Abraham Lioui, Professor of Finance at EDHEC Business School, talks to Jason Mitchell about the future of ESG outperformance.
Is ESG outperformance coming to end? Is an ESG asset bubble brewing? Listen to Jason Mitchell talk with Abraham Lioui, Professor of Finance at EDHEC Business School, about his latest paper, Chasing the ESG Factor. In this deep-dive episode, the two discuss how investors may be mispricing ESG, the implications of identifying ESG risk premia and why more rigorous approaches are absolutely essential to understanding ESG in the context of quantitative finance.
Recording date: 09 July 2021
A Sustainable Future: Prof. Abraham Lioui, EDHEC Business School, on the Shrinking ESG Premium
Abraham Lioui is Professor of Finance at EDHEC Business School. He’s also Co-Director of the Scientific Beta Advanced Factor & ESG Investing Research Chair.
Note: This transcription was generated using a combination of speech recognition software and human transcribers and may contain errors. As a part of this process, this transcript has also been edited for clarity.
I'm Jason Mitchell, co-head of responsible investment at Man Group. You're listening to A Sustainable Future, a podcast about what we're doing today to build a more sustainable world tomorrow.
Hi, everyone. Welcome back to the podcast. And I hope everyone is staying safe and well. There's an emerging line of ESG research and commentary that's beginning to pose some uncomfortable questions. Like, what if the outperformance over the past several years is over? How much of those returns were predicated on capital inflows, particularly, those from passive investors? If we're talking flows, how appropriate are comparisons to the Nifty 50 in the 1970s? And more to the point, are we headed for an ESG bubble? Look, I get it, these are contentious questions. They throw a curveball at the narrative that ESG isn't a fad. It's not about short-term crowding effects that instead, it's about a long-term structural rebalancing as ESG investors come to represent a bigger part of the market.
But should we be asking these questions anyways? I think so. And if you know anything about me, you know I'm an absolute sucker for a good, provocative piece of research on ESG. Especially one that offers a counterpoint to what seems at times an almost infallible, unquestioning faith in ESG investing. Which is why when I saw a recent article in the Financial Times on this research, I jumped at the opportunity to interview Abraham Lioui. We talk about how investors may be currently mispricing ESG, what the implications are for identifying ESG risk premia, and why more rigorous approaches are absolutely essential to understanding ESG in the context of quantitative finance. Abraham is professor of finance at EDHEC Business School. He's also co-director of the Scientific Beta Advanced Factor & ESG Investing Research Chair. Welcome to the podcast, Abraham Lioui. It's great to have you here. And thanks so much for taking the time.
It's great to be here. Thank you for inviting me.
I'm excited about this. Great. So Abraham, we have a lot of great stuff to talk about. I'd like to start out with your new paper called Chasing The ESG Factor, which was featured most recently in the Financial Times. Why don't you start by giving a synopsis about the findings and some context?
Yes. So I will start by the context, and give you in a nutshell what did we find stable. So the motivation for these papers prompted for... I'm working on ESG for many years, and I found the need to work on this paper three years ago, roughly speaking, where I saw papers from very big shock in the academia, professors from Harvard and the other big schools in the US and the UK. And what impressed me is that the two studies were using exactly the same data and they reached exactly the opposite conclusion. Which was, one study said that highly polluting firm outperform firms which pollute less, while the other study reached exactly the opposite. And then I asked myself, how is it possible? They use exactly the same data? And then they came to the point that... In fact there is a deep difference in the methodologies that have been used by these two papers.
And these methodologies are methodologies that were encountered in academic studies, and as a matter of fact, I told myself that this may be also one source of the confusion in the literature on ESG. In fact, the only literature, we know it's common place, everyone knows it, is very mixed. We don't know if ESG generate high performance or underperformance, though we have many papers that say you can be ESG and overperform, and other papers will say you can be ESG but you will probably underperform. So I am not entering the whole debate, the very specific point we wanted to address in this paper is how the methodology used to beat the ESG factor can lead to different outcomes. So this is the context of the paper.
