WEBVTT - Dan Ariely on How To Win Big by Betting on Human Capital 0:00:10.760 --> 0:00:15.120 Hello, and welcome to another episode of the Odd Lots Podcast. 0:00:15.200 --> 0:00:19.640 I'm Joe Wisenthal and I'm Tracy all Away. So, Tracy, 0:00:19.760 --> 0:00:23.479 something I've noticed before is we talked about E s 0:00:23.520 --> 0:00:27.440 G investing sometimes, and of course, as we've discussed in 0:00:27.480 --> 0:00:30.440 the past, you know, it's become this huge industry. The 0:00:30.480 --> 0:00:32.280 thing that I always think about in E s G 0:00:32.600 --> 0:00:35.480 is the E the environment. There's a lot of talk 0:00:35.560 --> 0:00:38.360 about climate and so forth, and it really is like 0:00:38.400 --> 0:00:40.240 I don't really know anything about what the s of 0:00:40.320 --> 0:00:43.000 the G are all about. Yeah, that would be the 0:00:43.159 --> 0:00:46.760 social governance part of the equation. And it's absolutely true 0:00:46.760 --> 0:00:48.920 when you think about the E s G space, most 0:00:48.960 --> 0:00:51.440 of the action tends to be on doing things to 0:00:51.520 --> 0:00:56.320 combat climate change. There's quite a bit um, but not 0:00:56.440 --> 0:01:01.480 as much about things like gender equality. But it beyond that, 0:01:01.600 --> 0:01:05.200 like this sort of social aspect you just don't hear 0:01:05.240 --> 0:01:08.360 about that much. Yeah, exactly right. And this sort of 0:01:08.400 --> 0:01:13.280 like the connection between these social aspects and returns is 0:01:13.319 --> 0:01:15.680 always interesting and it's always sort of one of the 0:01:15.760 --> 0:01:17.640 central questions of all the sort of E s G 0:01:17.959 --> 0:01:21.920 is like how much is it about investing with one's values? 0:01:22.200 --> 0:01:25.360 People want to invest in companies where they feel comfortable 0:01:25.360 --> 0:01:28.320 with the values of the company versus searching out sort 0:01:28.319 --> 0:01:31.560 of E. S G signals that can actually deliver alpha 0:01:31.760 --> 0:01:36.080 and somehow generate higher returns. Yeah. I think that's a 0:01:36.080 --> 0:01:38.560 really good way of framing it. And it's sort of 0:01:38.680 --> 0:01:41.679 unclear at this moment of time whether or not you 0:01:41.680 --> 0:01:46.240 would expect out performance of a good company because more 0:01:46.280 --> 0:01:50.840 people hopefully are interested in, you know, participating in a 0:01:50.960 --> 0:01:54.280 stock of a company that's doing good things, or if 0:01:54.320 --> 0:01:57.680 the company itself is sort of outperforming because it's more 0:01:57.800 --> 0:02:02.480 conscientious on things like the environment or gender or wealth 0:02:02.520 --> 0:02:05.800 inequality and things like that. Exactly right, So let's just 0:02:05.840 --> 0:02:07.760 get right into it. I'm really excited today we're gonna 0:02:07.760 --> 0:02:11.160 be talking about the S in E S G. And 0:02:11.280 --> 0:02:15.000 we have a world famous guest that I'm excited to 0:02:15.040 --> 0:02:17.880 be speaking with. We're gonna be speaking with um Dan 0:02:18.000 --> 0:02:22.679 ari Eli. He is a famous behavioral economist at Duke University, 0:02:22.720 --> 0:02:25.360 and he is also the co founder of a firm 0:02:25.400 --> 0:02:29.600 called Irrational Capital, who has formed five years ago, and 0:02:29.680 --> 0:02:33.760 it pursues the idea of looking at a company's human 0:02:33.960 --> 0:02:37.480 capital factor as a as something that could drive out 0:02:37.520 --> 0:02:40.359 performance in an investment. So I don't even know what 0:02:40.440 --> 0:02:43.400 the human capital factor is, but I'm excited to hear 0:02:43.960 --> 0:02:47.640 Dan talk about it and what he's learned in five years. Dan, 0:02:47.639 --> 0:02:50.960 thank you so much for coming on odd lot my pleasion. 0:02:51.040 --> 0:02:53.239 Nice to be here. Well, first of all, I love 0:02:53.280 --> 0:02:56.880 the name irrational I love the name irrational capital. It's perfect. 0:02:56.880 --> 0:02:59.800 But I'm curious what is the founding story of this? 0:03:00.400 --> 0:03:02.840 So the founding story is I'm a I'm a university professor. 