1 00:00:10,960 --> 00:00:14,360 Speaker 1: Hello, and welcome to another episode of the Odd Lots podcast. 2 00:00:14,400 --> 00:00:19,479 Speaker 1: I'm Joe Wisenthal and I'm Tracy Hallaway. So I don't 3 00:00:19,480 --> 00:00:22,720 Speaker 1: know what day people are going to be listening to 4 00:00:22,840 --> 00:00:26,240 Speaker 1: this episode, but um, you know, the stock market hit 5 00:00:26,239 --> 00:00:29,760 Speaker 1: a record high yesterday. Yeah, it's true. So all the 6 00:00:29,800 --> 00:00:33,040 Speaker 1: losses that we saw during the COVID crisis have basically 7 00:00:33,080 --> 00:00:36,720 Speaker 1: been raised and markets are back where they were before 8 00:00:37,040 --> 00:00:42,920 Speaker 1: all of this happened. Yeah, it's essentially six months from 9 00:00:42,920 --> 00:00:45,280 Speaker 1: the pre crisis peak to this one. So I think 10 00:00:45,320 --> 00:00:49,479 Speaker 1: the SUP peaked at um February on February fifteen, and 11 00:00:49,520 --> 00:00:53,640 Speaker 1: then we saw the the new peak yesterday, August eighteen. 12 00:00:54,160 --> 00:00:57,920 Speaker 1: And in a sense, it really feels like we've compressed 13 00:00:57,960 --> 00:01:01,760 Speaker 1: this sort of gigantic cycle into an extremely short period 14 00:01:01,760 --> 00:01:04,160 Speaker 1: of time. Yeah, that's true. And I was looking at 15 00:01:04,200 --> 00:01:07,760 Speaker 1: the latest fund managers survey from Bank of America and 16 00:01:07,800 --> 00:01:11,560 Speaker 1: it showed that I think fund managers have completely flipped 17 00:01:11,959 --> 00:01:15,040 Speaker 1: from thinking that we're in a recession to thinking that 18 00:01:15,080 --> 00:01:17,959 Speaker 1: we're in the early stages of a fresh economic cycle. 19 00:01:18,040 --> 00:01:21,319 Speaker 1: And if they're right to your point, it does suggest 20 00:01:21,319 --> 00:01:23,280 Speaker 1: that we've just seen, you know, one of the shortest 21 00:01:23,360 --> 00:01:26,600 Speaker 1: recessions of all time. Yeah. I mean, you could make 22 00:01:26,640 --> 00:01:29,360 Speaker 1: the argument that the recession recession in terms of the 23 00:01:29,440 --> 00:01:31,640 Speaker 1: shrinking of growth was done by the end of March, 24 00:01:32,000 --> 00:01:35,880 Speaker 1: when most data points started turning up. And while the 25 00:01:36,040 --> 00:01:39,319 Speaker 1: overall level of economic activity is still very depressed, and 26 00:01:39,360 --> 00:01:42,560 Speaker 1: of course unemployment rate is still above ten percent, so 27 00:01:42,640 --> 00:01:46,720 Speaker 1: hardly time to be declaring victory. We have seen steady 28 00:01:46,760 --> 00:01:50,360 Speaker 1: improvement on a host of economic data points basically since 29 00:01:50,520 --> 00:01:54,559 Speaker 1: end of March early April. That's true, but I also 30 00:01:54,600 --> 00:01:57,400 Speaker 1: feel like there's something kind of weird going on with 31 00:01:57,440 --> 00:02:01,120 Speaker 1: the data. Like there's the old stock first flow argument, 32 00:02:01,200 --> 00:02:04,320 Speaker 1: which we're seeing everywhere, but particularly in p M I. 33 00:02:04,480 --> 00:02:07,240 Speaker 1: So even when we get a big rebound in p 34 00:02:07,440 --> 00:02:09,720 Speaker 1: m I s, it doesn't necessarily mean that we're getting 35 00:02:09,720 --> 00:02:13,320 Speaker 1: back to the levels that we saw pre crisis. But 36 00:02:13,400 --> 00:02:17,560 Speaker 1: you're also seeing just sort of weird indicators that are 37 00:02:17,600 --> 00:02:20,600 Speaker 1: happening simultaneously. And I think one of our colleagues pointed 38 00:02:20,600 --> 00:02:23,880 Speaker 1: out a really good one recently, and that was intentions 39 00:02:24,000 --> 00:02:27,520 Speaker 1: to buy a house surging at the same time as 40 00:02:27,600 --> 00:02:32,520 Speaker 1: mortgage delinquencies, which I mean never happens in an economic crisis. 41 00:02:33,000 --> 00:02:35,720 Speaker 1: Now it's really weird. But I think because of all 42 00:02:35,760 --> 00:02:38,880 Speaker 1: the weirdness that we're seeing this sort of contrary indicators, 43 00:02:39,240 --> 00:02:43,760 Speaker 1: because there's this weird gap between pieces of change which 44 00:02:43,800 --> 00:02:46,840 Speaker 1: have been very fast and unexpected versus levels which are 45 00:02:46,840 --> 00:02:49,320 Speaker 1: still very bad levels. And then also just the fact 46 00:02:49,320 --> 00:02:53,359 Speaker 1: that it's so compressed, there's probably never been more demand 47 00:02:53,680 --> 00:02:57,080 Speaker 1: for sort of alternative real time data points and this 48 00:02:57,200 --> 00:02:59,919 Speaker 1: feeling that the official economic data points that we get 49 00:03:00,360 --> 00:03:04,800 Speaker 1: monthly jobs report, monthly retail sales report, they just there's 50 00:03:04,880 --> 00:03:07,560 Speaker 1: not enough of them. They're not timely enough to get 51 00:03:07,560 --> 00:03:10,360 Speaker 1: a sense of what's going on, given how fast the 52 00:03:10,480 --> 00:03:14,040 Speaker 1: changes have been both on the downturn and the rebound. Yeah. Absolutely, 53 00:03:14,080 --> 00:03:16,320 Speaker 1: and I mean, just on a very simple basis, everyone 54 00:03:16,360 --> 00:03:19,040 Speaker 1: wants to know what's going on with the recovery, right, 55 00:03:19,080 --> 00:03:22,760 Speaker 1: and everyone's tracking to what degree the economy has reopened, 56 00:03:22,840 --> 00:03:25,799 Speaker 1: and some of the most useful indicators for that are 57 00:03:25,919 --> 00:03:31,760 Speaker 1: arguably alternative economic indicators like um like open table reservations, 58 00:03:32,520 --> 00:03:35,040 Speaker 1: things like that. Yeah, totally. I mean, that's like one 59 00:03:35,080 --> 00:03:36,960 Speaker 1: of the things we've been watching the most. It's like 60 00:03:37,400 --> 00:03:41,280 Speaker 1: open table they could keep track of people making reservations 61 00:03:41,400 --> 00:03:43,560 Speaker 1: or doing in seeding dining, so if you want to 62 00:03:43,560 --> 00:03:47,360 Speaker 1: sort of understand how behavior has changed or how people 63 00:03:47,440 --> 00:03:51,200 Speaker 1: are doing different things. Um, due to the virus, that's 64 00:03:51,240 --> 00:03:53,800 Speaker 1: been one of the sort of key data points, not 65 00:03:54,040 --> 00:03:57,840 Speaker 1: something that people were really tracking before as far as 66 00:03:57,840 --> 00:04:00,440 Speaker 1: I know, on a meaningful level. So I think that's 67 00:04:00,480 --> 00:04:03,640 Speaker 1: really important. I mean, I think obviously real time alternative 68 00:04:03,720 --> 00:04:07,600 Speaker 1: data has never been more in demand than what we've 69 00:04:07,640 --> 00:04:10,440 Speaker 1: seen over the last six months. But I don't think 70 00:04:10,480 --> 00:04:12,680 Speaker 1: it's going away now. It's kind of another one of 71 00:04:12,720 --> 00:04:16,560 Speaker 1: these things where real time data points of a range 72 00:04:16,560 --> 00:04:18,920 Speaker 1: of things will sort of be part of the conversation 73 00:04:19,080 --> 00:04:22,240 Speaker 1: for a long time, even if and when we get 74 00:04:22,240 --> 00:04:25,400 Speaker 1: back to something resembling a normal economy. Yeah, I think 75 00:04:25,400 --> 00:04:28,520 Speaker 1: that's right. So today we're going to be talking all 76 00:04:28,560 --> 00:04:32,599 Speaker 1: about alternative data, what it's showing, and more importantly, how 77 00:04:32,720 --> 00:04:36,840 Speaker 1: investors actually use it in their process. And so we're 78 00:04:36,880 --> 00:04:39,200 Speaker 1: going to be speaking with Ben Brightholtz. He's a data 79 00:04:39,320 --> 00:04:42,880 Speaker 1: scientist at Arbor Data Science, which is part of Arbor 80 00:04:42,960 --> 00:04:46,359 Speaker 1: Research and Trading. I've been following their stuff. They do 81 00:04:46,400 --> 00:04:51,000 Speaker 1: some really interesting