Philip Brittan: Hello everybody, and welcome to our next session. Really appreciate everyone dialing in. I'm Philip Brittan, I'm the CEO of Crux joining you from the mountains of Montana again. And very excited about this next session, where we are joined by Mamadou-Abou Sarr and Julia Asri Meigh. And I'll let them introduce themselves in a second. But first, I just wanted to point a couple things out. This session is the third of six that we have going throughout the day. I hope you can join us for as many sessions as your schedules allow. We'll be recording everything so that if you can't make to a particular session, you will be able to see that later. But we want to have you live and submitting your questions as we're speaking. 

The session will run about 45 minutes. And we encourage you to submit questions throughout the discussion. You'll see that on the webinar platform on the left side, there is a box for you to submit questions, and we'll be able to see those. We'll try to get to as many of them as we possibly can although we may not be able to get to all of them during the session. There's also an audience chat, which is a live chat for all of the audience members to sort of talk to each other and help you become aware of each other in a way that's difficult on a remote platform like this. I encourage everyone to introduce yourself and say where you're dialing in from and what you do as well. So, you can get to know each other as well as get to know our esteemed speakers. Let's get started. Welcome. Thank you very much, Mamadou and Julia. And I'd love to start by having you introduce yourselves. Maybe we'll start with Julia.

Julia Asri Meigh: Great. Thanks. I actually work on the research team at Neudata. And we are an alternative data intelligence company. Our primary function is to help hedge funds and investment managers find a more sort of unique and interesting datasets that they can use in their investing strategies. And I actually specialize in our ESG data research. And I would actually say it's a really interesting time to be working in the intersection of alternative data and ESG. Because the recent developments in technology and AI has really sort of led to the growing availability of ESG data. So, I'm really glad to be here today so that I can hopefully share some interesting insights from our ESG research.

Philip Brittan: Mamadou, do you mind introducing yourself?

Mamadou-Abou Sarr: Yes. Well, hello, everybody. I think we got people dialing from around the world, so morning, afternoons, and evenings here. I'm the founder and President & CEO V-Square Quantitative Asset Management, which is a global quantitative shop focusing on sustainability. We are a part of the Valor Equity Partners, an equity firm. And our ultimate objective as an investment firm is to look at how to integrate ESG with intentionality. And ultimately, for us, it's just part of our utility function, how we think about portfolios. And I'm excited to be joined by Julia and you Philip here today to look into the datasets and dig into the topic of ESG data and how do you decipher that universe.

Philip Brittan: Terrific. Thank you both again. I thought maybe we'll start by kind of setting the context. Because you know, ESG, it's an acronym, it gets thrown around a lot. And my guess is a lot of people kind of stretch it to mean what they want it to mean. And there are a lot of folks and maybe some in the audience who are not very familiar yet with ESG. So why don't we start by defining it. I'd love to hear how each of you think about what is and is not ESG. And maybe define a little bit, you know, try to agree a little bit on a lexicon here. Maybe I'll throw that back to Mamadou to start. 

Mamadou-Abou Sarr: Sure. And I think we've done a great job of confusing the industry with all the acronyms and the definition. I'll try to make it simple and use actually a standard definition that was stated by the USC, sort of sustainable investment forum. And it states that sustainable, responsible, and impact investing, it is an investment discipline that actually integrates environmental, social, and governance factors to actually generate positive investment returns alongside societal positive impact. And the reason why I'm using a standard definition is that we can all come up with our own feel for what ESG means ultimately. But as an industry, if we actually refer back to a standard definition, it will allow us to level set and agree on that lexicon. From my perspective, all ESG issues are important, but not all of them are financially material. And I know we're going to dive into that shortly after, but it's a very important caveat as we think about how to integrate ESG into portfolios. Julia?

Julia Asri Meigh: Yeah, I know. We broadly agree. I guess at Neudata we sort of see the definition of ESG as encompassing a lot of different factors as well. And I think I really do agree with you when you say that agreeing on a lexicon and a set of standard rules is going to be very helpful in terms of progressing and integrating ESG standards. But I suppose our research is sort of firmly of the belief that we don't really think that industry consensus will actually ever be reached on how to define ESG. Because ESG is very subjective in nature. How you define good and bad business practices really differs across markets and sectors and even groups of people and time periods as well. ESG is very dynamic and will sort of constantly change and reflect the main social concerns of the day. So yeah, I think, really, industry consensus on this topic is not really going to be reached. So yeah, I don't know what kind of lexicon that you'd like to follow, Mamadou. What's sort of your feel on that?

