers or the circumstances that are allowing mag

202
00:09:56,080 --> 00:09:59,120
Speaker 2: seven to continue to grow in a phenomenal pace but

203
00:09:59,160 --> 00:10:01,959
Speaker 2: also just become the giant cash flow monsters.

204
00:10:02,720 --> 00:10:05,920
Speaker 4: Yeah, well, it has to do with the digitization of

205
00:10:06,000 --> 00:10:09,120
Speaker 4: the economy and how much technology is at the core

206
00:10:09,160 --> 00:10:12,280
Speaker 4: of how companies run. You know, JP Morgan right now

207
00:10:12,360 --> 00:10:15,400
Speaker 4: is a firm. We've got three hundred different language model

208
00:10:15,400 --> 00:10:19,520
Speaker 4: and ALI projects going on in and all of that

209
00:10:19,720 --> 00:10:21,559
Speaker 4: requires them one way or another, a lot of these

210
00:10:21,920 --> 00:10:23,560
Speaker 4: mag seven coming. I think we have to put Tesla

211
00:10:23,600 --> 00:10:26,120
Speaker 4: aside for a minute for reasons that we should discuss.

212
00:10:26,679 --> 00:10:29,760
Speaker 4: But technology is really at the foundation. And I think

213
00:10:30,000 --> 00:10:32,960
Speaker 4: a lot of times when people compare and they say, oh,

214
00:10:33,080 --> 00:10:35,400
Speaker 4: US docs are expensive relatives the rest of the world,

215
00:10:35,840 --> 00:10:38,880
Speaker 4: they're missing a couple of important things. They're missing the

216
00:10:38,920 --> 00:10:41,559
Speaker 4: fact that the US has a massive weight to technology

217
00:10:41,920 --> 00:10:45,120
Speaker 4: and in Europe, you know, basically it's it's a value market, right,

218
00:10:45,120 --> 00:10:50,120
Speaker 4: It's a lot of energy, financials, and consumer staples, and

219
00:10:50,200 --> 00:10:53,520
Speaker 4: so the US is really the dominant player in large

220
00:10:53,520 --> 00:10:58,080
Speaker 4: cap tech. There's an amazing chart. Mario Draghi, you know

221
00:10:58,240 --> 00:11:00,959
Speaker 4: who he is, obviously wrote a piece at the beginning

222
00:11:01,000 --> 00:11:03,840
Speaker 4: of the year what can we possibly I'm poweraphrasing, what

223
00:11:03,880 --> 00:11:07,320
Speaker 4: can we possibly do to reinvigorate entrepreneurship in Europe? And

224
00:11:07,360 --> 00:11:09,480
Speaker 4: he had a chart in there that we recomputed in

225
00:11:09,480 --> 00:11:11,720
Speaker 4: our own way, and it's a bubble chart that shows

226
00:11:11,720 --> 00:11:15,520
Speaker 4: the number of new companies created in the US versus

227
00:11:15,559 --> 00:11:18,320
Speaker 4: Europe since the year two thousand and Again, it's like

228
00:11:18,400 --> 00:11:20,960
Speaker 4: East and West Berlin. They are completely different. Yeah.

229
00:11:21,240 --> 00:11:23,839
Speaker 1: One of the things that struck me recently we did

230
00:11:23,840 --> 00:11:27,800
Speaker 1: an episode with the CEO of a women's clothing company,

231
00:11:27,880 --> 00:11:31,000
Speaker 1: M M. Lafleur. We're mostly talking about tariffs, but the

232
00:11:31,040 --> 00:11:33,240
Speaker 1: thing that I keep thinking about is she said that

233
00:11:33,320 --> 00:11:35,560
Speaker 1: when she started the company a little I think when

234
00:11:35,640 --> 00:11:38,520
Speaker 1: was it over ten years ago? Tracy, Yeah, her customer

235
00:11:38,600 --> 00:11:43,280
Speaker 1: acquisition costs on a platform like Instagram, where she was

236
00:11:43,520 --> 00:11:46,360
Speaker 1: like twelve dollars a customer and now it's like over

237
00:11:46,400 --> 00:11:50,560
Speaker 1: two hundred dollars. And it just struck me that one

238
00:11:50,600 --> 00:11:53,000
Speaker 1: of the things that must be going on in the

239
00:11:53,080 --> 00:11:58,959
Speaker 1: economy is everybody's margin becoming Facebook's profits, or everybody's margins

240
00:11:59,000 --> 00:12:03,720
Speaker 1: eventually becoming Alphabets profits. That part of this gap that

241
00:12:03,760 --> 00:12:06,760
Speaker 1: we're seeing is there is this digital real estate that exists.

242
00:12:07,080 --> 00:12:11,160
Speaker 1: It's where everyone is moving, and they could collect massive

243
00:12:11,200 --> 00:12:13,760
Speaker 1: rents from owning that real estate from anyone who wants

244
00:12:13,760 --> 00:12:14,680
Speaker 1: to do commerce there.

245
00:12:14,840 --> 00:12:17,560
Speaker 4: Yeah, and I think rent is the right word here,

246
00:12:18,080 --> 00:12:21,439
Speaker 4: and for reasons both good and bad. You know, Amazon

247
00:12:21,640 --> 00:12:23,559
Speaker 4: requires if you want to be if you want to

248
00:12:23,600 --> 00:12:26,079
Speaker 4: sell to prime customers, you have to agree to use

249
00:12:26,120 --> 00:12:29,840
Speaker 4: Amazon shipping. If you make an in purchase app in

250
00:12:29,920 --> 00:12:32,600
Speaker 4: it some kind of a game like Fortnite on an

251
00:12:32,600 --> 00:12:35,079
Speaker 4: Android phone, you have to make that purchase through the

252
00:12:35,160 --> 00:12:37,320
Speaker 4: Google Play Store. These are some of the issues that

253
00:12:37,320 --> 00:12:39,600
Speaker 4: have been litigated in some of the and so I'm

254
00:12:39,720 --> 00:12:43,920
Speaker 4: not one hundred percent in favor of some of the

255
00:12:44,040 --> 00:12:46,920
Speaker 4: tying that essentially they get. If JP Morgan said if

256
00:12:46,960 --> 00:12:48,160
Speaker 4: you want a credit card, you got to do your

257
00:12:48,160 --> 00:12:50,840
Speaker 4: mortgage with us, we get slapped down pretty fast, and

258
00:12:50,840 --> 00:12:53,280
Speaker 4: for good reason. And so there's a lot of product

259
00:12:53,360 --> 00:12:55,880
Speaker 4: tying that takes place in big tech. You know that

260
00:12:55,920 --> 00:12:57,719
Speaker 4: a lot of companies are beginning to challenge and the

261
00:12:57,800 --> 00:12:59,000
Speaker 4: DJ is looking at as well.

262
00:13:14,960 --> 00:13:17,800
Speaker 2: So we've been talking about the dominance of US tech

263
00:13:18,040 --> 00:13:20,160
Speaker 2: and you know, there are a lot of numbers to

264
00:13:20,200 --> 00:13:21,920
Speaker 2: support this, and you've been laying some of them out.

265
00:13:22,040 --> 00:13:26,440
Speaker 2: But before we had all these teriff related market freakouts,

266
00:13:26,480 --> 00:13:29,640
Speaker 2: the freakout that everyone was focused on was deep seek,

267
00:13:29,880 --> 00:13:33,400
Speaker 2: of course, and the idea that suddenly China has invented

268
00:13:33,520 --> 00:13:36,959
Speaker 2: a low cost AI model that is very competitive with

269
00:13:37,000 --> 00:13:40,359
Speaker 2: what we've seen from the giant companies like the Microsoft's

270
00:13:40,360 --> 00:13:43,360
Speaker 2: and the Googles of the world. Looking back at that,

271
00:13:44,320 --> 00:13:47,320
Speaker 2: was that freakout overstated or just was okay?

272
00:13:47,520 --> 00:13:47,640
Speaker 3: Right?

273
00:13:47,960 --> 00:13:50,520
Speaker 4: A couple of reasons. One, I think it's become clear

274
00:13:50,640 --> 00:13:53,640
Speaker 4: and I wrote a piece at the time. They were

275
00:13:53,679 --> 00:13:56,920
Speaker 4: not completely transparent about how they accomplished some of their

276
00:13:56,960 --> 00:14:01,600
Speaker 4: goals imitations. This is serious form of flattery. They actually

277
00:14:01,880 --> 00:14:05,360
Speaker 4: used i think open AI models to train their models,

278
00:14:05,840 --> 00:14:09,439
Speaker 4: and so they did a lot of piggybacking. That isn't

279
00:14:09,480 --> 00:14:14,200
Speaker 4: necessarily a sign of some kind of productivity breakthrough. Also,

280
00:14:14,240 --> 00:14:17,080
Speaker 4: the price at which they are willing to sell something

281
00:14:17,200 --> 00:14:20,760
Speaker 4: is not necessarily reflective of its true cost, and so

282
00:14:21,360 --> 00:14:23,520
Speaker 4: there's some loss leading that's going on in there. That

283
00:14:23,680 --> 00:14:28,680
Speaker 4: again is different that said, cheaper AI models, if that's

284
00:14:28,760 --> 00:14:32,960
Speaker 4: where we're going, is probably going to make adoption even easier.

285
00:14:33,360 --> 00:14:37,040
Speaker 4: So if this can all be done with less energy

286
00:14:37,080 --> 00:14:40,400
Speaker 4: at lower token cost, that's a bullish story for AI,

287
00:14:40,560 --> 00:14:41,560
Speaker 4: not a negative one.

288
00:14:41,800 --> 00:14:44,000
Speaker 1: It also occurred to me when that happened that if

289
00:14:44,040 --> 00:14:46,760
Speaker 1: there is this existential battle between the US and China

290
00:14:46,800 --> 00:14:50,560
Speaker 1: for AI supremacy, probably spend more on it. You know

291
00:14:50,960 --> 00:14:54,160
Speaker 1: it you know, it becomes a nuclear bomb type race,

292
00:14:54,280 --> 00:14:56,280
Speaker 1: then the impulse is going to be to spend more

293
00:14:56,360 --> 00:14:59,920
Speaker 1: on it. But actually, you mentioned things like energy cost

294
00:15:00,320 --> 00:15:03,400
Speaker 1: per token and so forth, and I'm curious, from your

295
00:15:03,440 --> 00:15:07,040
Speaker 1: perspective or you know, even doing this for a long time,

296
00:15:07,360 --> 00:15:09,480
Speaker 1: how did you learn about that? How did you educate

297
00:15:09,520 --> 00:15:14,960
Speaker 1: yourself on understanding the sort of the key economic variables

298
00:15:15,000 --> 00:15:17,480
Speaker 1: that you have to get wrap your head around in

299
00:15:17,600 --> 00:15:19,920
Speaker 1: order to say informed things about this new topic.

300
00:15:21,600 --> 00:15:23,400
Speaker 4: Well, there's a bunch of things.

301
00:15:23,560 --> 00:15:25,240
Speaker 1: We're trying to figure that out ourselves. You know, we

302
00:15:25,280 --> 00:15:27,560
Speaker 1: probably should be doing more about AI and talking so

303
00:15:27,600 --> 00:15:29,880
Speaker 1: we need to learn about how you tell yourself about it.

304
00:15:30,240 --> 00:15:33,080
Speaker 4: I think you have to budget your time. And my

305
00:15:33,280 --> 00:15:37,200
Speaker 4: most important thing is reading about twentusand to twenty five

306
00:15:37,280 --> 00:15:40,480
Speaker 4: hundred pages of research each week, So that comes first,

307
00:15:40,800 --> 00:15:42,880
Speaker 4: other than spending time with my wife and my dog.