In this paper, roughly you have two main methodologies that are used in the academia, they are state-of-the-art. What we wanted to check in this paper is how can we make these two methodology consistent between each other, and is there one which dominate the other? And we wanted to apply it in the ESG factor place. Why? Because the problem is that, when you use the two methodologies, and if you believe that ESG rating matters for the factor return, you cannot compare the methodology unless you're guaranteed that the two portfolio that you're comparing have the same ESG rating. So the purpose of this paper was to take these two methodologies, to adapt them such that they measure the return premium of the factor which has exactly the same rating in both cases. And then say one may be better than the other. And as a matter of fact, we found that one is better than the other for a very simple reason, because one methodology is able in a very flexible and simple way to account for other sources of difference in return between stocks beyond ESG.
Our findings are the following, with the methodology, we are able to compute an ESG premium on a monthly basis, not something which is based on 20 years of data, which is likely to smooth out a lot of very important variation impact when our factor premium is measured every month. Once we've assumed the factor premium every month, we extracted the outperformance of the factor. Outperformance of the factor while controlling for other sources of risk which we know are important like size, value of profitability, momentum, et cetera. And we found a very interesting factor into the outperformance of the ESG factor. The ESG factor is long-high-ESG stocks and short-low-ESG stocks. And what we found is that this factor had an alpha at the beginning of our sample, in the beginning of 2000, which was positive and economically important. And this alpha steadily decreased, and to nowadays, our sample ends in 2020, it became negative.
I want to stop you there because I want to dig into the methodology and mechanics around this work, which I think is incredibly important. But I want to stop you there and jump ahead, and I want to ask you this big picture question which is, what are the implications for ESG investors coming out of this paper? I guess another way of saying it is what is the say about, in your mind, the level of sophistication of ESG investors relative to value, growth, work, momentum-style investors?
I think it's the same level of sophistication, heavy sales points. We can say the point in a very simple word. If you think of factors, suppose you have a firm which has an ESG rating of 10, and another firm which has a ESG rating of one, the ESG factor will go long with the ten and short with the one. Now, suppose these two stocks, now the rating is four and six, ESG factor as they have been built so far, they will go long the six, and short the four. What we are saying is that this may not be the right way to go. We pretend that the spread in the rating is important. Why? Because it's likely to drag the premium. When the spread is one to 10, and when the spread is four to six, you cannot expect the same premium from ESG in both cases.
The stronger the possibility factors, that is long-short, they use the characteristic in general and the ESG rating is a rating factor, they use the ESG rating as an ordinal variable. You take ESG, you long the high and you short the low. What we are saying is, no, you should take seriously the level of the rating itself. That is, you need to adjust to investment into the high ESG and low ESG depending on the spread between the two ratings. That's the whole point of the data. And this is, I think, it's very tricky, it's like value stocks. 20 years ago, the spreading the book value to equity, to the market value, was let's say, five, six in lock downs between value and growth stocks and this spread decreased to be one or two. So the question is, should we maintain the same factor or should we adjust the investment in each leg of the factor as a function of the level of the spread? And that's our point.
Let me dig into this because, I mean, it's a pretty powerful assertion, and effectively in my mind, the holy grail of ESG right now. Which is to say, does ESG as a quantifiable factor exist? When I look at the findings from third party data providers and others, I often find that what is purported to be an ESG factor is nothing more than a sloppy amalgam of a lot of other common risk factors. I mean, namely, quality, size, low vol, et cetera. It runs into this definitional, circularity problem, which is, we're calling something ESG, the reality is, it's just a sloppy combination of other risk factors. So it gets to this point of what are we really measuring? So I want to press you on this point of how uncorrelated, how orthogonal is the ESG fact that you have produced?
Fantastic question. That's also one strength of the methodology. The ESG factor that we built has that exposure to the size, which is zero, and exposure to value, which is zero, and exposure to momentum, which is zero. So our ESG factor is pure or plain, that is our portfolio, the weight of each stock in our ESG factor are such that the exposure of this factor to the size characteristic that... We didn't take tons of characteristics. We took four or five, just the common ones. But obviously, you can extend as more, given some technical problems. But that's the point, that our ESG factor has no exposure to this ESG to the other characteristics, hence it's a pure ESG factor.