Mamadou-Abou Sarr: I mean, I agree with you that the complexity of it is that you're lacking standard. But it was attempts that you see at the EU level with a taxonomy, but also by actually practitioners such as SASB with a materiality map and actually use the end outcome as a way to define what ESG means ultimately. When it comes to value-based, we will never level-set. You can have your own set of values, it could be linked to your belief as an investor. But when it comes to good and bad, I think there's an ask for greater consensus. And if you think about a broader investment world, ultimately, there is no so much of a consensus and we made a living out of it, meaning arbitraging set of data and definition. And so, I know when it comes to ESG the belief or the holy grail will be, can all come together and agree what bad is bad and what good is good. But even when the bad and good, we've seen that with basically everything that's happening today from a political standpoint. Even good and bad could be extremely nuanced across the board. So, I really view that it's going to be hard to achieve that. But ultimately, the question mark I actually post to the audience, do we need to get to their level or do we need actually more overarching bodies to actually put a definition that one could use and follow and apply at a portfolio level?

Philip Brittan: I was just going to ask, is there a risk in this landscape having a lack of standards and being dynamic, to use Julia's word there? Is there a risk or does it just sort of...? Do we feel like it naturally evolves into a good place over time? Mamadou, do you have a thought on that? Or sorry, Julia.

Mamadou-Abou Sarr: Julia, you were about to comment, I'll let you start and then I'll add on. 

Julia Asri Meigh: Yeah, sure. I think definitely, there's some level of risk there because then you get companies sort of greenwashing their sustainability report and really putting a spin on how their business practices are good and ethical. But at the same time, I think we see the lack of standardization and definitions as an opportunity as well because it creates opportunities for hedge funds and ESG rating agencies to sort of carve out their own definition of ESG and sort of differentiate themselves from their competitors in that way. So I think there's good and bad to it. And I suppose the one thing I've always thought about is, you know, I think everyone compares ESG ratings to credit ratings. And credit ratings are quantitative, they're robust, they're standardized and they're measuring something very specific, where measuring ESG, you're measuring something very subjective and qualitative. So, is it sort of the right approach to try and get ESG standards and definitions in line with credit ratings as well or perhaps we're missing a trick there?

Mamadou-Abou Sarr: I think you're making a great point around comparing the rating methodology. But it all came down to having data providers coming up with a rating at a first place. Because ultimately, once you try to capture in essence, multiple layers into a score, and the score could be normalized, could be kind of set as such. It brings it back to what do you put in your recipe? And I think that's where you've seen a bit of ask from investors ultimately to ask, well, which one is the best? Do I need to use an MSCI scoring methodology, a Morningstar, an Arabesque, a Refinitiv and so on? And then you ended up having a plethora of different layers of basically scores and methodologies being thrown out at the markets. 

Now, if you think about the overall investment world, the likes of us are using, let's say, the IBIS database that looks into estimate and valuation. And we have tons of analysts giving their perspective on valuation across the board. No one is questioning if they need to be aligned, they don't need to be aligned. Ultimately, they are providing estimates and valuation. So, my question mark is, do we need to have actually a similar field, which is more open architecture and ultimately, each data provider will have his own biases ultimately, can bring his view and say, I define my ESG layers or score as such. And that's how I see Company A versus Company B. And just something to think about and question to the industry. But I think that's where there is a bit of a wriggle room to improve here.

Philip Brittan: Julia, any thoughts on that?

Julia Asri Meigh: No, I think I agree with what Mamadou said there. I think having that overarching framework is helpful, perhaps a bit more wiggle room than other sorts of standardized ratings and evaluation methodologies. But yeah, definitely, I think I'm in agreement with there. 