308
00:15:43,480 --> 00:15:45,880
Speaker 4: Those that's the most important thing I do because that's

309
00:15:45,920 --> 00:15:50,000
Speaker 4: where whether it's on energy topics or politics, or healthcare

310
00:15:50,600 --> 00:15:54,560
Speaker 4: or biotech like, it's an information laid in world. And

311
00:15:54,760 --> 00:15:58,520
Speaker 4: the great thing about JP Morgan is I can call

312
00:15:58,560 --> 00:16:02,560
Speaker 4: just about anybody say Hi, I'm the chief strategist on

313
00:16:02,600 --> 00:16:05,160
Speaker 4: the asthmatters, Miss JP Morgan. I'd like to talk to

314
00:16:05,160 --> 00:16:07,240
Speaker 4: you about something. Would you talk to me? I can

315
00:16:07,280 --> 00:16:10,960
Speaker 4: count on one hand the number of times that somebody

316
00:16:10,960 --> 00:16:13,360
Speaker 4: had just flat out said no. And one of those

317
00:16:13,400 --> 00:16:19,000
Speaker 4: people was the bundler for Bernie Madoff. And that's the

318
00:16:19,040 --> 00:16:22,320
Speaker 4: great thing about the halo effect that Jamie has brought

319
00:16:22,400 --> 00:16:25,720
Speaker 4: to JP Morgan. And remember I worked at JP Morgan

320
00:16:25,760 --> 00:16:28,520
Speaker 4: for many years when there was no halo, and you

321
00:16:28,560 --> 00:16:31,120
Speaker 4: know that halo effect is very powerful and benefits me

322
00:16:31,160 --> 00:16:33,120
Speaker 4: and a whole bunch of other people to company. And

323
00:16:33,440 --> 00:16:35,200
Speaker 4: I recognize that every day. So I spent a lot

324
00:16:35,240 --> 00:16:39,360
Speaker 4: of time working you know, deeply on some of these topics,

325
00:16:39,640 --> 00:16:42,640
Speaker 4: and energy as an example, I needed to learn about

326
00:16:42,760 --> 00:16:45,640
Speaker 4: energy from somebody that understood energy better than anybody else.

327
00:16:46,080 --> 00:16:48,920
Speaker 4: So fifteen years ago I made a pilgrimage to Manitoba

328
00:16:48,960 --> 00:16:51,760
Speaker 4: in February, right, which was can you prove yourself because

329
00:16:51,800 --> 00:16:55,320
Speaker 4: it was cold, and I established a relationship with vaclub Smill,

330
00:16:55,360 --> 00:16:59,440
Speaker 4: who for over a decade was my personal technical shirpa

331
00:16:59,560 --> 00:17:01,480
Speaker 4: on standing and learning about energy.

332
00:17:01,760 --> 00:17:03,800
Speaker 1: So it's someone we haven't had on the podcast.

333
00:17:04,080 --> 00:17:09,280
Speaker 4: He is a hermit, he's he's check, he's over eighty oh okay,

334
00:17:09,400 --> 00:17:13,280
Speaker 4: and he's a hermit, and he he hates the political

335
00:17:13,400 --> 00:17:15,840
Speaker 4: dynamics in the United States. So he will never get

336
00:17:15,880 --> 00:17:16,360
Speaker 4: it him here.

337
00:17:16,600 --> 00:17:19,000
Speaker 1: We'll go to Manitoba, we'll try.

338
00:17:19,080 --> 00:17:21,640
Speaker 2: I'm still stuck on the idea of the made off guy,

339
00:17:21,880 --> 00:17:24,440
Speaker 2: like answering the phone and being like, oh, we would

340
00:17:24,480 --> 00:17:27,760
Speaker 2: prefer not to talk in depth about why our line

341
00:17:27,840 --> 00:17:29,800
Speaker 2: just keeps going up in a very smooth way.

342
00:17:30,440 --> 00:17:34,240
Speaker 4: It was an eleven compound, yeah turn with two percent vault.

343
00:17:34,560 --> 00:17:38,480
Speaker 4: It was an unprecedented return history, like in the history

344
00:17:38,520 --> 00:17:39,160
Speaker 4: of edge funds.

345
00:17:39,200 --> 00:17:42,040
Speaker 1: So well, also, Tracy, I don't expect you to comment

346
00:17:42,080 --> 00:17:44,320
Speaker 1: on this, Michael. But Tracy, if I recall, one of

347
00:17:44,320 --> 00:17:48,280
Speaker 1: the people who was initially raised alarms about Madoff was

348
00:17:48,320 --> 00:17:50,880
Speaker 1: actually a guy man Zames at JP Morgan. Oh yeah,

349
00:17:50,880 --> 00:17:53,040
Speaker 1: so perhaps there was a reason I don't we don't

350
00:17:53,040 --> 00:17:54,720
Speaker 1: want to talk to these JP Morgan fellows.

351
00:17:54,840 --> 00:17:57,680
Speaker 4: Actually, if you go back, if you go back even further,

352
00:17:57,920 --> 00:17:59,960
Speaker 4: in either two thousand and one or two thousand and two,

353
00:18:00,000 --> 00:18:02,840
Speaker 4: there was a guy named Harry Markoppolos, Yeah, in Boston,

354
00:18:03,320 --> 00:18:06,320
Speaker 4: and he wrote a piece basically saying Bernie Madoff as

355
00:18:06,320 --> 00:18:08,600
Speaker 4: a fraud as upon his scheme, and he sent it

356
00:18:08,640 --> 00:18:11,560
Speaker 4: to the SEC and they kind of poked around and

357
00:18:11,720 --> 00:18:12,200
Speaker 4: did nothing.

358
00:18:12,320 --> 00:18:15,159
Speaker 2: Did nothing. Yeah, Okay, So when it comes to I

359
00:18:15,160 --> 00:18:18,160
Speaker 2: guess learning about news sectors like AI, I think one

360
00:18:18,160 --> 00:18:21,879
Speaker 2: of the things that everyone us included is struggling with

361
00:18:21,920 --> 00:18:25,440
Speaker 2: at the moment is to understand like how much actual

362
00:18:25,560 --> 00:18:28,439
Speaker 2: corporate spend there is going to be on this. And

363
00:18:28,760 --> 00:18:33,159
Speaker 2: for obvious reasons, companies, you know, they're not necessarily that

364
00:18:33,359 --> 00:18:36,840
Speaker 2: vocal about how much they're spending on this technology, although

365
00:18:36,880 --> 00:18:39,280
Speaker 2: they do like to talk about the importance of AI

366
00:18:39,480 --> 00:18:42,600
Speaker 2: in a sort of very nebulous theoretical manner, but what

367
00:18:42,640 --> 00:18:46,720
Speaker 2: are you doing to understand the actual spend on this tech?

368
00:18:46,960 --> 00:18:51,200
Speaker 4: Right? So there's this multiple stages to this. The first

369
00:18:51,200 --> 00:18:54,600
Speaker 4: stage was understanding how much the hyperscalers were spending. And

370
00:18:54,920 --> 00:18:58,440
Speaker 4: once we looked at those numbers, we were astonished. And

371
00:18:58,640 --> 00:19:02,600
Speaker 4: for the reasons you're mentioning concerned, right, I mean, there's

372
00:19:02,600 --> 00:19:04,400
Speaker 4: not a lot of examples. If you look back at

373
00:19:04,440 --> 00:19:08,960
Speaker 4: cable and airlines and casinos. When you start seeing R

374
00:19:09,000 --> 00:19:11,080
Speaker 4: and D and CAPEX as a presera of revenue go

375
00:19:11,119 --> 00:19:14,320
Speaker 4: to twenty to thirty percent for a sustained period, you know,

376
00:19:14,560 --> 00:19:18,439
Speaker 4: usually that's turned out bad. So we said, here, okay,

377
00:19:18,440 --> 00:19:21,560
Speaker 4: what's going on here? The next thing to look at

378
00:19:21,800 --> 00:19:27,200
Speaker 4: was signs of corporate adoption. Now there's nobody that hates

379
00:19:27,280 --> 00:19:30,280
Speaker 4: McKinsey surveys more than me, but you know, we had

380
00:19:30,320 --> 00:19:36,040
Speaker 4: mckensey surveys, you have Baine surveys, you've got Census surveys,

381
00:19:36,200 --> 00:19:38,560
Speaker 4: a lot of which we're pointing in the same direction,

382
00:19:38,880 --> 00:19:42,680
Speaker 4: which is that corporate adoption has been accelerating, particularly over

383
00:19:42,680 --> 00:19:46,720
Speaker 4: the last six months. And CEO surveys, I mean, there's

384
00:19:46,840 --> 00:19:49,919
Speaker 4: too many of them that are all telling you the

385
00:19:49,960 --> 00:19:53,120
Speaker 4: same thing. For them to be wildly off the mark.

386
00:19:53,840 --> 00:19:57,000
Speaker 4: But again, that doesn't tell you how much they're willing

387
00:19:57,040 --> 00:20:02,000
Speaker 4: to spend on it, right, You know, everybody might enjoy lollipops,

388
00:20:02,000 --> 00:20:03,280
Speaker 4: but that doesn't mean they're going to spend a lot

389
00:20:03,280 --> 00:20:05,760
Speaker 4: of their personal income on them. Then the question is

390
00:20:06,240 --> 00:20:09,639
Speaker 4: how much are the big hyperscalers starting to earn on AI?

391
00:20:10,560 --> 00:20:14,600
Speaker 4: And so far Microsoft is really the only company that's

392
00:20:14,640 --> 00:20:19,359
Speaker 4: disclosing it very discreetly, and we had a chart. I

393
00:20:19,440 --> 00:20:23,080
Speaker 4: published a piece on this on May thirteenth, And Microsoft's

394
00:20:23,119 --> 00:20:25,840
Speaker 4: the only company so far that's really explicitly telling you

395
00:20:25,920 --> 00:20:28,280
Speaker 4: how much they're earning. And those numbers are growing at

396
00:20:28,320 --> 00:20:30,840
Speaker 4: a pretty rapid base from a low base. But they're

397
00:20:30,880 --> 00:20:31,400
Speaker 4: earning a lot.

398
00:20:31,520 --> 00:20:33,960
Speaker 2: One hundred and fifty percent over the past year is

399
00:20:33,960 --> 00:20:35,000
Speaker 2: what I see in the right.

400
00:20:35,040 --> 00:20:38,080
Speaker 4: Right, the other companies are still burying it in cloud revenue.

401
00:20:38,160 --> 00:20:40,720
Speaker 1: Yeah, there's a lot of questions just on this, so

402
00:20:40,880 --> 00:20:43,119
Speaker 1: get to them. It's one thing for a company to

403
00:20:43,200 --> 00:20:46,840
Speaker 1: spend on AI. You know a lot of pilot projects

404
00:20:46,920 --> 00:20:50,360
Speaker 1: on there. When you ask companies in terms of yes,

405
00:20:50,440 --> 00:20:55,000
Speaker 1: but is this actually saving money? Is this actually profitable investment?

406
00:20:55,040 --> 00:20:59,000
Speaker 1: It starts to get Cagier. And sometimes when you hear

407
00:20:59,080 --> 00:21:02,040
Speaker 1: people talk about the returns that they've got from AI spending,

408
00:21:02,280 --> 00:21:04,840
Speaker 1: it actually sounds like they're talking about more traditional machine

409
00:21:04,880 --> 00:21:07,479
Speaker 1: learning algorithm stuff that's been going on for a while.

410
00:21:07,840 --> 00:21:10,919
Speaker 1: But for this spend to continue, it's going to have

411
00:21:10,960 --> 00:21:13,680
Speaker 1: to flow through to the bottom line in some way.

412
00:21:14,000 --> 00:21:16,560
Speaker 1: And I'm curious what you're seeing on that in terms of, like,

413
00:21:16,760 --> 00:21:19,280
Speaker 1: here is a company that's not a tech company, maybe

414
00:21:19,280 --> 00:21:22,160
Speaker 1: it's an airline company, maybe it's a chemicals company, whatever,

415
00:21:22,320 --> 00:21:23,879
Speaker 1: where it's like, you know what they can point to

416
00:21:23,920 --> 00:21:28,159
Speaker 1: something that's not just algorithms, that is actually generative AI

417
00:21:28,800 --> 00:21:32,080
Speaker 1: that is making money for them or saving money for

418
00:21:32,160 --> 00:21:33,600
Speaker 1: them in some way. And what are you seeing on that?

419
00:21:33,760 --> 00:21:39,160
Speaker 4: Yeah, we're seeing anecdotes, anecdotes, We're seeing anecdotes, anacdata. Yeah,

420
00:21:39,200 --> 00:21:42,159
Speaker 4: and usually it has to do with faster throughput and

421
00:21:42,240 --> 00:21:46,520
Speaker 4: call centers with less people, Okay, stuff like that, faster

422
00:21:46,840 --> 00:21:51,359
Speaker 4: customer acquisition, fraud reduction. Right for big banks like JP

423
00:21:51,480 --> 00:21:55,600
Speaker 4: Morgan JENERAI, one of the huge potential payoffs is identifying

424
00:21:55,640 --> 00:21:59,040
Speaker 4: serial defaulters and mortgages and credit cards and things like that.

425
00:21:59,480 --> 00:22:03,600
Speaker 4: So what is interesting is that among the best surveys

426
00:22:03,600 --> 00:22:07,680
Speaker 4: that are done. The surveys are asking them explicitly for

427
00:22:07,920 --> 00:22:11,600
Speaker 4: generative AI, how much money are you saving? And most

428
00:22:11,640 --> 00:22:14,480
Speaker 4: of the answers are still in the lowest category zero

429
00:22:14,560 --> 00:22:18,080
Speaker 4: to ten percent okay by far, and then it's the above.