0:03:02.880 --> 0:03:05.840 I do research on a few things, but among the 0:03:05.960 --> 0:03:11.240 human motivation. And in my academic career, I've from time 0:03:11.280 --> 0:03:13.840 to time I go to a company and I change 0:03:13.880 --> 0:03:18.040 things around. I change bonuses, I try to increase productivity, 0:03:18.160 --> 0:03:21.000 try to get people to care more about work. And 0:03:21.040 --> 0:03:23.200 my my experience has been that it's always been very 0:03:23.240 --> 0:03:26.480 easy to come and improve what people do and how 0:03:26.520 --> 0:03:30.440 to increase motivation because most companies just don't think about 0:03:30.480 --> 0:03:33.720 it very carefully. You know, if you think about h R, 0:03:34.800 --> 0:03:37.920 h R is usually a function that is about legal 0:03:38.000 --> 0:03:43.720 issues and I don't know training modules, but but it's 0:03:43.720 --> 0:03:45.920 not really a function that says, let's just get the 0:03:45.960 --> 0:03:48.480 best out of people. Let's just think about how do 0:03:48.520 --> 0:03:50.480 we motivate people, how do we get people to come 0:03:50.840 --> 0:03:53.240 happy to work? So I've been doing this for a 0:03:53.240 --> 0:03:57.360 long time and it's it's easy to do, and it's 0:03:57.440 --> 0:04:00.760 it's it's helpful. But when I met Dave, my partner 0:04:01.680 --> 0:04:04.040 can ask me whether I think that we could also 0:04:04.240 --> 0:04:06.920 look at something broader than instead of one company at 0:04:06.920 --> 0:04:10.320 a time, whether there's some way to look at companies, 0:04:10.880 --> 0:04:13.240 see how they treat their employees, see how the employees 0:04:13.280 --> 0:04:17.440 feel about the company, and whether this could predict stock 0:04:17.480 --> 0:04:20.479 market return. And I said, I don't know, that's what 0:04:20.640 --> 0:04:23.719 academics answer. I don't know that's the standard answer, but 0:04:23.839 --> 0:04:26.279 we can try it out. So we went on the 0:04:26.360 --> 0:04:30.920 hunt for data to see whether this hypothesis would ald 0:04:31.000 --> 0:04:33.000 or not, and it turns out it holds very well. 0:04:33.960 --> 0:04:37.080 So can I just press you on one point, which 0:04:37.120 --> 0:04:41.760 is what exactly is the definition of human capital? What 0:04:41.800 --> 0:04:47.720 are you looking at as you know undertake this exercise? Yea, 0:04:47.839 --> 0:04:49.840 so in this, in this exercise, and there's lots of 0:04:49.839 --> 0:04:52.880 ways to think about it, you're absolutely right. In this exercise, 0:04:52.920 --> 0:04:56.480 we said, let's take everything we can measure, everything that 0:04:56.520 --> 0:04:59.960 we can have access to, and then let's see which 0:05:00.120 --> 0:05:05.600 one of those things correlate actually predict stock market return. 0:05:06.600 --> 0:05:09.159 And and we we had some theories from the beginning 0:05:09.920 --> 0:05:12.880 about what would matter and what would not, but we 0:05:12.960 --> 0:05:16.720 also gave the data an opportunity to speak. And most 0:05:16.760 --> 0:05:19.200 of the things we found are very, very consistent with 0:05:20.160 --> 0:05:22.839 what we find in social science. So, for example, we 0:05:22.960 --> 0:05:27.200 find that absolute salary levels of absolute salary don't matter 0:05:27.279 --> 0:05:30.880 so much to our performance in the stock market, but 0:05:31.000 --> 0:05:34.760 the perception of fairness of salary matters a lot. Right. 0:05:34.880 --> 0:05:41.000 We find that physical environment at work, things like tables, chairs, coffee, 0:05:41.680 --> 0:05:46.560 don't matter so much. The sense of being appreciated matters 0:05:46.560 --> 0:05:49.680 a lot. So so when we think about human capital, 0:05:49.760 --> 0:05:52.800 we we start with the academic definition of let's take 0:05:52.839 --> 0:05:56.960 everything we know about what creates motivation. And for example, 0:05:57.000 --> 0:06:00.599 I did a study in which we showed that pizza 0:06:01.080 --> 0:06:03.440 delivered to somebody at home it can be much more 0:06:03.480 --> 0:06:06.400 motivating than a bonus if you think about the good 0:06:06.440 --> 0:06:09.720 will that that results from that. Right, So we've did 0:06:09.920 --> 0:06:13.479 lots of research on appreciation, so we said, hey, appreciation 0:06:13.560 --> 0:06:15.880 is really important. Let's let's see if we can find 0:06:15.920 --> 0:06:18.839 signals that companies that treat people in a way that 0:06:18.880 --> 0:06:21.599 they feel appreciated actually do better in the stock market. 