1
00:00:02,520 --> 00:00:11,840
Speaker 1: Bloomberg Audio Studios, Podcasts, radio news. This is Masters in

2
00:00:11,920 --> 00:00:15,760
Speaker 1: Business with Barry Ritholts on Bloomberg Radio.

3
00:00:16,600 --> 00:00:20,720
Speaker 2: On the latest Masters in Business podcast. What a fascinating

4
00:00:20,760 --> 00:00:26,520
Speaker 2: conversation I sit down with Ben Hunt. He writes Epsilon Theory,

5
00:00:26,680 --> 00:00:31,360
Speaker 2: but he is also the president and co founder of Percion.

6
00:00:32,040 --> 00:00:37,919
Speaker 2: What a fascinating analytics story they've put together. They essentially

7
00:00:38,000 --> 00:00:42,200
Speaker 2: take feeds of everything that's published around the world, whether

8
00:00:42,240 --> 00:00:45,640
Speaker 2: it's in English or Chinese or Russian. They create these

9
00:00:45,760 --> 00:00:53,760
Speaker 2: large language models and use artificial intelligence to identify rising narratives.

10
00:00:53,800 --> 00:00:56,600
Speaker 2: In other words, they're looking for the things that will

11
00:00:56,640 --> 00:01:01,760
Speaker 2: become storylines, but haven't quite that yet. I found this

12
00:01:01,800 --> 00:01:04,880
Speaker 2: conversation to be absolutely fascinating, and I think you will also.

13
00:01:05,480 --> 00:01:10,280
Speaker 2: With no further ado, my conversation with Persians Ben Hunt.

14
00:01:10,959 --> 00:01:13,280
Speaker 2: Ben Hunt, welcome to Bloomberg.

15
00:01:13,920 --> 00:01:16,320
Speaker 3: Thanks for having me. Well, this is a love intro.

16
00:01:16,440 --> 00:01:17,520
Speaker 3: I gotta have you at all my.

17
00:01:17,680 --> 00:01:20,560
Speaker 2: That's right, I'm available for hire. I can introduce you

18
00:01:20,600 --> 00:01:24,120
Speaker 2: at weddings by mitzvahs. Wherever you're giving a toast, I'll

19
00:01:24,120 --> 00:01:27,120
Speaker 2: be happy to tee you up. This is long overdue.

20
00:01:27,120 --> 00:01:31,120
Speaker 2: I've followed your work for so long. I'm fascinated by

21
00:01:31,200 --> 00:01:35,080
Speaker 2: both what you put out in your blog Epsilon Theory,

22
00:01:35,240 --> 00:01:38,240
Speaker 2: thank you, and which is now a blog and a newsletter,

23
00:01:38,360 --> 00:01:41,840
Speaker 2: and the work you do at Percion. We're gonna get

24
00:01:42,000 --> 00:01:45,800
Speaker 2: to that stuff, but before we do, I gotta ask

25
00:01:46,160 --> 00:01:50,720
Speaker 2: PhD from Harvard you were a tenured political science professor.

26
00:01:51,160 --> 00:01:53,480
Speaker 2: Was academia the original career plan?

27
00:01:54,680 --> 00:01:58,000
Speaker 3: You know? It's interesting, So I Academia was always a

28
00:01:58,280 --> 00:02:01,040
Speaker 3: i'll call it away station for me. It ended up

29
00:02:01,040 --> 00:02:05,240
Speaker 3: being a ten ten year way station plus grad school.

30
00:02:05,960 --> 00:02:08,359
Speaker 2: But I that's a little more than a waste.

31
00:02:08,560 --> 00:02:10,280
Speaker 3: It's a little more than a way station. But I

32
00:02:10,600 --> 00:02:13,000
Speaker 3: bet this will be familiar for a lot of your listeners.

33
00:02:13,280 --> 00:02:16,600
Speaker 3: I always had an entrepreneurial bug, you know. I started

34
00:02:16,639 --> 00:02:19,600
Speaker 3: my first company when I was in grad school, started

35
00:02:19,639 --> 00:02:22,320
Speaker 3: another one when I was when I was a professor.

36
00:02:23,120 --> 00:02:26,960
Speaker 3: And as I know, you know a lot of your

37
00:02:27,639 --> 00:02:31,240
Speaker 3: listeners viewers know it is. It is a bug. It's

38
00:02:31,240 --> 00:02:33,720
Speaker 3: not a feature. Yes, for sure, you can't help yourself.

39
00:02:34,080 --> 00:02:35,560
Speaker 2: So let me ask you a question about that.

40
00:02:35,600 --> 00:02:38,239
Speaker 3: Academy is not the place to be an a for sure.

41
00:02:38,800 --> 00:02:41,919
Speaker 2: So I know why I've started a series of companies

42
00:02:42,320 --> 00:02:46,000
Speaker 2: I can't work for other people. Why did you have

43
00:02:46,040 --> 00:02:49,799
Speaker 2: that bug? What motivated you to say I gotta get

44
00:02:49,840 --> 00:02:50,799
Speaker 2: this out into the world.

45
00:02:52,440 --> 00:02:55,920
Speaker 3: I love playing games and solving problems. I have a

46
00:02:55,960 --> 00:03:00,079
Speaker 3: similar issue about working for other people, which fortunately the

47
00:03:00,120 --> 00:03:04,679
Speaker 3: academia solves that to a large degree. I mean, you

48
00:03:04,680 --> 00:03:08,640
Speaker 3: you're working on your own stuff. You follow your own

49
00:03:09,000 --> 00:03:15,480
Speaker 3: intellectual bliss in a way that I've really never rediscovered that.

50
00:03:15,639 --> 00:03:19,399
Speaker 3: The problem with academy, of course, is you know, it's

51
00:03:19,639 --> 00:03:21,840
Speaker 3: very very low stakes. It don't pay.

52
00:03:22,000 --> 00:03:26,760
Speaker 2: That's why, that's why the academic fights are so vicious.

53
00:03:26,360 --> 00:03:30,640
Speaker 3: Because there's nothing at stake, and that that is actually true.

54
00:03:31,040 --> 00:03:36,720
Speaker 3: That's actually true, and so you learn survival techniques and

55
00:03:36,800 --> 00:03:40,040
Speaker 3: that kind of jungle where nothing is really at stake,

56
00:03:40,160 --> 00:03:44,240
Speaker 3: at least monetarily. Because the goal of any sort of

57
00:03:44,320 --> 00:03:51,160
Speaker 3: academia conference or presentation like is to appear smart, right,

58
00:03:51,200 --> 00:03:53,760
Speaker 3: it's not to actually be smart. You're not actually listening

59
00:03:53,800 --> 00:03:58,880
Speaker 3: to a presentation to listen to it. What you learned

60
00:03:58,920 --> 00:04:02,200
Speaker 3: to do is you're listening to the presentation the whole

61
00:04:02,200 --> 00:04:05,760
Speaker 3: time you're trying to calculate what's the most devastating question

62
00:04:05,880 --> 00:04:06,480
Speaker 3: I can ask?

63
00:04:07,560 --> 00:04:09,600
Speaker 2: So you're going to rock. This guy back on his

64
00:04:09,680 --> 00:04:11,960
Speaker 2: heels with a devastating.

65
00:04:11,440 --> 00:04:16,440
Speaker 3: Devastating question, and boy, that gets old after a while, Barry,

66
00:04:16,440 --> 00:04:19,360
Speaker 3: I gotta tell you, yeah, it really does. I loved

67
00:04:19,520 --> 00:04:23,880
Speaker 3: the teaching. I loved the research because, like I say,

68
00:04:23,960 --> 00:04:28,200
Speaker 3: nobody tells you what to work on, but the church

69
00:04:28,440 --> 00:04:34,640
Speaker 3: of academia, the actual institution of academia. Hey, it's for

70
00:04:34,680 --> 00:04:37,039
Speaker 3: the birds, even back when I was doing right, and

71
00:04:38,760 --> 00:04:41,400
Speaker 3: I think it's gotten significantly worse. I can imagine. I

72
00:04:41,440 --> 00:04:42,720
Speaker 3: can imagine what's happening with that.

73
00:04:42,800 --> 00:04:45,120
Speaker 2: So but the question that this leads me to is,

74
00:04:45,600 --> 00:04:48,320
Speaker 2: you've had all these jobs when the world to finance,

75
00:04:48,720 --> 00:04:53,560
Speaker 2: how did you background in academia shape how you view investing,

76
00:04:53,720 --> 00:04:55,920
Speaker 2: risk management, allocation.

77
00:04:58,000 --> 00:05:04,520
Speaker 3: Barry, Starting from academ and then getting into our business

78
00:05:04,560 --> 00:05:07,279
Speaker 3: of investing, I think it was the best thing that

79
00:05:07,320 --> 00:05:09,880
Speaker 3: could have happened for me for when I got into it,

80
00:05:10,040 --> 00:05:13,240
Speaker 3: which is kind of later in life, right, because you know,

81
00:05:13,360 --> 00:05:16,520
Speaker 3: after I left academic, if I start a software company

82
00:05:17,160 --> 00:05:23,440
Speaker 3: and after we sold that software company, Yeah, a buddy

83
00:05:23,480 --> 00:05:26,960
Speaker 3: of mine, I think this happens a lot a buddy

84
00:05:27,000 --> 00:05:29,120
Speaker 3: of you know, you have a buddy who's you need

85
00:05:29,279 --> 00:05:31,479
Speaker 3: to be pretty smart? How would you like to apply.

86
00:05:31,760 --> 00:05:35,520
Speaker 3: We're always we're always talking about company X or technology.

87
00:05:35,600 --> 00:05:39,279
Speaker 3: Why why you know, why don't come in. Let's let's

88
00:05:39,640 --> 00:05:44,400
Speaker 3: give this a try. So that was my path, if

89
00:05:44,400 --> 00:05:46,920
Speaker 3: you want to call it a path. And what really

90
00:05:46,960 --> 00:05:51,839
Speaker 3: sold me on it was that markets. It's the biggest

91
00:05:51,880 --> 00:05:54,960
Speaker 3: game in the world for sure. And like I say,

92
00:05:54,960 --> 00:05:58,080
Speaker 3: I'm a game player. I love a game theorist.

93
00:05:59,160 --> 00:06:00,480
Speaker 2: Let's work that.

94
00:06:01,120 --> 00:06:02,880
Speaker 3: I don't like to talk about that because, you know,

95
00:06:02,960 --> 00:06:06,840
Speaker 3: because the real game theorists get angry. Well, yes, and

96
00:06:07,160 --> 00:06:10,240
Speaker 3: I am a real one because that was that was

97
00:06:10,320 --> 00:06:12,880
Speaker 3: my field for a lot and it's a real field

98
00:06:12,920 --> 00:06:16,800
Speaker 3: and it's a real thing, but it's been so trivialized

99
00:06:16,920 --> 00:06:20,400
Speaker 3: when some talking head will come on, well, let's look

100
00:06:20,400 --> 00:06:22,520
Speaker 3: at the game theory of this, and you just want

101
00:06:22,520 --> 00:06:25,400
Speaker 3: to just, you know, shoot yourself when somebody does this.

102
00:06:25,800 --> 00:06:29,680
Speaker 2: So the other part of your research, the other part

103
00:06:29,720 --> 00:06:33,880
Speaker 2: of your academic focus, was on narrative theory. And so

104
00:06:34,080 --> 00:06:38,680
Speaker 2: let's talk about how did that focus develop. And we'll

105
00:06:38,720 --> 00:06:41,480
Speaker 2: talk a little later about what you do today at

106
00:06:41,520 --> 00:06:45,279
Speaker 2: personant with the narrative machine.

107
00:06:45,279 --> 00:06:48,480
Speaker 3: But not believe it or not, it all ties together.

108
00:06:48,839 --> 00:06:51,320
Speaker 2: I doesn't don't doubt that for a sing, and it's

109
00:06:51,360 --> 00:06:54,760
Speaker 2: what initially led you down that rabbit hole when.

110
00:06:54,600 --> 00:06:56,760
Speaker 3: We think about kind of who's been an influence on

111
00:06:56,839 --> 00:07:02,240
Speaker 3: you in your life. At a very influential undergrad professor

112
00:07:02,600 --> 00:07:06,880
Speaker 3: in political science, and then I had a very influential

113
00:07:07,560 --> 00:07:12,800
Speaker 3: graduate advisor. Again, they don't call it political science up

114
00:07:12,840 --> 00:07:15,560
Speaker 3: at Harvard. They call it government or something like that,

115
00:07:16,200 --> 00:07:19,880
Speaker 3: but it's political science. And the reason I say they

116
00:07:19,880 --> 00:07:25,200
Speaker 3: were influential is that they really got me focused on

117
00:07:25,320 --> 00:07:36,000
Speaker 3: the science side of political science, and that science side. Yes,

118
00:07:36,120 --> 00:07:39,560
Speaker 3: it's kind of some of the typical terrible stuff you

119
00:07:39,560 --> 00:07:44,040
Speaker 3: see in all social sciences like economics, where you've got

120
00:07:44,040 --> 00:07:48,120
Speaker 3: to learn how to deal with structured data, right, and

121
00:07:48,120 --> 00:07:52,960
Speaker 3: and there was a lot to learn. And it's worth

122
00:07:53,000 --> 00:07:55,760
Speaker 3: talking about because I see the same mistakes being made

123
00:07:55,800 --> 00:08:01,800
Speaker 3: over and over again by people in our business who

124
00:08:01,880 --> 00:08:06,720
Speaker 3: want to try to, you know, apply math to data,

125
00:08:07,480 --> 00:08:10,720
Speaker 3: and there are some real pitfalls and some some real

126
00:08:10,840 --> 00:08:15,400
Speaker 3: intellectual capital. I think that you can achieve that within

127
00:08:15,480 --> 00:08:18,560
Speaker 3: academia that you can then bring in and apply to

128
00:08:18,920 --> 00:08:20,239
Speaker 3: the investment world.

129
00:08:20,440 --> 00:08:26,880
Speaker 2: Well, we've certainly seen amongst the quants a very successful

130
00:08:26,920 --> 00:08:32,199
Speaker 2: application of math theory to data. In fact, some of

131
00:08:32,240 --> 00:08:36,120
Speaker 2: the best performing hedge funds are quantitatively driven.

132
00:08:36,320 --> 00:08:39,719
Speaker 3: That's not where you're going, well, it's it's part of

133
00:08:39,760 --> 00:08:42,120
Speaker 3: where I'm going, right, So so there there there's a

134
00:08:42,200 --> 00:08:47,400
Speaker 3: transition in all of all of the sciences, honestly, but

135
00:08:47,480 --> 00:08:53,240
Speaker 3: certainly the social sciences, where where yes, you start with numbers,

136
00:08:53,280 --> 00:08:58,880
Speaker 3: structured numbers, right, uh, price over time, you know, follows

137
00:08:58,920 --> 00:09:01,920
Speaker 3: the things you can calculate, late and measure as those numbers.

138
00:09:02,760 --> 00:09:07,040
Speaker 3: But what was clear immediately in politics, and I think

139
00:09:07,120 --> 00:09:11,200
Speaker 3: has become increasingly clear in the world of investing, is

140
00:09:11,320 --> 00:09:14,160
Speaker 3: it's not just the numbers that you get on your

141
00:09:14,320 --> 00:09:20,760
Speaker 3: Bloomberg terminal. It's also the words and the stories and

142
00:09:20,800 --> 00:09:25,439
Speaker 3: the narratives that are told to us. Politicians have known

143
00:09:25,440 --> 00:09:28,679
Speaker 3: this forever, right, So the story of politics is the

144
00:09:28,720 --> 00:09:34,079
Speaker 3: story of people suggesting laws or policies and then having

145
00:09:34,120 --> 00:09:36,440
Speaker 3: to present it in a way that gets them elected

146
00:09:36,840 --> 00:09:40,240
Speaker 3: or keeps them in power or whatever that is. So

147
00:09:40,280 --> 00:09:43,080
Speaker 3: there's always been a focus, I'll say, more of a

148
00:09:43,080 --> 00:09:50,840
Speaker 3: focus in political science than in economics. With words, economics

149
00:09:50,920 --> 00:09:55,760
Speaker 3: is almost seen as a sideline, right, it's somehow lesser

150
00:09:55,800 --> 00:10:03,680
Speaker 3: than the numbers. So what I was kind of early

151
00:10:03,800 --> 00:10:11,439
Speaker 3: on was applying the same techniques that we have for

152
00:10:11,679 --> 00:10:20,000
Speaker 3: understanding matrices and structured data, but apply it to unstructured data,

153
00:10:20,360 --> 00:10:22,959
Speaker 3: which you know, full circle, this is at the heart

154
00:10:23,160 --> 00:10:26,560
Speaker 3: of all of the generative AI and the.

155
00:10:27,120 --> 00:10:30,560
Speaker 2: AI you're getting way ahead of me now with generative

156
00:10:30,600 --> 00:10:33,200
Speaker 2: I will circle back to that, but it's.

157
00:10:33,080 --> 00:10:38,600
Speaker 3: All the math has not changed in thirty five years

158
00:10:38,640 --> 00:10:43,400
Speaker 3: since I started working with network math around unstructured data.

159
00:10:43,480 --> 00:10:47,360
Speaker 3: I mean, we didn't call it natural language processing back then,

160
00:10:47,480 --> 00:10:50,960
Speaker 3: and we didn't call it, you know, large language modeling,

161
00:10:51,240 --> 00:10:53,120
Speaker 3: but that's exactly what we were doing.

162
00:10:53,320 --> 00:10:57,000
Speaker 2: So what was what was the moment or the catalyst

163
00:10:57,080 --> 00:11:00,280
Speaker 2: for you to say, Hey, I'm working in all these

164
00:11:00,320 --> 00:11:05,760
Speaker 2: other areas. But the narratives continue to pop up over

165
00:11:05,760 --> 00:11:09,199
Speaker 2: and over again on all sorts of different data sets,

166
00:11:09,720 --> 00:11:12,960
Speaker 2: and I think in the financial markets, I can use

167
00:11:13,920 --> 00:11:19,120
Speaker 2: a novel approach to identifying narratives and anticipate where the

168
00:11:19,120 --> 00:11:23,480
Speaker 2: market's going. What led to that sort of insight.

169
00:11:26,800 --> 00:11:30,880
Speaker 3: Not the Great Recession, right, but the aftermath.

170
00:11:30,880 --> 00:11:34,800
Speaker 2: Meaning the twenty tens following the Great Financial Crisis.

171
00:11:34,320 --> 00:11:37,520
Speaker 3: Starting in two thousand and nine, in the recovery that

172
00:11:37,559 --> 00:11:40,520
Speaker 3: we had out of two thousand and nine, and in

173
00:11:40,559 --> 00:11:46,600
Speaker 3: particular when Ben Bernaki and the Federal Reserve moved towards

174
00:11:47,160 --> 00:11:53,720
Speaker 3: a very explicit effort to use their words to impact markets.

175
00:11:53,920 --> 00:11:55,760
Speaker 2: So let's talk a little bit about that, because I

176
00:11:55,880 --> 00:12:01,719
Speaker 2: have some really specific memories of the low of the

177
00:12:02,440 --> 00:12:07,000
Speaker 2: run up afterwards, all the noise that was going on.

178
00:12:10,400 --> 00:12:12,559
Speaker 2: Some of the phrases that have come out of that era,

179
00:12:12,800 --> 00:12:16,439
Speaker 2: like financial repression and other such things are just the

180
00:12:16,600 --> 00:12:23,319
Speaker 2: tip of the narrative, Iceberg. So walk us through your insight.

181
00:12:23,679 --> 00:12:27,559
Speaker 2: It's two thousand and nine. Yep, the market bottoms really

182
00:12:27,640 --> 00:12:30,160
Speaker 2: kind of a v bottom and took off from that.

183
00:12:30,200 --> 00:12:32,760
Speaker 2: What was that March ninth, March seventh, something like that,

184
00:12:33,320 --> 00:12:38,120
Speaker 2: nine and there was no turning back. What were you seeing?

185
00:12:38,280 --> 00:12:42,959
Speaker 2: How were you integrating that into a concept of let's

186
00:12:43,000 --> 00:12:47,199
Speaker 2: identify narratives in order to anticipate market most well.

187
00:12:47,200 --> 00:12:53,400
Speaker 3: So I was co founded a long short fund inside

188
00:12:53,440 --> 00:12:58,880
Speaker 3: of a larger asset manager going back in five and five,

189
00:12:59,200 --> 00:13:05,839
Speaker 3: six seven, and we did well, like everyone else did well.

190
00:13:06,360 --> 00:13:07,640
Speaker 3: And then in O eight we did great.

191
00:13:08,280 --> 00:13:10,320
Speaker 2: Really you know, would wait, I think want to say

192
00:13:10,360 --> 00:13:11,880
Speaker 2: SMP down thirty seven percent?

193
00:13:12,000 --> 00:13:14,320
Speaker 3: Yeah, we were up twenty something net.

194
00:13:14,440 --> 00:13:17,560
Speaker 2: Anything in the green not in the red is amazing.

195
00:13:18,200 --> 00:13:19,800
Speaker 3: Now, I'll tell you, and we can come back to this,

196
00:13:19,920 --> 00:13:23,720
Speaker 3: like the real question you should ask is that, given

197
00:13:24,640 --> 00:13:28,760
Speaker 3: what we believed, why wasn't why weren't we up forty percent?