430
00:22:18,080 --> 00:22:21,600
Speaker 4: Those levels have fewer responses. We're still in the exploratory

431
00:22:21,600 --> 00:22:23,399
Speaker 4: phase here, and I don't think we're out of the woods,

432
00:22:23,440 --> 00:22:25,639
Speaker 4: and there may be a day of reckoning at some point.

433
00:22:25,680 --> 00:22:28,920
Speaker 4: I think, because of how absurdly profitable these companies are,

434
00:22:29,080 --> 00:22:31,040
Speaker 4: the markets are going to give them another eighteen months

435
00:22:31,080 --> 00:22:34,720
Speaker 4: at least for the proof statement to bear out.

436
00:22:35,160 --> 00:22:37,040
Speaker 2: This was going to be my next question was, like,

437
00:22:37,160 --> 00:22:39,800
Speaker 2: how much of a leash are investors allowing some of

438
00:22:39,840 --> 00:22:40,920
Speaker 2: these company As.

439
00:22:40,760 --> 00:22:44,320
Speaker 4: Long as as long as the cloud revenue and the

440
00:22:44,359 --> 00:22:49,239
Speaker 4: AI revenue specifically are growing at fifty percent plus and

441
00:22:49,320 --> 00:22:52,160
Speaker 4: as long as the free cash flow margins don't fall

442
00:22:52,200 --> 00:22:55,600
Speaker 4: below some critical level, I think, you know, the markets

443
00:22:55,640 --> 00:22:57,159
Speaker 4: will still feel okay about it.

444
00:22:57,800 --> 00:23:01,640
Speaker 2: I wanted to also ask a question about your overall career.

445
00:23:01,760 --> 00:23:03,520
Speaker 2: As we said in the intro, you've been at JP

446
00:23:03,640 --> 00:23:07,960
Speaker 2: Morgan since nineteen eighty seven, which is a phenomenal, phenomenal

447
00:23:08,000 --> 00:23:11,320
Speaker 2: amount of time. Looking back, what was the hardest or

448
00:23:11,440 --> 00:23:15,040
Speaker 2: most challenging sector to wrap your head around over that

449
00:23:15,080 --> 00:23:17,639
Speaker 2: period of time, Because, as we've been discussing, you jump

450
00:23:17,680 --> 00:23:20,400
Speaker 2: from thing to thing and you basically become an expert

451
00:23:20,400 --> 00:23:23,199
Speaker 2: on whatever is important to investors or the market at

452
00:23:23,200 --> 00:23:24,639
Speaker 2: that particular moment in time.

453
00:23:26,200 --> 00:23:29,600
Speaker 4: Yeah, and I usually get airlifted into things that nobody

454
00:23:29,640 --> 00:23:32,439
Speaker 4: you know that are pain like I became for the

455
00:23:32,480 --> 00:23:34,439
Speaker 4: asset manager business and even for the broader firm. I

456
00:23:34,480 --> 00:23:37,000
Speaker 4: ended up being our COVID point person, and I put

457
00:23:37,000 --> 00:23:42,440
Speaker 4: together a medical advisory committee of biophysicians and mathematical epidemiologists

458
00:23:42,440 --> 00:23:45,600
Speaker 4: and people like that from La Joya Institute of Imminology,

459
00:23:45,680 --> 00:23:49,320
Speaker 4: and so I get airlifted into things. I would say,

460
00:23:49,400 --> 00:23:52,960
Speaker 4: oddly enough, of all the sectors and subtectors, I can

461
00:23:53,000 --> 00:23:55,600
Speaker 4: think of, the one that drives me nuts the most

462
00:23:56,119 --> 00:24:00,080
Speaker 4: is health care and managed care because you know how

463
00:24:00,119 --> 00:24:02,440
Speaker 4: they snuck this provision into the energy bill three years

464
00:24:02,480 --> 00:24:04,840
Speaker 4: ago where the government can negotiate drug prices. Finally I

465
00:24:04,840 --> 00:24:07,600
Speaker 4: can be able to do that. On the face of it,

466
00:24:07,880 --> 00:24:12,960
Speaker 4: that should be negative for the big farmer companies, but

467
00:24:13,320 --> 00:24:15,840
Speaker 4: because they can play around with all sorts of discounting

468
00:24:15,880 --> 00:24:20,440
Speaker 4: mechanisms and distribution arrangements, a lot of which aren't super public.

469
00:24:21,000 --> 00:24:23,639
Speaker 4: The actual impact on their margins so far has been

470
00:24:23,680 --> 00:24:28,840
Speaker 4: somewhat negligible, So there's a there's an impenetrability to parts

471
00:24:28,960 --> 00:24:32,800
Speaker 4: of the way the healthcare system function that drives me

472
00:24:32,840 --> 00:24:37,640
Speaker 4: a little bananas. That said, large cap pharma and biotech

473
00:24:37,840 --> 00:24:42,520
Speaker 4: right now are among the cheapest sectors if you compare

474
00:24:42,600 --> 00:24:45,680
Speaker 4: price to book to projected ROE or something like that.

475
00:24:46,040 --> 00:24:48,119
Speaker 4: We have this giant map of all those sectors and

476
00:24:48,160 --> 00:24:52,680
Speaker 4: subsectors and industries and where they trade relative to price

477
00:24:52,760 --> 00:24:55,640
Speaker 4: to book and ROA and ROE, and you know, healthcare

478
00:24:55,880 --> 00:24:59,040
Speaker 4: is cheaper. But it would normally make me want to

479
00:24:59,160 --> 00:25:01,879
Speaker 4: dive in and kind of come up with an asset

480
00:25:01,880 --> 00:25:05,600
Speaker 4: allocation recommendation for our investment process, and it's just very

481
00:25:05,640 --> 00:25:06,240
Speaker 4: difficult to do.

482
00:25:06,600 --> 00:25:09,040
Speaker 2: This is how I feel about US health insurance. By

483
00:25:09,080 --> 00:25:11,800
Speaker 2: the way, as someone who did not grow up that

484
00:25:11,920 --> 00:25:15,080
Speaker 2: much in the US, I still struggle to wrap my

485
00:25:15,119 --> 00:25:17,280
Speaker 2: head around how all of this works. It is truly

486
00:25:17,359 --> 00:25:18,600
Speaker 2: impenetrable to me.

487
00:25:18,960 --> 00:25:21,919
Speaker 1: I think many people would agree with that assessment. I

488
00:25:21,960 --> 00:25:25,200
Speaker 1: want to go back to actually AI for a second,

489
00:25:25,640 --> 00:25:30,200
Speaker 1: because you're talking about the hyperscalers and this booming business

490
00:25:30,280 --> 00:25:32,199
Speaker 1: for them. It's still actually a little bit hard for

491
00:25:32,240 --> 00:25:37,640
Speaker 1: me to understand, because for some companies, their AI revenue

492
00:25:37,680 --> 00:25:41,720
Speaker 1: is very easy to understand. Right, So Microsoft sells various

493
00:25:41,800 --> 00:25:45,080
Speaker 1: things that go along with its sweet that whatever, whether

494
00:25:45,119 --> 00:25:47,440
Speaker 1: it's coding tools or so forth. Okay, they're selling an

495
00:25:47,480 --> 00:25:50,679
Speaker 1: AI service. Maybe Google one day figure out how to

496
00:25:50,720 --> 00:25:54,080
Speaker 1: do that too. For a company like Meta, their model

497
00:25:54,160 --> 00:25:57,320
Speaker 1: is an open source model. They don't sell enterprise LAMA

498
00:25:57,359 --> 00:26:00,879
Speaker 1: as far as I know, yet, they spend like crazy

499
00:26:00,960 --> 00:26:03,119
Speaker 1: on AI and they talk about how much they spend.

500
00:26:03,840 --> 00:26:09,520
Speaker 1: Can you articulate for me what Meta's AI business model

501
00:26:09,800 --> 00:26:11,879
Speaker 1: is because all I hear is about their spending and

502
00:26:11,920 --> 00:26:14,040
Speaker 1: I don't even know what revenue they're deriving.

503
00:26:13,760 --> 00:26:16,320
Speaker 4: From the yeh. When you when you try to understand that,

504
00:26:16,359 --> 00:26:20,080
Speaker 4: what you learn is that Meta basically considers itself a

505
00:26:20,200 --> 00:26:24,119
Speaker 4: generative AI company and that everything they do and the

506
00:26:24,160 --> 00:26:27,080
Speaker 4: success of all of its algorithms and social media tools

507
00:26:27,840 --> 00:26:31,560
Speaker 4: maintain their market dominance because of how well they function

508
00:26:32,200 --> 00:26:35,240
Speaker 4: and the way that they function is because they're driven

509
00:26:35,320 --> 00:26:38,680
Speaker 4: by AI. And I and I agree. I think they

510
00:26:38,760 --> 00:26:44,440
Speaker 4: thought that the open source LAM model would become an

511
00:26:44,440 --> 00:26:47,639
Speaker 4: indirect profit center as people build tools and applications on

512
00:26:47,720 --> 00:26:49,920
Speaker 4: top of it. And it seems like they're stepping away

513
00:26:49,920 --> 00:26:52,520
Speaker 4: from that, huh right now like they a year ago.

514
00:26:52,560 --> 00:26:54,119
Speaker 4: That was the kind of message you were hearing in

515
00:26:54,119 --> 00:26:57,600
Speaker 4: not so much right now, but they you know that,

516
00:26:57,640 --> 00:26:58,760
Speaker 4: I think they would say.

517
00:26:58,640 --> 00:27:02,840
Speaker 1: Are somehow all of this spending which they talk as

518
00:27:02,960 --> 00:27:05,320
Speaker 1: AI spending specifically, even though.

519
00:27:05,280 --> 00:27:06,359
Speaker 4: They're selling AI.

520
00:27:07,000 --> 00:27:10,280
Speaker 1: Here's how, somehow it's accruing to their fundamental business of

521
00:27:10,320 --> 00:27:12,440
Speaker 1: selling ads or selling content selling.

522
00:27:12,240 --> 00:27:14,840
Speaker 4: Here, And here's how I would I would articulate that

523
00:27:15,080 --> 00:27:17,320
Speaker 4: if you go back three four years, there were these

524
00:27:17,400 --> 00:27:22,000
Speaker 4: really interesting cell side presentations that talked about, well, Google's

525
00:27:22,000 --> 00:27:24,040
Speaker 4: going to try to compete with Apple here, Apple's going

526
00:27:24,080 --> 00:27:26,520
Speaker 4: to try to compete with Nvidia here, and video is

527
00:27:26,560 --> 00:27:28,840
Speaker 4: going to try and compete with Opening out here. And

528
00:27:29,040 --> 00:27:32,159
Speaker 4: guess what, a lot of those motes are still in place,

529
00:27:32,720 --> 00:27:36,479
Speaker 4: and a lot of the encroachment that people were looking

530
00:27:36,560 --> 00:27:42,920
Speaker 4: for hasn't happened. So Meta's AI spend is primarily focused

531
00:27:43,160 --> 00:27:47,120
Speaker 4: on maintaining an impenetrable mode for other companies that might

532
00:27:47,160 --> 00:27:49,960
Speaker 4: otherwise want to compete with their core business. And that's

533
00:27:50,000 --> 00:27:53,840
Speaker 4: how they see it, which means that you basically will

534
00:27:53,880 --> 00:27:57,040
Speaker 4: want to look at top line, free cash flow and

535
00:27:57,119 --> 00:28:00,280
Speaker 4: other profitability metrics for the company as a whole to

536
00:28:00,400 --> 00:28:04,679
Speaker 4: understand the benefits of all that spent. I think investors

537
00:28:04,680 --> 00:28:07,080
Speaker 4: are right to be skeptical. It's only three years ago

538
00:28:07,320 --> 00:28:09,080
Speaker 4: that they took a ton of money, put it in

539
00:28:09,119 --> 00:28:13,400
Speaker 4: a pile, and set it on fire. Yeah, with the metaverse, right,

540
00:28:13,640 --> 00:28:18,520
Speaker 4: so they they have had a history of misreading where

541
00:28:18,520 --> 00:28:19,160
Speaker 4: things are going.