things with looking at Google search trends 82 00:04:51,040 --> 00:04:54,640 Speaker 1: for lots of different keywords and trying to divine an 83 00:04:54,720 --> 00:04:58,359 Speaker 1: economic significance from them. So let's talk more about that. Ben, 84 00:04:58,360 --> 00:05:00,520 Speaker 1: thank you very much for joining us. Yeah, thank you 85 00:05:00,640 --> 00:05:03,039 Speaker 1: very much. Joe, happy to be here. So let's just 86 00:05:03,080 --> 00:05:05,919 Speaker 1: start a big picture. What do you do? What is 87 00:05:06,120 --> 00:05:09,720 Speaker 1: arbor data science? Talk to us a little bit about 88 00:05:09,760 --> 00:05:15,320 Speaker 1: your work. Sure, So over the years we've gotten more 89 00:05:15,360 --> 00:05:18,920 Speaker 1: and more into essentially this idea of filling the gaps 90 00:05:19,000 --> 00:05:22,600 Speaker 1: between latent economic data and the econ data that can 91 00:05:22,600 --> 00:05:25,680 Speaker 1: be distorted like we've seen with unemployment data as of late, 92 00:05:25,920 --> 00:05:29,279 Speaker 1: and also really trying to help our customers and the 93 00:05:29,320 --> 00:05:32,720 Speaker 1: investment space in general deal with surveys that have been 94 00:05:32,839 --> 00:05:36,080 Speaker 1: more or less leading indicators for quite some time. They've 95 00:05:36,160 --> 00:05:38,159 Speaker 1: kind of fallen flat on their face. And this is 96 00:05:38,160 --> 00:05:42,280 Speaker 1: something that's taken place well before UM even the current 97 00:05:42,279 --> 00:05:46,120 Speaker 1: episode we're going through now, looking back to the financial crisis, 98 00:05:46,120 --> 00:05:51,200 Speaker 1: with really the polarization of the country and the world 99 00:05:51,279 --> 00:05:54,279 Speaker 1: on a political space, and really the advent of social 100 00:05:54,320 --> 00:05:59,240 Speaker 1: media has created really this bifurcation and in sentiment it 101 00:05:59,279 --> 00:06:01,480 Speaker 1: could be republic and democrat or it can be more 102 00:06:01,560 --> 00:06:05,640 Speaker 1: or less group think. Based on UM, the use of Facebook, Twitter, 103 00:06:05,800 --> 00:06:09,039 Speaker 1: we create all these small microcosms we essentially live within 104 00:06:09,520 --> 00:06:13,400 Speaker 1: and that is ultimately distorted the ability of survey data, 105 00:06:13,440 --> 00:06:16,840 Speaker 1: for example, to have this leading nature that it used 106 00:06:16,839 --> 00:06:19,720 Speaker 1: to have really for decades um. And that's posed a 107 00:06:19,760 --> 00:06:22,880 Speaker 1: significant problem for investors that are in putting this either 108 00:06:22,920 --> 00:06:25,640 Speaker 1: on a subjective level or within their own modeling to 109 00:06:25,760 --> 00:06:28,200 Speaker 1: then project board where they think financial markets will go 110 00:06:28,600 --> 00:06:31,600 Speaker 1: in the future. I have a really basic question, which 111 00:06:31,640 --> 00:06:35,240 Speaker 1: is what's the difference between big data and a large 112 00:06:35,400 --> 00:06:39,720 Speaker 1: set of data? So big data is such a misnomer 113 00:06:39,839 --> 00:06:42,920 Speaker 1: and um nasty term, you know most I think big 114 00:06:43,000 --> 00:06:45,920 Speaker 1: data is a term that's kind of slowly gone away 115 00:06:46,520 --> 00:06:49,000 Speaker 1: that I think the initial idea is that it's it's 116 00:06:49,120 --> 00:06:53,280 Speaker 1: unstructured data. That's for example, you can find all this 117 00:06:53,360 --> 00:06:56,680 Speaker 1: wonderful information on a Bloomberg terminal, all right, and it 118 00:06:56,800 --> 00:06:59,279 Speaker 1: comes you can download it via a p I or 119 00:06:59,320 --> 00:07:02,680 Speaker 1: access at via via your your Windows or your terminal, 120 00:07:02,760 --> 00:07:05,840 Speaker 1: all nice, clean and easy to use, ready to input. 121 00:07:06,360 --> 00:07:09,320 Speaker 1: And big data um to me. Uh, this day and age, 122 00:07:09,400 --> 00:07:11,680 Speaker 1: especially with alternative data, has to do with more or 123 00:07:11,760 --> 00:07:15,480 Speaker 1: less unstructured kind of ugly data. So this, for example 124 00:07:15,520 --> 00:07:18,120 Speaker 1: could be all just like us talking right now, or 125 00:07:18,160 --> 00:07:20,280 Speaker 1: when you are all on TV, you have all of 126 00:07:20,280 --> 00:07:23,600 Speaker 1: this this text, this closed captioning that exists out there, 127 00:07:23,800 --> 00:07:26,520 Speaker 1: and let's say, for example, it's in fifteen second increments, 128 00:07:26,880 --> 00:07:29,480 Speaker 1: and it can be ugly, it can be have plenty 129 00:07:29,480 --> 00:07:32,480 Speaker 1: of errors within the data within the closed captioning um 130 00:07:32,560 --> 00:07:37,080 Speaker 1: And essentially we have to use algorithms and different processes 131 00:07:37,160 --> 00:07:39,320 Speaker 1: in order to take that unstructured data and make it 132 00:07:39,440 --> 00:07:42,280 Speaker 1: something useful and really turn it into something that's more 133 00:07:42,320 --> 00:07:46,280 Speaker 1: or less numerical in order to benchmark against financial markets, 134 00:07:46,320 --> 00:07:49,760 Speaker 1: econ data, overall sentiment and so on. So, you know, 135 00:07:50,200 --> 00:07:52,920 Speaker 1: big data is kind of a word. I think that's 136 00:07:52,960 --> 00:07:55,559 Speaker 1: somewhat going away. But to me, again, it means somewhat 137 00:07:55,600 --> 00:07:59,600 Speaker 1: of an unstructured data set. So I'm thinking about what 138 00:07:59,680 --> 00:08:04,600 Speaker 1: you scribed as the problem with surveys, and uh, you know, 139 00:08:04,760 --> 00:08:07,840 Speaker 1: I think it's either the University of Michigan Consumer Sentiment 140 00:08:07,880 --> 00:08:10,640 Speaker 1: survey or the conference board one. There's one of these 141 00:08:10,920 --> 00:08:13,000 Speaker 1: data points that we have it on the Bloomberg terminal, 142 00:08:13,280 --> 00:08:15,080 Speaker 1: and it's like they say, it is now a good 143 00:08:15,120 --> 00:08:17,800 Speaker 1: time to buy a washing machine? Is now a good 144 00:08:17,840 --> 00:08:20,360 Speaker 1: time to buy a car. There's even one that's one 145 00:08:20,360 --> 00:08:22,080 Speaker 1: of my favorites. Is now a good time to buy 146 00:08:22,200 --> 00:08:25,880 Speaker 1: a vacuum cleaner. But I guess what you're doing is 147 00:08:26,000 --> 00:08:28,280 Speaker 1: you don't have to ask people is now a good 148 00:08:28,320 --> 00:08:31,760 Speaker 1: time to buy a vacuum cleaner, because in if you 149 00:08:31,840 --> 00:08:34,200 Speaker 1: know how to find the data, you can just look 150 00:08:34,200 --> 00:08:37,600 Speaker 1: at searches for vacuum cleaners and that's presumably a lot 151 00:08:37,640 --> 00:08:40,160 Speaker 1: more reliable than asking people into survey whether or now 152 00:08:40,240 --> 00:08:42,880 Speaker 1: is a good time to buy a vacuum cleaner. Right, So, 153 00:08:42,920 --> 00:08:46,599 Speaker 1: the within surveys, there's and there's a plenty of studies 154 00:08:46,800 --> 00:08:51,319 Speaker 1: on this as of late showing that respondents will not 155 00:08:51,520 --> 00:08:56,080 Speaker 1: provide really honest answers relating to their financial hardships. So 156 00:08:56,120 --> 00:08:59,360 Speaker 1: there's there's large gaps and you know, our things better 157 00:08:59,480 --> 00:09:02,280 Speaker 1: now or worse? Are you going to spend do you 158 00:09:02,320 --> 00:09:05,319 Speaker 1: have the money to spend here moving forward on a vacuum, 159 00:09:05,400 --> 00:09:08,000 Speaker 1: on a new washing machine? And so on? And there's 160 00:09:08,000 --> 00:09:11,079 Speaker 1: always been a gap, for example, example, between the web 161 00:09:11,120 --> 00:09:14,320 Speaker 1: based responses and phone based and we saw this too 162 00:09:14,320 --> 00:09:17,160 Speaker 1: with the election. That's a whole another