George Serafeim: It's a great pleasure to be here with you. As Lynn mentioned, we'll be discussing a little bit about an emerging regulation that is going to be extremely important in the future as many jurisdictions are developing regulations to define what is a sustainable activity. And let me just take a step back. Of course, we all know that the field of sustainable investing has grown tremendously over a couple of years, with trillions of dollars going into either private funds or public market funds across many asset classes. But the question that gets generated is, is basically this What is a sustainable activity and how we can actually define that? And one of the things that we have found from some of my previous work is that we are actually having a really hard time agreeing on what is that sustainable activity and I just want to show you here a couple of evidence around that. This is basically a model that looks at how much ESG ratings disagree with each other because there are many rating providers. So the extent to which they disagree with each other and then we're regressing that on basically a time time trend variable. And what you see there is that basically the coefficient is positive and significant, very significant, which basically shows that over time, basically rating agencies disagree more and more. And the fascinating thing about that is that this is happening despite the fact that we have more and more disclosure, as you can see in the second model.

So basically, over time, disclosure on all of those issues is going up. And what we would expect from research in credit rating agencies, for example, or in analysts forecasts, we would expect that with more disclosure, there will be more agreement. And instead we're actually finding that with more disclosure, there is more disagreement. And you see that on the right hand side as well. When we look at basically across all ranges, I would say, of ESG ratings companies that are rated low, companies that are rated generally in the medium range or companies that are rated in the high range, you see that as disclosure is going up, this agreement actually increases. So this is a fascinating finding because it suggests that disclosure on itself cannot solve this, cannot help us to actually agree on what? On whether a building we should call it a sustainable building or not. So what is happening as a result is regulation is kicking in. And some of the some of that regulation is happening across many jurisdictions. For example, there is a taxonomy for sustainable activities also in China, But the most significant one and the most widespread one is the one that was released by the European Union. And that has implications both for companies and for investors. For companies, it has implications because it basically mandates the disclosure of capital expenditures, operating expenditures and revenues that are aligned with the taxonomy as a sustainable activity.

And for investors, it has implications because investors need to aggregate that data and that data needs to be validated in order for funds to be aligned with the Sustainable Finance Disclosure Regulation, which allows for funds to be designated either as Article eight funds or Article nine funds, Article eight funds funds that are aligned with sustainable activities or Article nine funds funds that are contributing to sustainable activities. So it has implications both for companies and for investors. As you can see there. And the EU taxonomy is the classification system for economic activities that, as you can see on the left hand side, are is influencing what companies are reporting and on the right hand side is influencing what classification system and what designation funds can can obtain. So let me just say one thing about what the taxonomy is not and what it is. So the EU taxonomy is, as I said before, is a classification system for individual activities based on technical screening criteria. And I will I will mention a few of them so you can see them. That is basically saying whether an activity is aligned with with a specific objective. In this case, we'll be looking at climate change mitigation and climate change adaptation as the two objectives that the taxonomy has classified already and it's supposed to. Tilt basically capital flows towards those activities in order to accelerate the technological developments and the diffusion of those activities.

It's technology neutral. It basically doesn't say you should adopt this technology or that technology to adopt the objective. It's an outcome based classification system. And the other important piece to note is that it is what it is not. So it's not saying to investors what to invest in or what not to invest in. And it's not a rating of how green might be a certain company or a certain fund. It just basically says which activities actually contribute towards a sustainability objective. So let me just put it in context. It has three criteria. One is an activity needs to substantially contribute to one of the six objectives. The second one, it needs to not do any significant harm. So, for example, if you have an activity such as, for example, hydropower that is generating basically low carbon electricity, but at the same time it's causing water damage in the area, that wouldn't necessarily qualify. So you need to pass that that threshold. And then it has certain compliance requirements with minimum safeguards. And there will be six objectives to have been already quantified. And those two are around basically reducing carbon emissions or actually investments that are helping people to adapt to climate change. And of course there is no no better year talking about adaptation than this year that we have had so many natural disasters everywhere around the world and so much so many people that have suffered because of that. And it has specific technical screening criteria that are per economic activity and objective.