198
00:13:29,280 --> 00:13:32,760
Speaker 3: That that's actually that that's actually a question you can ask.

199
00:13:32,840 --> 00:13:37,360
Speaker 2: But I'm going to say a forty seven percent relative

200
00:13:37,679 --> 00:13:38,360
Speaker 2: price wing.

201
00:13:39,040 --> 00:13:41,360
Speaker 3: I'll take had A had A had a great year

202
00:13:41,400 --> 00:13:42,160
Speaker 3: in O eight.

203
00:13:42,000 --> 00:13:45,400
Speaker 2: And and did that continue in nine?

204
00:13:45,920 --> 00:13:48,760
Speaker 3: Flat lined right from O nine, So from March of

205
00:13:48,760 --> 00:13:52,480
Speaker 3: O nine, so we did well in January February the

206
00:13:52,480 --> 00:13:55,840
Speaker 3: first quarter right the rest from from March of nine,

207
00:13:56,640 --> 00:13:59,960
Speaker 3: our returns flat lined. So we never we never lost

208
00:14:00,160 --> 00:14:01,760
Speaker 3: money for our clients and our fund.

209
00:14:01,800 --> 00:14:03,560
Speaker 2: But you didn't catch that recovery though.

210
00:14:03,440 --> 00:14:10,600
Speaker 3: Did not catch the recovery, absolutely did not. And the

211
00:14:10,600 --> 00:14:12,880
Speaker 3: the recovery was interesting, right, you're you're right, there was

212
00:14:12,880 --> 00:14:15,680
Speaker 3: a v but there were there were starts and stops

213
00:14:15,679 --> 00:14:19,760
Speaker 3: to it. So the the big move up from the

214
00:14:19,800 --> 00:14:23,880
Speaker 3: bottom in late March going into April, it's like, all right,

215
00:14:24,520 --> 00:14:26,920
Speaker 3: that actually we caught a little bit of that, and

216
00:14:26,960 --> 00:14:32,280
Speaker 3: that that made sense. There was a second leg to

217
00:14:32,320 --> 00:14:37,120
Speaker 3: the rally April May, and then in June. In June

218
00:14:37,200 --> 00:14:41,960
Speaker 3: there was a ferocious rally, but June in particular it

219
00:14:42,240 --> 00:14:47,360
Speaker 3: was a classic crap rally. Right. It was a low quality,

220
00:14:47,920 --> 00:14:52,760
Speaker 3: low quality stuff right the end of March going into

221
00:14:52,800 --> 00:14:56,280
Speaker 3: April rally. It was right, this makes sense, Right, we've

222
00:14:56,320 --> 00:15:02,000
Speaker 3: bottomed the June rally. Wow, No, it was we didn't.

223
00:15:02,040 --> 00:15:02,360
Speaker 3: We didn't.

224
00:15:02,360 --> 00:15:05,080
Speaker 2: Really. Oh that's oh yes, fascinating.

225
00:15:05,120 --> 00:15:07,280
Speaker 3: Go back, go back and look at it, right, And

226
00:15:07,360 --> 00:15:08,160
Speaker 3: let me tell you I don't.

227
00:15:08,760 --> 00:15:11,880
Speaker 2: Just so you know, I have a vivid recollection of

228
00:15:11,960 --> 00:15:15,840
Speaker 2: chatting with Jim Bianco about this, and we were both bullish,

229
00:15:15,880 --> 00:15:19,560
Speaker 2: but for completely different reasons. To me, anytime US equity

230
00:15:19,560 --> 00:15:22,800
Speaker 2: markets are cut in half, I'm a buyer and people

231
00:15:22,800 --> 00:15:24,680
Speaker 2: say nineteen twenty nine, I'm like, great, you got to

232
00:15:24,680 --> 00:15:27,080
Speaker 2: go back a century to find the exception that proves

233
00:15:27,080 --> 00:15:31,920
Speaker 2: the rule. But Jim was early on in the Tina trade. Hey,

234
00:15:31,960 --> 00:15:35,120
Speaker 2: the fit has made everything cash trash bonds are They're

235
00:15:35,160 --> 00:15:39,960
Speaker 2: forcing you into equities, which is what led my post,

236
00:15:40,080 --> 00:15:43,080
Speaker 2: which everybody stole the line, this is the most hated

237
00:15:43,120 --> 00:15:44,480
Speaker 2: bull market in history.

238
00:15:44,760 --> 00:15:46,680
Speaker 3: And I wrote that up, I.

239
00:15:46,680 --> 00:15:49,880
Speaker 2: Send that out, and I heard everybody bar of that.

240
00:15:50,520 --> 00:15:54,760
Speaker 2: But I'm curious as to where the June rally took you.

241
00:15:54,920 --> 00:15:57,080
Speaker 3: Well this and this is where I'm going about the

242
00:15:57,160 --> 00:16:03,000
Speaker 3: role of forward guidance in Jim Beyonc's point about because

243
00:16:03,760 --> 00:16:14,320
Speaker 3: what the FED did wasn't just it's policies around interest rates.

244
00:16:14,400 --> 00:16:16,600
Speaker 3: You know, they took them to zero and that's where

245
00:16:16,600 --> 00:16:17,440
Speaker 3: we stayed.

246
00:16:17,120 --> 00:16:19,360
Speaker 2: Started buying mortgage backs, and.

247
00:16:19,320 --> 00:16:23,600
Speaker 3: They balance balance sheet operations.

248
00:16:22,320 --> 00:16:22,440
Speaker 2: Right.

249
00:16:23,960 --> 00:16:27,080
Speaker 3: Actually, you know, actually, and I look, I think que won,

250
00:16:27,280 --> 00:16:28,640
Speaker 3: I think it saved the world.

251
00:16:28,760 --> 00:16:30,880
Speaker 2: This is what and what was that a trillion dollars

252
00:16:31,000 --> 00:16:34,560
Speaker 2: something something? Yeah, something billions, something like eight hundred billion,

253
00:16:34,720 --> 00:16:35,640
Speaker 2: unthinkable number.

254
00:16:35,720 --> 00:16:38,840
Speaker 3: So I think that So those were specific actions took.

255
00:16:38,880 --> 00:16:43,680
Speaker 3: But but even if you people often say things when

256
00:16:43,680 --> 00:16:48,960
Speaker 3: they're leaving office. So so Bernanki's last speech is valedictory.

257
00:16:48,160 --> 00:16:51,320
Speaker 2: Address, more honest than intended.

258
00:16:51,520 --> 00:16:53,440
Speaker 3: Much more so. And you see this all the time, right,

259
00:16:53,480 --> 00:17:00,400
Speaker 3: George Washington leaves Eisenhower, I mean when when, when when

260
00:17:00,440 --> 00:17:05,040
Speaker 3: freaking Eisenhower warns you against the defense industrial complex? Yeah,

261
00:17:05,480 --> 00:17:09,040
Speaker 3: you might, you might. I'm just saying, that's right. General,

262
00:17:09,160 --> 00:17:13,280
Speaker 3: Like to me, what Bernanki said when he's leaving his

263
00:17:13,480 --> 00:17:17,440
Speaker 3: terms of office. He said, look, we had we had

264
00:17:17,480 --> 00:17:22,840
Speaker 3: two toolkits. One was traditional stuff, interest rates, down to zero.

265
00:17:23,320 --> 00:17:25,600
Speaker 3: At the time, we didn't know we could have nkeadventure straight.

266
00:17:25,680 --> 00:17:28,040
Speaker 3: So you know, that's where we were. Second were the

267
00:17:28,080 --> 00:17:33,600
Speaker 3: balance sheet operations, large scale asset purchases QE quantitative easy.

268
00:17:34,640 --> 00:17:37,480
Speaker 3: He said, you know Q one was great. Did what

269
00:17:37,640 --> 00:17:41,680
Speaker 3: we hoped he would do. Q E two Eh, operation

270
00:17:41,840 --> 00:17:45,320
Speaker 3: twist q E three. This is BRENANKI saying. He said,

271
00:17:45,560 --> 00:17:48,359
Speaker 3: I actually think that might have been a little counterproductive. Huh,

272
00:17:49,320 --> 00:17:54,240
Speaker 3: he said, But we had another toolkit, and that was

273
00:17:54,320 --> 00:17:59,200
Speaker 3: our communication policy. That was forward guidance. We started using

274
00:17:59,240 --> 00:18:03,919
Speaker 3: our words not to communicate to the market what we

275
00:18:04,000 --> 00:18:08,520
Speaker 3: actually felt. We started using our words and coordinating our

276
00:18:08,560 --> 00:18:14,560
Speaker 3: words to change the market, to change market behavior. This

277
00:18:14,640 --> 00:18:17,480
Speaker 3: is what I mean about making a conscious effort to

278
00:18:17,600 --> 00:18:22,359
Speaker 3: tell a story. It's not that it was necessarily lying,

279
00:18:23,280 --> 00:18:26,840
Speaker 3: but they they were using their words and choosing their

280
00:18:26,840 --> 00:18:28,679
Speaker 3: words for effect.

281
00:18:28,520 --> 00:18:31,760
Speaker 2: To shape perception of their shape.

282
00:18:31,600 --> 00:18:35,760
Speaker 3: Market Babe, to shape market behavior. And he said that

283
00:18:35,840 --> 00:18:39,439
Speaker 3: worked better than we had any hope that it could.

284
00:18:40,760 --> 00:18:42,280
Speaker 3: And that's where we are now.

285
00:18:42,640 --> 00:18:46,399
Speaker 2: So for the young'ins listening, I have to point out,

286
00:18:46,480 --> 00:18:49,520
Speaker 2: and you and I are old enough to remember back

287
00:18:49,560 --> 00:18:53,359
Speaker 2: in the days where there were no minutes released. There

288
00:18:53,960 --> 00:18:57,560
Speaker 2: wasn't an announcement, forget a press conference. You had to

289
00:18:57,600 --> 00:19:00,879
Speaker 2: be watching the bond market to figure out what the

290
00:19:00,920 --> 00:19:05,359
Speaker 2: Fed just did. Like today, there's we're holding the having

291
00:19:05,359 --> 00:19:08,879
Speaker 2: this conversation. There was a FED the October meeting, a

292
00:19:09,000 --> 00:19:12,600
Speaker 2: quarter point rate cut, a conversation about all sorts of stuff.

293
00:19:12,600 --> 00:19:14,679
Speaker 2: I really didn't pay a lot of attention to it.

294
00:19:16,440 --> 00:19:18,960
Speaker 2: Lack of clarity, no data, blah blah blah.

295
00:19:20,200 --> 00:19:25,119
Speaker 3: Worse than that, Greenspan would be intentionally vague and obtuse.

296
00:19:25,320 --> 00:19:28,880
Speaker 2: If you understood what I said, then then you miss unders.

297
00:19:29,080 --> 00:19:32,399
Speaker 2: You misinterpreted it exactly. You think you understand what I'm saying.

298
00:19:33,119 --> 00:19:35,960
Speaker 3: You know who led the committee to make all that change,

299
00:19:36,560 --> 00:19:39,440
Speaker 3: Janet Yelling, which is a Spice chair yep.

300
00:19:39,359 --> 00:19:41,679
Speaker 2: Back back in the financial crisis.

301
00:19:41,720 --> 00:19:46,280
Speaker 3: Yep. So this was it was a concerted effort. Bernanki

302
00:19:46,640 --> 00:19:49,800
Speaker 3: yelling to. This is when they also started going, putting

303
00:19:49,840 --> 00:19:54,080
Speaker 3: all the FED governors on a common calendar, right and when,

304
00:19:55,280 --> 00:19:57,840
Speaker 3: and assigning the fake you're going to speak this day,

305
00:19:57,920 --> 00:20:00,560
Speaker 3: You're going to speak that day. That's when all the started.

306
00:20:00,920 --> 00:20:03,920
Speaker 3: It was an intentional effort. And again this is something

307
00:20:03,960 --> 00:20:08,080
Speaker 3: that politicians have known forever, right, politicians craft the message

308
00:20:08,119 --> 00:20:11,879
Speaker 3: and use their words. So I knew the tools to

309
00:20:11,920 --> 00:20:15,800
Speaker 3: try to understand this, but what I wasn't prepared for

310
00:20:16,119 --> 00:20:19,720
Speaker 3: was how it neither was Berneki was how powerful this

311
00:20:19,800 --> 00:20:24,159
Speaker 3: would become to the point where today it's not just

312
00:20:25,320 --> 00:20:29,960
Speaker 3: central bankers using their words as their main policy toolkit,

313
00:20:30,640 --> 00:20:33,919
Speaker 3: but is every ceo? Is every ceo now? I mean

314
00:20:33,920 --> 00:20:37,520
Speaker 3: you go on this network or one of the other networks,

315
00:20:38,560 --> 00:20:41,639
Speaker 3: and what makes for a good CEO is can you

316
00:20:41,720 --> 00:20:45,760
Speaker 3: tell the story? Can you tell the narrative of your

317
00:20:45,840 --> 00:20:50,400
Speaker 3: company to get a multiple? Because a multiple is a narrative,

318
00:20:50,600 --> 00:20:51,679
Speaker 3: a multiple is a story.

319
00:20:52,320 --> 00:20:54,879
Speaker 2: Well, look at stories, look at some of the most

320
00:20:54,920 --> 00:20:59,880
Speaker 2: successful CEOs throughout history. I would throw Jack Welsh into

321
00:20:59,880 --> 00:21:04,399
Speaker 2: that a pile because he was a fabuloust fabulist. He

322
00:21:04,520 --> 00:21:08,560
Speaker 2: was a fabulous fabulouist because the stories he told were great,

323
00:21:09,080 --> 00:21:11,720
Speaker 2: right up until the point where we found out that

324
00:21:11,800 --> 00:21:15,679
Speaker 2: he was running a hedge fund with g Capital and

325
00:21:15,760 --> 00:21:17,560
Speaker 2: they magically always beat by a penny.

326
00:21:17,640 --> 00:21:23,000
Speaker 3: So I remember vividly when ge was coming to our shop.

327
00:21:27,119 --> 00:21:32,320
Speaker 3: What they wanted was a financial financial multiple, right, so

328
00:21:32,359 --> 00:21:35,600
Speaker 3: they were making they they wanted an old world even

329
00:21:35,640 --> 00:21:39,240
Speaker 3: though they're an industrial this that this was They wanted

330
00:21:39,240 --> 00:21:43,919
Speaker 3: to tell a story that they should be seen as

331
00:21:43,960 --> 00:21:47,360
Speaker 3: and get the multiple of a financial That's what GE

332
00:21:47,440 --> 00:21:52,600
Speaker 3: was all about in those years leading up to the GFC.

333
00:21:53,440 --> 00:21:57,840
Speaker 3: So hey, my my poster child for this is is

334
00:21:57,880 --> 00:22:04,359
Speaker 3: Mark Binnioff Salesforce because he's Prabay. You often see this

335
00:22:04,480 --> 00:22:07,199
Speaker 3: with people who come out of sales like like mar

336
00:22:07,359 --> 00:22:08,080
Speaker 3: I did.

337
00:22:08,440 --> 00:22:10,320
Speaker 2: But it's all about storytelling.

338
00:22:10,720 --> 00:22:14,840
Speaker 3: It's all about storytelling. And what isn't that true?

339
00:22:14,920 --> 00:22:20,480
Speaker 2: For go through the great look Steve Jobs, Reid Hastings,

340
00:22:22,200 --> 00:22:27,280
Speaker 2: Larry Ellison, an oracle to some degree, Steve Ballmer at

341
00:22:27,280 --> 00:22:30,439
Speaker 2: Microsoft who wasn't a great CEO, but he was a

342
00:22:30,440 --> 00:22:33,000
Speaker 2: great cheerleader and a great storyteller.

343
00:22:32,440 --> 00:22:37,080
Speaker 3: Great storyteller. What all of those companies have in common

344
00:22:37,240 --> 00:22:40,760
Speaker 3: is that they're great storytellers, and those are trillion dollar

345
00:22:40,840 --> 00:22:44,440
Speaker 3: companies today. What I would say to you is that

346
00:22:44,680 --> 00:22:48,800
Speaker 3: you don't remember or hear about the companies that did

347
00:22:48,840 --> 00:22:51,320
Speaker 3: not have CEOs who are great storytellers.

348
00:22:51,320 --> 00:22:54,680
Speaker 2: Well, Ken Lay was a great storyteller until you found

349
00:22:54,760 --> 00:22:57,440
Speaker 2: out that it was all nonsense. And you could say

350
00:22:57,480 --> 00:22:59,720
Speaker 2: the same thing about folks like Bernie Madoff.

351
00:23:00,240 --> 00:23:03,000
Speaker 3: There are a lot of stories that get told that

352
00:23:03,160 --> 00:23:08,800
Speaker 3: are not true, right, what And and I think even

353
00:23:08,840 --> 00:23:12,439
Speaker 3: today people think of this word narrative. They have a

354
00:23:12,480 --> 00:23:18,840
Speaker 3: pejorative sense to it. It's like, oh, for for sure,

355
00:23:18,840 --> 00:23:21,679
Speaker 3: that's just your narrative, man, you know, the big Lebowski.

356
00:23:21,720 --> 00:23:25,560
Speaker 3: You know, that's that's the And it's not that a

357
00:23:25,720 --> 00:23:31,080
Speaker 3: story is a lie. It's that the story is constructed

358
00:23:31,240 --> 00:23:36,640
Speaker 3: for effect. It's it's not and it's presented to you

359
00:23:36,760 --> 00:23:40,600
Speaker 3: as if this is my true and inner thoughts. But

360
00:23:42,359 --> 00:23:47,800
Speaker 3: the construction of the intentionality behind these stories phenomenal. Bniel,

361
00:23:47,880 --> 00:23:51,480
Speaker 3: for example, you know, created the metrics by which you

362
00:23:51,680 --> 00:23:56,480
Speaker 3: wanted salesforce to be judged, not metrics of profitability, but

363
00:23:56,680 --> 00:24:00,560
Speaker 3: metrics of what he called pro forma net revenue growth,

364
00:24:01,240 --> 00:24:04,480
Speaker 3: whatever the hell that means, right, Because if you can

365
00:24:04,520 --> 00:24:08,400
Speaker 3: construct the story, you can construct it in a way

366
00:24:08,440 --> 00:24:11,720
Speaker 3: that yes, I can beat and raise pretty much every quarter.

367
00:24:12,359 --> 00:24:17,760
Speaker 3: So there was there were three I think big changes

368
00:24:17,840 --> 00:24:23,640
Speaker 3: that happened to make the role of narrative overwhelming as

369
00:24:23,640 --> 00:24:26,359
Speaker 3: it is today. Whereas before it's always been there, to

370
00:24:26,400 --> 00:24:30,639
Speaker 3: your point, it's always been there. Today it's overwhelming. And

371
00:24:30,920 --> 00:24:34,520
Speaker 3: I think it's it's not just the success that first

372
00:24:34,560 --> 00:24:38,639
Speaker 3: central bankers and then CEOs. I mean, Wall Street's the

373
00:24:38,640 --> 00:24:42,800
Speaker 3: greatest copying machine. Wall Street copies what works. Sure, so

374
00:24:43,000 --> 00:24:46,000
Speaker 3: when you see the something's working, oh, they're getting a

375
00:24:46,080 --> 00:24:49,600
Speaker 3: multiple by telling the story and going on Kramer, you know,

376
00:24:49,760 --> 00:24:50,640
Speaker 3: four times a year.

377
00:24:50,880 --> 00:24:55,040
Speaker 2: It's endless iteration. You're just constantly tweaking it and if

378
00:24:55,080 --> 00:24:57,479
Speaker 2: it works, do more of it, and if it doesn't,

379
00:24:57,560 --> 00:24:58,240
Speaker 2: toiss it out.

380
00:24:58,359 --> 00:25:01,640
Speaker 3: So it was the fact that it works to tell

381
00:25:01,640 --> 00:25:06,680
Speaker 3: a story, and people got good at telling stories. It's

382
00:25:06,720 --> 00:25:09,040
Speaker 3: the growth of twenty four to seven. I'm want to

383
00:25:09,119 --> 00:25:11,120
Speaker 3: use air quotes here. I'm glad we're taking this.

384
00:25:11,160 --> 00:25:14,120
Speaker 2: Well, it's medius, social media news.

385
00:25:13,920 --> 00:25:17,919
Speaker 3: Right now, news like news liked. That wasn't the case.

386
00:25:18,560 --> 00:25:21,879
Speaker 2: So I want to annotate what you said slightly, okay,

387
00:25:22,080 --> 00:25:26,399
Speaker 2: because I think CEOs have always been storytellers. But they

388
00:25:26,440 --> 00:25:32,359
Speaker 2: were storytellers to their boards, to their employees, to their shareholders.

389
00:25:32,440 --> 00:25:36,320
Speaker 2: They're always you're hitting now on the modern world of

390
00:25:36,920 --> 00:25:40,320
Speaker 2: twenty four to seven media. Telling a story in a

391
00:25:40,359 --> 00:25:44,000
Speaker 2: boardroom is very different than sitting in a TV studio

392
00:25:44,040 --> 00:25:48,000
Speaker 2: and talking about, Hey, here's why our new chip is

393
00:25:48,080 --> 00:25:51,160
Speaker 2: going to catch up to video and it's the greatest thing. Ever, Yep,

394
00:25:51,440 --> 00:25:52,600
Speaker 2: that's a different skill set.