Philip Brittan: Great. And maybe, you know, given that, it'd be great to maybe hear from you. When you're looking at ESG data, what is it you're trying to solve for? What is it you're looking for? Or what are investors trying to optimize for in this data? How do they think about the data as a look through all these different sources, emerging sources and different standards around it? What is it that they're really focused on? Julia?

Mamadou-Abou Sarr: We all come at it from different angles, right? 

Julia Asri Meigh: Yeah.

Mamadou-Abou Sarr: On my side, we're managing portfolios. We have a fiduciary responsibility and the ultimate goal is to actually deliver better risk adjusted performance for our investors. So, as we are thrown with data, we have to decipher and put them in what I call two buckets. On one side, we have what we call soft data, good to know, good to have, but immaterial from an investment standpoint. And then you have hard data that ultimately can be part of what people our utility function. When we run our optimization process, we are pretty much give and take, right? You're saying, given the level of risk that I'm willing to take, what type of factors that I need to get exposed to deliver the return that I'm actually aspiring for. So, in my utility function, if I have ESG data, I'm actually starting with intentionality. 

And at V-Square, we actually narrowed down the puzzle to only four areas. Out of all the ESG topics you could pick, we decided to look into materiality, i.e., across every sector, which ESG issues tend to be more financially material, that's number one. Number two, we looked into actually, climate change as an overarching both global macro issue but also linked to bottom-up model from pure quant perspective. Number three, we looked into corporate governance, how corporate governance impacts companies bottom line to start with. And the last part is actually diving into what I call the deeper dive of ESG across the board, which is long-termism. Because we think about ESG as being, of course, an important part of the puzzle, but you have to think about what are we trying to solve here. And unless we are thinking about the long term, it will be just a quarter to quarter, a month a month, a day to day, an intraday an intraday feel that will give you different outcomes and a bit of a random world. And so, we were very deliberate in other than the scope or ESG topics. We are not saying that all of the others are not important, we're saying, in the portfolios of V-Square, these are the ultimate factors that we'll actually double click on and have a better focus in our utility function. 

Philip Brittan: Julia? As you go, I mean, that's a really fascinating framework that you've laid out there, Mamadou. Julia, as you add new data, advise clients on different datasets and what they should be looking for or not looking for? How do you think about that framework in terms of what is important, what's not important? How much of it is sort of objective and how much of it is subjective or specific to the mandate of a particular investor?

Julia Asri Meigh: Yeah, I know. That's a great question. It's a very broad topic as well. I suppose because I'm not actually the one making investment decisions, so when I speak to data providers, I'm looking for something very different. I suppose what I'm looking for is just the truth from them about how long their historical dataset is, if the dataset is point in time or not. And it also, I suppose, it varies across our client base as well. So I think quants have a very different approach in what they're looking for. So, history, point in time, frequency, lag, those are much more important to quant investors. I think, broadly speaking, at Neudata, I've actually really focused on ESG research and ESG data scouting towards the metrics that investors have found really difficult to find. Because not all ESG data is a regulatory disclosure, you're going to get a lot of data gaps. I've really focused my research on trying to find datasets to fill those data gaps. And a lot of those datasets are sort of relevant to the social dimensions or human capital factors. So, it's actually quite an interesting--

Philip Brittan: Can you give me a couple of examples around just to kind of bring it home. It's hard to find around social. 

Julia Asri Meigh: Yeah, sure. I guess the social dimensions really relate to welfare practices around employees, suppliers, customers. And so it's particularly across the employee dimensions that there's been a lot of research, that these factors are actually quite financially material, you know. So, if your employees are happy and they're satisfied, they're working harder, they're more motivated, their tenure will tend to be longer. I think a very overused example in the alternative data space here would be the use of something like Glassdoor data that sort of quantifies quite imperfectly employee satisfaction and the employee experience at certain companies. And you sort of break that down into different categories like career development and the CEO rating as well and diversity. And this is really something that a lot of investors had sort of considered to be quite intangible. And so I think Glassdoor does kind of offer a solution quite imperfectly to quantify something that has been sort of assumed to be intangible or unmeasurable, in a sense. I guess that with our research, Neudata's research, I've really just been focusing on looking for datasets that investment managers really lock in their ESG evaluations.