So where do you see this leading towards? I mean, if the co-movement relative to other common risk factors isn't there, how do you continue to evolve this? Because one thing just from a practical point of view for practitioners, when we talk about performance attribution or risk exposure, you don't find these kinds of things within your traditional Brinson attribution analysis, right? You look at performance attribution alongside industry, region, or stock attribution, but there's nothing that I've at least seen commercially available that points to an orthogonal ESG factor.
First of all, our factor is unlikely to be orthogonal to the factors that you're citing. Because the factors that you're citing are the long-short ones. And our point is to say that these long-short factors, maybe they don't capture what they are supposed to capture, that's one. Secondly, our contribution is to say the following, people from the last decade of 20 years, of two or three decades, and maybe the academia, is guilty for having created this perception. They think of factor as a free lunge. You buy your small stock, you sell big stocks, you get the factor premium. [crosstalk 00:13:19]. Sorry, I didn't hear you.
I was going to say risk premia.
Exactly. [ 00:13:22] don't understand this risk premium. What does it mean a risk premium, you are taking risk, it may be there, it may not be there, in terms of the idealization. That's one. Here, it's completely different this is where the theory was super helpful to us. What we are saying is that, we cannot even tell you whether high ESG should occur from low ESG or the reverse. Why? Because we have two opposites, mechanism that are at play. On the one hand, if people value ESG, they're willing to pay more, given the cash flows, it takes time for them to change, if they pay more, they will get less return. So probably also ESG regenerate if you want, a negative alpha.
However, when there is a high sentiment toward ESG, people start buying like crazy ESG stock, in the position between now that the sentiment is high [ 00:14:26] because through their portfolio, the prices will increase. So on the one hand, people like ESG they are willing to perform is friendship performance hence negative alpha. On the other hand, there is a buzz around ESG positive shock to be given for ESG hence positive alpha, negative versus positive. You will have the mixed evidence which is exactly what the academic literature documented.
I guess this is the most interesting provocative question now that... I mean, it would seem to point to the fact that ESG investors are relatively, and this is a generalization, but unsophisticated. They're buying high not buying low. And they're directed by their preferences, obviously, for a lot of different reasons in terms of normative screening, et cetera, but that seems to be directing the negative return element.
Yes. No, the positive return, because if there is a high demand, the prices go up. So those who are selling now, they will generate various positive return. So I would say, if you say that ESG investors are unsophisticated, I would like to add that they are consistent. Why? Because, in fact, if on average, they will understand that there may be a cost to buying these stocks high, and they still buy them, for me, that's the proof that this guy are really, really looking for an impact. Because in fact, when you talk about financial performance to both, sustainable investing, you're perfectly hypocrite. You're not consistent.
Why? Because if you want to do sustainable investing, by definition, the financial performance should not be what you care about. You care about the positive impact on the planet. It may be, of course, yes, that it would cost [ 00:16:22] here. So for me, if you describe these investors as unsophisticated, because they behind, I would like to add that they are also consistent. Because they know what they're looking after. And if they lose money that they're ready and prepared to lose double it. So they may be not that unsophisticated than one may think.
That's actually a really good point. I had this conversation with Camp Harvey earlier about that, and I guess, his contention is that these capital flows into ESG stocks are nothing more than crowding effects, short-term flows that would ultimately unwind. And I guess I'm wondering, from your perspective, is there a counter-argument to say that, actually, it's not necessarily short-term crowding effects and short-term capital flows, it is in fact, a longer term structural rebalancing of the markets, as more, "ESG investors," become a bigger portion of the market overall?
I think it's a very tough question. But what they can tell you is that what bothers me in this flow of funds to the SGE investing, and this is maybe what the financial industry has maybe to come up with something more convincing. What bothers me is the following, people want to invest in sustainable companies. A company of a high rated ESG, they undertake projects that are profitable but friendly, et cetera, et cetera. But what bothers me is that, the whole industry is marketing as we have a sustainable fund, we have a sustainable ETF, et cetera. Why? When you look at the data you see the... In fact, it's called the trivia, [ 00:18:08] should broke it through as sustainable or irresponsible. And for me, this is the source of irresponsibility from the financial industry to tell that they are doing something, and it's not clear at all that is exactly what they are doing. And because investors are putting their money there, so there is a problem, there is a paradox.