other topic, but 163 00:09:17,559 --> 00:09:20,760 Speaker 1: um on a web based survey and individuals are typically 164 00:09:20,840 --> 00:09:24,679 Speaker 1: much more honest than they are regarding financial hardship than 165 00:09:24,720 --> 00:09:28,080 Speaker 1: they are on the telephone or basically being put on 166 00:09:28,120 --> 00:09:33,320 Speaker 1: the spot. So the idea here between behind search activity 167 00:09:33,360 --> 00:09:35,400 Speaker 1: and this is something that I think that has improved 168 00:09:35,960 --> 00:09:38,320 Speaker 1: in most recent years, is yes, we can get ahead 169 00:09:38,400 --> 00:09:42,160 Speaker 1: of this intention of consumers and we're not necessarily we're 170 00:09:42,160 --> 00:09:45,679 Speaker 1: not really going to lie to that little window on Google. Um. 171 00:09:45,720 --> 00:09:48,360 Speaker 1: You know, we might lie maybe sometimes to our girlfriends 172 00:09:48,480 --> 00:09:51,840 Speaker 1: or our boyfriends or husbands or wives. Um, but you know, 173 00:09:51,880 --> 00:09:54,160 Speaker 1: what we put into that search window is really truly 174 00:09:54,200 --> 00:09:56,079 Speaker 1: what we're seeking and what we're actually trying to query. 175 00:09:56,120 --> 00:09:59,280 Speaker 1: There's no no one really looking over our shoulder. So 176 00:09:59,400 --> 00:10:02,840 Speaker 1: our belief is that search activity, um, really, over the 177 00:10:02,880 --> 00:10:07,160 Speaker 1: past five six years has become kind of a great 178 00:10:07,640 --> 00:10:11,280 Speaker 1: estimate or indication of the consumers intentions of what they 179 00:10:11,320 --> 00:10:14,280 Speaker 1: plan to do. Am I going to buy a wash machine? 180 00:10:14,400 --> 00:10:16,280 Speaker 1: Or if I'm in distress, what does it mean that 181 00:10:16,360 --> 00:10:19,400 Speaker 1: by default on my credit card payment or I don't 182 00:10:19,400 --> 00:10:21,360 Speaker 1: pay my credit card payment? Or what if I need 183 00:10:21,400 --> 00:10:23,839 Speaker 1: to go out and search and find a bankruptcy lawyer. 184 00:10:24,160 --> 00:10:26,440 Speaker 1: These are the type of things we can pick up on, uh, 185 00:10:26,760 --> 00:10:29,480 Speaker 1: you know, within this information to then create a kind 186 00:10:29,480 --> 00:10:32,800 Speaker 1: of um, you know, overall look at the consumer. And 187 00:10:32,800 --> 00:10:34,360 Speaker 1: this can be all the way from the you know 188 00:10:34,480 --> 00:10:37,000 Speaker 1: up towards the United States, the complete um, you know, 189 00:10:37,040 --> 00:10:39,640 Speaker 1: country level, it can be worldwide, and it can be 190 00:10:39,720 --> 00:10:42,880 Speaker 1: drilled down all the way down to a metropolitan area UM. 191 00:10:43,200 --> 00:10:45,679 Speaker 1: And again, the whole idea there is trying to get 192 00:10:45,679 --> 00:10:49,240 Speaker 1: the most honest representation of the individual. And I'll also 193 00:10:49,320 --> 00:10:53,440 Speaker 1: say that the growth UM in the Internet and really 194 00:10:53,520 --> 00:10:56,360 Speaker 1: access to the Internet, both mobile and on the PC, 195 00:10:56,559 --> 00:10:59,560 Speaker 1: has been a big boon for search activity, so that 196 00:10:59,600 --> 00:11:04,319 Speaker 1: you now have fifty of the world having Internet access 197 00:11:04,559 --> 00:11:06,679 Speaker 1: and using it on an active basis. That's more than 198 00:11:06,720 --> 00:11:10,120 Speaker 1: four and a half billion individuals, which has really doubled, 199 00:11:10,120 --> 00:11:12,439 Speaker 1: if not tripled, since the financial crisis. So I think 200 00:11:12,480 --> 00:11:16,480 Speaker 1: early efforts of using search activity UM is, I know, 201 00:11:16,520 --> 00:11:19,080 Speaker 1: a lot of it pre crisis kind of fell on 202 00:11:19,160 --> 00:11:21,480 Speaker 1: its face and kind of faded away. Google used to 203 00:11:21,520 --> 00:11:24,520 Speaker 1: have these curated indices um. I think they had twenty 204 00:11:24,600 --> 00:11:27,160 Speaker 1: five of them, kind of showing how the economy, economy 205 00:11:27,280 --> 00:11:30,360 Speaker 1: was moving um here and there. I think that that 206 00:11:31,559 --> 00:11:33,920 Speaker 1: what didn't work as well because we didn't have the 207 00:11:34,000 --> 00:11:37,520 Speaker 1: ubiquity of Google searches and really Internet access. And as 208 00:11:37,559 --> 00:11:41,480 Speaker 1: that improves, this type of information becomes that much more important. 209 00:11:41,640 --> 00:11:45,200 Speaker 1: I think to the investing process. How much do you 210 00:11:45,200 --> 00:11:50,000 Speaker 1: think the the unusual or the extreme circumstances surrounding the 211 00:11:50,080 --> 00:11:55,319 Speaker 1: coronavirus crisis are are distorting survey responses? And I asked 212 00:11:55,320 --> 00:11:57,760 Speaker 1: that because again, I've seen a lot of criticism of 213 00:11:57,800 --> 00:12:00,040 Speaker 1: the p m I s recently, and one of of 214 00:12:00,160 --> 00:12:03,880 Speaker 1: things people are saying about those surveys at the moment 215 00:12:04,080 --> 00:12:08,040 Speaker 1: is that respondents aren't really judging their experiences on a 216 00:12:08,080 --> 00:12:11,440 Speaker 1: month to month basis, but they're sort of responding by 217 00:12:11,480 --> 00:12:16,000 Speaker 1: comparing now to a period of relative normality. So everything's 218 00:12:16,040 --> 00:12:19,240 Speaker 1: getting skewed. Do you think the unusual nous of of 219 00:12:19,240 --> 00:12:24,240 Speaker 1: our current circumstances might be skewing survey data as well? Yes, 220 00:12:24,440 --> 00:12:26,440 Speaker 1: I think so. I think it's it's a combination. Like 221 00:12:26,480 --> 00:12:29,640 Speaker 1: you said earlier with stock flow, it's what type of 222 00:12:29,760 --> 00:12:32,600 Speaker 1: reaction have we had over the past couple of months. 223 00:12:32,640 --> 00:12:35,920 Speaker 1: UM I think has been more reflective within the survey data, 224 00:12:35,920 --> 00:12:39,199 Speaker 1: and we're seeing that breakdown between surge activity UM and surveys. 225 00:12:39,200 --> 00:12:41,679 Speaker 1: And we also have this big group think are almost 226 00:12:41,720 --> 00:12:44,680 Speaker 1: circular reference that occurs within a lot of the sentiment data. 227 00:12:44,679 --> 00:12:47,040 Speaker 1: So we all look to the equity market. We all 228 00:12:47,040 --> 00:12:49,640 Speaker 1: know that we can use the equity market essentially forecast 229 00:12:49,720 --> 00:12:53,640 Speaker 1: where consumer a confidence will be for the next month UM, 230 00:12:54,200 --> 00:12:56,720 Speaker 1: and I think a lot of that's feeding into some 231 00:12:56,800 --> 00:12:59,839 Speaker 1: of the more rosy consumer confidence numbers as well as 232 00:12:59,840 --> 00:13:03,280 Speaker 1: the UM eyes and again that's some somewhat of distortion 233 00:13:03,559 --> 00:13:06,560 Speaker 1: UM and why we seem to't like to rely on 234 00:13:06,600 --> 00:13:25,240 Speaker 1: the search activity for the most part. So let's talk 235 00:13:25,280 --> 00:13:28,960 Speaker 1: a little bit more about that search activity. How do 236 00:13:29,080 --> 00:13:31,960 Speaker 1: you take how do you get the data? First of all, 237 00:13:32,080 --> 00:13:34,920 Speaker 1: what does Google make available? And then how do you 238 00:13:35,040 --> 00:13:38,640 Speaker 1: present it in a form so that it's usable because 239 00:13:38,640 --> 00:13:43,120 Speaker 1: there's obviously seasonality factors the you know, you can't just 240 00:13:43,200 --> 00:13:45,959 Speaker 1: look at searches for a vacation and see whether they 241 00:13:45,960 --> 00:13:48,319 Speaker 1: go up or down because people don't vacation at the 242 00:13:48,360 --> 00:13:51,320 Speaker 1: same uh at the same pace all year round. So 243 00:13:51,400 --> 00:13:53,640 Speaker 1: how do you get the data from Google? What's that 244 