542
00:28:19,440 --> 00:28:22,000
Speaker 1: Tracy, by the way, I actually really admire how quick

543
00:28:22,040 --> 00:28:24,080
Speaker 1: they were willing to pivot off that, because you think,

544
00:28:24,119 --> 00:28:27,000
Speaker 1: you know, you make this big bed. It's embarrassing you

545
00:28:27,160 --> 00:28:30,320
Speaker 1: rename your company meta off of this, and then you're like,

546
00:28:30,480 --> 00:28:33,040
Speaker 1: we're an a company. I actually that raised my admiration

547
00:28:33,200 --> 00:28:36,160
Speaker 1: for Zuckerberg as a CEO, that he didn't like succumb

548
00:28:36,240 --> 00:28:38,000
Speaker 1: to some big sunk cost fallacy there.

549
00:28:38,040 --> 00:28:40,800
Speaker 4: Anyway, Well, look, I would.

550
00:28:40,640 --> 00:28:43,640
Speaker 2: Change my business model. That's my version of the Canes question.

551
00:28:43,880 --> 00:28:47,360
Speaker 4: He has demonstrated a mental agility to pivot as well

552
00:28:47,800 --> 00:28:50,080
Speaker 4: earlier this year, and I'm not going to make a

553
00:28:50,160 --> 00:28:53,000
Speaker 4: value judgment on this, but I was struck by him

554
00:28:53,000 --> 00:28:56,600
Speaker 4: saying that he needed to reinject the dose of masculinity

555
00:28:56,640 --> 00:28:58,960
Speaker 4: and oh yeah, yeah, I think that's a sign of

556
00:28:58,960 --> 00:29:03,440
Speaker 4: somebody that yeah, this is audio, so you can't see

557
00:29:03,440 --> 00:29:05,800
Speaker 4: my eyes rolling, but that's but you know, that's a

558
00:29:05,840 --> 00:29:08,480
Speaker 4: sign of somebody that's that is willing to kind of

559
00:29:08,560 --> 00:29:10,240
Speaker 4: change horses when yeah.

560
00:29:10,760 --> 00:29:14,760
Speaker 2: Oh, dear, I have thoughts on AI as a masculine endeavor.

561
00:29:14,800 --> 00:29:17,760
Speaker 2: But anyway, one thing I wanted to ask is Okay,

562
00:29:17,800 --> 00:29:20,680
Speaker 2: so at the beginning of this conversation, you described AI

563
00:29:20,920 --> 00:29:22,400
Speaker 2: as the bet of the century.

564
00:29:22,440 --> 00:29:22,719
Speaker 4: It is.

565
00:29:23,080 --> 00:29:25,080
Speaker 2: And when I hear you know, when we're talking about

566
00:29:25,080 --> 00:29:28,320
Speaker 2: the dominance of the big US tech players and the

567
00:29:28,360 --> 00:29:31,680
Speaker 2: moats and the billions, if not trillions of dollars that

568
00:29:31,720 --> 00:29:35,600
Speaker 2: people are spending on this now, it feels like almost

569
00:29:35,600 --> 00:29:38,240
Speaker 2: certainly it's the big players that are going to emerge

570
00:29:38,480 --> 00:29:40,960
Speaker 2: as the winners from this. It seems very very hard

571
00:29:41,000 --> 00:29:44,000
Speaker 2: for smaller entrance to get into this. Is that right?

572
00:29:44,160 --> 00:29:47,000
Speaker 2: Is it just you know, a fate a'll complete. I

573
00:29:47,000 --> 00:29:48,400
Speaker 2: guess the big guys are.

574
00:29:48,320 --> 00:29:50,960
Speaker 4: Going to be hyperscalar level, yes, but in a whole

575
00:29:50,960 --> 00:29:54,800
Speaker 4: bunch of other industries not necessarily plenty of room. And

576
00:29:54,840 --> 00:29:57,200
Speaker 4: but look, you're seeing these unicorns all the time that

577
00:29:57,280 --> 00:30:01,760
Speaker 4: are chipping away at different aspects of you know, machine

578
00:30:01,840 --> 00:30:04,440
Speaker 4: learning and generative AI. And one of the most fascinating

579
00:30:04,480 --> 00:30:07,920
Speaker 4: ones is look at Palenteer, right, I mean, Palenteer is

580
00:30:08,040 --> 00:30:12,880
Speaker 4: taking a run at the large defense contractors primarily through

581
00:30:12,920 --> 00:30:17,560
Speaker 4: its nimbleness and better application and understanding of the intersection

582
00:30:17,720 --> 00:30:21,440
Speaker 4: between generative AI and defense weaponry. And you know, for

583
00:30:21,440 --> 00:30:25,200
Speaker 4: everybody that thinks about data centers, the massive concentration of

584
00:30:25,280 --> 00:30:29,280
Speaker 4: data centers in Virginia and in the pgm isoregion more

585
00:30:29,320 --> 00:30:33,520
Speaker 4: broadly is very clear indication of just how the national

586
00:30:33,520 --> 00:30:36,840
Speaker 4: security and defense stuff is tied into generative AI and

587
00:30:36,880 --> 00:30:40,040
Speaker 4: who their big users are. Right, if that didn't matter,

588
00:30:40,480 --> 00:30:42,880
Speaker 4: there would be a lot more data centers built in

589
00:30:42,920 --> 00:30:46,160
Speaker 4: the Midwest where twenty to thirty percent of the time

590
00:30:46,320 --> 00:30:49,920
Speaker 4: electricity prices are actually negative. Right. Think about this. You

591
00:30:49,960 --> 00:30:53,760
Speaker 4: have certain applications that are so reliant on not there

592
00:30:53,800 --> 00:30:56,719
Speaker 4: being a millisecond of lag that they'd be willing to

593
00:30:56,720 --> 00:31:01,320
Speaker 4: pay more for power and proximity than to have cheaper

594
00:31:01,360 --> 00:31:03,000
Speaker 4: power that has just a tiny bit of lag.

595
00:31:03,000 --> 00:31:23,520
Speaker 1: Built one company where there's some question about the moat

596
00:31:23,800 --> 00:31:27,280
Speaker 1: is alphabet and you know, people are like, well, could

597
00:31:27,360 --> 00:31:31,920
Speaker 1: open Ai just become Peopil's default homepage or the first

598
00:31:31,960 --> 00:31:34,240
Speaker 1: pages that people go to when they want to learn something.

599
00:31:34,520 --> 00:31:39,600
Speaker 1: And you've seen Alphabet shares underperformed into some of the rebound.

600
00:31:39,680 --> 00:31:42,680
Speaker 1: I think from a multiple standpoint, it's trading at some

601
00:31:42,720 --> 00:31:46,840
Speaker 1: of its cheapest levels in a very long time. There's

602
00:31:46,880 --> 00:31:48,880
Speaker 1: a lot of questions about the degree to which it

603
00:31:48,920 --> 00:31:52,920
Speaker 1: can productize its own models. Historically it's not so good

604
00:31:52,920 --> 00:31:55,880
Speaker 1: at that. It makes great tech. The models are very good.

605
00:31:56,320 --> 00:31:59,040
Speaker 1: I'm still discovering new alphabet models that I never knew

606
00:31:59,080 --> 00:32:02,760
Speaker 1: existed because they're very under some URL. Is there a

607
00:32:02,880 --> 00:32:07,120
Speaker 1: risk to their business model, specifically that perhaps their distribution

608
00:32:07,560 --> 00:32:10,360
Speaker 1: their ownership of digital real estate risk to the rent

609
00:32:10,440 --> 00:32:11,520
Speaker 1: that they can extract from it.

610
00:32:12,600 --> 00:32:14,720
Speaker 4: Yeah, but this is the kind of thing where I

611
00:32:14,800 --> 00:32:17,760
Speaker 4: would continue to just look at their market share. You know,

612
00:32:18,800 --> 00:32:21,200
Speaker 4: for a while Bing was taking a run at.

613
00:32:21,040 --> 00:32:23,280
Speaker 1: Them, and then.

614
00:32:22,720 --> 00:32:26,240
Speaker 4: Nothing really happened. And okay, So in Europe, when you

615
00:32:26,320 --> 00:32:29,160
Speaker 4: buy a device, the Apple is no longer able to

616
00:32:29,160 --> 00:32:31,600
Speaker 4: make Google the default engine, so you have to set

617
00:32:31,600 --> 00:32:34,960
Speaker 4: it up yourself. People pay Google anyway, so you could

618
00:32:34,960 --> 00:32:38,080
Speaker 4: actually argue that Google is overpaying Apple for the default

619
00:32:38,160 --> 00:32:41,360
Speaker 4: status on their devices because people would be willing to

620
00:32:41,880 --> 00:32:45,480
Speaker 4: pick it anyway. I don't know. I think obviously the

621
00:32:45,600 --> 00:32:49,400
Speaker 4: tools open a I personally love perplexity, but I use

622
00:32:49,480 --> 00:32:53,600
Speaker 4: them when necessary. And I wouldn't bet against Google's generic

623
00:32:53,720 --> 00:32:57,160
Speaker 4: market share for search until you actually saw them beginning

624
00:32:57,160 --> 00:32:59,840
Speaker 4: to lose market share. It reminds me for the last

625
00:33:00,040 --> 00:33:04,080
Speaker 4: fifteen years people have been talking and screaming about the

626
00:33:04,200 --> 00:33:07,000
Speaker 4: end of the dollar as the world's reserve currency. And

627
00:33:07,640 --> 00:33:10,560
Speaker 4: every year now we look at the We look at

628
00:33:10,600 --> 00:33:13,360
Speaker 4: the data on the share of farm exchange reserves investment

629
00:33:13,360 --> 00:33:16,200
Speaker 4: in dollars, the share of corporate debt and equity debt

630
00:33:16,200 --> 00:33:20,240
Speaker 4: issuance in dollars, particularly offshore, the share of swift transaction payments,

631
00:33:20,600 --> 00:33:23,959
Speaker 4: and the dollar share of all these things BIS intercompany

632
00:33:24,000 --> 00:33:28,240
Speaker 4: bank loans across border. The dollar share is kind of unchanged.

633
00:33:28,840 --> 00:33:30,920
Speaker 4: So it's you know, well, let's you have to wait

634
00:33:30,960 --> 00:33:31,920
Speaker 4: until you see stuff.

635
00:33:32,400 --> 00:33:35,920
Speaker 2: Network effects are definitely a thing. One thing I wanted

636
00:33:35,920 --> 00:33:37,480
Speaker 2: to ask you, or one of the reasons we wanted

637
00:33:37,480 --> 00:33:39,800
Speaker 2: to have you on is, as you say, you read

638
00:33:39,840 --> 00:33:42,280
Speaker 2: a lot, you talk to a lot of people, a

639
00:33:42,320 --> 00:33:45,960
Speaker 2: lot of clients, and I'm curious, what's the sort of

640
00:33:46,160 --> 00:33:50,480
Speaker 2: time split or the question split between I guess Trump

641
00:33:50,960 --> 00:33:55,080
Speaker 2: and policy versus you know, other market questions like AI.

642
00:33:57,240 --> 00:33:59,000
Speaker 4: It's hard. I mean, it's hard to say, right. I

643
00:33:59,000 --> 00:34:03,520
Speaker 4: mean the tariff, the whole tariff drama was took an

644
00:34:03,640 --> 00:34:06,440
Speaker 4: enormous amount of time because I would say that a

645
00:34:06,480 --> 00:34:10,319
Speaker 4: lot of CEOs, and I've described this recently, as I

646
00:34:10,320 --> 00:34:12,239
Speaker 4: went to college in the Boston area, and every time

647
00:34:12,239 --> 00:34:14,439
Speaker 4: it snowed, some of the local kids would come out

648
00:34:14,560 --> 00:34:15,920
Speaker 4: and when there's a lot of snow and ice on

649
00:34:15,960 --> 00:34:17,960
Speaker 4: the road, they would hold onto the back of trucks

650
00:34:17,960 --> 00:34:20,440
Speaker 4: and go skitching. And I had never heard of skitching

651
00:34:20,480 --> 00:34:22,920
Speaker 4: until I saw people doing it. I never did it because,

652
00:34:22,960 --> 00:34:25,600
Speaker 4: like after you stop skitching, you tend to slam face

653
00:34:25,600 --> 00:34:28,320
Speaker 4: first into the ice. A lot of CEOs are sketching

654
00:34:28,400 --> 00:34:31,480
Speaker 4: right now on this administration because they thought that what

655
00:34:31,520 --> 00:34:35,280
Speaker 4: they were going to get was deregulation oil and gas

656
00:34:35,280 --> 00:34:38,560
Speaker 4: pipelines and lower corporate tax rates. That's what they were

657
00:34:38,560 --> 00:34:41,160
Speaker 4: supporting this guy for. And they come out of the

658
00:34:41,200 --> 00:34:46,240
Speaker 4: gate with deportations and tariffs, and the productivity shock supply

659
00:34:46,360 --> 00:34:49,120
Speaker 4: side agenda got put in the back seat for a while.