Because either the investors are not aware of the very large disagreement that data vendors about ESG rating, or they are aware and then they count maybe too much of the financial industry to make their mind on which data vendor is correct, and therefore raise the portfolio. So for me, the deepest problem right now is not exactly easy that people don't you tell them, "Do you want to put someone in social responsibility?" Everyone say yes. So there is good faith on the part of the investor, which is very good [ 00:19:11], broadly speaking. But for me, the big problem is that so far, we like good data. Because even exposed, you don't have a way to check whether the manager say, "I have a responsible for you." You don't have a way to check it, because the asset manager will always find the data source of code for a pullback, messaging of the data internally such that she or he justify that the portfolio was really responsible, at least a decision of this position.
So how do you reconcile two very different approaches to solving the question around ESG? Which is to say from a quantitative finance perspective, you are looking top-down trying to identify and define an ESG factor across markets, industries, regions, and that it runs counter very much against the discretionary, bottom-up, fundamental, approach, which is not about the breadth of a portfolio, it's about depth. It's about a concentrated portfolio, it tends to be anecdotally-driven, it tends to be narrated around case studies, it tends to be data driven, but in a much more personal storytelling way. And I guess what I'm saying is or asking is, how can those two approaches, this top-down approach that you're following, and this bottom-up approach coexist?
Thank you for this very, very important question. So I think that there is room for the two approaches today because the ESG investing didn't reach maturity. I think that when we reach maturity, the ESG factor will be an irrelevant question. Because this is not what investors are after. And instead of saying, how is ESG factor outperform or underperform, I think it is the wrongest question that we can ask about ESG. I think we should reach, for me, because we are in a position so the financial performance is still the concern for average investors, which is completely legitimate. But in my view, it will reach maturity whenever it would be possible to talk about ESG investing and ESG portfolios, whenever we'll be able to have metrics which have nothing to do with financial performance. They have to do with how much water you had to clean, by how much you reduced the emissions, et cetera, et cetera.
That's, for me, the proof that we will have reached the maturity in terms of ESG, not with financial indicators. But what ESG support that is indicative of how good we have dealt with them, these emissions, you helped countries to check problems with water or whatever, or helping countries which have problems with the labor of [ 00:22:27], et cetera. I think we should reach the stage where a portfolio we'd have clear metrics which are not financial on the E, on the S and on the G. Unless we will be there, there will always be room for these papers who will say ESG outperform or underperform. For me, that's the wrong question. And what you're saying when you say that these people seem unsophisticated because they may have maybe the other way around, is to say these people, they don't care about financial performance, they're willing to buy high because they know that investment will help the planet they're willing to include the cost.
So I hope that we will reach a situation in the medium-term, the long-term, I don't know, where the financial performance will no longer be a criterion or an indicator that will be used when building ESG factors. And this will call for... I don't remember who said this, maybe it's President Macron or the minister of economy in France, but someone says that, probably more than 50% of the jobs in 10 years do not exist today. And we see the asset management industry the same way that we had equity analysts, which were able to take the balance sheet and other accounting document, and from there to generate indicators which will help pick up stocks or pick up an industry or whatever.
I think we will need these kinds of people much more sophisticated not to look at the rating of some company [ 00:24:15], we will need these people who will exist within the asset management industry, and they will be able to do a pullback analysis, but not the way we knew it so far. That is there's no balance sheet, financial ratios, blah, blah, blah, but [ 00:24:32] on ESG component of the firm activity. And this is superb. It means that there is other additional competencies that financial people investing on the stock market should acquire in the coming years. I am a teacher so it's also important for me to try to forecast what will be the next jobs in five to 10 years to adapt the programs largely before the students come to the market?