00:13:53,720 --> 00:13:56,240 Speaker 1: process like? And then what do you do to actually 245 00:13:56,280 --> 00:13:59,440 Speaker 1: put it in a format such that it's not just 246 00:14:00,000 --> 00:14:04,079 Speaker 1: OLiS for investors, like just describe it overall? How work? Sure, 247 00:14:04,320 --> 00:14:07,000 Speaker 1: so we are able to access just like anybody else 248 00:14:07,120 --> 00:14:10,040 Speaker 1: via Google Trends, which there is an API to be 249 00:14:10,120 --> 00:14:13,480 Speaker 1: able to grab that information, and what we do is 250 00:14:13,520 --> 00:14:16,280 Speaker 1: we avoid using the specific search terms. So if we're 251 00:14:16,320 --> 00:14:19,240 Speaker 1: just going to say wash machine or vacuum UM, that 252 00:14:19,280 --> 00:14:23,120 Speaker 1: will include specifically that exact term UM. And we know 253 00:14:23,160 --> 00:14:26,680 Speaker 1: that there can be multiple variations of those actual text terms, 254 00:14:26,880 --> 00:14:28,360 Speaker 1: and so we want to pick up on that. The 255 00:14:28,400 --> 00:14:32,240 Speaker 1: beauty is Google curates and creates two different types of 256 00:14:32,320 --> 00:14:36,280 Speaker 1: groupings of search activity. And they do this for you know, 257 00:14:36,440 --> 00:14:38,840 Speaker 1: each and every country essentially, which is going to take 258 00:14:38,880 --> 00:14:41,480 Speaker 1: care of the major language barriers and issues that we'd 259 00:14:41,520 --> 00:14:44,920 Speaker 1: run into as well. And so that is they create categories, 260 00:14:45,640 --> 00:14:49,760 Speaker 1: which there are roughly und forty plus different categories, everything 261 00:14:49,800 --> 00:14:52,760 Speaker 1: from accounting services all the way out to urban transportation 262 00:14:52,760 --> 00:14:55,440 Speaker 1: which would be things like uber and lift. And then 263 00:14:55,480 --> 00:14:59,480 Speaker 1: they have topics and that can be anything from inflation 264 00:14:59,840 --> 00:15:03,480 Speaker 1: or those talking about disinflation, or gold bugs or bitcoin um. 265 00:15:03,520 --> 00:15:05,960 Speaker 1: And that's going to then be more encompassing and based 266 00:15:06,000 --> 00:15:09,720 Speaker 1: on their mapping of a numerous new it could be hundreds, 267 00:15:09,800 --> 00:15:13,000 Speaker 1: if not thousands of thousands in certain cases, of different 268 00:15:13,000 --> 00:15:17,040 Speaker 1: search terms and phrases that then get housed underneath those 269 00:15:17,080 --> 00:15:20,680 Speaker 1: individual UM topics. We can I stop you and ask 270 00:15:20,680 --> 00:15:23,080 Speaker 1: you a quick question right there, Sure the data the 271 00:15:23,160 --> 00:15:26,360 Speaker 1: year able to draw. Just to make clear, is that 272 00:15:26,440 --> 00:15:30,840 Speaker 1: the granular within those hundreds or thousands of terms, you're 273 00:15:30,880 --> 00:15:32,640 Speaker 1: able to get data for each one of those. You 274 00:15:32,640 --> 00:15:35,760 Speaker 1: could see beyond just the sort of general category. Yeah, 275 00:15:35,880 --> 00:15:37,640 Speaker 1: so we can drill down. There are ways to drill 276 00:15:37,680 --> 00:15:41,320 Speaker 1: down within the individual categories. We understand what the actual 277 00:15:41,400 --> 00:15:45,760 Speaker 1: searches are within those categories, but in order to create 278 00:15:45,800 --> 00:15:50,600 Speaker 1: a more encompassing indication of what the consumer business is 279 00:15:50,640 --> 00:15:53,960 Speaker 1: looking for or thinking about, we do then pull in 280 00:15:54,040 --> 00:15:57,960 Speaker 1: that search trend. Essentially, that's going to be an aggregation 281 00:15:58,280 --> 00:16:01,520 Speaker 1: of all of all those searches underneath a given topic 282 00:16:02,040 --> 00:16:05,400 Speaker 1: or underneath a given category. And like I said that, 283 00:16:05,520 --> 00:16:08,080 Speaker 1: one of the greatest things about the way that Google 284 00:16:08,160 --> 00:16:10,560 Speaker 1: set this up is that you are then able to say, 285 00:16:10,640 --> 00:16:14,920 Speaker 1: let's look at urban transportation uber and LIFT, and let's 286 00:16:14,920 --> 00:16:17,160 Speaker 1: look at it not just here in the US. Let's 287 00:16:17,200 --> 00:16:21,600 Speaker 1: go to UH somewhere like Germany, let's go to Australia, 288 00:16:21,720 --> 00:16:25,480 Speaker 1: or let's go to Japan UM and they take care of, 289 00:16:25,600 --> 00:16:29,000 Speaker 1: fortunately a lot of the language barriers in that urban 290 00:16:29,000 --> 00:16:33,320 Speaker 1: transportation that is translated into into um, you know, Japanese 291 00:16:33,440 --> 00:16:37,360 Speaker 1: or um you know whatever is being German, and so on. UM, 292 00:16:37,560 --> 00:16:40,640 Speaker 1: so getting into the course of how we then digest 293 00:16:40,720 --> 00:16:42,880 Speaker 1: and use that information. Like you said, is there's a 294 00:16:42,920 --> 00:16:45,280 Speaker 1: high degree of seasonality. Of course, it could be like 295 00:16:45,320 --> 00:16:48,040 Speaker 1: with clothing with back to school, or can be accounting 296 00:16:48,080 --> 00:16:51,680 Speaker 1: services coming into March, April and October. UM. So we 297 00:16:51,760 --> 00:16:56,480 Speaker 1: do decomposition where we'll we'll break down each individual topic 298 00:16:56,760 --> 00:17:00,520 Speaker 1: or categories search activity into three components, and that is 299 00:17:00,640 --> 00:17:02,720 Speaker 1: it's overall trend component. You think of it as like 300 00:17:02,760 --> 00:17:06,879 Speaker 1: as kind of a slower moving average trend of that 301 00:17:06,920 --> 00:17:09,640 Speaker 1: search activity. And then we have the seasonality that we're 302 00:17:09,680 --> 00:17:12,000 Speaker 1: able to then strip out. And then we also have 303 00:17:12,160 --> 00:17:15,119 Speaker 1: this thing we call the residual or the shock. What's 304 00:17:15,240 --> 00:17:18,919 Speaker 1: interesting about the experience that we've seen here in UM 305 00:17:18,920 --> 00:17:21,919 Speaker 1: with COVID nineteen is we were never so interested in 306 00:17:22,000 --> 00:17:25,560 Speaker 1: the shock component and the very quick um shifts in 307 00:17:25,640 --> 00:17:29,320 Speaker 1: search activity either positive or negative until COVID hit, when 308 00:17:29,320 --> 00:17:33,480 Speaker 1: we saw it's just substantial breaks from these trends and 309 00:17:33,520 --> 00:17:36,359 Speaker 1: what would be expected by seasonality. That can be anything 310 00:17:36,400 --> 00:17:39,959 Speaker 1: from the searching for you know, physical policy news, economic news, 311 00:17:40,359 --> 00:17:43,399 Speaker 1: how individuals are searching on the line, then for groceries 312 00:17:43,880 --> 00:17:47,639 Speaker 1: UM and making those type of consumer staples purchases. But 313 00:17:47,760 --> 00:17:49,600 Speaker 1: getting back to it, the idea is to break it 314 00:17:49,640 --> 00:17:51,720 Speaker 1: down into those three components that we get idea of 315 00:17:51,760 --> 00:17:54,360 Speaker 1: what is the you know, the long term trend UM 316 00:17:54,440 --> 00:17:57,800 Speaker 1: and shift really potentially in search activity. How does that 317 00:17:57,880 --> 00:18:00,440 Speaker 1: relate then to to what we're seeing within financial markets 318 00:18:00,520 --> 00:18:03,840 Speaker 1: and overall economic data. And then what are these shock 319 00:18:03,920 --> 00:18:08,600 Speaker 1: components and regarding those big distortions or shifts away from 320 00:18:08,600 --> 00:18:11,160 Speaker 1: those underlying trends, what does that have to tell us 321 00:18:11,760 --> 00:18:14,600 Speaker 1: about how things may be abruptly changing in the near 322 00:18:14,720 --> 00:18:17,199 Speaker 1: term UM and what that can mean? I mean, of 323 00:18:17,240 --> 00:18:21,240 Speaker 1: course for potential volatility UM and equity markets, uncertainty in 324 00:18:21,280 --> 00:18:24,600 Speaker 1: general UM from the consumer base UM and so on. 325 00:18:25,000 --> 00:18:27,160 Speaker 1: And so what we do is we pull down those 326 00:18:27,160 --> 00:18:30,200 Speaker 1: three pieces of information that then gets used within our 327 00:18:30,720 --> 00:18:33,320 Speaker 1: written content as well within our own models and our 328 00:18:33,320 --> 00:18:37,280 Speaker 1: clients models UM and so on. So correct me if 329 00:18:37,280 --> 00:18:40,040 Speaker 1: I'm wrong. But the data that you're using is mostly 330 00:18:40,640 --> 00:18:44,240 Speaker 1: public data. If investors all have access to the same data, 331 00:18:44,760 --> 00:18:49,919 Speaker 1: how are they using that to actually generate outperformance? How 332 00:18:49,920 --> 00:18:52,840 Speaker 1: do they differentiate how they're using the data versus how 333 00:18:53,000 --> 00:18:56,920 Speaker 1: another fund or another investor might be using the data? Right? 