0:06:21.960 --> 0:06:24.640 And and of course our exercise is to take the 0:06:24.760 --> 0:06:28.400 human capital in company X in year one and predicts 0:06:28.440 --> 0:06:31.520 the stock markets returned the following year. Right, we're not 0:06:31.560 --> 0:06:34.360 trying to estimate at that moment. We're trying to say, 0:06:34.480 --> 0:06:38.200 how is this human capital going to basically translate into 0:06:38.240 --> 0:06:42.240 better procedures, better products, more more innovation and so on. 0:06:42.640 --> 0:06:44.880 So so there's lots of things like that. And we 0:06:45.200 --> 0:06:47.400 have data that goes back to two thousand and six 0:06:47.440 --> 0:06:51.560 which we based a lot of our analysis on. But 0:06:51.640 --> 0:06:54.760 we also went and and double down during COVID. We 0:06:54.760 --> 0:06:59.280 studied fourteen hundred companies during COVID and and everything we 0:06:59.440 --> 0:07:03.680 found that was important before COVID became even more important 0:07:04.640 --> 0:07:07.960 during COVID and And the reason is think about kids, 0:07:08.680 --> 0:07:11.440 and if a kid is in the classroom, the teacher 0:07:11.480 --> 0:07:14.520 can control them to some degree. Right, it's straight, don't 0:07:14.520 --> 0:07:17.400 look at your phone, focus on this. And the same 0:07:17.440 --> 0:07:19.520 thing is true for us in the office. Right, nobody 0:07:19.560 --> 0:07:23.440 says it's straight, but but we're being seen by other people. 0:07:23.560 --> 0:07:25.720 We go to meetings, we have to focus. We can't 0:07:25.720 --> 0:07:29.480 do completely different things all of a sudden when people 0:07:29.480 --> 0:07:33.720 work from home. The role of intrinsic motivation is increasing dramatically, 0:07:34.960 --> 0:07:37.880 and because the kind of things we test are basically 0:07:37.880 --> 0:07:40.320 the things that lead to goodwill, the things that lead 0:07:40.360 --> 0:07:44.960 to extra performance, extra motivation, the role of everything we 0:07:45.080 --> 0:07:50.480 studied has increased quite dramatically during COVID. So this raises 0:07:50.480 --> 0:07:54.920 the obvious question, which is with traditional factors. Of course, 0:07:54.960 --> 0:07:57.760 you can look at the ten k or learning his 0:07:57.840 --> 0:08:02.760 report with environment fair there's increasingly companies are putting out 0:08:02.840 --> 0:08:07.720 some sort of separate perhaps an environmental scorecard that talks 0:08:07.720 --> 0:08:11.560 about climate and what they're doing out sustainability. There, how 0:08:11.600 --> 0:08:14.880 do you gather data in a systemic way on the 0:08:14.960 --> 0:08:17.840 types of things you're talking about, such as fairness, perception, 0:08:18.040 --> 0:08:22.120 and employee appreciation. Yeah, so there's lots of ways to 0:08:22.160 --> 0:08:25.120 think about data. You can think about data from linked In, 0:08:25.240 --> 0:08:27.200 and you can think about data from glass Door, and 0:08:27.240 --> 0:08:30.840 you can think about those. There's also companies that collect 0:08:30.920 --> 0:08:34.240 surveys about how employees feel about the company, and we 0:08:34.280 --> 0:08:37.760 take all of them. And in the first few years 0:08:37.840 --> 0:08:42.360 of our endeavor of irrational capital, we basically said, let's 0:08:42.360 --> 0:08:45.080 just get as much data as possible and see what 0:08:45.320 --> 0:08:48.160 matters and what doesn't matter. But once we found out 0:08:48.640 --> 0:08:51.760 what matters, now we can focus. And I want to 0:08:51.800 --> 0:08:53.440 kind of give you an example of why this is 0:08:53.440 --> 0:08:58.040 so important. So think about gender equality. What an important topic? Right? 