This is a really refreshing discussion. So I agree with you that perhaps unsophisticated is the wrong word, but enlightened might be a better description. But what I do wonder is where does ESG stand relative to the rest of quantitative finance? Is it an outlier, an anomaly in the sense of conventional quantitative finance tools? So if I were an allocator, as an example, and I had allocated into a large-cap Japan value fund, and I met with the investment team, the first part of that meeting would be looking at the returns. And those would be framed around, as you would expect, Japan, large-cap value, if suddenly the discussion veer towards US small-cap growth might probably redeem, right? I mean, they were outside their mandate. How do you have that discussion in the ESG world? Or do you think that that discussion doesn't necessarily take place? Because your work around this ESG factor would really support that discussion towards a common understanding of how to allocate to this kind of risk in the context of an allocators overall risk budget?
Thank you very much for this question. In fact, here is the part. Let's take one step back. So today, we tend to believe that the best on ESG is here and has been here forever. But in fact, when you look at the data, that's a recent birth, two, three, four years. And it's very interesting that this just came... I am an academic so you may think I'm objective, but it's unbelievable how this just came exactly with the burgeoning of an interesting literature, in the academia. Let me be precise, during the last five years, there has been a lot of papers that have been published in the academia saying, "Hey, maybe the factors disappear."
In fact, when we look at size, value, et cetera, it seems there are no more anomalies because we are unable to show that in the last decade, for example, [ 00:27:27] value with different growth, et cetera, et cetera. So if these factors are disappearing, then what can the asset management industry do to justify the salary. One way is to move to the passive stuff, ETFs. But in ETFs, it's the race to the lowest cost hence less revenue to the asset management industry. So if in the asset management industry you will stick to do something which looks active, but reducing the burden of having to show that you outperform ESG is [ 00:28:06]. What is the thing that you can think about which is simpler than checking the box, I have high ESG stuff. If you have the high ESG stuff, it seems it'll forgive you for whatever performance you do in the future.
So as an investor, we need to be careful, because if the asset management industry is pushing from the ESG buzz to attract formed flows, and say, "Well, look." And know how to pick up the high ESG stocks, et cetera, but it's only a way to distract from the fact that, yes, ESG is unable to generate abnormal performance. We as investors, broadly speaking, we should be very careful to put the bar higher for the asset management industry. True that as an ESG investor, I care less about financial performance, I am willing to compromise. But I will promise you [ 00:29:11] that my portfolio is having an impact. So because you use a rating properly, that's fine. But maybe at some point as time pass, maybe investors will require much more sophistication on the part of the asset managers to convince that they are really doing for the organization to pick up a portfolio, which if we knew we'd have a real impact on the economy.
I want to pick up on one thing, you mentioned passive investing. And I'm wondering within the context of the paper, to what degree to the rise of passive ESG strategies, have they skewed your findings? It seems like ETFs in particular have experienced the most pronounced inflows over the last several years and influenced the return expectations for ESG. Is there an argument to say that passive approaches have methodologically, this might not be fair, but dumbed down ESG as a style factor by simply applying third party, ordinal ratings, top quartile or quintile, at face value and calling it a day?
Yes, definitely, I guess that the passive investing, because it's very likely that the possibility as we said does not have these agreements. So if these passive investors are based on data vendors, so default flow mechanically to the ESG flow to the companies. And I completely agree with you that this may have accelerated the process. This is exactly where our papers are interesting, I'm not saying that the results are imbued of any critique or whatever, it's a mechanical war. So there is a lot of more to do, et cetera. But it's very interesting that the negative alpha zone only show up in the last one or two years. This is these ESG flows... ETF flows, sorry, has probably accelerated the transmission of the best for ESG into the market prices. I know we are reaching some maturity, and know that we have maturity that people have to think. There is no free lunge in life. If there is no free lunge, if you want to be responsible, and you have a highly responsible firm, it's very likely to come at a cost.
What is the cost? The cost is probably that you will get lower return. It's not a fatality. Because remember, as an investor you are buying the stock, the stock has two components, the cash flow and the discount rate. The discount rate means that investors are requiring less return of the company because the company is responsible. However, suppose the company is investing into clean production processes, and the company is about to speed up the adaptation of the new technology by the company itself to produce sufficiently, to get to the breaking point and therefore start getting the rent from the fact that the company is using the clean technology. If the company is able to toss them quickly, all the investment done into the technology to be responsible into cash flows, this may offset the negative alpha for the investors. Because on the one hand, they require a lower return, everything else equal. But if the company give them a boost by generating higher cash flow, then bingo, there may be no cost to be increased.