395
00:25:52,800 --> 00:25:55,560
Speaker 3: It changes the time horizon. It is a very different

396
00:25:55,600 --> 00:25:58,200
Speaker 3: skill set because you're not telling the story of oh

397
00:25:58,200 --> 00:26:02,400
Speaker 3: I'm getting another you know turn of leverage and our operations,

398
00:26:02,520 --> 00:26:07,240
Speaker 3: or you know, our capacity utilization and this factory went

399
00:26:07,320 --> 00:26:09,600
Speaker 3: up by five percent, which are the kind of stories

400
00:26:09,640 --> 00:26:13,760
Speaker 3: you would tell even on earnings call or certainly to

401
00:26:13,840 --> 00:26:17,000
Speaker 3: a board. Now, this is the story where you've got

402
00:26:17,000 --> 00:26:19,440
Speaker 3: a segment a little bit, Yes, you've got a segment.

403
00:26:19,880 --> 00:26:23,159
Speaker 3: You're going on Kramer, you say bye bye bye. You

404
00:26:23,240 --> 00:26:26,679
Speaker 3: got at most four minutes, right, How are you going

405
00:26:26,720 --> 00:26:31,439
Speaker 3: to tell that story that sings to that audience? Enormous

406
00:26:31,560 --> 00:26:38,760
Speaker 3: change change, structural change in our media, both quote unquote

407
00:26:38,880 --> 00:26:43,000
Speaker 3: news media, but also financial news media. The Wall Street

408
00:26:43,080 --> 00:26:46,480
Speaker 3: Journal today is a twenty four to seven news financial

409
00:26:46,520 --> 00:26:47,480
Speaker 3: news organization.

410
00:26:48,240 --> 00:26:51,879
Speaker 2: What is printing times in back to Washington Post? They

411
00:26:51,920 --> 00:26:54,520
Speaker 2: all have websites that can update it around the clock.

412
00:26:54,640 --> 00:26:58,760
Speaker 3: And here's the thing, there's not enough hard news to

413
00:26:58,880 --> 00:27:02,200
Speaker 3: fill the time or fill the space. So what takes

414
00:27:02,240 --> 00:27:06,679
Speaker 3: the space? Opinion story you channel glory takes the place

415
00:27:06,800 --> 00:27:07,240
Speaker 3: you were.

416
00:27:07,160 --> 00:27:11,119
Speaker 2: Channeling Michael Crichton from twenty five years ago. Most of

417
00:27:11,160 --> 00:27:16,320
Speaker 2: what you see in the media is speculation, opinion, and theory,

418
00:27:16,880 --> 00:27:17,640
Speaker 2: not news.

419
00:27:17,840 --> 00:27:20,760
Speaker 3: I've written so much about Criton in his study that.

420
00:27:20,920 --> 00:27:22,280
Speaker 2: I know that's why I threw that back.

421
00:27:22,640 --> 00:27:25,600
Speaker 3: He says he was he was so far ahead, a.

422
00:27:25,640 --> 00:27:27,920
Speaker 2: Quarter century ahead of what of what took place.

423
00:27:29,800 --> 00:27:31,800
Speaker 3: There's a third piece, though, give us.

424
00:27:31,680 --> 00:27:33,919
Speaker 2: The third piece before we go to our next.

425
00:27:34,040 --> 00:27:37,879
Speaker 3: Third piece has changed everything is our smartphones.

426
00:27:38,000 --> 00:27:41,080
Speaker 2: That you're you're walking around with not only a studio,

427
00:27:41,840 --> 00:27:47,159
Speaker 2: but a dopamine device that you're constantly it's.

428
00:27:47,760 --> 00:27:51,520
Speaker 3: Machine and I we do it to ourselves. It's not

429
00:27:51,560 --> 00:27:55,199
Speaker 3: that someone forces us to hear these stories over and

430
00:27:55,240 --> 00:27:57,960
Speaker 3: over again. We do it to ourselves. I mean, I

431
00:27:58,000 --> 00:28:00,760
Speaker 3: get a little nervous if I you know, pat, where's

432
00:28:00,760 --> 00:28:05,000
Speaker 3: my phone? Where's I get a little nervous. And it's

433
00:28:05,040 --> 00:28:12,400
Speaker 3: a it's It is absolutely a neurotransmitter addiction. I think

434
00:28:12,400 --> 00:28:15,640
Speaker 3: it's so important to keep that from our kids.

435
00:28:16,760 --> 00:28:19,800
Speaker 2: That's a whole another there's a whole depression situation with

436
00:28:19,840 --> 00:28:22,240
Speaker 2: teenagers today, and it all traces back to the phone

437
00:28:22,280 --> 00:28:22,960
Speaker 2: and social media.

438
00:28:23,080 --> 00:28:26,040
Speaker 3: These are three I think real secular changes we've had.

439
00:28:27,800 --> 00:28:35,639
Speaker 3: Markets become this political utility, the success of constructing a

440
00:28:35,720 --> 00:28:41,720
Speaker 3: story structural changes in social media and the devices that

441
00:28:41,760 --> 00:28:46,400
Speaker 3: we insist on caring with ourselves all the time.

442
00:28:46,880 --> 00:28:50,479
Speaker 2: Huh, absolutely fascinating. Coming up, we continue our conversation with

443
00:28:50,600 --> 00:28:55,880
Speaker 2: Ben Hunt, president and co founder of Percion, explaining how

444
00:28:55,920 --> 00:29:00,720
Speaker 2: he's using AI to identify narratives in real time.

445
00:29:01,320 --> 00:29:02,400
Speaker 3: I'm Barry Redhults.

446
00:29:02,480 --> 00:29:18,280
Speaker 2: You're listening to Masters in Business on Bloomberg Radio. I'm

447
00:29:18,320 --> 00:29:21,040
Speaker 2: Burry rid Halts. You're listening to Masters in Business on

448
00:29:21,080 --> 00:29:24,880
Speaker 2: Bloomberg Radio. My special guest this week is Ben Hunt.

449
00:29:25,360 --> 00:29:32,440
Speaker 2: He is a academic, fund manager, risk manager, entrepreneur, tech

450
00:29:32,560 --> 00:29:37,840
Speaker 2: startup person. He is currently co founder and president at percent,

451
00:29:38,000 --> 00:29:43,600
Speaker 2: which applies AI tools to map and measure market narratives

452
00:29:44,000 --> 00:29:47,680
Speaker 2: in real time. It's really more than market narratives. It's politics,

453
00:29:47,680 --> 00:29:50,920
Speaker 2: it's economics, it's markets. You cover up a whole lot

454
00:29:50,920 --> 00:29:51,280
Speaker 2: of stuff.

455
00:29:51,840 --> 00:29:53,720
Speaker 3: So what we're able to do today and this is

456
00:29:53,760 --> 00:29:57,080
Speaker 3: the crazy change in the world. Back from when I

457
00:29:57,200 --> 00:30:06,200
Speaker 3: was doing this on microfiche, back in yeah, exactly, we

458
00:30:06,360 --> 00:30:11,680
Speaker 3: get we have access to everything that's published publicly in

459
00:30:11,760 --> 00:30:15,120
Speaker 3: the world, and there are a couple of big data

460
00:30:15,200 --> 00:30:22,760
Speaker 3: aggregators dal Jones one, Lexis and Nexus another. Everything that

461
00:30:22,880 --> 00:30:26,440
Speaker 3: gets published in the world, all these languages. It's available

462
00:30:26,480 --> 00:30:30,440
Speaker 3: to you, and it's it's not cheap, but it's not

463
00:30:30,600 --> 00:30:33,560
Speaker 3: crazy expensive like it used to be. It's always getting cheaper.

464
00:30:34,160 --> 00:30:36,240
Speaker 3: So we're able to take everything in the world that

465
00:30:36,240 --> 00:30:39,760
Speaker 3: gets published, all the newspapers, all the websites, all the transcripts,

466
00:30:39,800 --> 00:30:45,120
Speaker 3: everything's published publicly. We can pull in and then we

467
00:30:45,240 --> 00:30:51,040
Speaker 3: can process it, process it with Really is the same

468
00:30:51,200 --> 00:30:54,600
Speaker 3: math that I was using thirty years ago. Nobody's invented

469
00:30:54,640 --> 00:30:56,000
Speaker 3: cold fusion here.

470
00:30:55,880 --> 00:30:59,120
Speaker 2: But the software tools are faster, stronger, better.

471
00:30:59,320 --> 00:31:05,280
Speaker 3: And infinite. So the calculations here are not particularly complicated,

472
00:31:05,960 --> 00:31:07,880
Speaker 3: but you have to do them at enormous scale.

473
00:31:08,120 --> 00:31:09,680
Speaker 2: It's a ton of volume.

474
00:31:09,600 --> 00:31:15,440
Speaker 3: So I mean, it's crazy, the just the scale of.

475
00:31:15,360 --> 00:31:17,720
Speaker 2: The abytes, terabytes, just crazy.

476
00:31:17,880 --> 00:31:20,120
Speaker 3: I mean, yeah, in the last couple of months we've

477
00:31:20,160 --> 00:31:26,320
Speaker 3: processed over two hundred billion tokens. Billion to is how

478
00:31:26,400 --> 00:31:28,840
Speaker 3: much a token is, like a word or a folk

479
00:31:29,680 --> 00:31:31,800
Speaker 3: And so that's the kind of the unit that you

480
00:31:31,840 --> 00:31:35,320
Speaker 3: talk about when you're putting through, when you're putting something

481
00:31:35,360 --> 00:31:39,480
Speaker 3: through a linguistic calculator, which is which is what all

482
00:31:39,560 --> 00:31:45,440
Speaker 3: of the all the llms are, they're linguistic calculators and

483
00:31:45,520 --> 00:31:51,320
Speaker 3: so you know, we've processed, you know, several hundred billion tokens. Again,

484
00:31:51,360 --> 00:31:56,520
Speaker 3: it's not complex, but it is at scale. And what

485
00:31:56,520 --> 00:32:00,360
Speaker 3: we're doing with that is we're reading the world's use

486
00:32:01,400 --> 00:32:06,440
Speaker 3: to understand the world's stories and narratives. And you're right,

487
00:32:06,520 --> 00:32:11,760
Speaker 3: it's much bigger than or it's much more focused that.

488
00:32:11,880 --> 00:32:16,280
Speaker 3: We have much higher resolution than just saying, oh, I'm

489
00:32:16,320 --> 00:32:21,000
Speaker 3: bullish on financials. I mean, that's a narrative, sure, but

490
00:32:21,640 --> 00:32:26,000
Speaker 3: there are twenty different variations of that. You're bullish on financials.

491
00:32:26,440 --> 00:32:29,160
Speaker 3: Why so wait, let me let me we attract all

492
00:32:29,160 --> 00:32:31,440
Speaker 3: those We attract all those stories and how they wax

493
00:32:31,480 --> 00:32:32,440
Speaker 3: and wane over time.

494
00:32:32,600 --> 00:32:34,960
Speaker 2: So let me roll back to I want you to

495
00:32:35,080 --> 00:32:39,000
Speaker 2: explain what Persient is. Who are the clients. I don't

496
00:32:39,040 --> 00:32:42,160
Speaker 2: mean names, but what type of clients do you have

497
00:32:42,720 --> 00:32:46,320
Speaker 2: and what do they do with Persian's output?

498
00:32:47,240 --> 00:32:52,240
Speaker 3: So Starry Percient in twenty eighteen, my partner from we

499
00:32:52,240 --> 00:32:55,960
Speaker 3: were at an asset manager, spun out there to take

500
00:32:56,040 --> 00:32:59,400
Speaker 3: the the technology that I've been working on for years

501
00:33:00,720 --> 00:33:04,240
Speaker 3: and really been writing about with epsilon theory. So we

502
00:33:04,280 --> 00:33:08,440
Speaker 3: started that in twenty eighteen to do the basic research

503
00:33:09,240 --> 00:33:15,920
Speaker 3: into processing enormous amounts of financial news data and to

504
00:33:16,080 --> 00:33:21,600
Speaker 3: track the stories and how they rise and fall and

505
00:33:21,720 --> 00:33:24,960
Speaker 3: wax and wane over time. So that was that was

506
00:33:24,960 --> 00:33:25,560
Speaker 3: the story.

507
00:33:25,360 --> 00:33:29,840
Speaker 2: That was I love that description because there are a

508
00:33:29,840 --> 00:33:34,719
Speaker 2: lot of trades going on where the storyline changes on

509
00:33:34,760 --> 00:33:39,920
Speaker 2: a regular basis. Probably the mac daddy of that is crypto. First,

510
00:33:39,960 --> 00:33:43,800
Speaker 2: it's hey, you know there's fiat currency, this is outside

511
00:33:43,840 --> 00:33:47,200
Speaker 2: of the system. It's DeFi. Then it's a hedge for deflation.

512
00:33:47,280 --> 00:33:51,040
Speaker 2: Then it's a hedge for inflation. Now it's scarcity.

513
00:33:51,200 --> 00:33:56,680
Speaker 3: And digital gold dig Gold's there are no fundamentals right

514
00:33:56,720 --> 00:33:59,280
Speaker 3: with with crypto and people will say there are, but

515
00:33:59,280 --> 00:34:03,320
Speaker 3: there aren't. It's it's driven by the waxing and waning

516
00:34:03,360 --> 00:34:08,200
Speaker 3: of stories. Uh huh and do they find purchase? Do they?

517
00:34:08,800 --> 00:34:11,600
Speaker 3: Or do they? Kind of people get tired of them.

518
00:34:11,880 --> 00:34:14,880
Speaker 2: Well, they got tired of the DeFi story and then Warnce,

519
00:34:15,800 --> 00:34:19,200
Speaker 2: JP Morgan and Black Rocks started creating it like the

520
00:34:19,520 --> 00:34:23,320
Speaker 2: by bit is the fastest etf to one hundred billion

521
00:34:23,360 --> 00:34:26,920
Speaker 2: dollars And so the old story of DeFi is gone

522
00:34:26,920 --> 00:34:29,000
Speaker 2: and a new story is, oh no, this is an

523
00:34:29,000 --> 00:34:31,160
Speaker 2: asset class at Wall Street's embracing that's why you have

524
00:34:31,200 --> 00:34:33,600
Speaker 2: to own it that story.

525
00:34:33,680 --> 00:34:36,640
Speaker 3: That story was and very that was a very similar story,

526
00:34:36,680 --> 00:34:39,800
Speaker 3: by the way, or a transition in story from physical

527
00:34:39,840 --> 00:34:43,960
Speaker 3: gold to GLD when that ETF came out, right, which

528
00:34:45,400 --> 00:34:50,120
Speaker 3: was a very similar pattern, because once it became a

529
00:34:50,160 --> 00:34:53,600
Speaker 3: Wall Street a table at the Wall Street casino, right,

530
00:34:54,320 --> 00:35:00,400
Speaker 3: then it takes on a different meaning, specially around go

531
00:35:00,400 --> 00:35:05,000
Speaker 3: gold change from being okay, something that you bury in

532
00:35:05,000 --> 00:35:07,080
Speaker 3: your back, your heart or you're having your vault, you know,

533
00:35:07,160 --> 00:35:11,520
Speaker 3: along with mo and seeds for when the hard times.

534
00:35:11,160 --> 00:35:14,880
Speaker 2: Coddled water right balls ready to eat gold and lead.

535
00:35:14,840 --> 00:35:19,520
Speaker 3: It becomes a security and it's meaning changes from that,

536
00:35:20,040 --> 00:35:24,920
Speaker 3: you know, apocalyptic modeled water. And the meaning of gold

537
00:35:24,920 --> 00:35:29,760
Speaker 3: today is as an insurance policy, a security against central

538
00:35:29,760 --> 00:35:34,040
Speaker 3: bank error or government error. That's the meaning of gold today.

539
00:35:35,080 --> 00:35:38,799
Speaker 2: And is that the dominant narrative that you're identifying as

540
00:35:38,920 --> 00:35:41,280
Speaker 2: gold rallied over through four thousand?

541
00:35:42,320 --> 00:35:45,319
Speaker 3: Absolutely so, And we've really been able to track that

542
00:35:45,360 --> 00:35:45,640
Speaker 3: one in.

543
00:35:46,040 --> 00:35:49,400
Speaker 2: So here's the really big here's the million dollar or

544
00:35:49,480 --> 00:35:55,120
Speaker 2: trillion dollar question. How do you identify a narrative and say, oh,

545
00:35:55,160 --> 00:35:57,120
Speaker 2: gold is going to double from here based on this

546
00:35:57,239 --> 00:35:59,280
Speaker 2: narrative or are we not there yet.

547
00:36:00,440 --> 00:36:03,680
Speaker 3: No, you you identify the narratives because the narrative the

548
00:36:03,719 --> 00:36:09,120
Speaker 3: stories don't ever change. So that the story that leave

549
00:36:09,120 --> 00:36:12,759
Speaker 3: Gold aside for a think about we're talking about your

550
00:36:12,800 --> 00:36:18,600
Speaker 3: bullish on company X y Z, because there are about

551
00:36:18,840 --> 00:36:21,400
Speaker 3: I don't know, depending on how again, how finally you

552
00:36:21,440 --> 00:36:25,800
Speaker 3: want to resolve that they're only about a dozen stories

553
00:36:25,880 --> 00:36:28,840
Speaker 3: for why you're bullish on something, right, They can be

554
00:36:29,320 --> 00:36:36,640
Speaker 3: management change, top line growth, opportunity, consolidation in the industry, catalyst.

555
00:36:36,920 --> 00:36:40,920
Speaker 3: So every catalyst story, new product is going to prove

556
00:36:40,960 --> 00:36:45,360
Speaker 3: the new drug, new molecule. So that story you just

557
00:36:45,480 --> 00:36:49,120
Speaker 3: change the name. That's the same story that's repeated over

558
00:36:49,239 --> 00:36:55,839
Speaker 3: and over again about any pharma or biotech company. The stories,

559
00:36:56,880 --> 00:37:02,960
Speaker 3: we think that they are amorphous and variable. The fact

560
00:37:03,040 --> 00:37:05,080
Speaker 3: is that the core of the story, what we call

561
00:37:05,120 --> 00:37:10,880
Speaker 3: the semantic signature, the meaning of a story, they're amazingly

562
00:37:12,320 --> 00:37:17,000
Speaker 3: constant over time. So what we're looking for is for

563
00:37:18,040 --> 00:37:20,680
Speaker 3: it could be dormant for a long time. But what

564
00:37:20,719 --> 00:37:24,000
Speaker 3: you want to know is when that story starts picking

565
00:37:24,080 --> 00:37:30,920
Speaker 3: up again, when someone starts playing that story, when it

566
00:37:31,080 --> 00:37:34,920
Speaker 3: appears on Kramer and starts happening in the financial press.

567
00:37:35,760 --> 00:37:38,480
Speaker 3: That's the stuff we can pick up with real precision.

568
00:37:38,800 --> 00:37:40,960
Speaker 3: So it's both the stories that are starting to fade,

569
00:37:41,920 --> 00:37:45,520
Speaker 3: but the I think the really interesting stories are the

570
00:37:45,560 --> 00:37:47,520
Speaker 3: stories that have been dormant for a long time and

571
00:37:47,560 --> 00:37:48,640
Speaker 3: they start picking up again.

572
00:37:49,000 --> 00:37:53,239
Speaker 2: You are reminding me of Campbell's hero's journey that there's

573
00:37:53,280 --> 00:37:54,440
Speaker 2: only so many.

574
00:37:54,719 --> 00:37:57,920
Speaker 3: My hero right, there are only so many stories right

575
00:37:58,000 --> 00:38:02,840
Speaker 3: in the world. And that's it now, I like to

576
00:38:02,840 --> 00:38:09,680
Speaker 3: talk about in Hollywood famously, there are only like five scripts, right, uh.

577
00:38:10,400 --> 00:38:13,080
Speaker 3: Tolstoy Is supposedly told Stoy he said, there are only

578
00:38:13,120 --> 00:38:16,200
Speaker 3: two stories that a man goes on a journey or

579
00:38:16,200 --> 00:38:18,839
Speaker 3: a stranger comes to town. Those are the only two

580
00:38:18,840 --> 00:38:19,640
Speaker 3: stories in the world.

581
00:38:20,040 --> 00:38:21,959
Speaker 2: And not quite but he's not quite.

582
00:38:22,000 --> 00:38:25,320
Speaker 3: But you're on the right track, right right. The point

583
00:38:25,360 --> 00:38:29,400
Speaker 3: is there's a finite number of stories you can drill down,

584
00:38:29,520 --> 00:38:31,600
Speaker 3: so you can get a couple of dozen about any

585
00:38:31,800 --> 00:38:34,440
Speaker 3: sector you want to talk about or like. But it's

586
00:38:34,560 --> 00:38:39,120
Speaker 3: it's a finite number. And so what we do, and

587
00:38:39,160 --> 00:38:42,320
Speaker 3: what I think is really interesting, is to track the

588
00:38:42,360 --> 00:38:45,880
Speaker 3: finite number of stories. And you're right, it's not just

589
00:38:45,920 --> 00:38:50,720
Speaker 3: around markets. We track several thousand of these stories today.