Let's look at two of them just to make it real. For example, buildings and I like to use buildings because we all live inside them, right? So we all understand them. The first one is you cannot say that a building is sustainable, not without actually looking at the insulation of a building, right? So the better the insulation of a building, the more likely it is that this will actually be sustainable because from a heating and cooling perspective is going to be more efficient. The second one is that it has to do with energy efficiency, how energy efficient might be a building given actually its energy requirements. So and you will see here that it has very, very specific technical screening criteria. For example, it needs to be 10%, not 11%, not 10%. Right. And the same thing for hydropower. You will see there that it has requirements about power density. It has requirements about life cycle, greenhouse gas emissions. So again, you go through many, many of those economic activities and you specify the technical screening criteria that an activity needs to satisfy. So it has been an enormous effort from a regulatory perspective that the EU has undertaken in collaboration with companies and investment firms to actually develop that regulation. And the regulation will of course evolve over time and change over time as the technical screening criteria will be changing. Of course, that doesn't come without controversy.

There is quite actually quite a bit of controversy, with many companies complaining that the regulation is too complex, too burdensome and too strict in many requirements. So there are complaints by that. Of course, there are complaints also at the country level because there is quite a bit of controversy about whether, for example, natural gas and nuclear energy should be designated as sustainable activities and the taxonomy designate both as sustainable activities, one of them as transitional, the other one as aligned. But as you can see there, there was quite a bit of controversy in Austria, actually filed a lawsuit against the European Commission for designating those. So there are there are there is significant, significant, I would say, political, economic and technological implications from the enforcement of those kinds of regulations. So what we are doing in this paper, we're taking a step back. We are using the first year that the data has been released. I believe I will be right to say that this is the first ever study to evaluate the EU sustainable taxonomy just because the data was just released. Basically this year there has been never been data of this nature. And the other thing that I want to mention about how important. More this is, is that this is the first time that we can actually use financial accounting data, audited financial accounting data to evaluate the sustainability claims of an activity. Before that, all of the time we were using self disclosed voluntary data that were non financial in nature to evaluate that.

So this is a very, very significant step towards increasing the credibility of the disclosures and also increasing the comparability of the disclosures and relying to the financial accounting system that pretty much we rely to make capital allocation decisions. The first thing that we're asking in this when we collected the data and our sample here is the 300, basically the Eurostoxx 300. So we looked at the largest European companies that need to disclose that kind of data, partly because these were the first companies to release the data and partly because they just have better accounting control systems. So their data is of higher quality. The first question that we asked is let's assume that you didn't have the regulation and you tried to estimate what is the percentage of revenues in this case, which is what I'm showing you here, The percentage of revenues of a company that would be aligned with that would be classified as sustainable activities that are coming from sustainable activities. And you would be using some kind of segment classification of those revenues and segment disclosures of the revenues because all companies need to provide as part of their business segment, disclosing disclosures, disclosures on assets, operating profits and revenues. And what you can see here is we're looking at data, estimated data from Bloomberg that then we are comparing them to the actual data that companies are disclosing.

And as you can see here, basically the models overestimate very, very significantly the amount of revenues that would be classified as coming from sustainable activities. So what you see as purple is across different sectors. The percentage of revenues that the model would assume that they are coming from sustainable activities. And what you see as green is the actual reported data that companies are releasing. And you can see here the first column is for all sectors. Then we're moving to communication services, consumer discretionary energy and so forth. So, for example, look at utilities. The models will imply that about 84% of the revenues would be coming from sustainable activities. The actual data that is getting reported from companies would suggest that half of those revenues are actually from sustainable activities. We're doing this test because we're taking a step back and we're saying, Do you actually need even the regulation or can we just use business business segment disclosure data to get to the same calculation? And the answer is you actually need it. You wouldn't be able the model wouldn't be able to actually classify that. The second thing that we're looking at is that the regulation is looking at creating a bucket of activities that is eligible activities and a bucket of activities within eligible activities that is aligned with the regulation, aligned with the technical screening criteria. On the right hand side, what you see is the eligible the percentage of revenues that are coming from eligible activities.