334 00:18:57,040 --> 00:18:59,119 Speaker 1: So I mean that's that's the question we get. We 335 00:18:59,280 --> 00:19:01,320 Speaker 1: probably get the most most is since we do deal 336 00:19:01,400 --> 00:19:04,040 Speaker 1: mainly again with with public forms of data, there's plenty 337 00:19:04,080 --> 00:19:06,880 Speaker 1: of alternative data that is private and the credit card 338 00:19:06,920 --> 00:19:10,360 Speaker 1: space and spending UM and so on. Is we try 339 00:19:10,400 --> 00:19:13,920 Speaker 1: to uncover data we think that is underutilized UM. And 340 00:19:13,960 --> 00:19:16,560 Speaker 1: in this case with all of our dealings specifically with 341 00:19:16,640 --> 00:19:20,960 Speaker 1: fixed income portfolio managers, pension fund managers UM and the like, 342 00:19:21,640 --> 00:19:24,520 Speaker 1: the use of search activity on a broader scale, on 343 00:19:24,600 --> 00:19:27,080 Speaker 1: a country by country, even a metro by metro level, 344 00:19:27,720 --> 00:19:31,360 Speaker 1: we believe has been under appreciated UM and non internalized 345 00:19:31,400 --> 00:19:33,600 Speaker 1: the extent that it could be. Now, like you said, once, 346 00:19:33,640 --> 00:19:37,720 Speaker 1: think something like this gets over used or gets used 347 00:19:37,760 --> 00:19:41,479 Speaker 1: as a key benchmark potentially to filling the latent gaps 348 00:19:41,520 --> 00:19:45,399 Speaker 1: between economic data. Potentially some that alpha creation could UM 349 00:19:45,560 --> 00:19:47,960 Speaker 1: evaporate UM and that would mean we have to move 350 00:19:47,960 --> 00:19:51,080 Speaker 1: on to some additional data sources for this time being. 351 00:19:51,480 --> 00:19:57,120 Speaker 1: In all our communications, the front offices of of investment managers, 352 00:19:57,600 --> 00:20:01,719 Speaker 1: banks and so on have not been heavy users of 353 00:20:01,760 --> 00:20:04,879 Speaker 1: the search activity. I think that early uses of it 354 00:20:05,359 --> 00:20:08,080 Speaker 1: prior to the crisis and during the crisis kind of 355 00:20:08,119 --> 00:20:13,360 Speaker 1: fell flat. Again. Maybe the ubiquity of actual Internet usage 356 00:20:13,440 --> 00:20:16,159 Speaker 1: and those young too old that we're using Google was 357 00:20:16,200 --> 00:20:19,760 Speaker 1: not there as of yet. And what we've seen over 358 00:20:19,840 --> 00:20:22,639 Speaker 1: the years, really since two thousand eleven two thousand and twelve, 359 00:20:22,680 --> 00:20:25,200 Speaker 1: search activities ability to fill the gap and really take 360 00:20:25,240 --> 00:20:29,720 Speaker 1: the place of surveys has improved markedly year after year UM. 361 00:20:29,720 --> 00:20:33,200 Speaker 1: And that's something we can measure UM statistically and VR 362 00:20:33,320 --> 00:20:37,520 Speaker 1: modeling for essentially those turning points as to when maybe 363 00:20:37,920 --> 00:20:41,440 Speaker 1: search activity loses its flare loses its ability UH to 364 00:20:41,600 --> 00:20:45,000 Speaker 1: then forecast and now cast via g d P retail 365 00:20:45,080 --> 00:20:48,720 Speaker 1: sales inflation UM and the like. But we're not there yet. 366 00:20:49,680 --> 00:20:54,080 Speaker 1: So obviously the demand for this data, and you mentioned 367 00:20:54,119 --> 00:20:57,880 Speaker 1: maybe search data is sort of relatively newly being incorporated 368 00:20:57,880 --> 00:21:00,720 Speaker 1: into investment processes, but for years we've been hearing about 369 00:21:01,760 --> 00:21:05,960 Speaker 1: satellite looking at parking lots in Walmart, or satellite looking 370 00:21:05,960 --> 00:21:09,560 Speaker 1: at trained or credit card data that's been out there 371 00:21:09,640 --> 00:21:13,639 Speaker 1: is a thing for a while. How intense is the 372 00:21:13,800 --> 00:21:18,480 Speaker 1: search basically for new data sources, either on the bi side, 373 00:21:18,520 --> 00:21:21,480 Speaker 1: the investor side, or you as sort of a data 374 00:21:21,560 --> 00:21:25,120 Speaker 1: vendor so to speak, to just constantly be coming up 375 00:21:25,160 --> 00:21:29,720 Speaker 1: with something that's relatively underappreciated. What does that process look 376 00:21:29,760 --> 00:21:34,399 Speaker 1: like the use of alternative data within the investment world. 377 00:21:34,480 --> 00:21:36,919 Speaker 1: You know, really the investment world was very late to 378 00:21:37,119 --> 00:21:41,760 Speaker 1: using alternative data UM, you know, compared to healthcare, even education, 379 00:21:42,400 --> 00:21:44,679 Speaker 1: UM and the like. And we initially saw this, you know, 380 00:21:45,000 --> 00:21:47,840 Speaker 1: in our routines of going out to big banks, for example, 381 00:21:48,200 --> 00:21:51,120 Speaker 1: and discussing with their teams, UM, you know, how they're 382 00:21:51,200 --> 00:21:54,520 Speaker 1: utilizing alternative data. It was almost always in the back office. 383 00:21:54,960 --> 00:21:56,760 Speaker 1: So it could be UM, you know, anything to do 384 00:21:56,880 --> 00:22:01,520 Speaker 1: with their customer relations. It could chatbots in terms of 385 00:22:01,560 --> 00:22:04,920 Speaker 1: creating natural better language, natural language processing and ployment data 386 00:22:04,960 --> 00:22:07,919 Speaker 1: for that. It could be trade matching all kinds of 387 00:22:07,960 --> 00:22:10,359 Speaker 1: different things that were done in the back office. They 388 00:22:10,359 --> 00:22:12,920 Speaker 1: were trying to basically bring in machine learning, bringing better 389 00:22:13,040 --> 00:22:15,639 Speaker 1: data to create better predictions. And it could have to 390 00:22:15,640 --> 00:22:18,080 Speaker 1: do again with their customers, which customers to call and 391 00:22:18,080 --> 00:22:21,879 Speaker 1: not call, who's going to potentially provide the best UM, 392 00:22:22,320 --> 00:22:25,320 Speaker 1: best avenue for new business UM and so on. But 393 00:22:25,359 --> 00:22:30,040 Speaker 1: what we've seen, i'd say, you know, starting roughly inen, 394 00:22:30,119 --> 00:22:33,000 Speaker 1: we started to see a UM with the advent of 395 00:22:33,720 --> 00:22:38,359 Speaker 1: more alternative data available via numerous vendors, the increase in 396 00:22:38,400 --> 00:22:43,119 Speaker 1: transfer to the front office has happened rapidly UM. And 397 00:22:43,160 --> 00:22:46,600 Speaker 1: I would say that now with COVID nineteen and the 398 00:22:46,680 --> 00:22:50,400 Speaker 1: inability for econ data to keep up with the actual 399 00:22:50,720 --> 00:22:54,440 Speaker 1: UM happenings of the economy UM, and really the needs 400 00:22:54,440 --> 00:22:56,679 Speaker 1: of investors to understand that just what's going on with 401 00:22:56,880 --> 00:23:01,200 Speaker 1: how rapidly things are changing. UM. The demand is just intense, 402 00:23:01,520 --> 00:23:03,800 Speaker 1: and so it calls to us and and calls I 403 00:23:03,880 --> 00:23:07,920 Speaker 1: know too many of our competitors in similar, similar alternative 404 00:23:08,000 --> 00:23:11,040 Speaker 1: data providers. UM has just shot to the moon and 405 00:23:11,160 --> 00:23:13,439 Speaker 1: you can see that again. Bloomberg of course offers some 406 00:23:13,480 --> 00:23:16,520 Speaker 1: of this alternative data. There's plenty of other repositories to 407 00:23:16,680 --> 00:23:19,720 Speaker 1: grab it, but I would say that the degree of 408 00:23:19,760 --> 00:23:23,840 Speaker 1: interest is increased tenfold UM since it's it's beginnings in 409 00:23:41,280 --> 00:23:45,160 Speaker 1: what's been your favorite alternative data set during the crisis, 410 00:23:45,240 --> 00:23:47,840 Speaker 1: Like what has either surprised you or what has been 411 00:23:48,000 --> 00:23:51,720 Speaker 1: most useful in judging the direction of the overall economy. 