660
00:34:49,719 --> 00:34:51,600
Speaker 4: They appeared to be wanting to get started on some

661
00:34:51,640 --> 00:34:53,520
Speaker 4: of that stuff, and I think it's instant we can

662
00:34:53,520 --> 00:34:55,600
Speaker 4: talk about it. It'll be interesting to see if they

663
00:34:55,640 --> 00:34:59,440
Speaker 4: can get the Constitution pipeline reinvigorated, which would bring Marcella's

664
00:34:59,440 --> 00:35:03,320
Speaker 4: shale gas to the northeast. But in the beginning, because

665
00:35:03,400 --> 00:35:06,280
Speaker 4: so much of the first stuff was tariffs and deportation,

666
00:35:06,320 --> 00:35:08,680
Speaker 4: there's a lot of focus on that, and tariffs in

667
00:35:08,719 --> 00:35:12,640
Speaker 4: particular require a lot of work. There's twelve thousand HTS

668
00:35:12,640 --> 00:35:16,319
Speaker 4: codes across two hundred something countries, and they came up

669
00:35:16,400 --> 00:35:20,040
Speaker 4: with a patchwork of legislation that differentiate and so there's

670
00:35:20,080 --> 00:35:23,879
Speaker 4: just a lot of math involved in really understanding what

671
00:35:23,920 --> 00:35:24,640
Speaker 4: the impacts are.

672
00:35:25,360 --> 00:35:27,360
Speaker 1: By the way, I see you into Tufts and so

673
00:35:27,400 --> 00:35:29,520
Speaker 1: I think you're the first person to ever say you

674
00:35:29,560 --> 00:35:31,759
Speaker 1: went to a Boston area college and that not being

675
00:35:31,800 --> 00:35:35,319
Speaker 1: code word for Harvard, it actually you actually did go

676
00:35:35,360 --> 00:35:38,840
Speaker 1: to a Boston area college. Excellent university. There's lots of them, No,

677
00:35:38,960 --> 00:35:41,319
Speaker 1: I know there's lots, but at all you know, you

678
00:35:41,360 --> 00:35:45,359
Speaker 1: weren't being kloy when you said Boston area college. As

679
00:35:45,360 --> 00:35:47,960
Speaker 1: you mentioned. You know, another huge theme over the years,

680
00:35:48,000 --> 00:35:49,880
Speaker 1: and you write a lot about it, is energy and

681
00:35:50,200 --> 00:35:52,719
Speaker 1: learned a lot from Folkloff's mill. And you know, we

682
00:35:52,719 --> 00:35:54,719
Speaker 1: could talk hours just about energy. I would love to

683
00:35:54,760 --> 00:35:58,239
Speaker 1: talk hours about energy at some point, but in the

684
00:35:58,320 --> 00:36:02,879
Speaker 1: interest of time. How much has the AI story over

685
00:36:02,880 --> 00:36:07,680
Speaker 1: the last few years changed how you thought about the

686
00:36:07,719 --> 00:36:12,640
Speaker 1: future of North American energy needs and you know, energy

687
00:36:12,680 --> 00:36:15,279
Speaker 1: production and just the energy story in general.

688
00:36:15,719 --> 00:36:18,279
Speaker 4: You know, not until very recently, because it was a

689
00:36:18,320 --> 00:36:22,160
Speaker 4: negligible part of over electricity consumption, you know, sub one

690
00:36:22,200 --> 00:36:25,000
Speaker 4: percent until a couple of years ago, and has been

691
00:36:25,000 --> 00:36:28,880
Speaker 4: growing slowly. There are three big components to this electricity story.

692
00:36:29,760 --> 00:36:35,479
Speaker 4: Ones AI, Another one's electrification of transportation, you know ease yeah.

693
00:36:35,520 --> 00:36:37,479
Speaker 4: And the third one that people should focus on because

694
00:36:37,480 --> 00:36:40,600
Speaker 4: it's just as big, is the potential for electrification of

695
00:36:40,680 --> 00:36:46,279
Speaker 4: winter heating in homes, office buildings, and industrial locations. One

696
00:36:46,320 --> 00:36:49,879
Speaker 4: of the hardest things when you when you break down

697
00:36:50,000 --> 00:36:53,279
Speaker 4: energy consumption, an enormous amount of it, enormous found of

698
00:36:53,320 --> 00:36:57,680
Speaker 4: fossil fuel consumption is used for heat. Some of the heat,

699
00:36:58,320 --> 00:37:00,680
Speaker 4: around around a third of all the heat. Heat that's

700
00:37:00,800 --> 00:37:04,560
Speaker 4: used is used at temperatures below two hundred degrees integrade,

701
00:37:04,640 --> 00:37:08,280
Speaker 4: and that's where heat pumps are extremely efficient and measured

702
00:37:08,320 --> 00:37:10,640
Speaker 4: by something called the coefficient to performance, which is the

703
00:37:10,760 --> 00:37:12,920
Speaker 4: units of heat you get per unit of electricity you

704
00:37:12,920 --> 00:37:16,319
Speaker 4: put in. And so there's three components roughly equal, right,

705
00:37:16,360 --> 00:37:22,640
Speaker 4: So data centers, electrification of heating, electrication, transport, the wide

706
00:37:22,680 --> 00:37:27,239
Speaker 4: eyed you know, kind of fairy tail projections of the

707
00:37:27,320 --> 00:37:31,400
Speaker 4: Rocky Mountain Institute and green tech media people, Yeah, are

708
00:37:31,840 --> 00:37:36,319
Speaker 4: looking at twenty to twenty five percent growth in electricity

709
00:37:36,360 --> 00:37:39,000
Speaker 4: demand over the next ten years. I think that's way

710
00:37:39,040 --> 00:37:41,719
Speaker 4: too high adding up those three components. The lower end

711
00:37:41,800 --> 00:37:46,239
Speaker 4: is closer of like five to seven percent. BNF, which

712
00:37:46,280 --> 00:37:48,080
Speaker 4: I think does very good work on these topics. Is

713
00:37:48,120 --> 00:37:48,560
Speaker 4: somewhere in.

714
00:37:48,480 --> 00:37:50,200
Speaker 2: Between, thank you, thank you for the plug.

715
00:37:50,320 --> 00:37:53,160
Speaker 4: Yeah. Again, I know all of the sources, yeah, rigor

716
00:37:53,200 --> 00:37:56,880
Speaker 4: of their work, and so so I think we will need,

717
00:37:58,160 --> 00:38:03,160
Speaker 4: you know, seven or eight percent additional electricity capacity for

718
00:38:03,280 --> 00:38:06,440
Speaker 4: generation over the next decade or so. Electricity conception of

719
00:38:06,440 --> 00:38:08,440
Speaker 4: the United States has been roughly flat for twenty years,

720
00:38:08,480 --> 00:38:10,600
Speaker 4: but before that it went up a lot. The issue

721
00:38:10,600 --> 00:38:13,000
Speaker 4: back then was we were meeting it with nuclear and

722
00:38:13,080 --> 00:38:17,440
Speaker 4: large natural gas bets. It's a lot harder today because

723
00:38:18,120 --> 00:38:19,560
Speaker 4: most of the country's going to try to do it

724
00:38:19,560 --> 00:38:22,320
Speaker 4: with renewals, which is just very difficult. And I'm also

725
00:38:22,880 --> 00:38:25,520
Speaker 4: for reasons I wrote about in our annual energy paper,

726
00:38:26,040 --> 00:38:29,440
Speaker 4: I'm a skeptic of small modular reactors. Great idea on paper,

727
00:38:30,040 --> 00:38:31,760
Speaker 4: very complicated in terms of execution.

728
00:38:32,080 --> 00:38:35,920
Speaker 1: By the way, Tracy, it's interesting the heat as the

729
00:38:36,040 --> 00:38:38,520
Speaker 1: one of the big consumption drivers of energy. You know,

730
00:38:38,560 --> 00:38:40,919
Speaker 1: there's a lot of Northeastern chauvinism where they say people

731
00:38:40,920 --> 00:38:43,480
Speaker 1: shouldn't be living out in the out in Arizona and

732
00:38:43,480 --> 00:38:47,520
Speaker 1: all these like air conditioned places, but it really is heat, Like, really,

733
00:38:47,520 --> 00:38:49,800
Speaker 1: if we want to be more energy efficient, my understanding

734
00:38:49,840 --> 00:38:51,520
Speaker 1: is we should all leave the Northeast and go to

735
00:38:51,560 --> 00:38:55,080
Speaker 1: the desert and live in air conditioning instead of heating

736
00:38:55,080 --> 00:38:55,399
Speaker 1: our home.

737
00:38:55,600 --> 00:38:58,360
Speaker 2: No, for real, As someone with an old house in Connecticut,

738
00:38:58,400 --> 00:39:00,640
Speaker 2: I will push back on that narrative. You know, I

739
00:39:00,680 --> 00:39:02,600
Speaker 2: have a personal bias on this, but I was very

740
00:39:02,680 --> 00:39:05,520
Speaker 2: excited about heat pumps. But then we looked into it

741
00:39:05,680 --> 00:39:08,440
Speaker 2: for a home that was built in eighteen fifty, and

742
00:39:08,600 --> 00:39:10,759
Speaker 2: it turns out you would have to like reinsulate the

743
00:39:10,920 --> 00:39:13,480
Speaker 2: entire house for this thing to have any effect.

744
00:39:13,960 --> 00:39:16,920
Speaker 4: Oh yeah, No. The biggest problem with heat pumps is

745
00:39:16,960 --> 00:39:21,919
Speaker 4: the fact that while it's more efficient you're buying electricity

746
00:39:22,000 --> 00:39:24,440
Speaker 4: instead of natural gas, and per megadule of energy it

747
00:39:24,480 --> 00:39:26,279
Speaker 4: can be depending on the state you live in, three

748
00:39:26,360 --> 00:39:29,480
Speaker 4: or four times more expensive. So it's great for the climate,

749
00:39:29,560 --> 00:39:33,160
Speaker 4: it's just economically, you know, challenging. And that's why I

750
00:39:33,160 --> 00:39:35,239
Speaker 4: think some of those Rocky Mountain Institute and Green Tech

751
00:39:35,280 --> 00:39:38,040
Speaker 4: media forecasts are way too aggressive, because I think these

752
00:39:38,080 --> 00:39:40,200
Speaker 4: transitions are going to take place over a much longer

753
00:39:40,239 --> 00:39:40,799
Speaker 4: period of time.

754
00:39:57,080 --> 00:39:59,160
Speaker 2: Since we have you, since you've been in the business

755
00:39:59,200 --> 00:40:02,759
Speaker 2: for almost four decades now. If I was reading one

756
00:40:02,800 --> 00:40:04,960
Speaker 2: of your notes or one of your research pieces in

757
00:40:05,080 --> 00:40:07,560
Speaker 2: let's say nineteen eighty seven or nineteen eighty eight, I'm

758
00:40:07,600 --> 00:40:09,279
Speaker 2: not sure what exactly you were doing.

759
00:40:09,520 --> 00:40:11,479
Speaker 4: It was an analyst, oh okay, and I was sitting

760
00:40:11,560 --> 00:40:14,640
Speaker 4: someplace behind a pile of pizza boxes and nobody was

761
00:40:14,680 --> 00:40:15,839
Speaker 4: talking to me. I don't know.

762
00:40:16,400 --> 00:40:19,799
Speaker 2: So how different would your research then be versus your

763
00:40:19,840 --> 00:40:20,480
Speaker 2: research now.

764
00:40:21,040 --> 00:40:26,279
Speaker 4: I'm definitely somebody that has taken advantage of massive quantities

765
00:40:26,320 --> 00:40:28,720
Speaker 4: of information and research right, it was a lot harder.

766
00:40:28,760 --> 00:40:30,480
Speaker 4: I used to have to go to the mid Manhattan

767
00:40:30,520 --> 00:40:34,960
Speaker 4: Research Library to find things, and I used to go

768
00:40:35,120 --> 00:40:38,040
Speaker 4: and you pull out microfiche and then you would go

769
00:40:38,160 --> 00:40:40,080
Speaker 4: and you would try to you know, you put in

770
00:40:40,120 --> 00:40:42,120
Speaker 4: a quarter and they would let you print it. I mean,

771
00:40:42,440 --> 00:40:45,160
Speaker 4: it was hard to get historical data back then, and

772
00:40:45,239 --> 00:40:47,160
Speaker 4: I'm a data hound, and so the way the world

773
00:40:47,160 --> 00:40:49,759
Speaker 4: works today is much easier for me to draw parallels.