590
00:38:51,040 --> 00:38:55,480
Speaker 2: How So, let's let's delve into that. So, so, how

591
00:38:55,480 --> 00:38:58,799
Speaker 2: do you you have this massive database you're sucking in

592
00:38:58,960 --> 00:39:06,200
Speaker 2: every news feed, every magazine, newspaper, everything that you could

593
00:39:06,280 --> 00:39:09,360
Speaker 2: quantify and run into a linguistics model.

594
00:39:10,040 --> 00:39:11,960
Speaker 3: What's the process.

595
00:39:11,320 --> 00:39:15,920
Speaker 2: For analyzing this, how to use artificial intelligence to go

596
00:39:16,000 --> 00:39:18,480
Speaker 2: through this? And how do you make sense out of

597
00:39:18,520 --> 00:39:22,000
Speaker 2: that heap of how do you find signal amidst all

598
00:39:22,040 --> 00:39:22,600
Speaker 2: that noise?

599
00:39:23,480 --> 00:39:26,600
Speaker 3: The crucial thing is you can't just ask AI an

600
00:39:26,600 --> 00:39:31,279
Speaker 3: open ended questions, say what are the narratives in this

601
00:39:31,600 --> 00:39:36,120
Speaker 3: for this company or for this sector. Don't do that right,

602
00:39:36,160 --> 00:39:38,040
Speaker 3: And this is a mistake that people make all the time.

603
00:39:38,080 --> 00:39:42,560
Speaker 3: They ask open ended questions of chat GPT or whoever

604
00:39:43,400 --> 00:39:45,560
Speaker 3: the problem is. Chat gpt will give you an answer,

605
00:39:46,280 --> 00:39:48,560
Speaker 3: just not a good one, not a good one. It'll

606
00:39:48,600 --> 00:39:52,520
Speaker 3: hallucinate a lot. Right, It'll go out, it'll find its

607
00:39:52,560 --> 00:39:58,360
Speaker 3: own data. The secret to using AI successfully is to

608
00:39:58,400 --> 00:40:02,040
Speaker 3: take this magic genie. It's a magic genie, and you

609
00:40:02,120 --> 00:40:05,359
Speaker 3: stuff it into that bottle. You do not let it out.

610
00:40:05,880 --> 00:40:08,880
Speaker 3: You constrain it dramatically. You don't let it go out

611
00:40:08,920 --> 00:40:13,520
Speaker 3: and find data. You give it the data. Crucial thing

612
00:40:14,239 --> 00:40:19,239
Speaker 3: you don't allow it to think you tell it how

613
00:40:19,320 --> 00:40:22,600
Speaker 3: to think. So the most important step that we do

614
00:40:23,080 --> 00:40:27,280
Speaker 3: is we don't ask AI what are the narratives that Okay,

615
00:40:27,760 --> 00:40:30,719
Speaker 3: we tell it. This is human direction. You have to

616
00:40:30,760 --> 00:40:31,920
Speaker 3: have human control.

617
00:40:32,040 --> 00:40:36,560
Speaker 2: I'm hearing data set is controlled by you as well

618
00:40:36,600 --> 00:40:37,800
Speaker 2: as the thinking.

619
00:40:39,440 --> 00:40:42,920
Speaker 3: Yes, and it's more than prompt, right, So the it

620
00:40:43,440 --> 00:40:48,360
Speaker 3: includes prompt. But the phrase that's used in this world

621
00:40:48,440 --> 00:40:54,600
Speaker 3: is called not prompt engineering, but context engineering. So you

622
00:40:54,640 --> 00:40:59,880
Speaker 3: want to you want to control everything around the A

623
00:41:00,920 --> 00:41:04,960
Speaker 3: because you want to limit it to being that linguistic calculator.

624
00:41:05,360 --> 00:41:08,640
Speaker 3: You want it to be your operating system. That's really

625
00:41:08,680 --> 00:41:11,480
Speaker 3: the way the thing. And if you do that, then

626
00:41:12,239 --> 00:41:15,040
Speaker 3: it will give you the same answer twice for the

627
00:41:15,080 --> 00:41:18,279
Speaker 3: same inputs and the same question. That's the crucial thing.

628
00:41:18,360 --> 00:41:20,719
Speaker 3: So for this to be true, for it to be consistent,

629
00:41:20,840 --> 00:41:23,560
Speaker 3: for it to be real signal. So this is a

630
00:41:23,680 --> 00:41:27,360
Speaker 3: human directed process. You can't ask AI an open ended question.

631
00:41:27,840 --> 00:41:30,759
Speaker 3: You have to control all the inputs. You have to

632
00:41:30,800 --> 00:41:33,759
Speaker 3: control the output, meaning you judge it, you run it

633
00:41:33,800 --> 00:41:36,759
Speaker 3: back through a different AI system to say had they

634
00:41:36,800 --> 00:41:39,600
Speaker 3: do did they go off the rails here? But the

635
00:41:39,640 --> 00:41:43,480
Speaker 3: most important thing is you have to give it the scaffolding.

636
00:41:43,560 --> 00:41:47,120
Speaker 3: You have to give it the the skeleton. You have

637
00:41:47,200 --> 00:41:50,160
Speaker 3: to tell it these are the thoughts you are allowed

638
00:41:50,200 --> 00:41:53,839
Speaker 3: to think about, and those are the signatures.

639
00:41:54,040 --> 00:41:59,120
Speaker 2: I've kind of learned. I have to avoid asking questions

640
00:41:59,160 --> 00:42:06,560
Speaker 2: that have a human emotional subtext, like tell me what

641
00:42:06,680 --> 00:42:09,360
Speaker 2: was most surprising about this? It doesn't know what a

642
00:42:09,440 --> 00:42:12,799
Speaker 2: surprise is. Tell me what was most interesting about this?

643
00:42:13,719 --> 00:42:16,440
Speaker 2: It's not able to do that. You really have to

644
00:42:16,600 --> 00:42:18,240
Speaker 2: treat it like it's a dumb machine.

645
00:42:18,520 --> 00:42:20,920
Speaker 3: Well, that's right this way I mean you treat it.

646
00:42:21,200 --> 00:42:25,080
Speaker 3: You need to treat it as an operating system. You

647
00:42:25,160 --> 00:42:28,319
Speaker 3: need to constrain every bit about it, particularly in how

648
00:42:28,400 --> 00:42:30,920
Speaker 3: you allow it to think, because it wants to please

649
00:42:30,960 --> 00:42:33,799
Speaker 3: you so badly it does. So if you ask it

650
00:42:33,840 --> 00:42:36,920
Speaker 3: what's interesting, it will look back at its history of

651
00:42:36,960 --> 00:42:40,800
Speaker 3: communication with you, and I'll think, what will Berry find interesting?

652
00:42:41,520 --> 00:42:43,120
Speaker 3: And it will give that answer to you. And if

653
00:42:43,120 --> 00:42:45,520
Speaker 3: you can't find it easily, it'll make it up. It'll

654
00:42:45,520 --> 00:42:48,000
Speaker 3: make up an answer that you will find interesting.

655
00:42:48,440 --> 00:42:52,920
Speaker 2: I find when I try and prompt, Hey, tell me

656
00:42:52,960 --> 00:42:57,279
Speaker 2: about Ben Hunt's background and give me the timeline of

657
00:42:57,320 --> 00:43:00,879
Speaker 2: his career that it's good at? Hey, what was Ben

658
00:43:00,960 --> 00:43:04,280
Speaker 2: Hunt really good at? It has got no idea.

659
00:43:04,960 --> 00:43:07,920
Speaker 3: So what you're able to do, if you're able to

660
00:43:08,280 --> 00:43:12,880
Speaker 3: again put the genie in the bottle, tell it. How

661
00:43:12,920 --> 00:43:16,120
Speaker 3: to think about a problem is you're able to identify.

662
00:43:17,200 --> 00:43:19,720
Speaker 3: And this does go back to the work from thirty

663
00:43:19,760 --> 00:43:23,840
Speaker 3: five years ago. The type of stories we tell that

664
00:43:23,880 --> 00:43:28,879
Speaker 3: we humans tell about stocks or politics. We tell two

665
00:43:28,920 --> 00:43:35,279
Speaker 3: types of stories. We tell descriptive stories. Oh, the you

666
00:43:35,320 --> 00:43:40,160
Speaker 3: know the Fed cut rates by twenty five basis points today,

667
00:43:40,239 --> 00:43:44,080
Speaker 3: a descriptive story. And we can also tell the description

668
00:43:44,120 --> 00:43:45,160
Speaker 3: of the you know.

669
00:43:45,840 --> 00:43:51,680
Speaker 2: It's the because clause because we're seeing slowing, increasing layoffs

670
00:43:51,719 --> 00:43:52,840
Speaker 2: and slowing consumers.

671
00:43:53,120 --> 00:44:01,040
Speaker 3: And that's high resolution and very descriptive. Right. The Powell

672
00:44:01,120 --> 00:44:07,600
Speaker 3: was surprisingly hawkish today, and he was, that's a description.

673
00:44:08,480 --> 00:44:12,080
Speaker 3: There's another type of story we tell, Barry, and that's prescriptive,

674
00:44:13,360 --> 00:44:19,240
Speaker 3: meaning meaning the FED should be hawkish, the Fed should

675
00:44:19,360 --> 00:44:23,840
Speaker 3: cut by twenty five basis points. Those are the stories

676
00:44:24,400 --> 00:44:29,880
Speaker 3: that are indicative of an effort being made to move

677
00:44:30,040 --> 00:44:35,600
Speaker 3: public opinion in a certain direction. That's like the forward

678
00:44:35,680 --> 00:44:40,480
Speaker 3: guidance that the FED still does right, using their words

679
00:44:40,520 --> 00:44:44,680
Speaker 3: for effect. They're using words to nudge you and how

680
00:44:44,719 --> 00:44:47,000
Speaker 3: you should think about the world. How to think about

681
00:44:47,000 --> 00:44:50,080
Speaker 3: the world. So the crucial thing when we're doing these,

682
00:44:50,640 --> 00:44:54,640
Speaker 3: when we're asking the AI to here's all the text

683
00:44:54,760 --> 00:44:57,879
Speaker 3: in the world. Here are the stories that we want

684
00:44:57,920 --> 00:45:01,160
Speaker 3: you to identify. We can all also boil that down

685
00:45:01,239 --> 00:45:07,920
Speaker 3: into identify the stories that are trying to tell the

686
00:45:07,960 --> 00:45:12,239
Speaker 3: reader how they should think or how policy should go.

687
00:45:12,960 --> 00:45:18,160
Speaker 3: That we find has a lot of predictive capability to it.

688
00:45:18,800 --> 00:45:24,919
Speaker 2: So you're in the business of analyzing the world's narratives

689
00:45:25,000 --> 00:45:25,719
Speaker 2: every day.

690
00:45:26,520 --> 00:45:27,840
Speaker 3: How is that even possible?

691
00:45:27,880 --> 00:45:31,440
Speaker 2: It seems like that is an impressing crazy crazy And.

692
00:45:32,760 --> 00:45:36,719
Speaker 3: It used to be it used to be crazy. I mean, Barry, I,

693
00:45:36,719 --> 00:45:39,760
Speaker 3: I really do remember in the academic days. I would

694
00:45:40,040 --> 00:45:43,279
Speaker 3: literally hire grad students and give them a cup of

695
00:45:43,360 --> 00:45:46,720
Speaker 3: dimes so they go down to the microfiche machine. Remember

696
00:45:46,760 --> 00:45:51,280
Speaker 3: those machines. Remember those machines, Yes, remember those. I would code,

697
00:45:51,400 --> 00:45:56,160
Speaker 3: hand code the or hand record the coded data. I

698
00:45:56,200 --> 00:46:02,440
Speaker 3: would type it into remote access for a digital equipment

699
00:46:02,560 --> 00:46:06,759
Speaker 3: mini frame and the next day, you know, something would

700
00:46:06,840 --> 00:46:11,480
Speaker 3: churn out for me. Today it's they say, we're processing

701
00:46:11,600 --> 00:46:15,440
Speaker 3: hundreds of billions of tokens. We get millions of documents

702
00:46:15,480 --> 00:46:19,680
Speaker 3: over an eight like that. We are infinite computing resources

703
00:46:19,719 --> 00:46:20,440
Speaker 3: available to us.

704
00:46:20,600 --> 00:46:22,880
Speaker 2: Is there anything you can't access that you wish you

705
00:46:22,880 --> 00:46:23,640
Speaker 2: had access to?

706
00:46:26,520 --> 00:46:31,000
Speaker 3: Any data source? So one of the things that's happened

707
00:46:31,040 --> 00:46:34,759
Speaker 3: on Wall Street is that the banks and the cell

708
00:46:34,880 --> 00:46:39,480
Speaker 3: side had become very jealous of their publications because they

709
00:46:39,520 --> 00:46:44,000
Speaker 3: tend to think, and they're wrong, that they're good at it, right,

710
00:46:44,320 --> 00:46:48,719
Speaker 3: that they're good at analysis. I personally don't think they are.

711
00:46:48,920 --> 00:46:51,960
Speaker 2: Well, let's just say some are better than others.

712
00:46:52,000 --> 00:46:55,320
Speaker 3: Some are better than others, but none of them are really.

713
00:46:57,680 --> 00:47:00,799
Speaker 3: If they were, if there was a call it kind

714
00:47:00,800 --> 00:47:05,279
Speaker 3: of significant alpha there, they wouldn't be a bank, that'd

715
00:47:05,280 --> 00:47:06,920
Speaker 3: be a hedge fund. Yeah, they wouldn't be a on

716
00:47:07,400 --> 00:47:11,279
Speaker 3: the cell side now. So I'm interested in reading the

717
00:47:11,320 --> 00:47:13,600
Speaker 3: cell side research not because I think there's some nugget

718
00:47:13,680 --> 00:47:16,560
Speaker 3: of truth in there, but because I want to see

719
00:47:16,600 --> 00:47:17,839
Speaker 3: what they're all talking about.

720
00:47:17,960 --> 00:47:22,759
Speaker 2: Right, it's reflective of, if not a consensus, certainly a

721
00:47:23,000 --> 00:47:25,239
Speaker 2: popular set of ideas.

722
00:47:24,800 --> 00:47:27,279
Speaker 3: And this is a crucial thing to talk about what

723
00:47:28,000 --> 00:47:30,920
Speaker 3: we do. I don't know what the truth is, right,

724
00:47:31,080 --> 00:47:31,800
Speaker 3: I have no idea.

725
00:47:31,840 --> 00:47:32,480
Speaker 2: Does it matter?

726
00:47:32,960 --> 00:47:36,600
Speaker 3: And I don't think it matters. I want to provide

727
00:47:36,680 --> 00:47:39,160
Speaker 3: this information to people who do have a view on

728
00:47:39,200 --> 00:47:42,480
Speaker 3: the truth. Let's say you're a you're a value investor, right,

729
00:47:42,480 --> 00:47:45,080
Speaker 3: you're running a fund. You've got your views, you've done

730
00:47:45,320 --> 00:47:48,000
Speaker 3: your homework, you've done your research.

731
00:47:47,840 --> 00:47:49,000
Speaker 2: You've got a good back test.

732
00:47:49,160 --> 00:47:51,600
Speaker 3: Yeah, yeah, yeah, you've got. So I say, I've got

733
00:47:52,080 --> 00:47:56,080
Speaker 3: Here are the companies where I think I've identified something special,

734
00:47:56,400 --> 00:48:01,520
Speaker 3: something that's valuable, that the market does not recognize, and

735
00:48:01,560 --> 00:48:03,840
Speaker 3: so I want to buy it. And then I'm just

736
00:48:03,880 --> 00:48:09,080
Speaker 3: gonna wait. I gotta wait until one day the market realizes,

737
00:48:09,120 --> 00:48:11,440
Speaker 3: the market comes to their senses and says, oh wow,

738
00:48:12,280 --> 00:48:14,600
Speaker 3: that should trade at a higher multiple or a higher

739
00:48:14,640 --> 00:48:19,200
Speaker 3: price because that special thing that you saw, that source

740
00:48:19,200 --> 00:48:25,319
Speaker 3: of value, the rest of the world comes to see that. Well,

741
00:48:25,520 --> 00:48:28,080
Speaker 3: what I can think I can show you is when

742
00:48:28,120 --> 00:48:29,920
Speaker 3: the rest of the world starts to wake up to

743
00:48:29,960 --> 00:48:31,040
Speaker 3: whatever it is you're looking for.

744
00:48:31,160 --> 00:48:35,200
Speaker 2: So you're catching the early lift off the bottom. I am.

745
00:48:35,280 --> 00:48:40,240
Speaker 3: So when a value investment only works when the market

746
00:48:40,320 --> 00:48:44,440
Speaker 3: recognizes it, and we are tracking when something like that

747
00:48:44,480 --> 00:48:46,000
Speaker 3: gets discovered by the market.

748
00:48:46,880 --> 00:48:50,280
Speaker 2: So before in video is five trillion, when it starts

749
00:48:50,360 --> 00:48:54,520
Speaker 2: ramping up to five hundred billion, Hey, something's going on here.

750
00:48:57,920 --> 00:49:02,239
Speaker 3: What I found in my experience as an investor, is

751
00:49:02,239 --> 00:49:05,200
Speaker 3: that you make the most money on a trade during

752
00:49:05,239 --> 00:49:07,680
Speaker 3: what I call is the discovery phase of a truck,

753
00:49:08,400 --> 00:49:11,800
Speaker 3: when the rest of the world wakes up to something

754
00:49:11,840 --> 00:49:15,960
Speaker 3: that you had noticed and identified before. That's when the

755
00:49:16,040 --> 00:49:17,840
Speaker 3: money is made. Once it gets.

756
00:49:17,640 --> 00:49:23,560
Speaker 2: Out there and it's reflected, then the market is eventually efficient.

757
00:49:23,680 --> 00:49:26,600
Speaker 3: It's a different it's a different risk and reward profile.

758
00:49:26,760 --> 00:49:28,719
Speaker 3: Let me let me put it that way. You you

759
00:49:28,760 --> 00:49:34,280
Speaker 3: get you get a lot more ups and downs post

760
00:49:34,400 --> 00:49:38,200
Speaker 3: that discovery phase than you do when you're enjoying the

761
00:49:38,239 --> 00:49:41,279
Speaker 3: discovery phase. It's a lot harder once you other now

762
00:49:42,080 --> 00:49:45,239
Speaker 3: we miss, oh, the other catalyst, and then say what.

763
00:49:45,400 --> 00:49:48,879
Speaker 2: But that discovery phase is I'm going to quote Doug

764
00:49:49,000 --> 00:49:53,080
Speaker 2: cast that's the there's a phrase he uses.

765
00:49:53,320 --> 00:49:53,880
Speaker 1: Uh.

766
00:49:54,360 --> 00:50:00,279
Speaker 2: It's like a contrariant perspective, a variant perspective perspect if

767
00:50:00,400 --> 00:50:03,759
Speaker 2: you have some insight that is not widely held right

768
00:50:03,960 --> 00:50:09,000
Speaker 2: and the market being mostly kind of sort of eventually efficient.

769
00:50:09,560 --> 00:50:12,760
Speaker 2: If there's a truth or at least a good story

770
00:50:13,160 --> 00:50:17,000
Speaker 2: in your variant perspectives.

771
00:50:15,520 --> 00:50:17,480
Speaker 3: All it takes is a story.

772
00:50:17,480 --> 00:50:18,560
Speaker 2: It's got to be a good story.

773
00:50:18,600 --> 00:50:20,200
Speaker 3: Though, Yeah, it's got to be a well i'll call

774
00:50:20,239 --> 00:50:22,279
Speaker 3: it it's got to be a compelling story, right right,

775
00:50:22,320 --> 00:50:27,399
Speaker 3: And that's that's why you're looking for a CEO who

776
00:50:27,440 --> 00:50:30,920
Speaker 3: can tell that compelling story. You're looking for the ability

777
00:50:30,960 --> 00:50:34,600
Speaker 3: to tell a compelling story, because that is what gives

778
00:50:34,640 --> 00:50:37,080
Speaker 3: you a multiple so is story.

779
00:50:37,280 --> 00:50:41,960
Speaker 2: So let me ask you a few more questions on

780
00:50:42,920 --> 00:50:46,440
Speaker 2: person and before we go into a few other areas. So, first,

781
00:50:47,000 --> 00:50:49,400
Speaker 2: you've been doing this for seven years. What are you

782
00:50:49,440 --> 00:50:54,439
Speaker 2: doing today that was unimaginable four, five, six, seven years ago.

783
00:50:55,000 --> 00:50:58,080
Speaker 3: I tell you something that was unimaginable really two years ago.

784
00:50:58,120 --> 00:51:02,839
Speaker 3: That's amazing, and that is the AI. So what we

785
00:51:02,840 --> 00:51:05,040
Speaker 3: were doing in our early days, we were basically doing

786
00:51:05,120 --> 00:51:08,360
Speaker 3: small language models. We were making the language models essentially

787
00:51:08,360 --> 00:51:12,719
Speaker 3: by hand. And you know, we did our first models

788
00:51:12,760 --> 00:51:14,960
Speaker 3: and our first versions of this, and we licensed it

789
00:51:15,120 --> 00:51:20,440
Speaker 3: some big banks. And here's the problem, Barry. We had

790
00:51:20,480 --> 00:51:24,080
Speaker 3: constructed a net and it was a pretty good net.