412 00:23:52,840 --> 00:23:57,000 Speaker 1: We've been benchmarking a lot. The mobility data that's available 413 00:23:57,119 --> 00:24:01,480 Speaker 1: via Apple UM and to cart lab is another one. 414 00:24:01,600 --> 00:24:05,160 Speaker 1: Google and Benjy benchmark in that off of search activity, 415 00:24:05,280 --> 00:24:09,600 Speaker 1: and I've been absolutely shocked at how well. Search activity 416 00:24:09,640 --> 00:24:12,600 Speaker 1: has been able to predict two things, and that's been 417 00:24:12,760 --> 00:24:15,560 Speaker 1: retail sales on a month over month basis and also 418 00:24:15,720 --> 00:24:18,560 Speaker 1: inflation on a month over month basis. A lot of 419 00:24:18,600 --> 00:24:22,919 Speaker 1: our kind of point forecasts looking forward, based on what 420 00:24:22,960 --> 00:24:25,440 Speaker 1: we believe are the most unique and important search activity 421 00:24:26,000 --> 00:24:30,480 Speaker 1: have done very well UM in predicting the rebound in May, 422 00:24:30,720 --> 00:24:36,600 Speaker 1: for example, the heavy damage done to transportation, energy, UM 423 00:24:36,640 --> 00:24:40,840 Speaker 1: and apparel within March and April. To CPI for example, 424 00:24:41,320 --> 00:24:44,959 Speaker 1: we had UM noticed the heavy degree of rebound in 425 00:24:45,320 --> 00:24:48,840 Speaker 1: all three of those categories, in particular UM within apparel, 426 00:24:49,200 --> 00:24:55,800 Speaker 1: which ultimately lead to rebound in overall apparel spending in May, 427 00:24:55,800 --> 00:25:00,159 Speaker 1: which then ultimately translated to higher inflation UM it was 428 00:25:00,200 --> 00:25:03,280 Speaker 1: reported in June. And so the search activity that we've 429 00:25:03,320 --> 00:25:05,560 Speaker 1: been able to use most utilized, which I think Joe 430 00:25:05,600 --> 00:25:08,360 Speaker 1: featured in a CHARTUM a number of weeks ago, has 431 00:25:08,400 --> 00:25:11,439 Speaker 1: to do with a series of key categories, and that 432 00:25:11,520 --> 00:25:14,480 Speaker 1: can be everything from beauty and fitness, which is we 433 00:25:14,520 --> 00:25:17,520 Speaker 1: found to be a highly leading indicator UM as well 434 00:25:17,520 --> 00:25:20,560 Speaker 1: as just the general public searching for economic news and 435 00:25:20,560 --> 00:25:24,640 Speaker 1: physical policy news revolving around welfare and unemployment and jobless 436 00:25:24,680 --> 00:25:28,719 Speaker 1: benefits welfare and unemployment. Unemployment itself has been a highly 437 00:25:28,800 --> 00:25:32,560 Speaker 1: leading indicator. And then also UM, one of the things 438 00:25:32,560 --> 00:25:35,880 Speaker 1: we picked up on very early was the incredible drive 439 00:25:36,119 --> 00:25:40,640 Speaker 1: for home improvement that really began in the final weeks 440 00:25:40,920 --> 00:25:44,440 Speaker 1: of March. UM. And what we had seen was this 441 00:25:44,680 --> 00:25:48,000 Speaker 1: effervent search activity UM, you know, looking across all the 442 00:25:48,040 --> 00:25:51,160 Speaker 1: major metros and all the major states of the United States, 443 00:25:51,600 --> 00:25:55,040 Speaker 1: heavy degree of need for our need, a desire to 444 00:25:55,200 --> 00:25:59,200 Speaker 1: place appliances, to paint their homes, to get a new roof, 445 00:25:59,320 --> 00:26:02,159 Speaker 1: new side, new carpeting, UM. And this is something that 446 00:26:02,200 --> 00:26:05,080 Speaker 1: really took place ahead of the Hares acting signed on 447 00:26:05,280 --> 00:26:08,720 Speaker 1: March seven, It began really two weeks before that, which 448 00:26:08,720 --> 00:26:10,960 Speaker 1: I think was a leading indicator that the consumer would 449 00:26:10,960 --> 00:26:15,119 Speaker 1: be stronger and potentially spend more UM than those uh, 450 00:26:15,200 --> 00:26:17,960 Speaker 1: the naysayers. And then that we had expected UM to 451 00:26:18,040 --> 00:26:21,080 Speaker 1: see given the calls for a recession and potential depression, 452 00:26:21,640 --> 00:26:25,000 Speaker 1: given the full stop to the economy. And it's really 453 00:26:25,080 --> 00:26:29,480 Speaker 1: striking just this week, UH, we've seen home depot and 454 00:26:29,600 --> 00:26:34,000 Speaker 1: lows post extraordinary sales. Home improvement has just been one 455 00:26:34,000 --> 00:26:38,640 Speaker 1: of the monster stories of this recovery. How much spending 456 00:26:38,880 --> 00:26:42,960 Speaker 1: and how sustained that's been I just want to drill 457 00:26:43,000 --> 00:26:45,080 Speaker 1: a little bit further down. I mean, it's clear that 458 00:26:45,119 --> 00:26:48,479 Speaker 1: like okay, if someone identified that trend at the end 459 00:26:48,520 --> 00:26:51,439 Speaker 1: of March and I saw what was going on, there 460 00:26:51,440 --> 00:26:53,920 Speaker 1: were huge investment opportunities because like again, like I said, 461 00:26:54,040 --> 00:26:56,680 Speaker 1: home depot lows, it's that are huge beneficiaries that their 462 00:26:56,680 --> 00:27:00,560 Speaker 1: stocks about extraordinary runs due to this, uh desire for 463 00:27:00,560 --> 00:27:03,200 Speaker 1: people that like renovate and fix things in their home 464 00:27:03,480 --> 00:27:07,359 Speaker 1: while they're working from home and so forth. How then, 465 00:27:07,760 --> 00:27:11,119 Speaker 1: do in your clients and when you talk to them, 466 00:27:11,160 --> 00:27:15,400 Speaker 1: how do they actually make a decision by or sell 467 00:27:15,920 --> 00:27:19,000 Speaker 1: based on the data and the context that you're giving 468 00:27:19,040 --> 00:27:20,919 Speaker 1: that What is the the you know, that's sort of 469 00:27:21,000 --> 00:27:23,960 Speaker 1: the last mild question, so to speak. They can get 470 00:27:23,960 --> 00:27:26,400 Speaker 1: the data from you, but then how are they actually 471 00:27:26,480 --> 00:27:29,920 Speaker 1: using it to form of you and take a risk 472 00:27:30,680 --> 00:27:33,640 Speaker 1: both on a subjective then also on an algorithmic basis. 