For example, not every activity has been classified. So buildings have been classified. As a result, they are eligible. But whether they are aligned, that depends. If they are hitting the technical screening criteria that I was showing you before. What you can see here is that across different sectors, the percentage of revenues that would be eligible very systematically, for example, they are very, very high in real estate and in utilities because pretty much all activities have been have been evaluated and they would be eligible because they would substantially contribute towards the objective of either climate change mitigation or adaptation. But what you can see there is that a much lower percentage is actually aligned. So that gives you a sense of the the distance to be traveled to actually get at full completion of the eligibility. And another way of looking at that is basically these two by. Two matrix, which basically shows you on the Y axis, the percentage of revenues that are eligible for for classification. Two the percentage of revenues that are actually aligned. And what you see here is that pretty much every industry is on the left hand side of that equation. And the distance of that dot to the 45 degrees line is basically what is the margin to improve? Right. So how much an industry could actually close that gap between eligibility and alignment? Right. So as you can see here, for example, when you are looking at automobiles or when you are looking at real estate management and development, there is a very, very significant still room to be traveled.

So those industries have a very significant distance to travel and would translate eligible activities to aligned activities. While some of the other industries, for example, energy, equipment and services, you can see they are very close to the 45 degrees line. So they just don't have that much capacity to translate eligible activities anymore to align activities. And that basically gap you can see here on the right hand side as well that we have we have codified. I think this is very important because when you are thinking about it from a capital allocation perspective and from an investment perspective, if you want to create transition portfolios, these are the industries where you would find more transition opportunities effectively. The other thing that is very important to note is we are looking at the coefficient of variation of those revenues within companies. So we are saying how much variability exists in eligible revenues versus aligned revenues per basically average unit of alignment or eligibility of the revenues. And the intuition here is basically do we observe more variability in aligned in aligned revenues relative to eligible revenues within an industry? And what you can see here is that pretty much most of the industries would fall below that 45 degrees line, which suggests that basically whether you are eligible or not is primarily defined by your industry membership.

Are you a real estate company? Are you a hotel? Are you a pharmaceutical company? Are you a bank? And so forth. But whether you have aligned aligned revenues, activities that are generating aligned revenues, that is much more of a firm specific strategy where you are observing significant amounts of variation across firms in the same industry. And we're going to take into account that that firm specific variation because it exists and we are going to ask the question about how capital expenditures and operating expenditures that companies are making is translating into revenues, right? So one of the things that we're doing here with these models, we're asking the question about once you actually have the companies that have more aligned capital expenditures and operating expenditures, do we observe them having more aligned revenues? And what is the rate of translation of this? And what you can see here is a couple of very interesting results. The first one is that we are finding an elasticity basically of translating capital expenditures and operating expenditure about 88% to revenues. That is not surprising. That is less than 100%, partly because there are some capital expenditures and operating expenditures that are not revenue generated, but they are risk mitigating, right? So you can actually think about the following example. If you are an automobile manufacturer and you are spending billions of dollars to create electric vehicles and create manufacturing plants and battery storage and so forth, that is going to be revenue generating for you.

But you might be spending also hundreds of millions of dollars to retrofit, for example, part of your operating plant with renewable energy power instead of using natural gas, for example. That is not going to generate revenues. That is just reducing your carbon emissions and mitigating risk. So you would expect less than 100% basically elasticity there because of that reason. But then we are decomposing that component and we're saying what seems to be driving more of the revenue. Where is the elasticity actually higher from a revenue perspective? And it's coming, as you can see here, primarily both for consumer discretionary sectors and for the materials sector, primarily from operating expenditures where you can actually see that the elasticity is more than 100%. So for example, in materials, basically these are primarily companies in the mining industry and the mining industry is an extremely important industry and that's why you see so much capital flow inside the industry because we need to actually extract a lot of cobalt and copper and lithium and so forth for the energy, for the energy transition. And you're finding that operating expenditures in that industry have an elasticity of about 130% related to revenues. So a company is making those investments and generating for every dollar basically that is investing in operating expenditures, $1.3 in revenues that are aligned from that.