791
00:51:24,280 --> 00:51:26,960
Speaker 3: I mean, when we would dip it into a data stream,

792
00:51:27,320 --> 00:51:30,440
Speaker 3: we'd catch a fish, meaning a signal, and the signal

793
00:51:31,080 --> 00:51:34,759
Speaker 3: that signal works, you know, good, good hit on the

794
00:51:35,080 --> 00:51:37,960
Speaker 3: signal that we put up. The problem was we were

795
00:51:37,960 --> 00:51:41,440
Speaker 3: missing too many damn fish. Our net was too small

796
00:51:41,520 --> 00:51:43,759
Speaker 3: and we'd say, oh, we got this one, and then

797
00:51:44,520 --> 00:51:46,960
Speaker 3: we'd look back at whatever it was we were designing

798
00:51:47,000 --> 00:51:49,040
Speaker 3: the models, so, well, how did we miss all these

799
00:51:49,040 --> 00:51:53,320
Speaker 3: other fish? And the answer was the small language model

800
00:51:53,400 --> 00:52:00,279
Speaker 3: we were constructing. It's incredibly complex if you're building these

801
00:52:00,320 --> 00:52:07,839
Speaker 3: models without the probabilistic approach that modern lms allow you

802
00:52:07,920 --> 00:52:10,840
Speaker 3: to take. So this is the whole notion of embedding.

803
00:52:10,960 --> 00:52:14,799
Speaker 3: So that there are a million ways to say I'm

804
00:52:14,840 --> 00:52:18,920
Speaker 3: bullish about the management change at company XYZ. There are

805
00:52:19,040 --> 00:52:21,719
Speaker 3: million ways you can say that. And if you're in

806
00:52:21,760 --> 00:52:25,000
Speaker 3: the business as we were, of kind of handcrafting, let's

807
00:52:25,040 --> 00:52:27,920
Speaker 3: write down all the ways you can say that. Really

808
00:52:28,040 --> 00:52:31,120
Speaker 3: you miss a lot. You've made a small net and

809
00:52:31,880 --> 00:52:35,640
Speaker 3: what you're trying to a small data set, and it's

810
00:52:35,640 --> 00:52:39,759
Speaker 3: a pretty small data set. So we rebuilt all of

811
00:52:39,800 --> 00:52:46,279
Speaker 3: our software and using AI as an operating system, context

812
00:52:46,320 --> 00:52:52,160
Speaker 3: engineering control, how you dole out the text data, how

813
00:52:52,200 --> 00:52:54,360
Speaker 3: you test it, how you allow it to think. And

814
00:52:54,360 --> 00:52:57,120
Speaker 3: we thought, all right, you know what, we're building a

815
00:52:57,120 --> 00:53:01,160
Speaker 3: bigger net. I bet we see five I've maybe even

816
00:53:01,280 --> 00:53:07,440
Speaker 3: six x improvement in our signal to over one hundred x.

817
00:53:07,520 --> 00:53:08,440
Speaker 2: That's unbelievable.

818
00:53:08,480 --> 00:53:10,600
Speaker 3: Over one hundred x it's over and now it's we've

819
00:53:10,600 --> 00:53:14,400
Speaker 3: had a multiple of that again, it's it's It is

820
00:53:14,440 --> 00:53:17,520
Speaker 3: hard to describe in in the in the the expansion

821
00:53:17,560 --> 00:53:21,520
Speaker 3: of the net is in so many different directions. Everything

822
00:53:21,520 --> 00:53:24,600
Speaker 3: we were doing before was just English language, right.

823
00:53:24,760 --> 00:53:27,160
Speaker 2: So now you speak of global and how many languages

824
00:53:27,239 --> 00:53:28,560
Speaker 2: are you pulling into the.

825
00:53:28,719 --> 00:53:33,120
Speaker 3: Anything anything that an AI has been trained on, we

826
00:53:33,239 --> 00:53:35,040
Speaker 3: read it and we can say.

827
00:53:35,080 --> 00:53:37,319
Speaker 2: We can say, got forty languages? Is that like?

828
00:53:37,360 --> 00:53:39,200
Speaker 3: How many? Well there are only you know, there are

829
00:53:39,239 --> 00:53:42,799
Speaker 3: about a dozen languages that are useful in.

830
00:53:43,000 --> 00:53:47,280
Speaker 2: Markets totally Japanese, Chinese, a variety of European languages.

831
00:53:47,320 --> 00:53:50,360
Speaker 3: So we can we can tell Indian, here's the here's

832
00:53:50,400 --> 00:53:55,319
Speaker 3: a story, here's the story about Chinese domestic markets. There's

833
00:53:55,360 --> 00:53:59,160
Speaker 3: a Western story, Western narratives, there's a domestic Chinese story,

834
00:53:59,200 --> 00:54:02,680
Speaker 3: and the domestic Chines story is very different. Oftun it's

835
00:54:02,800 --> 00:54:05,200
Speaker 3: very different. So we were able, for example, picking up

836
00:54:05,480 --> 00:54:09,959
Speaker 3: we were able to pick up way before it got

837
00:54:10,000 --> 00:54:17,200
Speaker 3: picked up in Western press. The demand for luxury goods

838
00:54:17,280 --> 00:54:23,279
Speaker 3: in China went off a clifliff in last November.

839
00:54:23,800 --> 00:54:26,680
Speaker 2: Listen, you can't go from a double digit GDP to

840
00:54:26,960 --> 00:54:29,960
Speaker 2: like load to mid single digits and not have an effect,

841
00:54:30,400 --> 00:54:31,640
Speaker 2: especially in a country like that.

842
00:54:31,680 --> 00:54:35,799
Speaker 3: Well, but it's interesting, right because because they've had you know,

843
00:54:36,000 --> 00:54:38,640
Speaker 3: ups and downs on business cycle before, but this is

844
00:54:38,719 --> 00:54:43,240
Speaker 3: the first time where you saw consumer behavior that really

845
00:54:43,320 --> 00:54:45,040
Speaker 3: was kind of similar to what you might find in

846
00:54:45,080 --> 00:54:48,480
Speaker 3: a Western consumer behavior. Point being, you didn't hear that

847
00:54:48,560 --> 00:54:53,960
Speaker 3: from LVMH or the Macau gaming guys until February, and

848
00:54:54,000 --> 00:54:57,400
Speaker 3: we were picking that up in November from the domestic

849
00:54:57,800 --> 00:54:58,560
Speaker 3: Chinese meatus.

850
00:54:58,760 --> 00:55:02,440
Speaker 2: So that raises a really interesting question. You know, at

851
00:55:02,440 --> 00:55:04,560
Speaker 2: the end of the day, clients want to be able

852
00:55:04,600 --> 00:55:09,200
Speaker 2: to make money on your research. How are people putting

853
00:55:09,239 --> 00:55:13,359
Speaker 2: this to work? How does what you're building help your

854
00:55:13,360 --> 00:55:14,920
Speaker 2: clients generate alpha.

855
00:55:15,600 --> 00:55:21,360
Speaker 3: What we think we've discovered is an entirely new source

856
00:55:21,520 --> 00:55:25,160
Speaker 3: of data, and what we're confident is that we've built

857
00:55:25,239 --> 00:55:28,600
Speaker 3: the systems that are very different from what you see

858
00:55:28,600 --> 00:55:32,200
Speaker 3: with this sort of analytics that are out there from

859
00:55:32,200 --> 00:55:35,239
Speaker 3: anyone else. Right, So this is not sentiment, right, We're

860
00:55:35,239 --> 00:55:37,120
Speaker 3: not tracking or using mean words or nice.

861
00:55:37,040 --> 00:55:38,440
Speaker 2: By the way, that was a big thing I don't

862
00:55:38,440 --> 00:55:42,680
Speaker 2: know is that ten years ago we're scanning Twitter to

863
00:55:42,760 --> 00:55:44,160
Speaker 2: identify investor sentiment.

864
00:55:44,280 --> 00:55:46,200
Speaker 3: Oh my god, I mean, but I was just looking

865
00:55:46,239 --> 00:55:48,880
Speaker 3: today and not to pick on Bloomberg, but you know

866
00:55:50,120 --> 00:55:55,960
Speaker 3: it was their live coverage of the FED statement, right,

867
00:55:56,000 --> 00:56:00,720
Speaker 3: and they were analyzing the sentences and the thing statement

868
00:56:00,840 --> 00:56:02,560
Speaker 3: for hawkish and dubvish sentiment.

869
00:56:02,680 --> 00:56:06,080
Speaker 2: How would change from the last meeting, And so so

870
00:56:06,120 --> 00:56:08,000
Speaker 2: you're saying there's not a whole lot of signal there.

871
00:56:09,239 --> 00:56:10,960
Speaker 3: I want to be careful with what I say, right,

872
00:56:11,080 --> 00:56:15,759
Speaker 3: which is that there are a number of i'll call

873
00:56:15,760 --> 00:56:20,400
Speaker 3: them high frequency stat ARP guys where if it's important

874
00:56:20,400 --> 00:56:25,279
Speaker 3: for you to note the difference in word choice on

875
00:56:25,480 --> 00:56:29,400
Speaker 3: a millisecond level, I think there. I think you can

876
00:56:29,400 --> 00:56:33,200
Speaker 3: get something out of that I do, and so there

877
00:56:33,200 --> 00:56:35,680
Speaker 3: are firms that can do that very well. That's not

878
00:56:35,760 --> 00:56:39,560
Speaker 3: our game, right, This is not the What we're trying

879
00:56:39,600 --> 00:56:42,880
Speaker 3: to identify is not just sentiment, not just word choice.

880
00:56:43,360 --> 00:56:46,160
Speaker 3: This is not Google trends and how many times did

881
00:56:46,960 --> 00:56:51,759
Speaker 3: they mention AI and the earnings report? Right, there's we're

882
00:56:51,800 --> 00:56:54,680
Speaker 3: able to track the actual stories that drive behavior.

883
00:56:55,040 --> 00:56:59,120
Speaker 2: So that that's the next question. And I'm gonna give

884
00:56:59,200 --> 00:57:04,520
Speaker 2: Dave not a credit for asking this. Does every narrative

885
00:57:04,760 --> 00:57:08,080
Speaker 2: turn into a decision? How do you know when something

886
00:57:08,239 --> 00:57:13,840
Speaker 2: is merely noisy versus where there's a significant tradable signal.

887
00:57:14,840 --> 00:57:20,560
Speaker 3: I don't, right, So this is my goal is not to, oh,

888
00:57:21,040 --> 00:57:25,120
Speaker 3: you know, do the trades. My goal is mister hedge

889
00:57:25,120 --> 00:57:31,400
Speaker 3: fun guy, mister asset allocator. You've got to you know China, right,

890
00:57:31,520 --> 00:57:36,600
Speaker 3: you know your companies, you know your commodities. Here is

891
00:57:37,560 --> 00:57:44,400
Speaker 3: data that I think you'll find useful. The efficacy of it, though,

892
00:57:44,480 --> 00:57:45,919
Speaker 3: is up to you. I don't know what you would

893
00:57:46,000 --> 00:57:47,920
Speaker 3: what you want to do with it. I want to

894
00:57:47,960 --> 00:57:52,800
Speaker 3: sell the picks and shovels, honestly for this vast new

895
00:57:52,880 --> 00:57:55,240
Speaker 3: data set that we all know is important, but we

896
00:57:55,360 --> 00:57:58,320
Speaker 3: haven't been able to measure it in a very predictive

897
00:57:58,360 --> 00:57:58,960
Speaker 3: way before.

898
00:57:59,280 --> 00:58:02,800
Speaker 2: So let's talk about a few things related to and

899
00:58:02,800 --> 00:58:03,360
Speaker 2: it's not just.

900
00:58:03,320 --> 00:58:07,240
Speaker 3: About using this for investment. So we have a product

901
00:58:07,280 --> 00:58:13,120
Speaker 3: for financial advisors. Uh huh, right, which is your client's

902
00:58:13,160 --> 00:58:17,720
Speaker 3: coming in? You've got a portfolio. You need to be

903
00:58:17,760 --> 00:58:21,520
Speaker 3: able to say, what what is what is my client? What?

904
00:58:21,520 --> 00:58:23,640
Speaker 3: What are they worried about, what are they nervous about?

905
00:58:23,640 --> 00:58:26,320
Speaker 3: What are they hopeful for? What are the stories they're reading?

906
00:58:26,360 --> 00:58:29,280
Speaker 2: And you're pulling this out of the flow of media, and.

907
00:58:29,240 --> 00:58:32,040
Speaker 3: We can tell you exactly for the portfolio you've got

908
00:58:32,040 --> 00:58:34,480
Speaker 3: for that client. Here's what they're going to be asking about,

909
00:58:35,280 --> 00:58:37,720
Speaker 3: worried about, and here are the answers you can give

910
00:58:37,760 --> 00:58:40,120
Speaker 3: them to show. This is when has happened before, when

911
00:58:40,120 --> 00:58:42,840
Speaker 3: this story has come up, Stay, the course is going

912
00:58:42,880 --> 00:58:45,160
Speaker 3: to be fine. This is the sort of stuff we

913
00:58:45,160 --> 00:58:46,520
Speaker 3: can do for financial advisors.

914
00:58:46,640 --> 00:58:48,880
Speaker 2: Huh, that's really interesting. Would I would love to see

915
00:58:48,880 --> 00:58:49,200
Speaker 2: some of that.

916
00:58:49,400 --> 00:58:52,240
Speaker 3: It's not just in markets, Berry, I got to tell you.

917
00:58:52,160 --> 00:58:56,840
Speaker 2: The policymakers, corporate executivesinkers.

918
00:58:57,040 --> 00:59:01,160
Speaker 3: So we did you know, we did before in the

919
00:59:01,280 --> 00:59:04,520
Speaker 3: run up to the Russian invasion of Ukraine, and we

920
00:59:04,560 --> 00:59:07,240
Speaker 3: published this on epsolon theory and this is my old

921
00:59:07,280 --> 00:59:11,400
Speaker 3: academic work book, getting to War. Before a country starts

922
00:59:11,400 --> 00:59:15,400
Speaker 3: a war, they mobilize public opinion. And so we were

923
00:59:15,400 --> 00:59:20,640
Speaker 3: looking at domestic Russian media looking as a friends, this

924
00:59:20,680 --> 00:59:21,560
Speaker 3: isn't going to be.

925
00:59:21,520 --> 00:59:23,200
Speaker 2: A limited thing.

926
00:59:23,320 --> 00:59:25,800
Speaker 3: This is happening. It's going to be a full scale invasion.

927
00:59:26,200 --> 00:59:29,720
Speaker 3: Because that was the messaging in domestic Russian media to

928
00:59:30,120 --> 00:59:36,439
Speaker 3: the Russian people. Wow, So brands, right, so you are

929
00:59:36,680 --> 00:59:40,000
Speaker 3: you know we're working with some some it sounds great,

930
00:59:40,080 --> 00:59:43,520
Speaker 3: but pro sports teams you want to tell a story

931
00:59:44,120 --> 00:59:49,720
Speaker 3: to build a stadium, some tickets build a stadium. Give

932
00:59:49,720 --> 00:59:52,320
Speaker 3: me some examples of people have told good stories and

933
00:59:52,400 --> 00:59:54,840
Speaker 3: how did that work for them? How can we do

934
00:59:54,880 --> 00:59:58,280
Speaker 3: the same thing? So looking at storytelling.

935
00:59:58,640 --> 01:00:02,000
Speaker 2: If Jeff Bezos were as a subscriber to this back

936
01:00:02,000 --> 01:00:06,440
Speaker 2: when he was trying to build a tax funded HQ

937
01:00:06,960 --> 01:00:10,440
Speaker 2: on the Hudson, had he had your data and you

938
01:00:10,480 --> 01:00:14,120
Speaker 2: were crunching all the New York City news stories about this,

939
01:00:14,400 --> 01:00:17,440
Speaker 2: might you have been able to give him advice that

940
01:00:18,560 --> 01:00:22,400
Speaker 2: he suffered. He suffered such a backlash.

941
01:00:21,760 --> 01:00:25,160
Speaker 3: Because you know we have today, we have polls right because.

942
01:00:24,920 --> 01:00:27,160
Speaker 2: We're terrible, which are terrible, which.

943
01:00:27,000 --> 01:00:30,200
Speaker 3: Are mostly terrible now, you know, and I love the

944
01:00:30,240 --> 01:00:32,240
Speaker 3: polymarket stuff and other things where you try to get

945
01:00:32,240 --> 01:00:34,640
Speaker 3: as many people as possible to put money on something right.

946
01:00:34,720 --> 01:00:37,919
Speaker 2: So you know, when you ask a personal question, you're

947
01:00:37,960 --> 01:00:40,800
Speaker 2: asking them, hey, what do you think you think and

948
01:00:40,840 --> 01:00:42,560
Speaker 2: what do you think you're going to do in the future.

949
01:00:42,880 --> 01:00:45,320
Speaker 2: When we know people are terrible at both of them terrible,

950
01:00:46,040 --> 01:00:47,680
Speaker 2: But if you put a little money on it, all right,

951
01:00:47,760 --> 01:00:50,680
Speaker 2: maybe they might be a little more circumspect.

952
01:00:50,800 --> 01:00:53,280
Speaker 3: That that that that helps the ton So in places

953
01:00:53,320 --> 01:00:55,840
Speaker 3: where you can kin make a bet. I think that

954
01:00:55,960 --> 01:00:59,080
Speaker 3: improves the kind of information you can get. It still

955
01:00:59,160 --> 01:01:02,000
Speaker 3: lends itself to a lot of manipulation and a lot of.

956
01:01:04,160 --> 01:01:08,240
Speaker 2: Poly markets and calshi and all those things. It's not

957
01:01:08,320 --> 01:01:11,640
Speaker 2: the bond market. It's not one hundred and something trillion dollars.

958
01:01:12,080 --> 01:01:13,880
Speaker 2: It's a couple of bucks on each of these, and

959
01:01:13,920 --> 01:01:16,960
Speaker 2: sometimes special million dollars move them. Half a million dollars

960
01:01:17,000 --> 01:01:17,600
Speaker 2: can move the market.

961
01:01:17,680 --> 01:01:19,160
Speaker 3: But for a lot of things. Let's say you're polling

962
01:01:19,200 --> 01:01:22,040
Speaker 3: for a political candidate, right, you can't ask them to

963
01:01:22,040 --> 01:01:25,480
Speaker 3: put money down on something, but you really want to

964
01:01:25,520 --> 01:01:29,960
Speaker 3: know these these policies that my candidate is thinking about

965
01:01:30,200 --> 01:01:33,920
Speaker 3: taking on. Is that popular, does it resonate? Is it

966
01:01:33,920 --> 01:01:37,920
Speaker 3: a compelling story? We can absolutely see if that is

967
01:01:38,000 --> 01:01:42,000
Speaker 3: true by looking at local media, local social media, all

968
01:01:42,040 --> 01:01:45,840
Speaker 3: of that. So it's it's pretty wild bare. I mean,

969
01:01:45,880 --> 01:01:50,720
Speaker 3: once you start looking at how important stories are, and

970
01:01:50,760 --> 01:01:53,000
Speaker 3: once you've got a tool where you can actually measure

971
01:01:53,040 --> 01:01:57,440
Speaker 3: them and visualize them, it's like it's like, I feel

972
01:01:57,440 --> 01:02:00,520
Speaker 3: like it was like when they invented whoever was oven hook,

973
01:02:00,560 --> 01:02:04,040
Speaker 3: whoever vinded the microscope, When you're actually to see something

974
01:02:04,080 --> 01:02:06,400
Speaker 3: that we all know is there, Well, now we do.

975
01:02:06,520 --> 01:02:09,880
Speaker 2: Back then, remember jump theory took what a century to

976
01:02:09,920 --> 01:02:10,280
Speaker 2: catch up?

977
01:02:10,480 --> 01:02:12,560
Speaker 3: It took a century, And it takes it takes seeing

978
01:02:12,600 --> 01:02:15,240
Speaker 3: it right, It takes the instrument to actually measure it

979
01:02:15,280 --> 01:02:19,320
Speaker 3: before you actually believe in it. And so I feel

980
01:02:19,320 --> 01:02:21,080
Speaker 3: like that's kind of where we are right now. It's

981
01:02:21,120 --> 01:02:24,320
Speaker 3: these early days. But to actually see and measure the

982
01:02:24,360 --> 01:02:28,960
Speaker 3: storytelling at this level of resolution, this magnification, it's pretty

983
01:02:28,960 --> 01:02:29,600
Speaker 3: freaking cool.

984
01:02:29,720 --> 01:02:32,400
Speaker 2: So so we're having this conversation with market at all

985
01:02:32,480 --> 01:02:37,240
Speaker 2: time highs and you've written about the ravine. Yes, tell

986
01:02:37,320 --> 01:02:40,040
Speaker 2: us a little bit about what is the ravine? How

987
01:02:40,080 --> 01:02:46,120
Speaker 2: does does your data identify that? Tell us what this means?

988
01:02:46,840 --> 01:02:50,120
Speaker 3: Well, there's there's there's clearly been a change in policy

989
01:02:50,320 --> 01:02:53,680
Speaker 3: regime out of Washington and.

990
01:02:54,120 --> 01:02:56,680
Speaker 2: New administration and are radically different.

991
01:02:56,480 --> 01:02:57,720
Speaker 3: Radically different.