0:08:58.160 --> 0:09:01.800 How important? It's incredibly important for investors to care about 0:09:02.600 --> 0:09:05.840 gender equality. And and there are two reasons. One is, 0:09:06.400 --> 0:09:08.839 you know, the moral issue, and the second one is performance. 0:09:09.000 --> 0:09:12.200 If you discriminate half of your workforce, that can't possibly 0:09:12.240 --> 0:09:16.080 be good for business. So there is something called the 0:09:16.120 --> 0:09:19.800 She Index, and the She Index counts how many women 0:09:19.840 --> 0:09:22.679 are in top positions and how many women are on 0:09:22.720 --> 0:09:26.440 the board. And if you think about what that means 0:09:26.840 --> 0:09:29.880 that the people who created this index had the thought 0:09:30.480 --> 0:09:32.679 that this is a good proxy for how a company 0:09:32.720 --> 0:09:35.439 is treating women. But if you look at the performance 0:09:35.440 --> 0:09:38.360 of the SMP of the SMP five compared to the 0:09:38.400 --> 0:09:44.080 SHE Index, the She Index is dramatically underperforming the dramatically 0:09:44.120 --> 0:09:49.280 every year systematically. Now is this because it's not a 0:09:49.280 --> 0:09:52.800 good idea to treat women equally? Of course not, but 0:09:52.800 --> 0:09:56.280 but it's a proxy that doesn't really capture the essence 0:09:56.920 --> 0:10:00.240 of what it means to be equal in country US. 0:10:00.400 --> 0:10:03.480 We took our data and in our data, we looked 0:10:03.520 --> 0:10:06.160 at how women feel compared to men on all kinds 0:10:06.200 --> 0:10:09.960 of things, and it turns out that the delta between 0:10:10.000 --> 0:10:13.439 women and men is incredibly important, you know it. It 0:10:13.760 --> 0:10:16.280 doesn't matter so much if you're a company treats everybody 0:10:16.320 --> 0:10:19.960 well or everybody badly, that doesn't matter so much. It's 0:10:20.000 --> 0:10:23.079 the inequality that is very hard to get to. Because 0:10:23.480 --> 0:10:26.240 if a company treats everybody by the way, it's good. 0:10:26.520 --> 0:10:29.319 I'm not recommending that people start treating their employees badly, 0:10:29.360 --> 0:10:32.800 but but in general, it matters less if you're equal 0:10:32.840 --> 0:10:35.959 offender because people get used to it and that's the 0:10:36.360 --> 0:10:39.640 level in which they evaluate themselves. But if you're a 0:10:39.679 --> 0:10:42.520 woman and you feel that somebody next to in the 0:10:42.559 --> 0:10:45.520 cubicle next to you is treated slightly better because they 0:10:45.520 --> 0:10:49.079 have a different chromosome structure, that's just bothers you, no 0:10:49.280 --> 0:10:52.440 end and it doesn't go away. So so as an example, 0:10:53.120 --> 0:10:55.720 if we take our data set from two thousand and 0:10:55.760 --> 0:11:01.400 six and every year we calculate which companies are the 0:11:01.440 --> 0:11:04.600 ones that are treating their women most equal to men, 0:11:05.480 --> 0:11:08.520 and we buy the top twenty percent companies every year 0:11:08.679 --> 0:11:11.560 who are doing that, and we construct a portfolio like this. 0:11:12.600 --> 0:11:16.240 That portfolio has a return of about five point four 0:11:16.280 --> 0:11:19.920 percent a year above the SNP just just by looking 0:11:19.960 --> 0:11:23.240 at equality nothing else. You're saying, let's just invest in 0:11:23.280 --> 0:11:27.720 the top twenty percent companies who are treating women the 0:11:27.800 --> 0:11:30.360 most equal, and you compare that to the she in 0:11:30.360 --> 0:11:35.119 the xtation the excess underperforming. So so it's incredibly important 0:11:35.440 --> 0:11:40.199 to actually do the research correctly and to start focusing 0:11:40.240 --> 0:11:43.320 on what's important to measure right. In the same way 0:11:43.320 --> 0:11:46.680 that I told you salary doesn't matter so much relative 0:11:46.720 --> 0:11:50.640 salary fairness in salary matters a lot. We need to 0:11:50.679 --> 0:11:52.800 figure out first what are the things that are actually 