774
00:40:49,920 --> 00:40:54,120
Speaker 4: I've also learned a lot of lessons in investing for

775
00:40:54,160 --> 00:40:56,480
Speaker 4: all these years. I was the chief investment officer just

776
00:40:56,520 --> 00:40:58,560
Speaker 4: in the private bank for a long time, and so

777
00:40:58,840 --> 00:41:01,160
Speaker 4: I've learned a lot of less And we're in the

778
00:41:01,160 --> 00:41:03,759
Speaker 4: process of putting together a twenty year on the Market retrospective,

779
00:41:03,800 --> 00:41:06,480
Speaker 4: and it's amazing to go back and publish some of

780
00:41:06,520 --> 00:41:09,520
Speaker 4: those pieces and see what we learned at the time.

781
00:41:09,719 --> 00:41:11,719
Speaker 2: Oh well, okay, give us a preview. What were the

782
00:41:11,719 --> 00:41:12,440
Speaker 2: big lessons?

783
00:41:12,600 --> 00:41:15,040
Speaker 4: The number one thing more than anything else. And we

784
00:41:15,120 --> 00:41:17,160
Speaker 4: have thirty topics we're going to show, and some of

785
00:41:17,200 --> 00:41:19,680
Speaker 4: them are in there kind of for humor or sarcasm,

786
00:41:19,920 --> 00:41:22,840
Speaker 4: it would or But the number one investment lesson was

787
00:41:23,680 --> 00:41:28,120
Speaker 4: equity's bottom before everything else. And so does credit, and

788
00:41:28,200 --> 00:41:30,200
Speaker 4: so does high yield, and so does real estate. In

789
00:41:30,239 --> 00:41:36,920
Speaker 4: other words, asset prices bottom so far in advance of

790
00:41:36,960 --> 00:41:41,880
Speaker 4: the related fundamentals. For instance, during the financial crisis in

791
00:41:41,880 --> 00:41:46,200
Speaker 4: two thousand and nine, bank stocks bottomed when only ten

792
00:41:46,239 --> 00:41:49,920
Speaker 4: percent of the ultimate bank failures had taken place. So

793
00:41:50,440 --> 00:41:53,279
Speaker 4: if you were listening to the really bariess voices at

794
00:41:53,280 --> 00:41:56,000
Speaker 4: the time, they would say, too soon to invest. There's

795
00:41:56,080 --> 00:41:59,680
Speaker 4: a giant title wave of bank failures coming. But the

796
00:42:00,080 --> 00:42:03,400
Speaker 4: KBW or whatever index you want to look at, bottomed

797
00:42:03,440 --> 00:42:05,799
Speaker 4: in advance of all of that. And the same thing

798
00:42:05,880 --> 00:42:10,319
Speaker 4: happened during the SNL crisis. Stocks bottomed way before all

799
00:42:10,320 --> 00:42:13,479
Speaker 4: the devaults eventually took place. And if you then take

800
00:42:13,520 --> 00:42:16,560
Speaker 4: the seven post war recessions, in six out of the

801
00:42:16,640 --> 00:42:20,200
Speaker 4: seven of them, equity is bottomed before you even saw

802
00:42:20,200 --> 00:42:25,759
Speaker 4: an upturn in payrolls, industrial production, housing, credit card delinquencies.

803
00:42:26,080 --> 00:42:29,960
Speaker 4: And so as an investments person, I have to acclimate

804
00:42:30,000 --> 00:42:32,920
Speaker 4: our clients and our investment committee to be willing to

805
00:42:32,960 --> 00:42:35,720
Speaker 4: take risk. When you look out the window and everything

806
00:42:35,719 --> 00:42:36,560
Speaker 4: looks terrible.

807
00:42:36,760 --> 00:42:38,479
Speaker 1: Yeah, oh this is a great This is a great

808
00:42:38,560 --> 00:42:40,920
Speaker 1: lesson because you know, people say, right now, oh, the

809
00:42:40,960 --> 00:42:43,800
Speaker 1: shortages from the tariffs haven't even hit. Maybe we won't

810
00:42:43,800 --> 00:42:46,680
Speaker 1: have them because the tariffs have been dialed back so much.

811
00:42:47,120 --> 00:42:50,440
Speaker 1: But the idea that maybe the market already bottomed for

812
00:42:50,520 --> 00:42:54,040
Speaker 1: this cycle on everything related to tariffs. I sometimes I

813
00:42:54,160 --> 00:42:56,200
Speaker 1: like to say hope is the only strategy, because if

814
00:42:56,239 --> 00:42:58,880
Speaker 1: you wait until all of the data confirms that things

815
00:42:58,920 --> 00:43:01,200
Speaker 1: are back, it's clear really going to be priced in

816
00:43:01,280 --> 00:43:04,080
Speaker 1: at that point. But I actually wanted to ask something else.

817
00:43:04,840 --> 00:43:08,239
Speaker 1: I am an SMR skeptic. Two, But I form my

818
00:43:08,320 --> 00:43:11,520
Speaker 1: opinions based on seeing five tweets that I like. Okay, yeah,

819
00:43:11,520 --> 00:43:13,200
Speaker 1: those sound good to me. I'm an SMR skeptic.

820
00:43:13,600 --> 00:43:14,120
Speaker 4: You don't.

821
00:43:14,160 --> 00:43:16,440
Speaker 1: You have a much more rigorous process. Can you explain

822
00:43:16,520 --> 00:43:20,040
Speaker 1: to us why you are a skeptic of small modular reactors?

823
00:43:20,080 --> 00:43:25,120
Speaker 4: Okay, So for the easiest one, is nobody's built one?

824
00:43:25,560 --> 00:43:28,400
Speaker 4: That's a good answer in the United States. Yeah, one

825
00:43:28,440 --> 00:43:31,520
Speaker 4: hasn't been built. Point number two, there are three or

826
00:43:31,560 --> 00:43:34,440
Speaker 4: four of them, depending on how you want to define

827
00:43:34,680 --> 00:43:38,240
Speaker 4: what small is that have been built in China and Russia,

828
00:43:38,719 --> 00:43:42,040
Speaker 4: and the costs have been multiples of the original projections,

829
00:43:42,120 --> 00:43:44,760
Speaker 4: and the time frames have been much longer, So failure

830
00:43:44,880 --> 00:43:48,920
Speaker 4: even in places that have success building regular.

831
00:43:49,040 --> 00:43:51,640
Speaker 1: He's what I figured China would have built one hundred

832
00:43:51,640 --> 00:43:52,000
Speaker 1: by the right.

833
00:43:52,480 --> 00:43:55,960
Speaker 4: Third point, there's no industrial project on Earth that is

834
00:43:56,000 --> 00:43:59,480
Speaker 4: more capital intentsive the nuclear power and so when nuclear

835
00:43:59,520 --> 00:44:02,480
Speaker 4: power for started to become popular in the late sixties early seventies,

836
00:44:02,480 --> 00:44:05,640
Speaker 4: the goal was how big can we make them, because

837
00:44:05,719 --> 00:44:08,200
Speaker 4: we want to spread some of the sunk costs over

838
00:44:08,239 --> 00:44:12,000
Speaker 4: the most megawats we can build. The notion that you

839
00:44:12,120 --> 00:44:14,560
Speaker 4: want to shrink a capital project to make it more

840
00:44:14,719 --> 00:44:18,800
Speaker 4: evisient makes no sense to me unless you can completely

841
00:44:18,840 --> 00:44:22,480
Speaker 4: commoditize them. The mistake I think people are making is

842
00:44:22,520 --> 00:44:28,440
Speaker 4: that there has been an incredible learning curve in polysilicon panels,

843
00:44:28,480 --> 00:44:30,799
Speaker 4: in to lesser extent in wind turbines, both on shore

844
00:44:30,840 --> 00:44:33,840
Speaker 4: and offshore. And the best example is lithium ion batteries,

845
00:44:34,040 --> 00:44:36,440
Speaker 4: best example of a learning curve you've ever seen, but

846
00:44:36,520 --> 00:44:40,040
Speaker 4: that's because millions and tens of millions of units are

847
00:44:40,080 --> 00:44:44,320
Speaker 4: being produced. I don't believe that you can from going

848
00:44:44,360 --> 00:44:48,880
Speaker 4: from one reactor to four reactors or six reactors develop

849
00:44:49,000 --> 00:44:52,640
Speaker 4: the kind of learning curve that would result in exponential

850
00:44:52,640 --> 00:44:56,319
Speaker 4: cost decliinents and so you know, I'll wait until I

851
00:44:56,360 --> 00:44:57,880
Speaker 4: see it, and I don't think we will. But I

852
00:44:57,880 --> 00:45:00,680
Speaker 4: would love to eat my hat if Aklan or some

853
00:45:00,800 --> 00:45:03,799
Speaker 4: of the other public companies or private companies are working

854
00:45:03,880 --> 00:45:08,279
Speaker 4: on this can deliver a small module or reactor in

855
00:45:08,440 --> 00:45:11,440
Speaker 4: anywhere of the zip code of the original projections. And

856
00:45:11,600 --> 00:45:14,480
Speaker 4: just as just one last thing as a reminder, the

857
00:45:14,840 --> 00:45:20,080
Speaker 4: new scale, you know, originally said three million dollars a megawatt.

858
00:45:20,480 --> 00:45:23,920
Speaker 4: When it hit twenty the plug was pulled. So the

859
00:45:23,960 --> 00:45:27,160
Speaker 4: most recent example of such efforts in the United States

860
00:45:27,160 --> 00:45:29,400
Speaker 4: failed when the budget hit six x.

861
00:45:29,880 --> 00:45:33,080
Speaker 1: That was a fantastic answer, very intuitive answer, too real quickly,

862
00:45:33,120 --> 00:45:35,319
Speaker 1: what about a geothermal do you think that could be

863
00:45:35,320 --> 00:45:37,280
Speaker 1: a meaningful contributor to our portfolio?

864
00:45:37,360 --> 00:45:40,080
Speaker 4: And I do? You know it takes effort. Chris Wright,

865
00:45:40,120 --> 00:45:43,759
Speaker 4: who's a second Advantagy's a big fan. Yeah, And you know,

866
00:45:43,760 --> 00:45:47,200
Speaker 4: there's different kinds of geothermal. The most interesting kind is

867
00:45:47,239 --> 00:45:51,200
Speaker 4: where you're going really deep and you're accessing like supercritical

868
00:45:51,280 --> 00:45:55,000
Speaker 4: fluids whose heat is comparable to nuclear power. But you

869
00:45:55,080 --> 00:45:57,400
Speaker 4: know that instrumentation has to be able to withstand very

870
00:45:57,440 --> 00:46:00,160
Speaker 4: high heat. You have to use plasma drilling bets. I

871
00:46:00,160 --> 00:46:03,880
Speaker 4: think there's opportunity for expansion of geothermal, but their long

872
00:46:03,960 --> 00:46:07,000
Speaker 4: term industrial projects, it'll take a long time to do,

873
00:46:07,440 --> 00:46:09,960
Speaker 4: not a quick fix. Yeah, you know. And the other

874
00:46:10,040 --> 00:46:13,600
Speaker 4: thing too is and this is amazing. I've been very

875
00:46:13,600 --> 00:46:16,080
Speaker 4: focused on the rising costs of wind and solar power,

876
00:46:16,800 --> 00:46:18,680
Speaker 4: and I should have been paying more attention to it.

877
00:46:18,719 --> 00:46:21,080
Speaker 4: I turned my head the other day. The cost of

878
00:46:21,120 --> 00:46:24,120
Speaker 4: a national gas combined cycle turbine has gone from twelve

879
00:46:24,200 --> 00:46:28,200
Speaker 4: hundred per kilowatt to twenty five hundred, like more than

880
00:46:28,239 --> 00:46:31,920
Speaker 4: double just in the last two years. Supply chains and

881
00:46:32,040 --> 00:46:34,680
Speaker 4: mostly a reflection of the fact that g E Vvernova

882
00:46:34,680 --> 00:46:39,080
Speaker 4: and Siemens and the other companies involved don't believe the

883
00:46:39,200 --> 00:46:44,120
Speaker 4: long term demand story enough to expand their production facilities.

884
00:46:44,280 --> 00:46:47,680
Speaker 4: So that's why classic energy phenomenon, right. So that's why

885
00:46:47,719 --> 00:46:50,080
Speaker 4: you go to Gefernova today. We're happy to deliver you

886
00:46:50,239 --> 00:46:52,359
Speaker 4: a plant. It's going to cost two times more than

887
00:46:52,360 --> 00:46:53,920
Speaker 4: it did a couple of years ago, and we'll give

888
00:46:53,960 --> 00:46:56,920
Speaker 4: it to you in five years. That's how AI bottleneck.