992
01:02:57,640 --> 01:03:00,160
Speaker 2: Even from the first Trump president.

993
01:03:00,160 --> 01:03:04,880
Speaker 3: From the first presidency. And so what we're able to measure,

994
01:03:05,240 --> 01:03:09,840
Speaker 3: really measure is how does that i'll say play, But

995
01:03:09,960 --> 01:03:17,760
Speaker 3: also what comes narratives never happened in a vacuum. There's

996
01:03:17,800 --> 01:03:21,400
Speaker 3: always a counter story. For every bull story, there's a

997
01:03:21,400 --> 01:03:25,600
Speaker 3: bear story, and they they they you can almost kind

998
01:03:25,600 --> 01:03:29,360
Speaker 3: of see the battlefield of ideas, the battlefield of stories

999
01:03:29,440 --> 01:03:35,880
Speaker 3: and how it emerges. So my strong sense, Parry, is

1000
01:03:35,920 --> 01:03:41,680
Speaker 3: that we are going towards politically in this country, towards

1001
01:03:43,080 --> 01:03:48,160
Speaker 3: trench warfare greater and greater. I'll call it narrative violence.

1002
01:03:50,160 --> 01:03:53,600
Speaker 3: And it if you look historically how that plays out

1003
01:03:53,640 --> 01:03:56,800
Speaker 3: in countries, it doesn't It doesn't play out well, the.

1004
01:03:56,880 --> 01:04:00,000
Speaker 2: Civil war, domestic political violence, things like that.

1005
01:04:01,280 --> 01:04:06,520
Speaker 3: Sadly, yes, exactly like that, exactly like that. So there's

1006
01:04:06,680 --> 01:04:09,640
Speaker 3: that element, and so that's a that's a sad one

1007
01:04:10,000 --> 01:04:12,640
Speaker 3: or a very troubling one and trying to well, how

1008
01:04:12,640 --> 01:04:16,800
Speaker 3: do we navigate that? But even when in markets, and

1009
01:04:16,840 --> 01:04:20,320
Speaker 3: I alluded to this earlier, how capital markets have become

1010
01:04:20,360 --> 01:04:27,160
Speaker 3: a political utility, uh, and the role of markets in

1011
01:04:27,200 --> 01:04:30,600
Speaker 3: our society, you can really see how that changes in

1012
01:04:30,680 --> 01:04:34,480
Speaker 3: the stories we tell ourselves about the role of markets,

1013
01:04:34,520 --> 01:04:35,680
Speaker 3: what it's there for.

1014
01:04:37,680 --> 01:04:40,600
Speaker 2: And I end of day options in speculation, what else

1015
01:04:40,680 --> 01:04:41,720
Speaker 2: is it supposed to be there for?

1016
01:04:44,160 --> 01:04:47,240
Speaker 3: But that but that's kind of what I'm getting at, right.

1017
01:04:47,720 --> 01:04:52,520
Speaker 3: You can see an enormous change in the meaning of markets,

1018
01:04:53,240 --> 01:04:57,080
Speaker 3: and it connects with a yes, they'll call it the

1019
01:04:57,120 --> 01:05:02,080
Speaker 3: speculation layer. But it also connects with financial needs. Yolo

1020
01:05:04,240 --> 01:05:12,080
Speaker 3: it it leads to a very I think, less attractive

1021
01:05:12,200 --> 01:05:17,640
Speaker 3: future for how we think about money and the role

1022
01:05:17,720 --> 01:05:20,600
Speaker 3: of markets and the role of capitalism.

1023
01:05:20,640 --> 01:05:25,080
Speaker 2: Coming up, we continue our conversation with Ben Hunt, president

1024
01:05:25,280 --> 01:05:30,240
Speaker 2: and co founder of person and discussing how money managers

1025
01:05:30,720 --> 01:05:35,560
Speaker 2: use their output to generate alpha. I'm Barry Ridults. You're

1026
01:05:35,600 --> 01:05:57,520
Speaker 2: listening to Masters in Business on Bloomberg Radio. I'm Barry Ridults.

1027
01:05:57,560 --> 01:06:00,640
Speaker 2: You're listening to Masters in Business on Bloomberg Radio. My

1028
01:06:00,840 --> 01:06:04,360
Speaker 2: extra special guest today is Ben Hunt. He is president

1029
01:06:04,440 --> 01:06:11,480
Speaker 2: and co founder of Percient, a data analytics, narrative storytelling

1030
01:06:12,160 --> 01:06:17,240
Speaker 2: large language model using artificial intelligence to find some signal

1031
01:06:17,760 --> 01:06:23,160
Speaker 2: amongst the noise. Firm their clients range everything from large

1032
01:06:23,880 --> 01:06:28,720
Speaker 2: money managers to hedge funds, to academics and corporate America.

1033
01:06:29,040 --> 01:06:33,240
Speaker 2: So you write epsilon theory, and some of it is

1034
01:06:33,400 --> 01:06:37,120
Speaker 2: for subscribers, some of it is public. One of the

1035
01:06:37,120 --> 01:06:43,560
Speaker 2: things you wrote is absolutely my favorite item from the

1036
01:06:43,600 --> 01:06:48,200
Speaker 2: first few months of the Trump presidency because I like

1037
01:06:48,240 --> 01:06:50,680
Speaker 2: to talk in probabilities because I don't know what's going

1038
01:06:50,760 --> 01:06:53,920
Speaker 2: to happen. But here's the best case scenario, here's the

1039
01:06:53,960 --> 01:06:57,800
Speaker 2: worst case here's all the middle scenarios. You wrote a

1040
01:06:57,840 --> 01:07:02,000
Speaker 2: piece the end of Paksamariana that I thought was the

1041
01:07:02,000 --> 01:07:08,120
Speaker 2: most cogent, imaginative, well thought out. Hey, here's the worst

1042
01:07:08,120 --> 01:07:15,440
Speaker 2: case scenario, and we are becoming dangerously flirting with this

1043
01:07:15,560 --> 01:07:18,720
Speaker 2: possible outcome, and I use that as all right, So

1044
01:07:18,760 --> 01:07:21,720
Speaker 2: here's what I think is high probability. Here's the best case.

1045
01:07:22,120 --> 01:07:24,480
Speaker 2: But if you really want to think about how this

1046
01:07:24,560 --> 01:07:27,280
Speaker 2: can go off the rails, check out what Ben wrote.

1047
01:07:27,680 --> 01:07:30,800
Speaker 2: And so tell us a little bit about how did

1048
01:07:30,960 --> 01:07:37,400
Speaker 2: Persian inform the end of Pax Americana, Because I get

1049
01:07:37,440 --> 01:07:42,720
Speaker 2: the sense that domestically that wasn't really where a how

1050
01:07:42,720 --> 01:07:44,920
Speaker 2: a lot of people were thinking, but I got the

1051
01:07:44,960 --> 01:07:48,520
Speaker 2: sense from overseas that was a much more common thought.

1052
01:07:48,760 --> 01:07:51,040
Speaker 2: Tell us a little bit about both the peace and

1053
01:07:51,440 --> 01:07:53,680
Speaker 2: how the data informed it.

1054
01:07:55,360 --> 01:07:58,680
Speaker 3: I'll start with I'll call it the international dimension. I

1055
01:07:58,680 --> 01:08:00,880
Speaker 3: want to come back to the domestics, and I think

1056
01:08:00,880 --> 01:08:06,120
Speaker 3: we've got some really interesting data recently to share. It

1057
01:08:06,160 --> 01:08:09,400
Speaker 3: goes back to this again. I cringe whiz I talk about,

1058
01:08:09,440 --> 01:08:12,280
Speaker 3: but but it is. It is game theory, right, and

1059
01:08:12,400 --> 01:08:17,640
Speaker 3: all game theory is is strategic interaction. It's that the

1060
01:08:18,160 --> 01:08:22,200
Speaker 3: United States, Yes, is the most powerful player on the

1061
01:08:22,240 --> 01:08:28,240
Speaker 3: world stage. But every country has some degrees of freedom

1062
01:08:28,280 --> 01:08:31,320
Speaker 3: and some autonomy in there in the policies that they

1063
01:08:31,360 --> 01:08:35,439
Speaker 3: that they implement, there is not a dominant strategy, meaning

1064
01:08:35,920 --> 01:08:39,880
Speaker 3: an outcome, an equilibrium outcome. Again, I hate using these words,

1065
01:08:39,920 --> 01:08:42,880
Speaker 3: but I'll use them anyway. All an equilibrium means is

1066
01:08:43,240 --> 01:08:47,920
Speaker 3: it's a balancing point where both parties, let's call them

1067
01:08:48,360 --> 01:08:52,880
Speaker 3: the United States and a European country, where do they

1068
01:08:53,000 --> 01:08:54,760
Speaker 3: end up where they all say, okay, I can live

1069
01:08:54,800 --> 01:08:57,840
Speaker 3: with this, Where I can live with this, and the

1070
01:08:57,920 --> 01:09:00,720
Speaker 3: America first set of policies them is all these set

1071
01:09:00,760 --> 01:09:07,440
Speaker 3: of policies. What you end up with is less potential,

1072
01:09:07,880 --> 01:09:13,200
Speaker 3: economic growth, trade, all all of these things.

1073
01:09:12,800 --> 01:09:15,160
Speaker 2: All the things that we enjoyed post World War two.

1074
01:09:15,760 --> 01:09:19,559
Speaker 2: That realignment that im due to our benefits so.

1075
01:09:19,520 --> 01:09:25,240
Speaker 3: Great enormously advantage, enormously to our advantage. It goes by

1076
01:09:25,240 --> 01:09:28,879
Speaker 3: the name of soft power. And you know the dollar system,

1077
01:09:29,080 --> 01:09:30,080
Speaker 3: the breton Wood system.

1078
01:09:30,160 --> 01:09:32,760
Speaker 2: All of this serve currency on and on the.

1079
01:09:32,720 --> 01:09:35,680
Speaker 3: Reserve currency to finance our definity. All of this is

1080
01:09:35,880 --> 01:09:39,360
Speaker 3: enormously to our advantage. And yes, there are free riders

1081
01:09:39,400 --> 01:09:42,920
Speaker 3: on that system. Yes there are costs to that, particularly

1082
01:09:42,920 --> 01:09:46,360
Speaker 3: on defense and some other some other areas. Tariffs are

1083
01:09:46,360 --> 01:09:49,439
Speaker 3: another area where there's absolutely free riders on that. So

1084
01:09:49,720 --> 01:09:52,759
Speaker 3: there's improvement that can be made in that system, for sure,

1085
01:09:53,720 --> 01:09:57,479
Speaker 3: But this is not This is a different system. This

1086
01:09:57,600 --> 01:09:59,559
Speaker 3: is a different set of rules of the road, and

1087
01:09:59,600 --> 01:10:03,679
Speaker 3: it leads to a different strategic interaction between between countries.

1088
01:10:04,280 --> 01:10:08,719
Speaker 3: So that's what the note was about, and it's it's

1089
01:10:09,600 --> 01:10:11,960
Speaker 3: I'm not trying to predict. I'm just I'm just trying

1090
01:10:11,960 --> 01:10:16,400
Speaker 3: to observe, which is a great line actually by George Sorows,

1091
01:10:16,400 --> 01:10:19,880
Speaker 3: which is say, I'm not predicting, I'm observing, which I

1092
01:10:19,960 --> 01:10:23,960
Speaker 3: love that as a line. I'm not and our technology

1093
01:10:24,000 --> 01:10:26,559
Speaker 3: is not there to try to predict the future, is

1094
01:10:26,560 --> 01:10:29,439
Speaker 3: trying to tell you what is true in the present today.

1095
01:10:31,040 --> 01:10:33,640
Speaker 2: I'm amazed how many people think they can forecast the

1096
01:10:33,640 --> 01:10:35,839
Speaker 2: future when they have no idea what's happening.

1097
01:10:35,960 --> 01:10:38,080
Speaker 3: I want to I want to now cast is what

1098
01:10:38,120 --> 01:10:40,519
Speaker 3: I want to do, not not forecast. I want to

1099
01:10:40,560 --> 01:10:44,360
Speaker 3: now cast. And I guess what I wanted to talk

1100
01:10:44,400 --> 01:10:51,360
Speaker 3: about domestic. So we started tracking the different narratives around

1101
01:10:51,520 --> 01:10:57,559
Speaker 3: immigration policy early last year, and what you saw in

1102
01:10:57,600 --> 01:10:58,759
Speaker 3: some of the kind of simple.

1103
01:10:59,320 --> 01:11:01,480
Speaker 2: This is during the run up to the election.

1104
01:11:01,360 --> 01:11:04,760
Speaker 3: Yes, exactly, and what you saw in really going back

1105
01:11:04,840 --> 01:11:06,760
Speaker 3: because we can take we take this stuff back for

1106
01:11:06,880 --> 01:11:12,000
Speaker 3: a decade or more. And once you absolutely saw up

1107
01:11:12,040 --> 01:11:15,400
Speaker 3: to last October. There's a there was an event from

1108
01:11:15,520 --> 01:11:18,559
Speaker 3: last October. Not the election, but an event from last October.

1109
01:11:19,120 --> 01:11:24,280
Speaker 3: You see a steady increase in regardless of your political affiliation.

1110
01:11:25,040 --> 01:11:29,640
Speaker 3: You know what immigration isn't working for us, isn't as Americans,

1111
01:11:29,920 --> 01:11:37,679
Speaker 3: So a steady increase starting with the they're eating the cats.

1112
01:11:37,680 --> 01:11:40,400
Speaker 2: They're eating the cats, They're eating their dogs, they're eating.

1113
01:11:40,160 --> 01:11:45,720
Speaker 3: The dogs, the Columbus mode when you that that moment.

1114
01:11:45,680 --> 01:11:49,759
Speaker 2: One of the most that's the real moments in debate history.

1115
01:11:49,840 --> 01:11:55,280
Speaker 3: And we see very clearly in our media data, and

1116
01:11:55,320 --> 01:11:58,360
Speaker 3: this is this is not mainstream media. This is everything

1117
01:11:58,720 --> 01:12:04,880
Speaker 3: that we're pulling up. We've seen a really significant decline

1118
01:12:05,160 --> 01:12:09,600
Speaker 3: in the volume and density of oh my god, immigration

1119
01:12:09,760 --> 01:12:13,160
Speaker 3: is a problem, we need mass deportations. On the contrary,

1120
01:12:13,240 --> 01:12:17,320
Speaker 3: we've seen an enormous increase again regardless of political affiliation,

1121
01:12:17,439 --> 01:12:22,400
Speaker 3: including Republicans. No immigration is a good thing for this country.

1122
01:12:23,560 --> 01:12:27,879
Speaker 3: The stories of America that are pro immigrant in immigration

1123
01:12:28,240 --> 01:12:31,120
Speaker 3: now you would not believe that if you were looking

1124
01:12:31,160 --> 01:12:35,040
Speaker 3: at the policies that the White House has implemented during

1125
01:12:35,080 --> 01:12:38,080
Speaker 3: these first you know, ten months of the administration.

1126
01:12:38,400 --> 01:12:45,280
Speaker 2: So wait, when did you first notice that, regardless of policy,

1127
01:12:45,320 --> 01:12:48,120
Speaker 2: we think immigration is a net benefit to America.

1128
01:12:48,160 --> 01:12:50,560
Speaker 3: When did that first start showing up? It started changing

1129
01:12:51,040 --> 01:12:54,320
Speaker 3: right in October because it was like a breach too far.

1130
01:12:54,439 --> 01:12:57,240
Speaker 3: It was like, this is just stupid. This is stilly

1131
01:12:57,280 --> 01:13:01,000
Speaker 3: and stupid, and it's been a steady reece, you know,

1132
01:13:01,080 --> 01:13:03,280
Speaker 3: and you wouldn't you wouldn't believe that if you were

1133
01:13:03,760 --> 01:13:08,960
Speaker 3: immersed in Twitter. But so this country is actually quite

1134
01:13:10,439 --> 01:13:11,400
Speaker 3: pro immigrant.

1135
01:13:12,600 --> 01:13:15,560
Speaker 2: I know that sounds but we are a nation of immigrants.

1136
01:13:16,000 --> 01:13:19,280
Speaker 2: Here here's another data point that's kind of mind blowing.

1137
01:13:19,320 --> 01:13:23,000
Speaker 2: We were talking about a different data point in the

1138
01:13:23,200 --> 01:13:27,080
Speaker 2: entirety of the US history. Twenty twenty five looks like

1139
01:13:27,200 --> 01:13:31,400
Speaker 2: the first year where the US population will decrease.

1140
01:13:32,040 --> 01:13:33,840
Speaker 3: So it's a decrease.

1141
01:13:33,360 --> 01:13:38,000
Speaker 2: In legal immigrants, not not only illegal immigrants, but legal immigrants.

1142
01:13:38,200 --> 01:13:41,320
Speaker 2: Add that with the deportations and just people staying away.

1143
01:13:41,479 --> 01:13:45,200
Speaker 3: Well, this gets back to the economic picture and from others.

1144
01:13:45,240 --> 01:13:48,600
Speaker 3: So what you've had is a clear and again we

1145
01:13:49,120 --> 01:13:52,479
Speaker 3: see this in our data from other countries, there's a

1146
01:13:52,560 --> 01:13:57,799
Speaker 3: clear effort to repatriate assets and funds away from the US.

1147
01:13:58,280 --> 01:14:01,800
Speaker 3: There's been a clear effort outside the US. You see

1148
01:14:01,920 --> 01:14:04,719
Speaker 3: this was central banks, but also with course hence gold,

1149
01:14:04,840 --> 01:14:10,360
Speaker 3: the gold gold treasuries used to be a safe haven asset.

1150
01:14:11,280 --> 01:14:13,439
Speaker 3: No more. That's that's it's if.

1151
01:14:13,439 --> 01:14:18,120
Speaker 2: You understand how significant that charge is that you're making.

1152
01:14:18,560 --> 01:14:23,560
Speaker 2: You're basically accusing the president well of submarining one of

1153
01:14:23,600 --> 01:14:25,639
Speaker 2: the single greatest assets America.

1154
01:14:25,680 --> 01:14:29,280
Speaker 3: Halts well that it's a it is built up to

1155
01:14:29,320 --> 01:14:31,880
Speaker 3: such an extent, right that the way I think of

1156
01:14:31,960 --> 01:14:35,679
Speaker 3: this is an iceberg that is melting. But it is melting.

1157
01:14:36,200 --> 01:14:40,519
Speaker 3: What we do not see we do not see capital flight. Right.

1158
01:14:40,560 --> 01:14:43,240
Speaker 2: We saw it for like a week in April, and

1159
01:14:43,240 --> 01:14:44,559
Speaker 2: then that we reversed.

1160
01:14:44,560 --> 01:14:46,639
Speaker 3: That was it. So there's there. There is there's no

1161
01:14:46,680 --> 01:14:50,599
Speaker 3: capital flight there, there are no leaving the country. There

1162
01:14:50,600 --> 01:14:52,679
Speaker 3: were fierce and we can track it. If that starts

1163
01:14:52,680 --> 01:14:57,600
Speaker 3: to happen, we'll see it immediately. That's not happening. Repatriation

1164
01:14:57,760 --> 01:15:03,840
Speaker 3: continues to happen mm hmm, slowly measured balanced melting iceberg. Right,

1165
01:15:03,880 --> 01:15:07,960
Speaker 3: Because there are limits to if you're a Bermuda reinsure.

1166
01:15:08,040 --> 01:15:10,920
Speaker 3: I mean you're kind of you you're stuck with treasury, right.

1167
01:15:10,920 --> 01:15:13,840
Speaker 2: I mean that you could buy some gold to offset it,

1168
01:15:13,920 --> 01:15:16,400
Speaker 2: but you can't sell one hundred billion dollars worth of treasuries.

1169
01:15:16,600 --> 01:15:18,920
Speaker 3: You can't. You can't. So it's a melting iceberg. But

1170
01:15:19,120 --> 01:15:22,599
Speaker 3: we can clearly see that the dog is not barking

1171
01:15:22,920 --> 01:15:27,240
Speaker 3: yet and maybe never will. Is capital flight. What we

1172
01:15:27,360 --> 01:15:30,640
Speaker 3: see in politically domestically is that actually, and this was

1173
01:15:30,760 --> 01:15:33,160
Speaker 3: validated by a gallop pool they've been doing twenty years,

1174
01:15:33,840 --> 01:15:37,240
Speaker 3: which also showed what we had saw started seeing a

1175
01:15:37,320 --> 01:15:44,719
Speaker 3: lot earlier, where immigration as a thing is actually pretty

1176
01:15:44,800 --> 01:15:46,759
Speaker 3: darn popular in the United States.

1177
01:15:47,160 --> 01:15:51,240
Speaker 2: Yeah, but so so is restricting assault rifles. And we

1178
01:15:51,280 --> 01:15:54,839
Speaker 2: can't get any change on that. I think sevent.

1179
01:15:56,439 --> 01:15:59,040
Speaker 3: Or the Act. I gotta tell you, I gotta tell

1180
01:15:59,080 --> 01:16:02,240
Speaker 3: you that depends very much and how that question is asked. Well,

1181
01:16:02,280 --> 01:16:04,960
Speaker 3: that's true for all polling. Well, this is the benefit

1182
01:16:05,040 --> 01:16:07,320
Speaker 3: of what I'm tied. So so when we're doing these

1183
01:16:07,400 --> 01:16:10,439
Speaker 3: semantic we call the semantic signatures, it has nothing to

1184
01:16:10,520 --> 01:16:13,679
Speaker 3: do with how you're phrasing the question. We're not doing polling.