Philip Brittan: Yeah, makes sense. Mamadou, as you are looking to use this data to maximize returns to your point, I'm curious to what extent you look at it in isolation versus mixing it with other types of data. I mean, alternative data is increasingly getting the reputation of well, it's interesting on its own, but it's way more interesting as an additional dimension to a model that has a lot of other things going on. Is that true of ESG data or how do you think about it?

Mamadou-Abou Sarr: Yeah. And I don't like to look at ESG data in silo, because ultimately, for me, it's not an overlay. And so it has to be part of basically the core of my portfolio. So, if I were to overlay that I could do it exposed or in thinking, well, it's basically taking securities out of the portfolio and that's your outcome. This is an exposed approach to that. But my take is an [inaudible 00:18:23] model where ultimately it has to speak to other datasets. What I'm looking for is cross-sectional volatility and also cross-sectional correlation between my ESG dataset and my other financial factors. Now, I don't like to use ESG as what people will call a non-financial dimension in my portfolio. If it's non-financial, I can leave it and do that on my Sundays. Actually, I think it's financial, [inaudible 00:18:47] bottom line. 

And so when I look at datasets, I have very strict, actually requirements that are the same for any type of financial data. The depth of the coverage, do I have enough data? The length of the track record, when did it start? If it's from yesterday, how to actually apply it in a portfolio. I need to look at the stability of the methodology. Are they changing methodologies every few years and therefore, it's hard for me to run it back test and apply it at a portfolio level? Then I have all sorts of timeliness. When do I receive the data? When is the data updated? And I would like to give an example. A lot of data providers don't appreciate the notion of rebalancing. So, they will give you data after the rebalancing cycle that you have for most providers. And so for me, that data cannot be applied on my portfolio level because they just missed the point where I'm rebalancing. And I can miss an opportunity to actually address that within the portfolio. 

And then you have other factors that are actually linked to the quality of data. Is it a disclosure information, i.e., there are biases, it's binary, it's yes and no. Do you disclose on issue A? Yes. If no, then you are of that universe. I'm very careful what I call disclosure ESG data because I considered them as being self-data. They don't tell me much about the quality of the information, they are just telling me that a company A has probably more resources, more people to feel these reporting, and therefore can claim that they are a more ESG, so to speak. And so there is a lot of data scrubbing that has to happen. It's cumbersome. But that's why we actually use our PMs in that context because we want them to be involved. If you have a separate ESG team that dictates their views and come to the PM and say, "That's how I see the world," a PM is likely to push back and say, "Well, I have a lot of other things to do today," unless they look at it as an issue within my portfolio. So, all our PMs have a combined role of being ESG savvy and aware so they can appreciate what is basically the cross-sectional correlation with other sets they have within their portfolios.

Julia Asri Meigh: What do you define as a hard ESG data, Mamadou? Just out of interest? 

Mamadou-Abou Sarr: Yeah, so for me a hard ESG data, if I look at governance, as per se, I will use the IIGCC definition of corporate governance or network of corporate governance, looking at the composition of the board, independence of the board, a set of other factors, i.e., do they have a qualified auditor opinion? Is all information reported? It's part of kind of the governance framework, right? We are a part of the international network on corporate governance, so we adhere to their standards of what they mean by corporate governance. When it comes to materiality, we actually use the SASB materiality map within every portfolio. We think that SASB has done a great job as actually addressing the issue and the big elephant in the room. What is material? Now, the hope is that the industry will actually follow through in what we call pre-listing requirements. Because once you are listed, it's pretty much a story of resources. Do you have enough people to do the job? But if you make it part of the prerequisites for any company becoming listed to disclose that information in a standardized manner, you will actually have a more standard in the industry.

Julia Asri Meigh: Yeah.

Philip Brittan: Something that you guys have both touched on lightly is, of course, the fact that E, S, and G are three different concepts. We put them together, I think, we all agree they started off as sort of these virtues. And they were grouped because we might see them as corporate virtues and increasingly, we now see them no, these are just well-run companies. Thoughtful, in-tune companies adhere to these and that's why companies that have good ESG ratings actually perform very well financially to Mamadou's earlier point. Julia, as you are scouting data, as you're advising clients, are you feeling from either the data providers or from the clients maybe even more importantly, that E, S, and G are sort of looked at independently or...? And if so, is there more of a focus in today's market on one of those three? Or is there any kind of interesting synthesis across that that people are more focused on?