473 00:27:33,640 --> 00:27:37,760 Speaker 1: We have many, many clients that are effectively now casting, 474 00:27:37,960 --> 00:27:42,240 Speaker 1: and so they're now casting either econ data, the econ environment, 475 00:27:42,720 --> 00:27:44,840 Speaker 1: and then as well the financial the impact on the 476 00:27:44,960 --> 00:27:47,880 Speaker 1: on the actual financial market in terms of producing their 477 00:27:47,920 --> 00:27:51,200 Speaker 1: own actual forecasts of where things will be one week 478 00:27:51,320 --> 00:27:53,920 Speaker 1: to six weeks to twelve weeks later. So the search 479 00:27:53,960 --> 00:27:59,400 Speaker 1: activity UM is one that we found provides a lead time. UM. 480 00:27:59,440 --> 00:28:02,479 Speaker 1: That's more you know, kind of medium term as opposed 481 00:28:02,480 --> 00:28:06,399 Speaker 1: to ultra high frequency short term. So uh, you know, 482 00:28:06,440 --> 00:28:10,000 Speaker 1: within the searches, just like survey data, UM, we're not 483 00:28:10,119 --> 00:28:13,439 Speaker 1: going to be able to help someone UM if if 484 00:28:13,560 --> 00:28:16,000 Speaker 1: effectively make a decision for that day. You know, what 485 00:28:16,119 --> 00:28:18,960 Speaker 1: is the next twenty four hours of economic activita? People 486 00:28:18,960 --> 00:28:21,639 Speaker 1: buying more watch machines, they buying more cars, m Are 487 00:28:21,640 --> 00:28:23,919 Speaker 1: they buying more apparel? It's not That's not exactly how 488 00:28:23,960 --> 00:28:27,000 Speaker 1: it works. It's a more immediate term, medium term focus 489 00:28:27,080 --> 00:28:30,840 Speaker 1: of UM. Varying lead times typically from one week to 490 00:28:30,920 --> 00:28:33,720 Speaker 1: eight weeks. So we have things, for example, like apparel 491 00:28:34,440 --> 00:28:37,240 Speaker 1: UM that will have a lead time of days to 492 00:28:37,400 --> 00:28:39,680 Speaker 1: a week UM, and then we'll have things like building 493 00:28:39,720 --> 00:28:42,880 Speaker 1: materials or roofing that will have a lead time of 494 00:28:42,920 --> 00:28:45,160 Speaker 1: seven to eight weeks. So then what our customers and 495 00:28:45,160 --> 00:28:48,840 Speaker 1: our clients are doing is taking that information in understanding 496 00:28:48,880 --> 00:28:51,160 Speaker 1: those lead times and then either in putting it to 497 00:28:51,160 --> 00:28:54,880 Speaker 1: their own subjective decision making process in order to affect 498 00:28:54,960 --> 00:28:57,800 Speaker 1: their decision. It could be a risk management one in 499 00:28:57,840 --> 00:29:00,680 Speaker 1: regards to their actual book or their position, determine if 500 00:29:00,680 --> 00:29:03,920 Speaker 1: there's something that could be disruptive to their position, or 501 00:29:03,960 --> 00:29:07,400 Speaker 1: it could be on the flip side, someone that's actually UM, 502 00:29:07,440 --> 00:29:10,840 Speaker 1: you know, using on a more tactical basis, that is, 503 00:29:10,880 --> 00:29:13,840 Speaker 1: and in putting it to their own now casting forecasting 504 00:29:13,840 --> 00:29:16,760 Speaker 1: process and that coming up with their own conclusion UM 505 00:29:16,960 --> 00:29:19,760 Speaker 1: of how will that supports or doesn't support their their 506 00:29:19,840 --> 00:29:23,320 Speaker 1: general idea. But UM with this data, along with a 507 00:29:23,320 --> 00:29:25,880 Speaker 1: lot of the natural language processing data that we work with, 508 00:29:26,320 --> 00:29:29,120 Speaker 1: UM does not have a high frequency basis. This is 509 00:29:29,120 --> 00:29:32,400 Speaker 1: something that's more medium term, if not long term in nature. 510 00:29:33,640 --> 00:29:35,360 Speaker 1: Then last thing like where do you see what's the 511 00:29:35,400 --> 00:29:38,360 Speaker 1: next big thing for you? In terms of I just 512 00:29:38,400 --> 00:29:40,800 Speaker 1: thinking back to when you said, Okay, at some point 513 00:29:41,160 --> 00:29:43,480 Speaker 1: the search data will get more used, the it will 514 00:29:43,480 --> 00:29:46,440 Speaker 1: get more modified, the alpha from having access to it 515 00:29:46,640 --> 00:29:50,520 Speaker 1: will theoretically diminish. What are the next frontiers in terms 516 00:29:50,560 --> 00:29:54,400 Speaker 1: of data that you think are interesting and potentially still 517 00:29:54,480 --> 00:29:57,880 Speaker 1: underappreciated or underutilized at this point. So I think the 518 00:29:57,960 --> 00:30:01,600 Speaker 1: advent of mobility data, for example with discard Dicart labs 519 00:30:01,640 --> 00:30:05,240 Speaker 1: that's able to zero in on specific retailers and look 520 00:30:05,280 --> 00:30:08,160 Speaker 1: at the actual foot traffic UM that's occurring coming to 521 00:30:08,280 --> 00:30:10,480 Speaker 1: them going away from them. It can be also down 522 00:30:10,480 --> 00:30:14,440 Speaker 1: to you know, parks UM specific locations within different metros 523 00:30:14,520 --> 00:30:17,880 Speaker 1: or rural areas. I think this mobility data, which we 524 00:30:17,920 --> 00:30:21,040 Speaker 1: don't have a high degree of historical data to work with, 525 00:30:21,640 --> 00:30:24,960 Speaker 1: is something that moving forward will become more and more 526 00:30:25,600 --> 00:30:29,480 Speaker 1: of the leading indicator that think individuals will seek for. Unfortunately, 527 00:30:29,760 --> 00:30:32,560 Speaker 1: you know, Apple, Google and the cart labs have you know, 528 00:30:32,600 --> 00:30:36,040 Speaker 1: they sell this data, so it's not necessarily publicly available. 529 00:30:36,560 --> 00:30:38,600 Speaker 1: But I think as they build a larger and larger 530 00:30:38,680 --> 00:30:41,960 Speaker 1: track record in order to benchmark that against anything begun 531 00:30:42,040 --> 00:30:45,920 Speaker 1: search activity, survey information of how consumers are operating, where 532 00:30:45,920 --> 00:30:48,120 Speaker 1: they're moving, and what they're doing. I think that is 533 00:30:48,360 --> 00:30:50,440 Speaker 1: more or less the kind of the cutting edge and 534 00:30:50,840 --> 00:30:54,000 Speaker 1: leading edge of understanding the consumer and now and then 535 00:30:54,040 --> 00:30:58,000 Speaker 1: how they're interacting with retail, interacting with people around them 536 00:30:58,160 --> 00:31:01,520 Speaker 1: using urban transportation UM and so on. And obviously in 537 00:31:01,520 --> 00:31:04,840 Speaker 1: this environment of COVID nineteen with how much we were 538 00:31:04,880 --> 00:31:07,479 Speaker 1: not moving around in March and April UM, I think 539 00:31:07,520 --> 00:31:10,880 Speaker 1: it'll be critical UM here moving forward to get a 540 00:31:10,880 --> 00:31:14,400 Speaker 1: better grasp on how much of a revival um economies. 541 00:31:14,400 --> 00:31:17,280 Speaker 1: There's economies are seeing and how mobile people have become 542 00:31:17,440 --> 00:31:19,360 Speaker 1: or it will be one of the fun ones we 543 00:31:19,400 --> 00:31:22,040 Speaker 1: didn't talk about, but I know it's uh, it's definitely um, 544 00:31:22,080 --> 00:31:24,360 Speaker 1: you know, fringe too, just because it's it's such a 545 00:31:24,400 --> 00:31:27,280 Speaker 1: strange space. Is that the Twitter sentiment is one that's 546 00:31:27,680 --> 00:31:31,240 Speaker 1: become I think, more and more useful UM in terms 547 00:31:31,280 --> 00:31:34,440 Speaker 1: of gauging actual investor sentiment. It's been pretty wild to 548 00:31:34,520 --> 00:31:38,600 Speaker 1: watch the number of economists and even formal central bankers 549 00:31:38,600 --> 00:31:41,440 Speaker 1: that have popped up on Twitter that use it pretty voraciously. 550 00:31:41,520 --> 00:31:43,920 Speaker 1: We even have like Christia Freeland, Um, you just took 551 00:31:43,960 --> 00:31:47,760 Speaker 1: over a finance minister in Canada. There's just such noteworthy individuals, 552 00:31:47,800 --> 00:31:51,040 Speaker 1: and it's it's become something that's become more and more predictive, 553 00:31:51,160 --> 00:31:54,160 Speaker 1: I think, not necessarily a direction of equity markets, but 554 00:31:54,240 --> 00:31:57,640 Speaker 1: more or less a gauge of uncertainty UM and you know, 555 00:31:57,680 --> 00:32:01,800 Speaker 1: financial market volatility. So we've built a lot of algorithms too. 556 00:32:02,040 --> 00:32:03,959 Speaker 1: It's just like another thing where it's kind of like 557 00:32:04,080 --> 00:32:06,960 Speaker 1: because I know people were like interested that ten years ago, 558 00:32:07,000 --> 00:32:09,920 Speaker 1: but it wasn't enough interest. There weren't enough people on 559 00:32:09,960 --> 00:32:13,560 Speaker 1: there for Twitter or social media to be representative, but 560 00:32:13,680 --> 00:32:15,320 Speaker 1: sort of like kind of like search where you can 561 00:32:15,320 --> 00:32:19,440 Speaker 1: actually get a big enough cross section that it's meaningful exactly. 