1185
01:16:14,120 --> 01:16:18,080
Speaker 3: We're seeing what people are actually the meaning of what

1186
01:16:18,200 --> 01:16:20,000
Speaker 3: they're talking about in media.

1187
01:16:20,240 --> 01:16:25,080
Speaker 2: So how does that play out if people are legitimately

1188
01:16:25,200 --> 01:16:29,000
Speaker 2: saying no, immigration is a good thing for America. Is

1189
01:16:29,080 --> 01:16:32,320
Speaker 2: there an impact on population and the economy. Is there

1190
01:16:32,320 --> 01:16:35,120
Speaker 2: an impact on markets? Is there an impact on policy

1191
01:16:35,160 --> 01:16:35,719
Speaker 2: and politics?

1192
01:16:35,760 --> 01:16:38,520
Speaker 3: I think there's an impact on the election the midterms

1193
01:16:38,600 --> 01:16:39,040
Speaker 3: next year.

1194
01:16:40,439 --> 01:16:45,000
Speaker 2: So we're we're talking literally twelve months from now.

1195
01:16:46,040 --> 01:16:50,759
Speaker 3: Yeah, because that's how this stuff gets on the political front.

1196
01:16:52,000 --> 01:16:56,240
Speaker 3: Narratives and opinions, they get cashed out in elections, right,

1197
01:16:56,439 --> 01:16:58,760
Speaker 3: Mark is a different You cash stuff out every day,

1198
01:16:59,120 --> 01:17:03,080
Speaker 3: right the market. That loop is so rapid with markets exactly.

1199
01:17:04,240 --> 01:17:07,200
Speaker 3: Politics is a different story. So hedy gets cashed out

1200
01:17:07,240 --> 01:17:07,679
Speaker 3: in the election.

1201
01:17:08,080 --> 01:17:11,439
Speaker 2: If you were to ask me before this conversation, what's

1202
01:17:11,479 --> 01:17:15,519
Speaker 2: the most significant impact on the midterm elections? There was

1203
01:17:15,680 --> 01:17:21,439
Speaker 2: just a Gallop poll yesterday, Gop questions on the economy.

1204
01:17:21,520 --> 01:17:25,040
Speaker 2: They were plus fourteen percent two years ago. They're minus

1205
01:17:25,120 --> 01:17:28,320
Speaker 2: four percent this year. I saw that, and that's an

1206
01:17:28,520 --> 01:17:32,960
Speaker 2: amazing swing. And I'm I'm saying to myself, you know, listen,

1207
01:17:33,080 --> 01:17:36,800
Speaker 2: the out of power party usually picks up, you know,

1208
01:17:36,960 --> 01:17:40,679
Speaker 2: ten to fifteen seats. You have the redistricting question, which

1209
01:17:40,760 --> 01:17:44,800
Speaker 2: may blunt that. But if the most important question during

1210
01:17:44,800 --> 01:17:47,719
Speaker 2: the election was on the economy and that's an eighteen

1211
01:17:47,760 --> 01:17:51,559
Speaker 2: percent swing, this is looking like a pretty substantial shift.

1212
01:17:51,680 --> 01:17:54,439
Speaker 2: What I'm hearing from you is, hey, this isn't just

1213
01:17:54,520 --> 01:17:58,280
Speaker 2: about the economy. It's not It's not so it's the economy,

1214
01:17:58,320 --> 01:17:58,959
Speaker 2: it's immigration.

1215
01:17:59,160 --> 01:17:59,559
Speaker 3: What else?

1216
01:18:00,320 --> 01:18:01,599
Speaker 2: What else are you seeing? That's interest?

1217
01:18:01,680 --> 01:18:04,040
Speaker 3: That's that's that's pretty now. There are other aspects though,

1218
01:18:04,240 --> 01:18:11,400
Speaker 3: where the stories it This is, this is people talking

1219
01:18:11,439 --> 01:18:14,080
Speaker 3: about immigration is the thing. Now, Whether now whether that

1220
01:18:14,200 --> 01:18:17,559
Speaker 3: gets translated into a political party position, because I got

1221
01:18:17,640 --> 01:18:22,320
Speaker 3: to tell you, the Democratic Party is enormously there. There

1222
01:18:22,320 --> 01:18:25,120
Speaker 3: are no there are no narratives that are being put

1223
01:18:25,200 --> 01:18:29,840
Speaker 3: forward by the Democrats that are powerful or or or

1224
01:18:30,000 --> 01:18:30,800
Speaker 3: or popular, so.

1225
01:18:30,960 --> 01:18:34,120
Speaker 2: I think, other than the mayoral contest in New York.

1226
01:18:34,439 --> 01:18:37,800
Speaker 3: Yeah, yeah, right, what is one. What I'm saying is

1227
01:18:37,840 --> 01:18:41,760
Speaker 3: that the is that the many of the policies that

1228
01:18:41,840 --> 01:18:47,680
Speaker 3: are being presented by this administration are unpopular, not just

1229
01:18:47,760 --> 01:18:53,800
Speaker 3: with Democrats but with Republicans and increasingly so, immigration being

1230
01:18:53,880 --> 01:18:58,320
Speaker 3: one of them, I think economy being another one.

1231
01:18:58,640 --> 01:19:03,160
Speaker 2: Tariffs aren't really popular amongst the but people perceive that

1232
01:19:03,240 --> 01:19:06,840
Speaker 2: as a tax increase or or a pressure on small business.

1233
01:19:07,400 --> 01:19:10,640
Speaker 3: Government shut down, right, So both calls and also in

1234
01:19:10,680 --> 01:19:13,800
Speaker 3: this it's it's it's that because the question, oh, well,

1235
01:19:13,840 --> 01:19:16,080
Speaker 3: the Democrats will get blamed, Well, that's not really true.

1236
01:19:16,120 --> 01:19:20,759
Speaker 3: That's not what's happening. Now, how that all pays the way.

1237
01:19:22,160 --> 01:19:26,519
Speaker 2: You're not a heart You're not a Democrat. God, like,

1238
01:19:26,960 --> 01:19:30,160
Speaker 2: I know you and your politics and the stuff you're saying.

1239
01:19:30,960 --> 01:19:33,000
Speaker 2: I could hear people shrugging and saying, well that Ben

1240
01:19:33,120 --> 01:19:35,880
Speaker 2: Hunt is just a liberal Democrat. I'm like, no, no, no,

1241
01:19:36,040 --> 01:19:39,840
Speaker 2: that's not who Ben is. No, no, it's not You're

1242
01:19:39,960 --> 01:19:41,960
Speaker 2: just talking about here's what I see in the data.

1243
01:19:42,120 --> 01:19:45,800
Speaker 3: I'm saying. I'm observing. I'm observing, And I think that

1244
01:19:45,880 --> 01:19:48,920
Speaker 3: a lot of times, if we are, like me, very

1245
01:19:49,040 --> 01:19:54,680
Speaker 3: online and on Twitter too much, you don't see the

1246
01:19:54,800 --> 01:20:01,280
Speaker 3: broader picture of what is happening on blog posts and

1247
01:20:01,600 --> 01:20:05,760
Speaker 3: local newspapers and and and everything else. So I think

1248
01:20:06,000 --> 01:20:14,000
Speaker 3: having the ability to read everything and track these stories

1249
01:20:15,240 --> 01:20:19,640
Speaker 3: that you know, they don't they they like say they

1250
01:20:19,720 --> 01:20:22,120
Speaker 3: wax and wane, but they don't. They don't ever go away.

1251
01:20:23,560 --> 01:20:27,240
Speaker 3: It's it's it's something we're very excited about to track

1252
01:20:27,320 --> 01:20:31,400
Speaker 3: the stories of America I'll give you another one, and

1253
01:20:31,439 --> 01:20:35,760
Speaker 3: this one, actually I don't know what to do with this.

1254
01:20:37,160 --> 01:20:40,679
Speaker 3: One of the stories of America is that America has

1255
01:20:43,600 --> 01:20:49,760
Speaker 3: raised the world's wealth, right, that that global standard of living,

1256
01:20:49,800 --> 01:20:52,439
Speaker 3: global center of living, That that America has been this

1257
01:20:52,600 --> 01:20:57,000
Speaker 3: powerful force to and capitalism in America have been this

1258
01:20:57,120 --> 01:21:00,720
Speaker 3: powerful force to raise people out of poverty. That's a

1259
01:21:00,800 --> 01:21:05,720
Speaker 3: story that in other periods of time has been out,

1260
01:21:05,760 --> 01:21:10,120
Speaker 3: that has been prominently talked about, very resonant. That story

1261
01:21:10,240 --> 01:21:11,559
Speaker 3: is non existent today.

1262
01:21:11,800 --> 01:21:15,599
Speaker 2: Well, you cut things like HIV a few million dollars

1263
01:21:15,840 --> 01:21:20,200
Speaker 2: to inoculate all of Africa from HIV, that's going to

1264
01:21:20,280 --> 01:21:21,160
Speaker 2: have repercussions.

1265
01:21:21,280 --> 01:21:26,240
Speaker 3: Well, it's it's not just that that that has the

1266
01:21:26,400 --> 01:21:30,600
Speaker 3: conquer those repercussions that you're describing. What I'm saying is

1267
01:21:30,760 --> 01:21:37,080
Speaker 3: that the story that would support that it's gone.

1268
01:21:37,960 --> 01:21:39,280
Speaker 2: Is it gone forever? No?

1269
01:21:39,880 --> 01:21:42,799
Speaker 3: No, doesn't. These things are never gone. These stories stories

1270
01:21:42,880 --> 01:21:46,840
Speaker 3: have they never die, right, They're they're they're waiting for

1271
01:21:47,880 --> 01:21:50,560
Speaker 3: a new force to give them this version of it.

1272
01:21:50,880 --> 01:21:53,880
Speaker 2: If US is not seen as a force for raising

1273
01:21:53,960 --> 01:21:58,000
Speaker 2: people out of poverty around the world, do people see

1274
01:21:58,120 --> 01:21:59,280
Speaker 2: China as that force?

1275
01:22:00,640 --> 01:22:02,240
Speaker 3: I haven't looked at that we've been looking at the

1276
01:22:02,520 --> 01:22:05,120
Speaker 3: US story. But I but I my my personal seness

1277
01:22:05,400 --> 01:22:13,160
Speaker 3: is that we've essentially seeded the field and a particularly

1278
01:22:13,280 --> 01:22:17,160
Speaker 3: in South Asia and Africa. Uh huh to to to China.

1279
01:22:18,040 --> 01:22:20,640
Speaker 3: I mean that that seems that seems pretty you know,

1280
01:22:20,840 --> 01:22:21,320
Speaker 3: pretty clear.

1281
01:22:21,439 --> 01:22:23,560
Speaker 2: This is I'm glad I asked the question about the

1282
01:22:23,680 --> 01:22:27,599
Speaker 2: end of Past Americana because I've just found that piece

1283
01:22:27,760 --> 01:22:30,840
Speaker 2: so insightful and so useful, and I could keep you

1284
01:22:31,000 --> 01:22:35,040
Speaker 2: here talking about this stuff for hours, but out of

1285
01:22:35,240 --> 01:22:39,439
Speaker 2: deference and respect for our listener's time, I'm going to

1286
01:22:39,560 --> 01:22:42,960
Speaker 2: jump to our favorite questions that I ask, oh all

1287
01:22:43,000 --> 01:22:46,040
Speaker 2: of our all of our guests. But before I do,

1288
01:22:47,000 --> 01:22:50,920
Speaker 2: there's a question that I have to pose to you,

1289
01:22:51,760 --> 01:22:56,479
Speaker 2: which is, so you get to see things before they

1290
01:22:56,560 --> 01:22:59,800
Speaker 2: sort of bubble up into the mainstream, before they be

1291
01:23:00,040 --> 01:23:04,240
Speaker 2: I'm a well understood narrative. What's a narrative that you're

1292
01:23:04,520 --> 01:23:08,679
Speaker 2: just starting to sniff out that most of the world,

1293
01:23:08,840 --> 01:23:11,519
Speaker 2: or most of the country, or most of Wall Street

1294
01:23:11,600 --> 01:23:14,000
Speaker 2: and Finance hasn't seen yet.

1295
01:23:15,920 --> 01:23:21,560
Speaker 3: It's a story that people are talking about since Tricolor

1296
01:23:21,840 --> 01:23:26,639
Speaker 3: and first Brands went out. So it's the story on credit,

1297
01:23:27,520 --> 01:23:28,920
Speaker 3: story of credit, and.

1298
01:23:29,360 --> 01:23:33,919
Speaker 2: I mean first Brands is legitimate. It sounds like felonies

1299
01:23:33,960 --> 01:23:35,839
Speaker 2: were happening there, so we'll have to see.

1300
01:23:35,680 --> 01:23:41,320
Speaker 3: What tricolor as well. But my point is that every

1301
01:23:41,439 --> 01:23:44,880
Speaker 3: day you have a new bank CEO come out and

1302
01:23:44,960 --> 01:23:48,479
Speaker 3: say no, no, everything's fine. So yesterday it was David

1303
01:23:48,560 --> 01:23:52,519
Speaker 3: Solomon at Golden Sacks saying no, no, it's fine. Today

1304
01:23:52,640 --> 01:23:55,920
Speaker 3: I was hearing the Brookfield guys saying no, no, it's fine.

1305
01:23:57,120 --> 01:24:02,320
Speaker 3: For every story, every interview, the guy says it's fine.

1306
01:24:03,040 --> 01:24:06,840
Speaker 3: You're getting two articles in the ft or the journal

1307
01:24:06,920 --> 01:24:08,160
Speaker 3: saying ain't fine.

1308
01:24:09,120 --> 01:24:12,040
Speaker 2: So they think thou dost protest this much.

1309
01:24:12,000 --> 01:24:15,240
Speaker 3: All of that. All I'm saying is I am observing.

1310
01:24:15,840 --> 01:24:19,839
Speaker 3: I'm not predicting. I am observing the level of volume

1311
01:24:20,479 --> 01:24:24,680
Speaker 3: for alternative asset managers, their exposure to private credit. This

1312
01:24:24,800 --> 01:24:28,160
Speaker 3: is problematic. Oh, I'm wondering what's going to happen when

1313
01:24:28,200 --> 01:24:29,040
Speaker 3: the music stops.

1314
01:24:29,520 --> 01:24:29,680
Speaker 1: Oh.

1315
01:24:29,760 --> 01:24:33,120
Speaker 3: I think that these this blows back into the commercial

1316
01:24:33,200 --> 01:24:39,720
Speaker 3: banking system. Those stories, those narratives are higher by an

1317
01:24:39,840 --> 01:24:42,439
Speaker 3: order of magnitude than they've been at any point in

1318
01:24:42,479 --> 01:24:46,439
Speaker 3: the last ten years, at any point including during COVID,

1319
01:24:46,640 --> 01:24:50,000
Speaker 3: during when you had similar concerns over Oh, my god,

1320
01:24:50,040 --> 01:24:51,040
Speaker 3: private credits COVID.

1321
01:24:51,040 --> 01:24:54,040
Speaker 2: It was easy to rationalize everything. Hey listen, we're all frozen.

1322
01:24:54,400 --> 01:24:59,439
Speaker 3: Just but I this is a story that has legs.

1323
01:25:00,280 --> 01:25:03,439
Speaker 3: It is growing in a way that I rarely see.

1324
01:25:04,120 --> 01:25:07,280
Speaker 3: And what's a story like this grows? It doesn't just

1325
01:25:07,400 --> 01:25:11,760
Speaker 3: go away, And it's not fixed by bernanke saying oh,

1326
01:25:11,920 --> 01:25:15,320
Speaker 3: subprime is contained immediately what I thought of as soon

1327
01:25:15,400 --> 01:25:18,760
Speaker 3: as you absolutely it does not get fixed by people

1328
01:25:18,880 --> 01:25:21,080
Speaker 3: saying don't worry, there's no problem.

1329
01:25:21,640 --> 01:25:25,280
Speaker 2: Jim Grant said he was right. It was contained to Earth.

1330
01:25:26,040 --> 01:25:29,200
Speaker 2: The rest of the solar system was fine, which turned

1331
01:25:29,200 --> 01:25:31,640
Speaker 2: out to be a very witty and clever observation.

1332
01:25:31,880 --> 01:25:36,040
Speaker 3: This story, this worry, this concern absolutely has legs. Huh.

1333
01:25:36,320 --> 01:25:41,800
Speaker 3: And we're seeing no signs of it easing off in

1334
01:25:42,160 --> 01:25:43,519
Speaker 3: financial media and press.

1335
01:25:43,680 --> 01:25:47,640
Speaker 2: So the big question is is it systemic or is

1336
01:25:47,720 --> 01:25:48,960
Speaker 2: it the specifics.

1337
01:25:49,320 --> 01:25:52,880
Speaker 3: I don't know the truth. I don't know reality. All

1338
01:25:52,920 --> 01:25:56,679
Speaker 3: I can tell you is we're seeing more of this. Well,

1339
01:25:57,080 --> 01:26:01,200
Speaker 3: we're seeing this is the story really really interesting.

1340
01:26:01,560 --> 01:26:06,120
Speaker 2: Let let's jump to our favorite questions, starting with who

1341
01:26:06,200 --> 01:26:09,280
Speaker 2: are your mentors who helped shape your career?

1342
01:26:10,080 --> 01:26:14,439
Speaker 3: It's good, well, I mentioned that the people both in

1343
01:26:14,560 --> 01:26:17,880
Speaker 3: undergrad and graduate school who turned me on to the

1344
01:26:18,000 --> 01:26:22,599
Speaker 3: science part of political science, and in particular in graduate school,

1345
01:26:22,680 --> 01:26:26,840
Speaker 3: Gary King, who runs the whole social science research center

1346
01:26:26,920 --> 01:26:29,559
Speaker 3: up there at Harvard, and as for a long time now,

1347
01:26:30,040 --> 01:26:33,200
Speaker 3: he wrote the original, really the book on inference. You

1348
01:26:33,280 --> 01:26:34,320
Speaker 3: hear about inference all.

1349
01:26:34,240 --> 01:26:35,960
Speaker 2: The time Aldini, of course.

1350
01:26:36,560 --> 01:26:40,439
Speaker 3: So the notion of inference, taking large data sets and

1351
01:26:40,560 --> 01:26:41,360
Speaker 3: pulling out.

1352
01:26:41,400 --> 01:26:44,120
Speaker 2: Oh, inference. I think you said influence, No, no, no, no,

1353
01:26:44,280 --> 01:26:45,440
Speaker 2: inference inference.

1354
01:26:45,720 --> 01:26:51,360
Speaker 3: Huh So the science of inference. I learned that from Gary,

1355
01:26:51,720 --> 01:26:54,680
Speaker 3: you know, thirty years ago. And it's so interesting to

1356
01:26:54,680 --> 01:26:56,920
Speaker 3: see that come full circle because that's at the core

1357
01:26:57,160 --> 01:27:00,240
Speaker 3: of you know, Jensen Wong is always talking about it

1358
01:27:00,320 --> 01:27:03,439
Speaker 3: and the whole notion of AI to that inference spind

1359
01:27:03,479 --> 01:27:06,000
Speaker 3: do they talk about. I was there at the beginning

1360
01:27:06,880 --> 01:27:10,760
Speaker 3: for how you what inference is and why it is

1361
01:27:10,840 --> 01:27:14,640
Speaker 3: so powerful. That was a huge influence on me.

1362
01:27:15,160 --> 01:27:16,280
Speaker 2: Let's talk about books.

1363
01:27:16,680 --> 01:27:17,760
Speaker 3: What are you reading now?

1364
01:27:17,880 --> 01:27:19,120
Speaker 2: What are some of your favorites.

1365
01:27:19,800 --> 01:27:21,240
Speaker 3: I'm a science fiction guy.

1366
01:27:21,320 --> 01:27:23,960
Speaker 2: Okay, I really talking to another sci fi guy. So

1367
01:27:24,120 --> 01:27:27,200
Speaker 2: let's so I give us something new and something classic.

1368
01:27:27,479 --> 01:27:31,519
Speaker 3: Well, something classic would be Luci Shin and the three

1369
01:27:31,560 --> 01:27:32,320
Speaker 3: Body Problem.

1370
01:27:33,800 --> 01:27:36,680
Speaker 2: The Netflix show on that was surprisingly watchable.

1371
01:27:36,840 --> 01:27:41,000
Speaker 3: I that was excellent before that, And you know, I

1372
01:27:41,080 --> 01:27:44,840
Speaker 3: named a company after him. Was the Foundation Trilogy as

1373
01:27:45,720 --> 01:27:48,320
Speaker 3: those books hold up less well, honestly, And I thought

1374
01:27:48,400 --> 01:27:52,080
Speaker 3: the series was. It had his moments. It was so

1375
01:27:52,200 --> 01:27:52,960
Speaker 3: pretty so so to me.

1376
01:27:53,720 --> 01:27:53,840
Speaker 2: Uh.

1377
01:27:54,680 --> 01:28:00,439
Speaker 3: Current science fiction, Rebecca Goes, she goes by R. F.