So these are for us are very, very important models because it's a it's a way of evaluating the effectiveness of those investments from a company perspective. And what you should see is that companies are becoming better and better at effectively increasing those coefficients over time. So you would want those coefficients to go up over time and companies are becoming more and more effective at translating those things into revenues. The the other thing that we're looking at and we're interested in is trying to understand what is the relationship of that data that is getting out from the EU taxonomy with business analysis and fundamental and accounting analysis data. And one of the things that we looked at is whether there is a dominant strategy right now. So is alignment a dominant strategy for a company within an industry, or is it basically that you're observing that companies are finding different ways of competing and some companies are transitioning and they are as competitive as the companies that are not transitioning right now. And this is basically what we're finding. We're not finding really any relationship either on the profitability margin side or with past sales growth. So we're not finding basically that those companies are having higher or lower margins or higher or lower sales growth, at least to this point. When we're looking at the data, it could be that in the future that of course changes. But the assumption that we might have that, for example, those companies, because they're investing more, might generate lower margins, doesn't seem to be the case or the assumption that they might have higher growth.

That doesn't seem to be the case. Again, it's very, very early days and we only have one year of data. So this doesn't seem that doesn't mean that it's not going to change. But here is what we're finding so far. And the other thing that we looked at, of course, is whether investors recognize the differences in any future growth or risk in relation to those activities. And again, the answer is no. We don't seem to find any differences in valuation ratios either on the price to earnings ratio or the price to book ratio. Again, that is early days. It might be that over time investors will better understand that data. Maybe we're asking for too much because the data was just released, but we wanted to actually understand about what is happening. The last piece that I'm going to touch on is we're using that data to understand what is the relationship between the alignment of those activities. So percentage, for example, of aligned revenues that we might be getting or percentage of your operating expenditures and capital expenditures that are aligned relative to environmental ratings that that that investors use. And as I mentioned before, environmental ratings are pretty much defined by non financial data. They don't use financial accounting data. So this is the first time that we get to run this cool experiment that we actually have an independent view that companies are providing.

And this is what you're observing. So we have mapped it against two of the big environmental rating providers. So on the Y axis is basically the environmental rating and on the x axis on the horizontal axis is the taxonomy aligned revenue percentage of the company. And I don't even need to run the correlation here. You can see the image basically, which shows you that there is very little actually correlation. And most strikingly, one of the things that you're finding is that even companies that are basically generating. Zero revenues, 0% aligned revenues can get pretty close to perfect scores, basically. Right. And that's what you're finding there. So we are calling that red region, the area of greenwashing risk. Basically when you're thinking about that, because you have very little basically revenues aligned with sustainable activities. And at the same time, you're getting a view that basically says that you're perfect. Again, that is not a perfect test of greenwashing risk because ratings are also reflecting some other considerations. Right? It's not only about climate change mitigation, climate change adaptation, it's about other environmental issues as well. But one of the things that we have done here to mitigate that effect is we are focusing only on industries in this kind of data where basically carbon emissions is the primary risk driver for that business, which is carbon emissions, the primary thing.

So we have tried to mitigate that effect. And and the and the other chart basically looks the same When we benchmark against another environmental rating, you're finding a very, very similar basically effect where most of the dots, the dots here are basically companies are sitting on that on that greenwashing risk area. So for us, when we're putting the data together, we're looking at that and we're saying, well, gee, like this is giving you basically a very different view about where the market stands right now relative to the view that you might be having before. So for us, there are five overarching almost conclusions that we arrive from this early research. The the first one is, as you are looking into the future, what you should be looking for, I think the first that you need to be looking for is what I showed you before. With the 45 degrees line. You should be looking for industries to be moving closer to that 45 degrees line, effectively trying to shrink the gap between eligible and aligned metrics, either on the operating expenditure or the capital expenditure or the revenue expenditure. The second piece is that think over time the percentage of activities will probably be increasing and we will be seeing that as more and more companies are making progress technologically also managerially. We should also expect those coefficients that I showed you, the translation, the elasticity coefficients between operating expenditures and capital expenditures and revenues to increase over time as companies are becoming better at translating investments into revenues and earning a margin on them.