0:11:52.840 --> 0:11:56.120 good proxies of motivation, and what are the things that 0:11:56.160 --> 0:12:00.240 we think are predictive of motivation but really have thing 0:12:00.240 --> 0:12:04.480 to do with it. We found that sometimes when companies 0:12:04.559 --> 0:12:09.400 appoint women to high positions, it actually backfires, And and 0:12:09.440 --> 0:12:13.080 the reason it backfires is because they make it clear 0:12:13.160 --> 0:12:17.480 in some way to the women are working there that 0:12:17.559 --> 0:12:20.560 they don't care about women issues. They're only doing things 0:12:20.600 --> 0:12:25.560 for pr purposes. I think, think about the women engineer 0:12:25.600 --> 0:12:29.280 at company X, and one day the company is appointing 0:12:29.600 --> 0:12:34.640 a wonderful woman to the board. If if nothing follows 0:12:34.960 --> 0:12:37.840 in terms of her daily life. She is going to 0:12:37.880 --> 0:12:42.520 basically say, this company doesn't really care. They're doing things 0:12:42.600 --> 0:12:46.400 for pr they're checking the box. That's that's not so 0:12:46.480 --> 0:12:49.840 it actually creates sometimes what we call window dressing, that 0:12:49.880 --> 0:13:08.760 it's just for outside purposes, and then it backfires. So 0:13:09.040 --> 0:13:13.040 I love the nuance that you just described in measuring 0:13:13.200 --> 0:13:16.800 gender inequality, the difference between just counting up the number 0:13:17.080 --> 0:13:21.080 of women in leadership positions or who are on the board, 0:13:21.280 --> 0:13:26.000 versus the discrepancy in how women and men perceived that 0:13:26.040 --> 0:13:29.720 they are actually being treated. And I guess this is 0:13:29.720 --> 0:13:31.360 a topic sort of close to my heart, but I 0:13:31.640 --> 0:13:35.040 really think people should think more about it. But in 0:13:35.080 --> 0:13:38.840 thinking about this issue, there are clearly different ways, um 0:13:38.880 --> 0:13:42.720 to ask the question of how do you feel the 0:13:42.760 --> 0:13:46.440 company is doing on gender equality? Doesn't treat men and 0:13:46.440 --> 0:13:49.679 women the same, how do you feel the company treats you. 0:13:50.320 --> 0:13:52.680 I'm curious, just on the data collection side, how do 0:13:52.720 --> 0:13:57.280 you control for different surveys that are, you know, being 0:13:57.320 --> 0:14:01.880 done by different consultancies, for are different firms, Because that 0:14:01.880 --> 0:14:05.440 seems like a pretty big challenge. Yeah. So, so, first 0:14:05.440 --> 0:14:07.640 of all, we have we've worked with all kinds of 0:14:07.720 --> 0:14:10.160 data providers and of course it's easiest to work on 0:14:10.400 --> 0:14:13.440 with one right because then the consistency of the question 0:14:13.559 --> 0:14:16.200 is is equal, and then with the questions are not 0:14:16.559 --> 0:14:20.400 the same. It does provide a challenge, but because a 0:14:20.440 --> 0:14:22.800 lot of these data providers work with lots of companies, 0:14:22.800 --> 0:14:27.200 there are statistical ways to try and calibrate them to 0:14:27.280 --> 0:14:31.400 each other. But but we don't ask women how do 0:14:31.440 --> 0:14:33.800 you feel women are treated? It's not about asking them 0:14:33.840 --> 0:14:37.720 explicitly about this, but it's for example, you look at 0:14:37.720 --> 0:14:41.560 the difference between how women feel that promotions are fair 0:14:41.680 --> 0:14:45.880 versus how men feel that promotions are fair. By the way, 0:14:45.920 --> 0:14:47.960 the same thing that is true about men and women 0:14:48.040 --> 0:14:51.640 is also true about employees versus management. If you look 0:14:51.680 --> 0:14:54.400 at the company and you say, you know, feeling appreciated 0:14:54.560 --> 0:14:57.720 ends up being important. What happens to a company. Let's 0:14:57.720 --> 0:14:59.960 just take a simple case. Imagine it's a five point scare. 