Julia Asri Meigh: Yeah, that's a great question. And I think it really just depends on the individual investment manager. I think people that are sort of starting their ESG journey, they just look at ESG as sort of one broad topic together. And as they go along, they sort of separate these factors. I think in the last few years, quite rightly, environmental dimensions have gotten a lot of attention, particularly as the sort of imminent risks that climate change poses. But what's really interesting is I think COVID-19 has actually sort of brought into the forefront some of the social issues and social factors. So, you know, policies regarding employee welfare and sick leave and those sorts of metrics. So, that's why ESG sort of and the importance of ESG factors sort of can change over time and is quite dynamic and it sort of focuses on the social concerns of the day as well. So yeah, I'm not sure Mamadou if you have any other thoughts on that?

Mamadou-Abou Sarr: No. And I think you're making a fair point here. My take is that any acronym needs to be "taken with a grain of salt," right? 

Julia Asri Meigh: Yes, sure. 

Mamadou-Abou Sarr: And so, an acronym basically kind of takes different layers that may or may not be actually all interlinked. For me, one of the best examples so to speak, is the BRICK markets, Brazil, you take Russia, you take India, you take all these countries that we decided that they have something in common as being emerging markets at some point, one could actually argue that they have not a lot in common in many ways. It's slightly different for E, S, and G. But for me, the central piece by luck is actually the middle piece, the S. And I don't think we should look at them in order, but if I were to put them from my perspective on what is the center piece that holds the two together, is actually human capital in the middle. And we lost sight of that for so long by just thinking, well, I will do my E part, great. Give me scope one and scope two, scope three Give me information on corporate governance. The S piece well is more philanthropy or community involvement or is actually more CSR. No, it is not. 

But human capital is actually the centerpiece that allows you to think about corporate governance in a certain way, but also address environmental issues on the other side, right? And so, if I were to reinvent the acronym, I will actually put S as a capital S and I'll put the E and the G on the side as kind of minor pieces that ultimately are all linked to human capital and could be modified as need B. I know it may be controversial or you could start with SEG as a new acronym, but ultimately, to centerpiece for me is human capital hence our focus at V-Square on human capital as part of our core ESG focus.

Philip Brittan: Fascinating. I'd like to mix in a question from the audience. We have a few coming in, but I'll throw one of them out here. What are the biggest errors or potholes you've seen with evaluating ESG data? Are there common issues specific to any particular types of data within the ESG landscape? Julia?

Julia Asri Meigh: I think one of the most common issues or challenges or perhaps mistakes is to really base your ESG evaluation solely on self-reported or disclosure data. There's a lot of problems with doing that because you're really just relying completely on the company and issuer for your information on what their business practices are. I think a really simplified way that I sort of explain this to our client base is sort of like creating a character profile for someone. You're trying to assess how good somebody is. You wouldn't just take the information they provided to you, you would also get sort of third-party references or ask the people that they have relationships with and interact with. So, it shouldn't really be different for corporations either. So, when you're integrating feedback data from suppliers and employees and customers, you're much more likely to generate a more objective and accurate ESG risk signal. So, if you don't and if you're just relying on just issuer disclosed information, it's obviously going to be at higher risk from bias. I think that's one of the common issues that I sort of see across the ESG data industry. Yeah.

Philip Brittan: Any thoughts on that, Mamadou?

Mamadou-Abou Sarr: Yeah, I'll just add to that, that ultimately, we need and that's why machine learning and AI are playing a critical role. We need also to use all those sets of data alongside what I call the stale ESG data. Because ESG definition or evolving what was a breach 10 years ago is slightly different than what is a breach today. And you need a bit of that flexibility in the middle. And a lot of the first datasets that were coming out in the market were giving you a rear mirror view of the issues of yesterday pricing as of today. So, you're saying a company A did A and B in the past two quarters and therefore that's the rating. By the time you provide a rating in a market place, well, a lot of new issues have come out and clients are asking you how solid is that information when I'm actually about to use that in my portfolio. And you need to look at it as a both side of the puzzle, you have the asset owners that may use ESG in their investment decision or they may actually use an asset manager. You have the custodian that actually has to report a lot of issues. You have the asset managers on the other side and the data providers. All that ecosystem is actually trading on information that may or may not be stale. 