Right? So the companies that are going to win in this space are going to be the companies that will have a higher rate of translation of those capital and operating expenditures into revenues over time. The fourth thing is looking at the business analysis that I showed you before, we should see if companies within industries that are able to align more are actually earning now higher margins and experiencing higher growth. So that will be another fourth test about what we might expect in the future if if the regulation is going to be successful. And the fifth one is going to be that you might observe over time a very different picture in the greenwashing risk view that I was showing you before. So you should expect to start seeing more of those dots moving towards the green space rather than the red space, basically. So as companies are making progress on that, we might see also some ratings readjustments where rating agencies might be looking back and might be saying, well, that was a little bit too optimistic and they might be starting to make some of those judgments based on data that is coming from the financial accounting space. So I hope that gives you a picture. We still have ten minutes. I will pass it back to you. Lee, thank you very much. Thank you.

Speaker2: Thank you, George. That's brilliant. We've got some questions on here, but I'm just going to open it up and see if anybody in the audience had one. There's one over there. I'm not sure. I think there's a microphone on its way to you. Yeah. Gentleman just here.

There you go.

Speaker3: Thank you very much for that very instructive talk. What should a middle of the fairway US asset manager do in this circumstance who's not running a fully blown responsible investing strategy, but more like an Article eight lite green strategy? So the data, as you said, is not really available. Msci is available, but the data is not accurate. So should they be using Bloomberg industry data instead? How do they craft something that's reasonable and is not so that they themselves are at least being not disingenuous in the whole process? Right.

George Serafeim: It's a great question. Right. So and I would say that we're now in this messy space, right, where we're trying to do things, but we don't have the perfect infrastructure to be doing many of those things. So what I would do if I were in your position is the first one is to try to understand what that new data is telling us about where industries might be might be sitting in right now. And as I showed you before, there will be some industries where basically the classifications that you might be deriving from business segment disclosures might be fairly good or close to what the picture might be for a company in an industry where you might not have the data. And then you will have industries where that gap is very big. So for the former, I wouldn't worry as much and I would rely on the business segment disclosure data and what I might have there for US companies. For the latter, I would worry a lot. Right. So that's where I would spend most of my time trying to dig in around kind of like different strategies that companies might have because that's where you might need to do a little bit of a second layer and a third layer analysis to differentiate between companies, because that's where you see a lot of the variation. That's where you have a very big gap between eligible and aligned activities.

George Serafeim: And and to the extent that that might be something that is part of the strategy, that's where you might also start asking for companies for to make some of those disclosures and give you a sense of where they are allocating capital, either on the operating expenditure side or on the capital expenditure side or on the revenue side of things. Right. I think and to me that is a useful framework, right? You can think about things that basically are long term oriented investments that need to be basically capitalized and depreciated over time. You can think about investments that don't meet the rules of basically recognition to an asset perspective, and they are more like operating expenditures in nature or early stage R&D that doesn't get classified. And then how are you generating your revenues, where they're coming from? And what would be I think a great question for a manager is like, what is your estimate of your revenues that might be subject to those technical screening criteria and how you would think about that? So I would I would try to break down the problem at the industry level first, understand which industry levels are making me more exposed and then start digging deeper into some of those company strategies to create at least approximation of the buckets where they might fall.

Speaker2: So let me ask you a question we had here. So. So in your Purpose and profit book, which the question says is highly recommended reading, by the way, you make the argument that drilling down to specifics of ESG is crucial. Yes. Is this true in this study, too?

George Serafeim: Absolutely. Absolutely. In many ways. Right. So we could make a broad argument and say, well, the regulation is helpful or the regulation is unhelpful. And both statements are wrong. Right. Because the natural reaction that I would have is for which industry, for which sector and for which operating metric. Are you talking about? Capital expenditures. Are you talking about revenues? Are you talking about utilities? Are you talking about consumer goods? Are you talking about automobiles? And I will give you a different answer. And I give that reaction always when we're talking about ESG. Right. That becomes something that ten, 15 years ago, nobody knew about. And now, like the presidential candidates know about. Right? Apparently so. Maybe. But everybody has an opinion about. Right. And I think I think if you want to be thinking about it from an investment professional perspective, you need to be asking the nuanced questions about what is it that you're trying to achieve, What is the metric that you're using and what does the metric basically telling you about the objective that you're trying to reach?