0:15:00.000 --> 0:15:03.760 All how appreciated do you feel? And let's say you 0:15:03.840 --> 0:15:07.840 have a company where the employees are at three and 0:15:07.880 --> 0:15:10.960 the management is at three and a half. All we 0:15:11.040 --> 0:15:15.560 have a company where the employees are at four, much 0:15:15.600 --> 0:15:19.880 higher and the management is at five. So the employees 0:15:19.880 --> 0:15:22.440 are better and the management is much better. The gap 0:15:22.560 --> 0:15:26.200 is different. Which of those companies is going to be 0:15:26.400 --> 0:15:29.880 more successful in the stock market. They're going to our results. 0:15:30.720 --> 0:15:34.320 It's the first one. Why because a lot of things 0:15:34.360 --> 0:15:39.120 are about relative happiness. Think about your own life, right, 0:15:39.160 --> 0:15:41.240 A lot of it is about where you feel compared 0:15:41.280 --> 0:15:45.320 to other people. And sources of injustice are usually asked 0:15:45.360 --> 0:15:49.040 compared to other people, men compared to women, women compared 0:15:49.080 --> 0:15:52.920 to men, management compared to the employees. So so having 0:15:53.640 --> 0:15:57.880 a consistent strategy is very important. And and one other 0:15:57.920 --> 0:16:01.960 thing that we found incredibly importan and for for motivation 0:16:02.840 --> 0:16:05.680 is so so I told you that feeling appreciated is 0:16:05.720 --> 0:16:08.640 one and of course during COVID it was harder to 0:16:08.680 --> 0:16:11.760 communicate to people that they matter, and the companies who 0:16:11.760 --> 0:16:15.080 did this did much better. But another important thing is 0:16:15.400 --> 0:16:19.680 feeling that people can make honest mistakes. And and this 0:16:19.720 --> 0:16:21.480 is why it's so important, you know, think about it. 0:16:21.560 --> 0:16:25.600 Every company wants people to innovate, but some companies want 0:16:25.600 --> 0:16:28.040 people to innovate, but they also get them to feel 0:16:28.040 --> 0:16:32.440 that if they will make a mistake, they will be punished, right, 0:16:32.480 --> 0:16:36.960 they might not get promoted, they might lose something. And 0:16:37.000 --> 0:16:39.280 every time you try something different, you take a risk. 0:16:40.240 --> 0:16:43.200 So on one hand, companies declare that they want people 0:16:43.240 --> 0:16:44.920 to take a risk. On the other hand, lots of 0:16:44.960 --> 0:16:48.320 companies behave as if if people take a risk and 0:16:48.400 --> 0:16:51.400 don't succeed, they will they will get penalized for it, 0:16:52.200 --> 0:16:54.920 and the company is that score high on that that 0:16:55.040 --> 0:16:57.280 people feel they will be penalized if they try something 0:16:57.320 --> 0:17:00.600 new and didn't succeed. Of course do much worse in 0:17:00.640 --> 0:17:04.879 the stock man. So you know, you you founded the 0:17:04.920 --> 0:17:08.199 firm five years ago with the intent to see if 0:17:08.240 --> 0:17:11.440 there was useful signal in searching out this data. And 0:17:11.880 --> 0:17:15.320 five years is not actually that long of a time 0:17:15.400 --> 0:17:19.920 really in terms of market history and what's a durable signal? 0:17:19.960 --> 0:17:22.200 And I know a lot of like approaches they talk about, 0:17:22.240 --> 0:17:24.879 like the idea of like out of sample data and 0:17:24.920 --> 0:17:26.560 so you're like, look at something, but you want to 0:17:26.600 --> 0:17:29.920 make sure you're not getting correlation and causation backwards or something. 0:17:29.960 --> 0:17:33.640 So you need time. How do you sort of as 0:17:33.680 --> 0:17:38.000 a as as an investor or someone selling investment products 0:17:38.000 --> 0:17:40.560 built on this, how do you establish that what you've 0:17:40.600 --> 0:17:44.600 identified these links between some of the survey data is uh, 0:17:44.720 --> 0:17:47.040 is this is real and not just you know, maybe 0:17:47.040 --> 0:17:48.960 a couple of years worth or something, and something that 0:17:49.000 --> 0:17:52.560 will actually stand the test of time. So so a 0:17:52.560 --> 0:17:54.960 couple of senses for this. The first one is, even 0:17:54.960 --> 0:17:57.520 though we started five years ago, we were lucky enough 0:17:57.560 --> 0:18:00.399 to get data that goes more backward in that so 0:18:00.440 --> 0:18:02.440 we have data going back to two thousand and six. 0:18:03.280 --> 0:18:05.840 So you know, it's slightly better than than five years, 0:18:05.880 --> 0:18:08.200 but you know, it doesn't solve the problem in general. 