Now you've seen a lot of new data providers improving that approach. You've seen the likes of TruValue Labs and others coming out and RepRisk giving you alternative feel for whether data could be complimentary. But I'm always putting a bit, you know, buy the be aware sign when you actually use pre-packaged score because ultimately, these scores are not as a kind of a flexible as you think. It's like if you buy a pre-packaged meal, you may enjoy it. It could be fantastic cuisine. I'm French and biased. I like to buy the ingredients piece by piece and make my own meal. And so I look at raw data as actually an easiest way for me to corn model to actually make my peak. But then, for a retail investor, knowing that you've used data from a good data provider may be a good feel for how they've actually created that package.

Philip Brittan: Something I've been wondering about is, you know, in the investing world, when you look at data there are big buckets of data that's good to trade on. If everyone's looking at the same data, certain technical signals are commonly viewed that way. And then there's the other view of data and you see this a lot in the alternative data space where people feel like you want very few people trading on this data. And the more novel the data is, the better. Otherwise, the alpha gets sucked out of it. How important is it that ESG data be novel or be, you know, sort of if you have more unique access to, is it good that everybody's looking at the same data? Or does it give you as an investor an edge if you're looking at data very few others have access to? Mamadou?

Mamadou-Abou Sarr: Yeah, happy to kind of answer first. And Julia, you can look at it from working with hedge funds as well and how they use the ESG data. For me, it's a tale of two stories because ultimately, we can have the same set of information together and look at it in many different ways. And so, it's the same with what we call factor investing, is the value of signal crowded because you have a lot of low volatility asset coin managers out there playing that premium. My take is that they will all have a different definition of what low vol is even though you may argue that a covariance gives you a feel for what it could be. Ultimately, every manager will use the ESG data for a lot of various reasons. I don't think that there is such an opportunity of saying, well, I have a new set of information, unless you're actually using it in different ways than others. 

I'll say the complementary value depends on why you use the ESG data. And for me, and when I think about my portfolios, I always think not so much about can I arbitrage the ESG signals versus other signals, it's more, how does it interact with other information I need to have in my portfolio? Because ESG is not, and I have to repeat it with capital letters, it is not a separate asset class. Meaning that you're not buying a new wave of actually, securities that were not there before. We're talking about the alpha bed, the Facebook of the world and all companies we're all familiar with, who were just adding another layer of information to pick them within the portfolio. If that's the case, I'm fishing in the same pool than everybody else. Now, I'm thinking, can I have more information to make a better investment decision ultimately? I believe the answer is yes. But if we put a myth out there that ESG is a new holy grail that we just find new companies that were never there before, it's actually pretty misleading, ultimately. And therefore, one has to think about actually the advocacy of ESG in the context of how do you invest in what is your investment style?

Philip Brittan: Julia, what are your thoughts on this?

Julia Asri Meigh: Yeah, I think I sort of share the same perspective. I think the only thing I'd really add is, I suppose that's why we come from this view that it's perhaps more beneficial that we all sort of assess ESG in a more nuanced way and a definition that makes sense to us. Because that's how investors can really get edge from ESG by creating their own sort of methodology and sort of standards behind it. And I suppose there is a risk of alpha decay if everybody is sort of using the same ESG datasets and using the same standardized model. But at the same time, there's always a growth in creation of new datasets all the time. And in fact, that is actually what keeps me in employment. So, I'm sort of there to scout for new and interesting, unique datasets in that sense. So, yeah. I think I broadly agree with Mamadou there as well.

Philip Brittan: I'm going to ask another question from the audience. And both of you have touched on fact or it seems to be a fact that the focus changes over time on different aspects of this. So, in that context, how do you think about materiality given that it changes with the times, for example, the asker says the increased focus on human capital given COVID? Anything about materiality in some sort of stable way? Julia, do you want to start with that? 