889
00:46:57,000 --> 00:47:01,120
Speaker 1: This is how falling demand can result in higher prices falling.

890
00:47:01,160 --> 00:47:04,560
Speaker 4: Right, because the two to three entities Mitsubishi as well,

891
00:47:04,760 --> 00:47:08,680
Speaker 4: that are involved in this sector don't feel comfortable enough

892
00:47:08,719 --> 00:47:11,120
Speaker 4: to go out and do a mega capital raise to

893
00:47:11,280 --> 00:47:13,640
Speaker 4: double and triple the size of their production facilities for

894
00:47:13,680 --> 00:47:15,839
Speaker 4: these kind of turbines, and that I think is going

895
00:47:15,880 --> 00:47:19,640
Speaker 4: to end up being a binding constraint for AI at

896
00:47:19,680 --> 00:47:20,200
Speaker 4: some point.

897
00:47:20,400 --> 00:47:22,840
Speaker 2: I just don't know western So one of the reasons

898
00:47:22,880 --> 00:47:24,759
Speaker 2: I'm a big fan of your research is because it

899
00:47:24,840 --> 00:47:27,880
Speaker 2: is so wide ranging, and again, you have a very

900
00:47:27,920 --> 00:47:30,640
Speaker 2: diverse diet when it comes to what you read and

901
00:47:30,680 --> 00:47:34,799
Speaker 2: the information and data that you consume. Last question for me,

902
00:47:34,920 --> 00:47:39,000
Speaker 2: but give us one under the radar thing that you

903
00:47:39,040 --> 00:47:44,759
Speaker 2: are watching at the moment. Hmm, what should basically, what

904
00:47:44,800 --> 00:47:45,920
Speaker 2: should we do an episode on?

905
00:47:46,120 --> 00:47:47,960
Speaker 1: Hah, yeah, what should we do an episode on?

906
00:47:50,680 --> 00:47:53,640
Speaker 4: Let's see, I mean, not in any kind of order importance.

907
00:47:53,680 --> 00:47:56,960
Speaker 4: I'm just going to get a stream of consciousness sometimes.

908
00:47:57,000 --> 00:47:59,200
Speaker 4: I heard you in the introduction talk about narratives that

909
00:47:59,280 --> 00:48:01,799
Speaker 4: come and go, and I agree with that totally. And

910
00:48:01,840 --> 00:48:04,040
Speaker 4: one of the narratives that came and went but might

911
00:48:04,080 --> 00:48:09,640
Speaker 4: come back is there are still five to seven hundred

912
00:48:09,640 --> 00:48:13,359
Speaker 4: billion dollars of unrealized interest rate related losses on bank

913
00:48:13,400 --> 00:48:17,359
Speaker 4: balance sheets that has not gone away. Now it only

914
00:48:17,440 --> 00:48:20,680
Speaker 4: goes away if you go back to you know, three

915
00:48:20,719 --> 00:48:23,719
Speaker 4: percent or some three percent on the tenure. But we

916
00:48:23,960 --> 00:48:28,160
Speaker 4: still have a banking system where many of the participants

917
00:48:28,880 --> 00:48:33,640
Speaker 4: went hog wild extending duration on treasuries and agencies at

918
00:48:33,680 --> 00:48:37,120
Speaker 4: a time of you know, historically low interest rates because

919
00:48:37,120 --> 00:48:40,120
Speaker 4: you had this deposit surge from all the FED Bazuka money,

920
00:48:40,680 --> 00:48:43,120
Speaker 4: and that hasn't gone away now the Fed, the FED

921
00:48:43,160 --> 00:48:47,120
Speaker 4: has proven itself willing to provide whatever emergency facilities. I

922
00:48:47,200 --> 00:48:52,800
Speaker 4: actually was, you know, personally against bailing out the VC

923
00:48:53,000 --> 00:48:56,160
Speaker 4: depositors in Silicon Valley Bank. You know, the average deposit

924
00:48:56,320 --> 00:48:58,000
Speaker 4: balance in that bank was two and a half million dollars.

925
00:48:58,000 --> 00:48:59,759
Speaker 4: I mean, it wasn't really a bank, it was something else,

926
00:49:00,160 --> 00:49:02,919
Speaker 4: but they decided for systemic reasons to do it. They'd

927
00:49:02,920 --> 00:49:05,239
Speaker 4: probably do it again. But that issue hasn't gone away. Yeah,

928
00:49:05,239 --> 00:49:07,839
Speaker 4: and the higher ten year rates go, the more those

929
00:49:07,920 --> 00:49:10,480
Speaker 4: losses get pronounced. So that's an issue that's kind of

930
00:49:10,520 --> 00:49:10,839
Speaker 4: still out.

931
00:49:10,920 --> 00:49:12,359
Speaker 1: I'm glad you brought that up because I remember we

932
00:49:12,400 --> 00:49:14,080
Speaker 1: traced and I did a bunch of episodes in like

933
00:49:14,120 --> 00:49:16,120
Speaker 1: twenty twenty one, twenty two. It's like, why isn't the

934
00:49:16,200 --> 00:49:18,360
Speaker 1: rate rise having more of an effect that everyone's like, oh,

935
00:49:18,400 --> 00:49:20,960
Speaker 1: you know, they got turned out. Everyone locked in low rates,

936
00:49:21,000 --> 00:49:23,319
Speaker 1: et cetera. And then I was like, hey, but now

937
00:49:23,400 --> 00:49:25,759
Speaker 1: rates are highering at some point, like they only locked

938
00:49:25,800 --> 00:49:26,680
Speaker 1: them in for so long.

939
00:49:26,760 --> 00:49:30,160
Speaker 4: I'm only I'm almost tempted to believe that the FED

940
00:49:31,400 --> 00:49:37,319
Speaker 4: essentially created this problem through financial repression, tempted some bad

941
00:49:37,360 --> 00:49:40,200
Speaker 4: asset liability managers at these regional banks to load up

942
00:49:40,200 --> 00:49:43,400
Speaker 4: at low rates and feel some sense of actual responsibility

943
00:49:43,400 --> 00:49:46,320
Speaker 4: for this mess because it was Fed monetary and congressional

944
00:49:46,320 --> 00:49:50,200
Speaker 4: fiscal policy that led to this mess. So yeah, so

945
00:49:50,239 --> 00:49:53,399
Speaker 4: I think they feel somewhat, you know, beholden to try

946
00:49:53,400 --> 00:49:55,960
Speaker 4: to manage around some of the risks. Another thing I

947
00:49:56,000 --> 00:49:59,720
Speaker 4: would say is, I don't think the tariff story is done. Okay,

948
00:50:00,120 --> 00:50:01,920
Speaker 4: you know, the average tariff rate at the end of

949
00:50:01,960 --> 00:50:03,720
Speaker 4: last year was something like two and a half percent.

950
00:50:03,719 --> 00:50:06,040
Speaker 4: I'm using round numbers. Yeah, if you did all the

951
00:50:06,040 --> 00:50:08,680
Speaker 4: math and you don't assume any import substitution. At the

952
00:50:08,719 --> 00:50:13,600
Speaker 4: peak of the kind of you know Lutnick cardboard cut out,

953
00:50:13,760 --> 00:50:17,439
Speaker 4: you know, presentation, we were looking at a twenty five

954
00:50:17,480 --> 00:50:21,319
Speaker 4: percent all in rate, again assuming no import substitution and

955
00:50:21,320 --> 00:50:24,800
Speaker 4: things like that. Now we're back down at like thirteen percent.

956
00:50:24,840 --> 00:50:27,359
Speaker 4: So the market's like, wow, we dodged a bullet. It

957
00:50:27,440 --> 00:50:30,919
Speaker 4: went from twenty something to fourteen, But fourteen is still

958
00:50:31,000 --> 00:50:32,279
Speaker 4: very high compared to two.

959
00:50:33,000 --> 00:50:35,120
Speaker 2: Isn't it like the highest since the nineteen thirties or

960
00:50:35,160 --> 00:50:35,960
Speaker 2: forties or something.

961
00:50:36,040 --> 00:50:40,280
Speaker 4: Yeah, now he may backtrack again, you know, we'll see.

962
00:50:40,719 --> 00:50:44,520
Speaker 4: But historically, whenever there's been a disconnect between hard data

963
00:50:44,520 --> 00:50:47,960
Speaker 4: and soft data, it takes about ninety days for that

964
00:50:48,000 --> 00:50:51,319
Speaker 4: to resolve itself. So some of the shocks that are

965
00:50:51,320 --> 00:50:54,359
Speaker 4: seeping through the system are going to take until July

966
00:50:54,480 --> 00:50:56,680
Speaker 4: or August to play out. One of the things I

967
00:50:56,760 --> 00:50:59,920
Speaker 4: wrote earlier this year that resonated with a lot of

968
00:51:00,080 --> 00:51:02,520
Speaker 4: people is the markets are the one that you can't

969
00:51:02,520 --> 00:51:06,840
Speaker 4: deport them. You can't intimidate them, you can't arrest them.

970
00:51:07,239 --> 00:51:09,840
Speaker 4: You know, at the end of the day, the markets

971
00:51:09,840 --> 00:51:12,040
Speaker 4: are going to be a binding constraint on what this

972
00:51:12,120 --> 00:51:16,120
Speaker 4: administration can do and can't do. And that's actually a

973
00:51:16,200 --> 00:51:18,839
Speaker 4: plus for investors because it means at some point they're

974
00:51:18,960 --> 00:51:22,120
Speaker 4: going to have to adhere to some generally acceptable level

975
00:51:22,120 --> 00:51:22,839
Speaker 4: of policy making.

976
00:51:23,120 --> 00:51:25,520
Speaker 1: I just have one last quick question going back to AI.

977
00:51:25,920 --> 00:51:28,320
Speaker 1: You hear a lot about inference versus training and people

978
00:51:28,320 --> 00:51:30,840
Speaker 1: wondering could there be some alternative to in Nvidia for

979
00:51:30,960 --> 00:51:33,520
Speaker 1: some of these services, because right now on the chip side,

980
00:51:33,719 --> 00:51:36,400
Speaker 1: all the profits seem to be accruing to in VideA.

981
00:51:37,040 --> 00:51:40,080
Speaker 1: Is that mode you know people talk about the Kuda ecosystem.

982
00:51:40,480 --> 00:51:43,120
Speaker 1: Is that an impenetrable moat? Or are there opportunities for

983
00:51:43,160 --> 00:51:46,880
Speaker 1: other chip companies to gain some real profits in the

984
00:51:46,920 --> 00:51:47,520
Speaker 1: AI world?

985
00:51:47,840 --> 00:51:50,440
Speaker 4: There are, you know, in the piece that we did

986
00:51:50,560 --> 00:51:53,040
Speaker 4: last year, we listed all the companies that are trying

987
00:51:53,080 --> 00:51:56,760
Speaker 4: to pick away in video mode. But in the asset

988
00:51:56,800 --> 00:52:02,640
Speaker 4: management business or when companies hire underwriters, if an underwriting

989
00:52:02,719 --> 00:52:06,360
Speaker 4: goes bad and you picked JP Morgan Goldman and Morgan Stanley,

990
00:52:06,760 --> 00:52:10,760
Speaker 4: nobody said to question you. If you pick a ranked

991
00:52:10,760 --> 00:52:14,040
Speaker 4: eight or eleven, you know underwriter and something goes wrong,

992
00:52:14,080 --> 00:52:17,040
Speaker 4: people will question you. So to some extent, the network

993
00:52:17,080 --> 00:52:18,960
Speaker 4: effect and the modes that you guys are referring to,

994
00:52:19,920 --> 00:52:22,880
Speaker 4: nobody's going to get blamed for spending extra money to

995
00:52:22,960 --> 00:52:25,400
Speaker 4: buy in video products and wait for a long time

996
00:52:25,680 --> 00:52:28,799
Speaker 4: until these other products have proven themselves. And so I

997
00:52:28,800 --> 00:52:30,719
Speaker 4: think it's going to take a long time for that

998
00:52:30,760 --> 00:52:35,560
Speaker 4: to play out. But yeah, eventually, if this AI story works,

999
00:52:36,000 --> 00:52:38,800
Speaker 4: we should be looking at eighty percent inference and twenty

1000
00:52:38,800 --> 00:52:42,400
Speaker 4: percent training in two to three years. Right at that point,

1001
00:52:42,680 --> 00:52:46,600
Speaker 4: corporate adoption, which drives the training models are supposed to

1002
00:52:46,680 --> 00:52:49,680
Speaker 4: be the overwhelming amount of what's going on. But again,

1003
00:52:50,560 --> 00:52:54,040
Speaker 4: Amazon claims to have been working on a foundational model.