0:18:09.040 --> 0:18:10.879 The other thing is that you know, there's lots of 0:18:11.280 --> 0:18:14.040 modeling approaches they're out there that are kind of black 0:18:14.080 --> 0:18:17.840 box models. Our approach is not a black box model. 0:18:18.359 --> 0:18:21.280 Like I start with things that are already proven in 0:18:21.320 --> 0:18:24.760 the social science world. Right, so when I when I 0:18:24.800 --> 0:18:29.800 looked for relative salary matters, fairness matters a lot much 0:18:29.840 --> 0:18:33.040 more than absolute level. This is not hey, I just 0:18:33.080 --> 0:18:35.439 discovered the feature in the data. This is based on 0:18:35.480 --> 0:18:38.680 the hypothesis that we have from lots of other sources. 0:18:39.800 --> 0:18:41.800 So you know, it's a little bit like like you 0:18:41.920 --> 0:18:44.920 create a vaccine out there, you use a lot of 0:18:45.040 --> 0:18:49.159 data about biology, You're not still doing a random model 0:18:49.200 --> 0:18:51.600 and creating a vaccine. So so we have a better 0:18:51.640 --> 0:18:55.600 starting point because we start with what we know matters. 0:18:55.600 --> 0:18:58.080 We have a model of human behavior. But by the way, 0:18:58.400 --> 0:19:02.440 the reason that Fern you don't matter, and the social 0:19:02.960 --> 0:19:06.520 retirement benefits don't matter, and health benefits don't matter. These 0:19:06.520 --> 0:19:09.639 are things that we we predicted up front, and the 0:19:09.680 --> 0:19:12.040 reason they don't matter is they just fade to the background. 0:19:12.960 --> 0:19:15.159 You know, you might be excited when you get hired 0:19:15.200 --> 0:19:17.400 by all the benefits you have, but but how many 0:19:17.480 --> 0:19:19.159 days do you wake up or go to work and 0:19:19.200 --> 0:19:21.880 think about your retirement. But I love our ice coffee 0:19:22.080 --> 0:19:24.439 here at the Bloom office, and every day it's the 0:19:24.480 --> 0:19:26.080 first thing I do when I get it, as I 0:19:26.119 --> 0:19:29.359 pour myself a big cup of the cold brew that 0:19:29.400 --> 0:19:31.000 we have on tap here. So I just want to say, 0:19:31.040 --> 0:19:32.480 I just want to give a shout out to our 0:19:32.480 --> 0:19:35.040 cold break. I I hope they will not listen to 0:19:35.040 --> 0:19:37.320 our show and take and take you aware your eyes 0:19:37.400 --> 0:19:41.240 coffee to my intention. And the third thing, when you 0:19:41.320 --> 0:19:44.800 think about what's causal and correlational, you can actually look 0:19:44.840 --> 0:19:47.800 at the data. So we, as I told you, in 0:19:47.840 --> 0:19:52.399 our modeling approach, we take employees state of motivation in 0:19:52.560 --> 0:19:54.679 year one and we predict and let's in two thousands, 0:19:54.680 --> 0:19:57.000 says we predict the stock market in two thousand seven. 0:19:57.000 --> 0:20:01.240 We take two thousand nineteen, we predict E twenty. But 0:20:01.320 --> 0:20:03.840 also it's a question of what matters. You know, when 0:20:03.840 --> 0:20:07.080 the company does does well, there are some things they 0:20:07.080 --> 0:20:11.640 can do better easily. They can increase employees retirement benefit, 0:20:11.720 --> 0:20:14.280 they can buy better furniture and so on, and we 0:20:14.359 --> 0:20:16.520 see that happening, But it turns out that's not the 0:20:16.520 --> 0:20:19.560 important thing. If you if you look at what we found, 0:20:19.640 --> 0:20:23.080 it matters, for example, the sense of feeling appreciated. That's 0:20:23.119 --> 0:20:25.679 a deep cultural thing, not easy to measure, but it's 0:20:25.720 --> 0:20:29.840 a deep cultural thing. And when companies do better financially, 0:20:30.080 --> 0:20:32.480 it doesn't necessarily mean that all of a sudden we're 0:20:32.520 --> 0:20:36.399 better at appreciating our employees. So between all of those 0:20:36.440 --> 