Julia Asri Meigh: Well, I don't think there is a sort of a stable and defined way that we can look at materiality. I think one of the things that I'm constantly having to do in my job is to just keep reading white papers and back test results to find where materiality is in ESG, I suppose. Mamadou, I'm not sure you have some thoughts on that.

Mamadou-Abou Sarr: No, I don't think we're moving the post as per se, as long as it's pretty fine within a [inaudible 00:34:58] information. If you think about how the Internet era changed the way we were interacting as practitioners, if you think about how we've seen a new wave of companies actually using the cloud to store information, ultimately, that will bring a new set of what we call materiality issues. We had a great panel session yesterday on the cloud and data usage and energy usage. That was just another panel that we'll have seen 10 years ago. My take is that we'll have to evolve. It has to be flexible to reflect the economy that we live in. Issues, though, I think that should remain critical, and I know that COVID pulled out basically some of the issues around healthcare and inequalities and diversity and inclusion within social unrest, they should be a sense of peace for me. We should not wait a pandemic to actually question our healthcare system or what is material in healthcare. And also, if diversity and inclusion is actually a corporate issue or societal prerequisite to actually live in harmony. And so for me, that's just a philosophical question, do we need to have these types of events that are, I don't think they are black swan events. They are likely to become more vivid unless we actually change the way we operate. 

And ultimately, if you think about a charter of a corporation, what is the benefit of a company? There are a lot of different perspectives on the matter. One could argue that stakeholder capitalism will become the norm. Because ultimately, as a company, not only you need to sell products, you need to obviously generate revenue. But then you operate in communities and you're servicing people that actually represent a broader set of what you actually see in your annual report. And I will actually think that we've reached a tipping point where it's out there for the world to see. And the fact that we all work from home makes us realize who are the haves and have nots in that ecosystem? Who has the luxury of having an internet connection and without the need to commute to work? But ultimately, how we all play a role in that society and that we have to be accounted for. I know it's more philosophical than practical, but I believe that stakeholder capitalism will actually become a new way of assessing companies.

Philip Brittan: But to put up a little bit of a practical edge on that, we have a question from the audience that touches on exactly this point about stakeholder views. The question is, given the prominence of social data, can you talk about how you deal with stakeholder risk data? How do you collect that data? Do you use sentiment data or something else? What do you think about stakeholder risk data?

Mamadou-Abou Sarr: Julia, do you want to look at it and then I can dive into our own assessment?

Julia Asri Meigh: Yeah, definitely. I think depending on what stakeholder it is, you can collect stakeholder feedback data. I think I referred to the example of Glassdoor earlier. And so Glassdoor is a crowdsource site and just sort of pulling employee feedback data from that crowdsource site. You can do something very similar with customers by setting customer welfare practices or customer feedback. You can do that from product reviews and ratings, and particularly look out for comments on the products quality and value and any sort of safety and health risks as well. With customer feedback data as well, you can also pull that from social media. I think it's much more common these days for customers to take to social media like Twitter, for example, to complain about the quality of a product or service on social media before that data sort of shows up anywhere else. So yeah, I suppose it really does depend on what stakeholder it is. Yeah.

Mamadou-Abou Sarr: I'll be very controversial here. Because my take is that employee satisfaction, I know now we're trying to force that into ESG. If that's the case, every company could argue that they are ESG for per se because if they sell products, they always assess how customers are happy and also, you can see from an employee standpoint, how employers engage. I call it more engagement and actually true factors here. When it comes down to health and safety and that type of assessment and quality of products and the sourcing and the value chain, for me, there are more measurable factors that we'll actually assess on my stakeholder value chain here. Because I can look through and say, if you're using body A, B, and C to create product D, all party A, B, and C also fully align in terms of health and safety, sourcing of material, and also highest part of an overall product set. 

Employee satisfaction for me, there's biases to it. If you look at reviews online, you wonder what are paid reviews or even Twitter I'll say maybe even more with the bot and all these information that comes out. It's hard to decipher who's who. But I'll say I like more hard data here and ultimately all like the health and safety, quality, and also audited information when it comes to stakeholder or dataset.