1378
01:28:01,120 --> 01:28:06,519
Speaker 3: Quog and she wrote a book called Babbel and she's

1379
01:28:06,520 --> 01:28:09,920
Speaker 3: got a new one out. I like Tower of Babel

1380
01:28:10,320 --> 01:28:12,439
Speaker 3: b A B E L. And she's got a new

1381
01:28:12,479 --> 01:28:15,840
Speaker 3: one out, Catabasis. I think it is. But science fiction

1382
01:28:16,200 --> 01:28:19,240
Speaker 3: dealing with language and linguistics. I love that stuff.

1383
01:28:19,640 --> 01:28:22,439
Speaker 2: That sounds like that's right in your your sweet spot.

1384
01:28:23,120 --> 01:28:26,479
Speaker 2: Let's talk about streaming. What are you watching or listening today?

1385
01:28:26,840 --> 01:28:28,599
Speaker 2: Podcasts or Netflix or whatever.

1386
01:28:28,840 --> 01:28:31,120
Speaker 3: So honestly, I don't listen to any podcasts. Isn't that

1387
01:28:31,240 --> 01:28:31,880
Speaker 3: terrible to admit?

1388
01:28:31,960 --> 01:28:34,080
Speaker 2: Well, if you host, I will say this, if you

1389
01:28:34,200 --> 01:28:38,599
Speaker 2: host a podcast, you're either preparing for a podcast, doing

1390
01:28:38,640 --> 01:28:41,800
Speaker 2: a podcast, or ordering a podcast, right, So it's like,

1391
01:28:42,000 --> 01:28:43,360
Speaker 2: all right, that's three of that.

1392
01:28:43,960 --> 01:28:44,720
Speaker 3: It is right.

1393
01:28:44,840 --> 01:28:47,439
Speaker 2: Every now and then I'll catch something because I want

1394
01:28:47,479 --> 01:28:50,600
Speaker 2: to either listen to this a specific guest like I,

1395
01:28:51,160 --> 01:28:53,800
Speaker 2: I will not listen to any podcasts that you are

1396
01:28:53,880 --> 01:28:56,320
Speaker 2: on before we do our podcast because I don't want

1397
01:28:56,360 --> 01:28:59,960
Speaker 2: to steal anybody else's or narrative or questions you'll line

1398
01:29:00,120 --> 01:29:04,680
Speaker 2: of thinking. So that's purposeful. But every now and then,

1399
01:29:04,840 --> 01:29:08,559
Speaker 2: so I listened to the things I listened to are

1400
01:29:08,720 --> 01:29:14,400
Speaker 2: like John Pizzarelli's Radio Deluxe, which is sort of like

1401
01:29:14,560 --> 01:29:17,560
Speaker 2: a podcast slash music series.

1402
01:29:17,320 --> 01:29:19,680
Speaker 3: Something a little different, exactly right.

1403
01:29:19,920 --> 01:29:21,160
Speaker 2: What about streaming?

1404
01:29:21,200 --> 01:29:23,400
Speaker 3: What do you well? You and I are both big

1405
01:29:23,439 --> 01:29:26,560
Speaker 3: Godfather fans, so we love all the Mobster Have you

1406
01:29:26,640 --> 01:29:27,720
Speaker 3: seen Mobland yet?

1407
01:29:27,880 --> 01:29:28,720
Speaker 2: With no? All?

1408
01:29:28,760 --> 01:29:33,000
Speaker 3: Right? Worth your time? Okay, Pierce Brosnan eats up every scene.

1409
01:29:33,320 --> 01:29:35,560
Speaker 3: Tom Hardy is awfully good. I just thank him in

1410
01:29:35,600 --> 01:29:38,320
Speaker 3: anything if you I don't know if you ever saw

1411
01:29:38,400 --> 01:29:39,920
Speaker 3: Peaky Blinders.

1412
01:29:39,600 --> 01:29:44,960
Speaker 2: Right, I kind of full. So my issue is I

1413
01:29:45,120 --> 01:29:47,480
Speaker 2: have to watch something that my wife tolerates.

1414
01:29:47,600 --> 01:29:50,200
Speaker 3: Like I have friends that's a struggle.

1415
01:29:50,200 --> 01:29:53,040
Speaker 2: Who like he goes into this room, she goes into

1416
01:29:53,080 --> 01:29:56,200
Speaker 2: that room. They don't watch anything together, Like, I don't

1417
01:29:56,240 --> 01:29:59,599
Speaker 2: know how that works. Like she will watch some stuff

1418
01:29:59,640 --> 01:30:03,200
Speaker 2: without me. I will watch some stuff with right, eighty

1419
01:30:03,320 --> 01:30:06,920
Speaker 2: percent of what we watch. So I have fallen down

1420
01:30:07,280 --> 01:30:12,599
Speaker 2: the British Upstairs, downstairs, Gilded Age six, So we watched

1421
01:30:12,640 --> 01:30:15,360
Speaker 2: the Crown, we watched the Gilded Age. We finally let

1422
01:30:15,400 --> 01:30:18,519
Speaker 2: me give you one you want back to Dalton Abbey,

1423
01:30:18,560 --> 01:30:20,400
Speaker 2: which I missed when I first rolled out.

1424
01:30:20,600 --> 01:30:23,360
Speaker 3: Check out, so my guilty pleasure and I did watch

1425
01:30:23,400 --> 01:30:24,920
Speaker 3: this with my wife. Is a Diplomat.

1426
01:30:25,479 --> 01:30:28,599
Speaker 2: The new series, just the new season, just and that's

1427
01:30:28,680 --> 01:30:29,799
Speaker 2: teed up for this weekend.

1428
01:30:29,840 --> 01:30:30,920
Speaker 3: I'm looking forward to that.

1429
01:30:31,280 --> 01:30:34,720
Speaker 2: Excellent that the first was the season two or season three?

1430
01:30:34,800 --> 01:30:35,960
Speaker 3: The first season was great.

1431
01:30:36,640 --> 01:30:39,040
Speaker 2: Yeah, and I'm going to give you if you like that.

1432
01:30:40,920 --> 01:30:43,880
Speaker 2: So my wife finds these really interesting. She got me

1433
01:30:43,960 --> 01:30:48,800
Speaker 2: into Killing Eve, which was a little more spyfair than Diplomacy.

1434
01:30:50,040 --> 01:30:52,560
Speaker 2: And if you get a chance to watch Slow Horses.

1435
01:30:52,400 --> 01:30:53,720
Speaker 3: Oh that's the one I've got to see.

1436
01:30:53,840 --> 01:30:57,600
Speaker 2: Yeah, it's really so the first season is great and

1437
01:30:57,720 --> 01:31:00,840
Speaker 2: people will told me the most recent season it's some

1438
01:31:01,000 --> 01:31:04,800
Speaker 2: people will say I don't like I know people who

1439
01:31:04,920 --> 01:31:08,240
Speaker 2: loved it all the way through, but it's it's really

1440
01:31:09,840 --> 01:31:12,000
Speaker 2: it's really an interesting, well told.

1441
01:31:12,080 --> 01:31:13,800
Speaker 3: Beautiful cast, fantastic.

1442
01:31:13,920 --> 01:31:19,000
Speaker 2: Yeah, you'll you'll love that. Our final two questions, Yeah,

1443
01:31:19,200 --> 01:31:21,880
Speaker 2: what sort of advice would you give a recent college

1444
01:31:21,920 --> 01:31:28,400
Speaker 2: grad interested in either investing or working with large language models,

1445
01:31:29,000 --> 01:31:33,360
Speaker 2: working with narrative analysis and artificial intelligence.

1446
01:31:35,800 --> 01:31:39,880
Speaker 3: Don't I say that tongue in cheek.

1447
01:31:40,000 --> 01:31:42,680
Speaker 2: I you had a lot of fits and starts going

1448
01:31:42,760 --> 01:31:46,240
Speaker 2: back seven eight years for sure. And I.

1449
01:31:50,200 --> 01:31:58,080
Speaker 3: I tell you what, I I think some important you

1450
01:31:58,240 --> 01:32:00,360
Speaker 3: have to would you would? I think think you can

1451
01:32:00,360 --> 01:32:04,080
Speaker 3: accomplish in academia, either grad school or whatever it is,

1452
01:32:04,320 --> 01:32:08,599
Speaker 3: is you build your intellectual capital. And so I think

1453
01:32:08,600 --> 01:32:12,479
Speaker 3: a lot of times when you get out of college

1454
01:32:13,320 --> 01:32:15,439
Speaker 3: you say, okay, I'm just ready to kind of live

1455
01:32:15,520 --> 01:32:18,400
Speaker 3: my life and start something, and you haven't built that

1456
01:32:18,520 --> 01:32:23,880
Speaker 3: intellectual capital yet. The issue, though, is that once you enter,

1457
01:32:24,800 --> 01:32:28,679
Speaker 3: particularly the investment world, where if you're responsible for managing

1458
01:32:28,720 --> 01:32:34,120
Speaker 3: other people's money. Brother, that's it, right, I mean, you're

1459
01:32:34,360 --> 01:32:38,680
Speaker 3: you're there's there's not you. That's got to be your

1460
01:32:39,680 --> 01:32:45,160
Speaker 3: your total focus. You are spending your intellectual capital. You're

1461
01:32:45,240 --> 01:32:49,599
Speaker 3: not gaining intellectual capital once you take on a role

1462
01:32:49,760 --> 01:32:54,320
Speaker 3: like that. So I my first advice is find a

1463
01:32:54,439 --> 01:32:58,599
Speaker 3: path where and that's often in in in academia where

1464
01:32:58,600 --> 01:33:02,000
Speaker 3: you're building intellectual capital, because once you leave that environment,

1465
01:33:02,840 --> 01:33:05,559
Speaker 3: then you're spending it down you're spending it down. Make

1466
01:33:05,640 --> 01:33:06,439
Speaker 3: sen spending it down.

1467
01:33:06,840 --> 01:33:09,439
Speaker 2: Our final question, what do you know about the world

1468
01:33:09,720 --> 01:33:16,080
Speaker 2: of filling the blank? Investing? Data, analytics, narrative storytelling today

1469
01:33:16,520 --> 01:33:19,000
Speaker 2: would have been helpful twenty five to thirty years ago.

1470
01:33:22,200 --> 01:33:31,280
Speaker 3: And investing. I wish I had understood the role of

1471
01:33:34,400 --> 01:33:36,920
Speaker 3: at least step back a second. I think that whether

1472
01:33:36,960 --> 01:33:39,919
Speaker 3: you're talking about data or whether you're talking about investing.

1473
01:33:42,600 --> 01:33:44,080
Speaker 3: And I'll speak for myself, but I think there are

1474
01:33:44,080 --> 01:33:46,240
Speaker 3: a lot of people like me. We think there's an

1475
01:33:46,360 --> 01:33:50,160
Speaker 3: answer with a capital a huh right there in the

1476
01:33:50,640 --> 01:33:53,479
Speaker 3: in the numbers, and if you just look hard enough,

1477
01:33:53,960 --> 01:33:57,920
Speaker 3: and if you work hard enough, you'll find that answer. Right,

1478
01:33:58,000 --> 01:34:00,480
Speaker 3: you'll find the secret the secret formula.

1479
01:34:00,960 --> 01:34:02,679
Speaker 2: And you have learned since then.

1480
01:34:03,120 --> 01:34:07,040
Speaker 3: Ain't no such thing? Right now, There is magic and

1481
01:34:07,120 --> 01:34:10,280
Speaker 3: there are patterns, but it's it's it's not in the

1482
01:34:10,400 --> 01:34:14,880
Speaker 3: structured data. It's not in the numbers. It's actually in

1483
01:34:15,000 --> 01:34:19,320
Speaker 3: the error. It's in the probabilitabilities. It's in we're calling,

1484
01:34:19,600 --> 01:34:24,839
Speaker 3: you know, the stochastic element, right, the role of chance,

1485
01:34:25,080 --> 01:34:29,000
Speaker 3: and understanding that there are patterns and there's real magic,

1486
01:34:29,160 --> 01:34:33,880
Speaker 3: and understanding that I didn't get that when I was

1487
01:34:34,000 --> 01:34:38,519
Speaker 3: either starting with data analysis or with investing. I was

1488
01:34:38,600 --> 01:34:41,759
Speaker 3: looking for the answer with the capital A as opposed

1489
01:34:41,800 --> 01:34:43,360
Speaker 3: to a process with a capital P.

1490
01:34:43,920 --> 01:34:48,840
Speaker 2: I love that answer because so my exercise in confirmation

1491
01:34:49,000 --> 01:34:54,200
Speaker 2: bias is I love the fact that you looking back,

1492
01:34:54,520 --> 01:34:57,680
Speaker 2: we all have the benefit of hindsight. Yeah, and unfortunately

1493
01:34:58,360 --> 01:35:00,760
Speaker 2: that can be a bias to certain p people. But

1494
01:35:00,880 --> 01:35:04,040
Speaker 2: when you're looking back at it, you know how it happens.

1495
01:35:04,800 --> 01:35:07,360
Speaker 2: At the moment when you don't know what the outcome is,

1496
01:35:08,200 --> 01:35:13,599
Speaker 2: thinking about it probabilistically is a much healthier approach than saying,

1497
01:35:14,160 --> 01:35:16,479
Speaker 2: here's a binary up or down, yes or no, and

1498
01:35:16,560 --> 01:35:20,200
Speaker 2: I'm either right or wrong. And I see people struggle

1499
01:35:20,240 --> 01:35:22,000
Speaker 2: with that constantly, constantly.

1500
01:35:22,760 --> 01:35:25,840
Speaker 3: When first started investing, we're again late. I got two

1501
01:35:25,880 --> 01:35:30,760
Speaker 3: pieces of good advice right. One was never go all in, right,

1502
01:35:31,360 --> 01:35:35,240
Speaker 3: which is which is really interesting because it's this, this

1503
01:35:35,479 --> 01:35:39,160
Speaker 3: business of investing. It's a wonderful life. You were solving puzzles,

1504
01:35:39,280 --> 01:35:41,800
Speaker 3: we meet interesting people, we get to have these sort

1505
01:35:41,800 --> 01:35:45,400
Speaker 3: of conversations. It's a it's a career for a lifetime.

1506
01:35:46,160 --> 01:35:51,200
Speaker 3: And the two pieces of advice early it never go

1507
01:35:51,360 --> 01:35:58,439
Speaker 3: all in. And also, reputation is absolutely the most important thing.

1508
01:35:59,240 --> 01:36:02,640
Speaker 3: And again, especially when you're young, you think, oh, well,

1509
01:36:02,720 --> 01:36:05,000
Speaker 3: that's not the most important thing. You know, being right's

1510
01:36:05,000 --> 01:36:09,760
Speaker 3: the most important thing. It ain't. It's playing the long

1511
01:36:09,880 --> 01:36:13,040
Speaker 3: game here would you want to? It's never go all

1512
01:36:13,120 --> 01:36:16,439
Speaker 3: in and the right never.

1513
01:36:17,720 --> 01:36:21,160
Speaker 2: By never go in, you mean never put everything at

1514
01:36:21,320 --> 01:36:23,360
Speaker 2: risks so that if it doesn't work out.

1515
01:36:23,280 --> 01:36:26,160
Speaker 3: You're you're out of the game. You're stay in the game.

1516
01:36:26,360 --> 01:36:29,120
Speaker 2: Gerald Loeb's book The Battle for Investment Survival.

1517
01:36:29,360 --> 01:36:32,800
Speaker 3: Stay in the game, avoid that risk of ruin. You

1518
01:36:33,360 --> 01:36:37,639
Speaker 3: just or don't ever risk ruin. That makes sense because

1519
01:36:37,960 --> 01:36:39,280
Speaker 3: once you don't think in those terms.

1520
01:36:39,360 --> 01:36:40,360
Speaker 2: But that's really it.

1521
01:36:40,520 --> 01:36:43,639
Speaker 3: All in means risk of ruin. It's the risk of ruin.

1522
01:36:46,760 --> 01:36:49,280
Speaker 3: And it also connects with the reputation because once.

1523
01:36:49,120 --> 01:36:52,639
Speaker 2: You once you blow up, it's tough to go back

1524
01:36:53,280 --> 01:36:54,400
Speaker 2: and you start.

1525
01:36:54,280 --> 01:36:57,519
Speaker 3: Saying, oh, I can fix it by taking this shortcut

1526
01:36:57,680 --> 01:37:00,680
Speaker 3: or doing this other thing, and once you do that,

1527
01:37:01,080 --> 01:37:04,840
Speaker 3: it never ends well all right, and you tell yourself, oh,

1528
01:37:05,040 --> 01:37:06,960
Speaker 3: just this one time. It's never this one time.

1529
01:37:07,360 --> 01:37:11,680
Speaker 2: Ben, This has been absolutely delightful. I'm so glad we've

1530
01:37:11,760 --> 01:37:17,439
Speaker 2: finally too gone around to doing this. I'm just entranced

1531
01:37:17,520 --> 01:37:21,720
Speaker 2: by your thought process. Some of the things like I

1532
01:37:21,880 --> 01:37:25,040
Speaker 2: love reading stuff of yours that I totally disagree with,

1533
01:37:25,760 --> 01:37:28,720
Speaker 2: and then because it forces me to say, well, he's

1534
01:37:28,840 --> 01:37:32,280
Speaker 2: not just making this up. I know how your brain works,

1535
01:37:32,720 --> 01:37:34,920
Speaker 2: and it's like, all right, if Ben is saying this,

1536
01:37:35,200 --> 01:37:38,759
Speaker 2: then I'm going to make this my worst case because

1537
01:37:38,800 --> 01:37:42,839
Speaker 2: it's easy to dismiss it. Everybody is out seeking confirming

1538
01:37:42,960 --> 01:37:46,840
Speaker 2: information and rather than being dismissive of it, all right,

1539
01:37:47,040 --> 01:37:50,880
Speaker 2: you claim to think probabilistically, where does this fit into

1540
01:37:50,960 --> 01:37:53,680
Speaker 2: the range of probabilities? And once you start thinking in

1541
01:37:53,720 --> 01:37:56,519
Speaker 2: those terms, it's like, oh, so the worst case scenario

1542
01:37:56,800 --> 01:37:59,880
Speaker 2: is worse than my worst case scenario. I got to

1543
01:38:00,120 --> 01:38:03,559
Speaker 2: move the bottom of my range down further because if

1544
01:38:03,600 --> 01:38:06,400
Speaker 2: this goes off the rails, this is really bad. That's

1545
01:38:06,439 --> 01:38:10,200
Speaker 2: what your past americanopies did with me, and really it

1546
01:38:10,360 --> 01:38:14,719
Speaker 2: really helped me figure out how to think about especially

1547
01:38:14,800 --> 01:38:18,759
Speaker 2: in that week between April twond and ninth where everyone

1548
01:38:18,880 --> 01:38:21,400
Speaker 2: was losing their mind. It's like, oh, no, so they're

1549
01:38:21,439 --> 01:38:24,519
Speaker 2: losing the mind because hey, this could happen, but maybe

1550
01:38:24,960 --> 01:38:26,080
Speaker 2: something good comes out of it.

1551
01:38:26,400 --> 01:38:28,919
Speaker 3: Trying to avoid tunnel vision, I think is so important

1552
01:38:29,000 --> 01:38:31,320
Speaker 3: in our business or any business, but especially.

1553
01:38:31,320 --> 01:38:35,519
Speaker 2: In our business in particular. Well, thank you for being

1554
01:38:35,600 --> 01:38:38,280
Speaker 2: so generous with your time. We have been speaking with

1555
01:38:38,400 --> 01:38:42,840
Speaker 2: Ben Hunts. He is the co founder and president of Prescients,

1556
01:38:42,920 --> 01:38:46,720
Speaker 2: and you can find his writing at Epsilon Theory. If

1557
01:38:46,800 --> 01:38:49,519
Speaker 2: you enjoy this conversation, well check out any of the

1558
01:38:49,640 --> 01:38:52,799
Speaker 2: five hundred and ninety three we've done over the past

1559
01:38:53,320 --> 01:38:58,920
Speaker 2: eleven and a half years. You can find those at Bloomberg, iTunes, Spotify,

1560
01:38:59,560 --> 01:39:03,040
Speaker 2: you two Tube, wherever you find your favorite podcasts. Be

1561
01:39:03,160 --> 01:39:06,120
Speaker 2: sure and check out my new book, How Not to

1562
01:39:06,280 --> 01:39:10,280
Speaker 2: Invest The Ideas, numbers and behaviors that destroy Wealth and

1563
01:39:10,400 --> 01:39:13,639
Speaker 2: How to Avoid Them at your favorite bookseller. I would

1564
01:39:13,680 --> 01:39:16,320
Speaker 2: be remiss if I not thank the crack staff that

1565
01:39:16,520 --> 01:39:21,040
Speaker 2: helps put these conversations together each week. Alexis Noriega is

1566
01:39:21,160 --> 01:39:26,080
Speaker 2: my video producer. Sean Russo is my researcher. Anna Luke

1567
01:39:26,200 --> 01:39:30,439
Speaker 2: is my podcast producer. Sage Bauman is the head of

1568
01:39:30,520 --> 01:39:35,160
Speaker 2: podcasts here at Bloomberg. I'm Barry Retults. You've been listening

1569
01:39:35,240 --> 01:39:38,280
Speaker 2: to Masters in Business on Bloomberg Radio