0:20:39.720 we do have more data than when we started. We're 0:20:39.720 --> 0:20:43.000 starting with a strong theory about what motivates you and being, 0:20:43.000 --> 0:20:46.720 and the data we explore it as much as possible 0:20:47.040 --> 0:20:50.440 to see what matters and not, and we feel quite 0:20:50.480 --> 0:20:57.080 confident that that it is a predictive model. So Joe 0:20:57.119 --> 0:21:00.640 and I started out this conversation by king about how 0:21:00.640 --> 0:21:02.720 the E and E s G tends to get a 0:21:02.800 --> 0:21:05.679 lot more attention. Why do you think that is? Is 0:21:05.720 --> 0:21:08.280 it a problem of data. This notion that maybe you 0:21:08.320 --> 0:21:11.840 can quantify more easily what a company is doing to 0:21:11.920 --> 0:21:15.879 reduce its carbon emissions, but quantifying what it's doing to 0:21:16.040 --> 0:21:19.560 make its employees feel a sense of well being and 0:21:19.680 --> 0:21:24.919 appreciated is much more difficult. I think, so you know, 0:21:25.200 --> 0:21:26.679 in the same way that I told you about the 0:21:26.720 --> 0:21:28.760 she in, because I think we often go to the 0:21:28.800 --> 0:21:31.520 things that are easy to measure rather the things that 0:21:31.560 --> 0:21:35.800 are important. Now, let's take let's take key and you 0:21:35.840 --> 0:21:38.080 mentioned it in the beginning, and you basically said there 0:21:38.080 --> 0:21:42.920 are two paths for it to be a good alpha strategy. 0:21:43.280 --> 0:21:45.520 One is more people would want that stock, so that 0:21:45.600 --> 0:21:49.000 stock will go up. That's one approach. The second approach 0:21:49.119 --> 0:21:55.119 is saying, somehow being environmental would create more motivation for employees, 0:21:55.720 --> 0:21:58.800 the employees will be more proud to work there. Well, 0:21:58.840 --> 0:22:02.199 that's actually true. We find that most of the effect 0:22:02.280 --> 0:22:07.040 of E comes from human motivation, right, And if you 0:22:07.080 --> 0:22:09.480 can measure human motivation directly, you do much better. And 0:22:09.520 --> 0:22:13.200 you could do You could be a an E company 0:22:13.280 --> 0:22:15.760 and the employees don't know about it, and therefore you're 0:22:15.760 --> 0:22:19.040 not increasing their motivation. And you can be an environmental 0:22:19.040 --> 0:22:21.760 company and the employees could be incredibly involved and connected 0:22:22.160 --> 0:22:24.800 and feel more proud and now you get also the 0:22:24.840 --> 0:22:28.360 benefit from that. So so I I look at companies 0:22:28.400 --> 0:22:30.760 as a mechanistic thing. I look at it's an engine. 0:22:31.480 --> 0:22:34.440 It's an engine of innovation and creativity and thought, and 0:22:34.760 --> 0:22:39.240 ask what what fuels that engine and what creates friction? 0:22:39.320 --> 0:22:43.240 And you know, dust and slows it down, and bureaucracy 0:22:43.280 --> 0:22:47.240 slows it down. And people feeling connected to the company, 0:22:47.280 --> 0:22:50.080 and feeling that their utility function is aligned with the company, 0:22:50.080 --> 0:22:55.000 and people wanting to be part of a successful company. 0:22:55.040 --> 0:22:57.360 All of those things are It's what fueled the machine. 0:22:58.040 --> 0:23:01.240 And from my perspective, the human capital party is to say, 0:23:01.320 --> 0:23:04.840 let's just quantify that, because you know, the reality is 0:23:04.880 --> 0:23:08.040 that lots of people are not happy enough it work, 0:23:08.800 --> 0:23:12.479 and it's kind of a waste. Partly, we're investing in 0:23:12.480 --> 0:23:15.040 this strategy because it has a good alpha, but partially 0:23:15.520 --> 0:23:17.480 I think it's just kind of the moral thing to do. 0:23:17.840 --> 0:23:20.880 You know, if people come to work unhappy, everybody loses. 0:23:21.119 --> 0:23:24.359 The people are miserable, management is miserable, shareholders are miserable. 0:23:24.400 --> 0:23:29.159 If people come to work happy, everybody benefits. Right. I 0:23:29.280 --> 0:23:31.199 I can't tell you what what it's like to to 0:23:31.280 --> 0:23:34.000