Julia Asri Meigh: Yeah. Definitely. I think I definitely agree with your points that there are some of these sorts of stakeholder intelligence sources that you can find harder datasets for. And I suppose with alternative data, just like any other dataset, no dataset is perfect. It's always going to have its problems. But I think it's much better to integrate the perspective of stakeholders into an ESG evaluation than to leave them out. And I think there are actually ways to improve a dataset like Glassdoor and to ensure that the sort of the comments or the ratings is valid. I think using NLP techniques to make sure the ratings are consistent with the reviews is one way. And I think they've done, can't remember exactly who did research on this, but I think it was [inaudible 00:41:05]. But they actually found that a lot of five-star rating spikes happened in October, which was before Glassdoor releases their top 100 best places to work. So, being aware of those factors as well and sort of integrating that into your methodology can really help improve the data set. I think there are always ways around it. But yes, I think harder data is always more preferable.

Philip Brittan: Yeah. And there was an audience question exactly to that point about bias in the data. You know, anything that's reported, obviously, you're going to have people who don't speak up who should. You're going to have people trying to game the system and a lot of different things. And you touched on that. Now, Julia, are there systematic ways that you can account for and help reduce the bias that we get in these kinds of very subjective, either social media or survey type data?

Julia Asri Meigh: Yeah, definitely. I think it depends on the dataset and what type of dataset it is. With something like crowdsource ratings and reviews, you know, using an NLP techniques to make sure the ratings and the reviews are consistent with each other is sort of just one way out of many. I think it's also important to note that there is bias in disclosure information as well. So, I think a lot of people actually have the assumption that disclosure data is more factual and harder. But in the sense corporations have, I guess, a motive when they're publishing ESG data as well. So that's where you get problems like, I guess large cap bias. So obviously, larger companies they are more resourced to be publishing and to be investing in sustainability literature and content and marketing material. I think it's important that whatever dataset that you're looking at, whether it's stakeholder intelligence or disclosure information, there is going to be a degree of bias. Unless it is the sort of datasets that Mamadou you were talking about earlier, the hard ESG metrics and data that's very, very objective.

Philip Brittan: Mamadou, any thoughts on removing bias?

Mamadou-Abou Sarr: Fully agree. I say just be careful with the false positive and false negative. It's like when you put your best suit on, you want to be seen and if you don't have a suit, you try to hide if you have a meeting that requires a suit. It's exactly the same with ESG datasets. If you're large and you're doing something great, you will actually highlight that. It's the same with the PRI reporting. Initially, I was part of the reporting working group, and it was not meant to be used as a marketing tool. Now, if you have a nice PRI score, you will tend to get a press release showing how great you are. But for me, that ultimately kind of poses a question around the false positive versus the false negative. The resources, small cap, micro cap, and even on the private market, certain companies don't have resources to actually put a nice CSR report around community involvement. But some of them are actually 100% wired to support actually certain communities. So that's the hardest part and you need other analysts who will put an overlay and assess these companies on merit. And so that's something that we are fully aware. We don't take data for granted. We always question, is it basically linked to how they are publishing information. I wonder how many analysts read a full CSR report? And if you do, it's fascinating. There is a lot of pages that are actually copied throughout the entire report, 400 pages how great companies are. And so it's a job in itself.

Philip Brittan: Yeah, makes sense. I'm afraid our time is up. I hate to do this because this has just been an incredibly wonderful conversation and I've learned a whole lot. I want to thank Julia and Mamadou very much for your time and your wisdom and insights here. I also want to thank the audience for your participation. I know we have way more questions than we were able to get to. Really excited by the level of engagement. We'll try to engage with you on those questions offline. Feel free to reach out to any of us. You can reach me or we can help connect you with either Julia or Mamadou if you write to data@cruxinformatics.com. And thanks again and hopefully you will all dial in for our next session coming up in a little while. So, thanks to our speakers. 

Mamadou-Abou Sarr: Thank you. 

Philip Brittan: Thank you very much. 

Mamadou-Abou Sarr: Thank you. Bye. 

Julia Asri Meigh: Thank you. Bye. 

Philip Brittan: Take care. Bye.