562 00:32:19,520 --> 00:32:23,200 Speaker 1: So that's the what's absolutely wild is the number of 563 00:32:23,760 --> 00:32:27,960 Speaker 1: people provide providing original content and the speed by which 564 00:32:27,960 --> 00:32:32,160 Speaker 1: they are actually tweeting has accelerated just demonstably. So we 565 00:32:32,200 --> 00:32:36,440 Speaker 1: saw this incredible crescendo in tweeting and Twitter activity in 566 00:32:36,520 --> 00:32:39,600 Speaker 1: fin twitt through really the middle of March, and it's 567 00:32:39,680 --> 00:32:42,720 Speaker 1: just held there ever since. With this COVID you know, 568 00:32:42,760 --> 00:32:46,920 Speaker 1: pandemic everyone at home, UM and really grasping for information. 569 00:32:47,600 --> 00:32:49,480 Speaker 1: So it's been fun to be able to break down 570 00:32:49,560 --> 00:32:52,120 Speaker 1: all the different opponents of Twitter, which we do into 571 00:32:52,400 --> 00:32:56,160 Speaker 1: is based on clustering prior to the financial crisis. We 572 00:32:56,200 --> 00:33:00,320 Speaker 1: break it down into primables, bears, primatists, UM, economy US 573 00:33:00,440 --> 00:33:03,040 Speaker 1: UM and the like UH, and then able to grab 574 00:33:03,040 --> 00:33:05,720 Speaker 1: out you know, how are they feeling about liquidity in 575 00:33:05,720 --> 00:33:07,880 Speaker 1: the market, how are they feeling about the equity market? 576 00:33:07,880 --> 00:33:11,440 Speaker 1: COVID nineteen UM, I had the consumer UM and so on, 577 00:33:11,560 --> 00:33:14,800 Speaker 1: And it's amazing pulling in the information, like you said, 578 00:33:14,880 --> 00:33:17,920 Speaker 1: prior to just three or four years ago, its ability 579 00:33:18,000 --> 00:33:21,240 Speaker 1: to actually get ahead of and forecast, you know, volatility 580 00:33:21,280 --> 00:33:24,000 Speaker 1: and maybe a little bit of financial market direction is 581 00:33:25,000 --> 00:33:28,440 Speaker 1: improved significantly. So it's it's it's been an interesting space 582 00:33:28,480 --> 00:33:32,320 Speaker 1: to dabble into. Ben. That was great, Ben bright Hole, 583 00:33:32,400 --> 00:33:35,440 Speaker 1: I really appreciate you joining us. This feels like such 584 00:33:35,440 --> 00:33:39,320 Speaker 1: a big area and there's such a clear explanation of 585 00:33:39,360 --> 00:33:41,880 Speaker 1: how it all worked. Thank you for coming on. All right, 586 00:33:41,880 --> 00:33:48,440 Speaker 1: Thanks Joe be so much fun. That's really interesting. Yeah, 587 00:33:54,920 --> 00:33:56,640 Speaker 1: I thought that was great. You know, I do feel 588 00:33:56,640 --> 00:34:00,200 Speaker 1: like just from us from a media perspective, we you've 589 00:34:00,280 --> 00:34:04,360 Speaker 1: never used alternative real time data as much as we 590 00:34:04,440 --> 00:34:06,600 Speaker 1: have over the last six months, and so I thought 591 00:34:06,600 --> 00:34:09,719 Speaker 1: it was great to hear how it's actually collected and 592 00:34:09,760 --> 00:34:14,720 Speaker 1: then how it's actually used to put into an investment process. Yeah, 593 00:34:14,760 --> 00:34:17,120 Speaker 1: it's funny like thinking back to this now, but I 594 00:34:17,160 --> 00:34:20,760 Speaker 1: remember in I guess it would have been February telling 595 00:34:20,840 --> 00:34:26,600 Speaker 1: someone about how we were tracking movie bookings in our 596 00:34:26,719 --> 00:34:30,640 Speaker 1: theater bookings in South Korea because of the COVID outbreak there, 597 00:34:30,760 --> 00:34:32,919 Speaker 1: and the person I was telling it to just thought 598 00:34:32,960 --> 00:34:35,200 Speaker 1: it was like so unusual and so amazing. But of 599 00:34:35,239 --> 00:34:38,840 Speaker 1: course now everywhere around the world and especially in the US, 600 00:34:38,880 --> 00:34:41,960 Speaker 1: people are looking at all sorts of those kinds of things, 601 00:34:42,000 --> 00:34:45,040 Speaker 1: from restaurant bookings to the mobility data that Ben was 602 00:34:45,080 --> 00:34:48,480 Speaker 1: talking about. Um, it's kind of become normal. Yeah, And 603 00:34:48,520 --> 00:34:52,600 Speaker 1: I'm really fascinated by the sort of you know, the 604 00:34:53,000 --> 00:34:55,880 Speaker 1: speed with which sort of alpha deterior rates. So you 605 00:34:55,880 --> 00:34:59,240 Speaker 1: can imagine the first person who really discovers that search 606 00:34:59,280 --> 00:35:02,960 Speaker 1: indications for certain terms has some predictive value. There's a 607 00:35:03,000 --> 00:35:05,879 Speaker 1: lot of money to be made in that. But look, 608 00:35:05,920 --> 00:35:08,040 Speaker 1: I mean we're talking about it on the podcast that 609 00:35:08,480 --> 00:35:12,000 Speaker 1: Ben's active, and pretty soon you have to figure that 610 00:35:12,080 --> 00:35:14,560 Speaker 1: will be table stakes that people will be searching for 611 00:35:14,840 --> 00:35:17,520 Speaker 1: the next, the next thing, that that is a process 612 00:35:17,560 --> 00:35:21,239 Speaker 1: that will essentially never stop. Yeah, I think that's right. 613 00:35:21,280 --> 00:35:24,040 Speaker 1: But also I think what becomes clearer from speaking with 614 00:35:24,120 --> 00:35:29,600 Speaker 1: Ben is that understanding the data, how it's collected, and 615 00:35:29,640 --> 00:35:32,719 Speaker 1: how you can actually apply it is really really important. 616 00:35:32,800 --> 00:35:36,000 Speaker 1: So even with something like the mobility data, it's very 617 00:35:36,080 --> 00:35:39,120 Speaker 1: useful at the moment, but I think it's benchmark to 618 00:35:39,960 --> 00:35:42,879 Speaker 1: early January or something like that. So it's really good 619 00:35:42,880 --> 00:35:45,640 Speaker 1: to be aware when the summer comes around that the 620 00:35:45,680 --> 00:35:49,160 Speaker 1: benchmark that you're comparing the data to might not be 621 00:35:49,200 --> 00:35:53,440 Speaker 1: you know, completely applicable to warmer weather. So there's all 622 00:35:53,480 --> 00:35:56,759 Speaker 1: these quirks in each data set that you really have 623 00:35:56,880 --> 00:35:59,160 Speaker 1: to get to know. Yeah, totally. I mean even with 624 00:35:59,239 --> 00:36:02,239 Speaker 1: the Google Day it a just having the sort of 625 00:36:02,320 --> 00:36:07,040 Speaker 1: experience to adjust for seasonality it takes. Those are all 626 00:36:07,120 --> 00:36:08,680 Speaker 1: things that if you were to say, if I were 627 00:36:08,719 --> 00:36:12,720 Speaker 1: to just look on Google trends and look at that vacations, 628 00:36:13,200 --> 00:36:15,280 Speaker 1: it would be hard for me to get much signal 629 00:36:15,320 --> 00:36:18,480 Speaker 1: unless I really like understood the data and had experience 630 00:36:18,560 --> 00:36:22,080 Speaker 1: working with it. Mm hmm, yeah exactly. All right, shall 631 00:36:22,120 --> 00:36:25,720 Speaker 1: we leave it there? Yeah, okay, this has been another 632 00:36:25,719 --> 00:36:28,680 Speaker 1: episode of the ad Thoughts podcast. I'm Tracy Alloway. You 633 00:36:28,680 --> 00:36:31,600 Speaker 1: can follow me on Twitter at Tracy Alloway and I'm 634 00:36:31,680 --> 00:36:34,440 Speaker 1: Joe Wisn't Thought. You can follow me at the Stalwarts, 635 00:36:34,520 --> 00:36:37,760 Speaker 1: and you should follow our guest Ben Brightholtz on Twitter. 636 00:36:38,080 --> 00:36:41,759 Speaker 1: He posts tons of interesting charts from the uh the 637 00:36:41,920 --> 00:36:46,120 Speaker 1: Arbor research work that they do. Follow him at Ben Brightholtz. 638 00:36:46,320 --> 00:36:50,560 Speaker 1: Follow our producer on Twitter, Laura Carlson. She's at Laura M. Carlson. 639 00:36:50,880 --> 00:36:54,200 Speaker 1: Followed the Bloomberg head of podcast, Francesco Leavi at Francesca 640 00:36:54,320 --> 00:36:57,719 Speaker 1: Today and check out all of our podcasts at Bloomberg 641 00:36:57,760 --> 00:37:00,560 Speaker 1: unto the handle at podcast I for listening.