1004
00:52:54,320 --> 00:52:59,160
Speaker 4: Haven't seen it yet. Microsoft has actually kind of hinted that, oh,

1005
00:52:59,200 --> 00:53:02,080
Speaker 4: we've created our own foundational model. There was this great

1006
00:53:02,120 --> 00:53:05,239
Speaker 4: story recently, if you like AI. The head of the

1007
00:53:05,280 --> 00:53:08,480
Speaker 4: Microsoft AI group was basically told to pound sand by

1008
00:53:08,480 --> 00:53:10,799
Speaker 4: the Open Eye people when he was asking them to

1009
00:53:10,920 --> 00:53:13,480
Speaker 4: fully disclose how some of the new reasoning models worked.

1010
00:53:13,840 --> 00:53:18,040
Speaker 4: So Microsoft is working on their own foundational model. Hasn't

1011
00:53:18,080 --> 00:53:22,160
Speaker 4: been released yet either, even internally within Microsoft. So again,

1012
00:53:22,239 --> 00:53:25,760
Speaker 4: so some of these motes, I wouldn't underestimate the amount

1013
00:53:25,800 --> 00:53:29,279
Speaker 4: of time that video is spending defending that moat and

1014
00:53:29,360 --> 00:53:32,839
Speaker 4: improving its products and its energy efficiency so that they

1015
00:53:32,840 --> 00:53:34,560
Speaker 4: don't give up market share prematurely.

1016
00:53:34,800 --> 00:53:37,160
Speaker 2: All right, Well, by the time this episode comes out

1017
00:53:37,200 --> 00:53:39,640
Speaker 2: in video, earnings will have been published, so maybe we'll

1018
00:53:39,680 --> 00:53:42,759
Speaker 2: get a little bit more insight. But Michael Symbolists, thank

1019
00:53:42,800 --> 00:53:45,239
Speaker 2: you so much for coming on. Yeah, that was a

1020
00:53:45,280 --> 00:53:45,880
Speaker 2: real pleasure.

1021
00:53:45,880 --> 00:53:47,960
Speaker 1: There's a crime that we hadn't had you before, but

1022
00:53:48,040 --> 00:53:49,680
Speaker 1: I guess you couldn't.

1023
00:53:49,400 --> 00:53:51,319
Speaker 4: Use I think I'm being put out of the back

1024
00:53:51,320 --> 00:53:52,840
Speaker 4: in the cage after this time.

1025
00:54:05,480 --> 00:54:08,480
Speaker 2: Joe. That was a real treat to get Michael finally

1026
00:54:08,760 --> 00:54:10,080
Speaker 2: finally on the podcast.

1027
00:54:10,160 --> 00:54:13,520
Speaker 1: It was so good, and it's like, I've always liked

1028
00:54:13,560 --> 00:54:16,400
Speaker 1: his notes, Yeah, I mean a I think everyone who

1029
00:54:16,480 --> 00:54:19,160
Speaker 1: hasn't read them will now understand why he has sort

1030
00:54:19,160 --> 00:54:21,960
Speaker 1: of this cult following on Wall Street where everyone devours

1031
00:54:22,440 --> 00:54:25,920
Speaker 1: his notes. But then even beyond that, hearing him, like

1032
00:54:26,000 --> 00:54:28,360
Speaker 1: in my mind, it's like, Okay, these notes are great,

1033
00:54:28,600 --> 00:54:31,799
Speaker 1: but there's a difference between comfort at like being able

1034
00:54:31,840 --> 00:54:33,840
Speaker 1: to put together this big research product or there's a

1035
00:54:33,840 --> 00:54:35,880
Speaker 1: big note product, but also just being able to talk

1036
00:54:35,920 --> 00:54:39,719
Speaker 1: about all these things extemporaneously. Absolutely, it's really impressive.

1037
00:54:40,200 --> 00:54:43,440
Speaker 2: There's so much interesting stuff to pick out there. But

1038
00:54:43,800 --> 00:54:46,279
Speaker 2: one of the things that struck me was, I guess

1039
00:54:46,320 --> 00:54:50,080
Speaker 2: the importance of geographical location for data centers. Yeah, because

1040
00:54:50,080 --> 00:54:52,600
Speaker 2: I hadn't thought about that. But he's absolutely right. You know,

1041
00:54:52,800 --> 00:54:55,239
Speaker 2: you could locate a bunch of data centers in the

1042
00:54:55,280 --> 00:54:58,040
Speaker 2: middle of nowhere where maybe there's cheaper energy costs, but

1043
00:54:58,239 --> 00:55:00,960
Speaker 2: people so far have not to do that. Yeah.

1044
00:55:01,000 --> 00:55:03,680
Speaker 1: I hadn't really thought about that either, because it's like,

1045
00:55:03,760 --> 00:55:06,560
Speaker 1: you know, we all know that Northern Virginia is this

1046
00:55:06,680 --> 00:55:11,399
Speaker 1: huge data center cluster, But why the data centers really

1047
00:55:11,440 --> 00:55:14,000
Speaker 1: need to be that close to the government and the

1048
00:55:14,040 --> 00:55:17,120
Speaker 1: defence industry and so forth. It's like, come on, It's like,

1049
00:55:17,440 --> 00:55:20,120
Speaker 1: isn't really that big of a deal, right to like

1050
00:55:20,200 --> 00:55:23,839
Speaker 1: you're an extra millisecond of or something. But I mean,

1051
00:55:24,239 --> 00:55:27,640
Speaker 1: you know, they're obviously there for a reason. I also

1052
00:55:27,719 --> 00:55:30,000
Speaker 1: just thought, like this idea of the moats you know

1053
00:55:30,040 --> 00:55:33,040
Speaker 1: obviously that the big tech companies have built up, and

1054
00:55:33,120 --> 00:55:36,000
Speaker 1: the amount of effort they have to sustain those moats

1055
00:55:36,440 --> 00:55:39,400
Speaker 1: and distribute through those modes and so forth, and like

1056
00:55:39,640 --> 00:55:41,880
Speaker 1: how powerful that is and how much that is the

1057
00:55:41,920 --> 00:55:44,640
Speaker 1: story of the fact that the biggest keep getting bigger

1058
00:55:44,640 --> 00:55:47,000
Speaker 1: at literally at the expense of anything else. It's such

1059
00:55:47,040 --> 00:55:49,560
Speaker 1: an important perspective have on all kinds of dimensions, whether

1060
00:55:49,560 --> 00:55:53,439
Speaker 1: you're thinking US versus international, the price of US equities

1061
00:55:53,600 --> 00:55:56,040
Speaker 1: and so forth. You just have to remember, like how

1062
00:55:56,120 --> 00:55:58,640
Speaker 1: extraordinarily unusual these companies.

1063
00:55:58,280 --> 00:56:01,640
Speaker 2: Are, absolutely and the fact that they continued to just

1064
00:56:01,760 --> 00:56:06,200
Speaker 2: throw off cash even while they're spending billions and billions

1065
00:56:06,200 --> 00:56:09,359
Speaker 2: of dollars on AI tech which were not entirely sure,

1066
00:56:09,440 --> 00:56:11,400
Speaker 2: like the degree to which it's actually going to be

1067
00:56:11,440 --> 00:56:13,880
Speaker 2: able to be monetized, but they still managed to do

1068
00:56:13,920 --> 00:56:15,160
Speaker 2: it while generating cash.

1069
00:56:15,200 --> 00:56:17,799
Speaker 1: The clarity of the way he spoke about energy is

1070
00:56:17,800 --> 00:56:20,120
Speaker 1: really impressive, and I really liked what he said about

1071
00:56:20,200 --> 00:56:21,720
Speaker 1: small modular reactors.

1072
00:56:21,760 --> 00:56:23,760
Speaker 2: You know, no one's done it yet in the US.

1073
00:56:23,960 --> 00:56:26,239
Speaker 1: Yeah, And when we talked to Jiggershaw, a body, he

1074
00:56:26,320 --> 00:56:28,560
Speaker 1: used an analogy which I thought was fantastic. Know he's

1075
00:56:28,600 --> 00:56:31,040
Speaker 1: more optimistic about them. He used an analogy that I

1076
00:56:31,040 --> 00:56:34,239
Speaker 1: thought was great, which is that the nuclear industry has

1077
00:56:34,280 --> 00:56:37,360
Speaker 1: to be in the business of building airplanes instead of airports,

1078
00:56:37,360 --> 00:56:41,200
Speaker 1: because airports are all one offs. But airplanes are obviously

1079
00:56:41,400 --> 00:56:44,000
Speaker 1: you know, they're built down a very complex assembly line,

1080
00:56:44,160 --> 00:56:47,000
Speaker 1: but you know, after you've built several the cost gets cheaper.

1081
00:56:47,480 --> 00:56:52,880
Speaker 1: And this idea that there's so much capital intensivity in

1082
00:56:53,120 --> 00:56:56,800
Speaker 1: building nuclear and the idea that there is no insight

1083
00:56:57,040 --> 00:56:59,919
Speaker 1: prospect for that learning curve where you do it over

1084
00:57:00,000 --> 00:57:02,560
Speaker 1: and over and over again for the cost to follow dramatically,

1085
00:57:02,840 --> 00:57:05,839
Speaker 1: that they haven't even done that in China. Maybe you're

1086
00:57:05,880 --> 00:57:08,680
Speaker 1: not going to ever really get the sort of learning

1087
00:57:08,719 --> 00:57:12,120
Speaker 1: curves that you get in solar or lithium ion batteries.

1088
00:57:12,480 --> 00:57:17,120
Speaker 1: Super very clear, intuitive explanation of why the opportunity may

1089
00:57:17,120 --> 00:57:17,640
Speaker 1: not be there.

1090
00:57:17,720 --> 00:57:19,040
Speaker 2: Yeah, we have to have them back on.

1091
00:57:19,280 --> 00:57:21,240
Speaker 1: Yeah, oh so much more we could talk about.

1092
00:57:21,280 --> 00:57:22,360
Speaker 2: All right, shall we leave it there?

1093
00:57:22,440 --> 00:57:23,120
Speaker 1: Let's leave it there.

1094
00:57:23,240 --> 00:57:25,840
Speaker 2: This has been another episode of the Outhoughts podcast. I'm

1095
00:57:25,880 --> 00:57:28,760
Speaker 2: Tracy Alloway. You can follow me at Tracy Alloway and.

1096
00:57:28,720 --> 00:57:31,320
Speaker 1: I'm joll Wisenthal. You can follow me at the Stalwart.

1097
00:57:31,600 --> 00:57:35,080
Speaker 1: Check out all of Michael Symbolist's notes. They're all online.

1098
00:57:35,360 --> 00:57:37,560
Speaker 1: Search for them. All of them are a must read.

1099
00:57:37,640 --> 00:57:39,640
Speaker 1: I'm gonna go back and read a bunch of them.

1100
00:57:39,760 --> 00:57:43,040
Speaker 1: Follow our producers Kerman Rodriguez at Kerman Arman, dash Ob

1101
00:57:43,040 --> 00:57:46,680
Speaker 1: Bennett at dashbod and cal Brooks at Kelbrooks. For more

1102
00:57:46,720 --> 00:57:49,760
Speaker 1: Oddlots content, go to Bloomberg dot com slash Odlocks. We

1103
00:57:49,800 --> 00:57:52,400
Speaker 1: have a daily newsletter and all of our episodes, and

1104
00:57:52,440 --> 00:57:54,480
Speaker 1: you can chat about all of these topics twenty four

1105
00:57:54,600 --> 00:57:58,720
Speaker 1: seven in our discord Discord dot gg slash odd.

1106
00:57:58,560 --> 00:58:01,120
Speaker 2: Locks and if you and joy all thoughts. If you

1107
00:58:01,280 --> 00:58:04,880
Speaker 2: like it when we talk to market veterans like Michael Symbolus,

1108
00:58:05,000 --> 00:58:07,520
Speaker 2: then please leave us a positive review on your favorite

1109
00:58:07,520 --> 00:58:11,320
Speaker 2: podcast platform. And remember, if you are a Bloomberg subscriber,

1110
00:58:11,360 --> 00:58:14,439
Speaker 2: you can listen to all of our episodes absolutely ad free.

1111
00:58:14,800 --> 00:58:16,920
Speaker 2: All you need to do is find the Bloomberg channel

1112
00:58:16,960 --> 00:58:20,720
Speaker 2: on Apple Podcasts and follow the instructions there. Thanks for listening.