Speaker2: Let's see if there's any more questions in the audience. I've got some more on here. But if anyone's got any. Don't be shy. All right. We'll go back to here. So you spoke of investors, but Hummel and Birkhofer. 2022 on taxonomy identify capital markets and bankers as crucial adopters of EU data. What is your view?

George Serafeim: It will be. It will be a very significant piece of regulation and banks will also adopt this. Some of the fastest growing instruments have been either on the use of proceeds instruments, much like green bonds and so forth. There are hundreds of billions of them out there in issuance or on the linked side, basically sustainability linked loans, sustainability linked bonds. Those are basically credit instruments that tie coupon rates or interest rates on some of those metrics. So I wouldn't be surprised if what is going to happen is you are going to see lending contract design that instead of relying on non financial KPIs, will start relying on financial KPIs, as I showed here, because they are more credible, they have been audited, they are comparable. So they have many of the information quality characteristics where you would want Tony to link your bonds to.

Speaker2: So I'm not sure you can answer this question or not. What was the one company that appeared as the outlier in the Green Zone on the last table?

George Serafeim: I have no idea. Okay.

Speaker2: That's what I figured I'd ask.

Speaker4: It's a good question. Yeah, it's a good question.

Speaker2: We have a question from the audience here.

Speaker5: Thank you very much, Professor Joe Seraphim. Thank you for that excellent presentation. An important paper. So I want to turn it practical to the long term investor. In your wisdom today, as you speak to long term investors in the room today, they'll be talking to their portfolio companies in the next week, month, quarter. What one question, based on your insights, would you encourage them to ask their portfolio company in the next meeting?

George Serafeim: It's a great question. Thank you for that. The question that I would ask is what is the rate at which you're translating capital expenditures and operating expenditures to revenues according to the alignment of those activities? So what I showed you in the models, I would actually ask that to a manager. Can you speak about the rate at which you are basically allocating capital, either on the capital expenditure side or the operating expenditure side? And how is that leading to basically revenues that might be aligned with sustainable activities across utilities, energy, automobiles, consumer discretionary, all of those are basically allocating real capital. When we're looking at the data, we're looking at like hundreds of billions of dollars that are basically allocating there. And the question for me is, one, about effectiveness and efficiency and not the only measure of effectiveness and efficiency is revenue translation, but it's a very, very important one. Are your products selling that you are developing and you're trying to sell and you see that in electric vehicles, in batteries, in renewable energy, across across many elements of. So I would ask that question.

Speaker2: So I think we have time for one more. Is anybody? Okay. So Race to zero demands, 10 to 15% emissions cuts annually through 2030. How did your study consider what company activities were cut to reduce pollution?

George Serafeim: So this is defined by the regulation, by the taxonomy. Basically, the regulations embedded basically model behind it is getting to net zero by mid-century, and then the technical screening criteria are coming on the back of that. So the embedded emission reductions are basically trying to get you to that, to that level, and that is defined by the regulation. That's why you see that, for example, on the energy efficiency side for buildings, that would be about 10%, because what is happening is that this is a little bit technical. So grab your coffee and just just just don't fall asleep. But basically what is happening is that you have a carbon budget out there in the world. The whole world has a carbon budget that gets allocated to economic activities and sectors. And then you project and you forecast based on technology and economic feasibility where you can cut more and where you can cut less. The way that I basically describe that is you can cut more in buildings because we have the technology and the infrastructure to do that. You can cut less in aviation because we really don't have commercially scale technologies to reduce emissions in aviation to a large extent.

Speaker2: George, we're out of time, unfortunately. But thank you. Fascinating. Thank you. And very important. Thank you very much.

Speaker4: Thank you. Thank you very much. Thank you. Thank you.