But this takes time, you understand? That it takes time when you adopt a new technology, it takes time, energy, you'll start getting some rent from this new technology. Let alone that there is another issue which you should be aware of. When you look at the data, what you see is that in general high ESG goes hand by hand with big companies. Big companies are companies which are well-established companies. And this companies is very unlikely that they will change their technology or their way of producing in the next one, two or three years. Why? Because it doesn't take time, and if they invested in the last five years, they would 10 or 15 years to get back their money. So they are not going to change their technology very quickly. So think about it.
Why I believe that the discount rate effect is likely to be the dominating one and therefore ESG investors are likely to impute a negative alpha, it's because if everyone goes to high ESG stocks, in fact, they go to big firms, and this high ESG stock of big firms, you will be surprised to know that when you go to desegregated data, these companies have high ESG stock score. But when you disentangle the score between strength and concerns, you will be surprised that they rank high profit strength and concerns. So that is, they have high concerns, I am the company which produce cars, by constriction, I am the producing company. And by constriction it will take me 30 years to change my production process. What should they do? Sit down and be treated as the less responsible firm? No. I can control the strength. So this will be the company which will donate a lot, give grants, participate to things that will increase their strength rating. Strength rating will be high, they will offset the concern and the raters will classify this company as high ESG.
But if you look a bit, you will see high ESG because the investor locked in strapped to offset high concerns. To come back to my point, any investor investing in high ESG firms is likely to incur a negative alpha. What can it tell about this negative alpha? One thing is what we have seen in the last two, three, five years that is an unexpected shock, a positive shock to the high ESG stocks. Another thing that can offset this negative alpha and which may make us see that high ESG firms still deliver in terms of financial performance is that we have an unexpected positive shock to the cash flows of this.
I mean, speaking of unexpected shocks, how do you think about exogenous factors around this model? I don't think it's something that you've specifically talked about in the paper, but regulation clearly is playing a major role in terms of directing capital flows, and specifically the EU Sustainable Finance Disclosure Regulation or SFDR, does two specific things. It provides safeguards against greenwashing, but it also in the most explicit way that I've ever seen in my working life, steers capital towards sustainable investments as defined by the EU taxonomy and away from unsustainable investments. With the SFDR having just launched this past March, why wouldn't we see sustained capital inflows into ESG and impact products? Which I would imagine perpetuate or even add to the ESG premium?
I think that you're giving too much credibility to EU taxonomy. I'll give you a very personal exception, if you give box to check to company to be labeled sustainable, you can be sure that in five years, 99% of the company will be sustainable. 99%. So think about rating in the fixed income market. Something incredible, if you take two, three raters of a bond issue, the correlation between the rating is the whole 90%, 95%. How come it's so high the correlation? Because the company's understood the game, and now they play the game. If they play the game, you will see that you will get exactly the rating that you want, and all the raters will give. So I'm not sure that the EU taxonomy is really the panacea. So I don't know. I'm not sure, that's it.
I'm less enthusiastic than all of those who are saying they will [ 00:38:15]. I'm not sure because it's an easy way for firms to check the box. Now, think about it, without the taxonomy, that is, even though we know that in reality we don't know what's the ESG level of an arbitrary firm, we don't know. And still we see tons of money flowing into the asset management. So will the EU taxonomy change the game or create a shift upward? I don't know. That's an [ 00:38:49] question. I'm not sure. I think that whoever wanted to shift the force toward ESG, we are likely to be at the end of the cycle. Or we don't have that much time left until everyone wanted to move to ESG has moved. So these people they are not going to reallocate their portfolio on a daily basis, right? Even though the ESG ratings are on a regular flow.
I'm not sure that the EU taxonomy will be something that will reshape the market and create some front flows toward... See what happened with Morningstar, when Morningstar changed their rating to an ESG, [ 00:39:33] because Morningstar had some credibility and there was a paper which just came out one or two years ago. And that showed that when they adjusted the rating with ESG, [ 00:39:49] there has been a lot of flows into form, which are highly rated in terms of ease. But I think the deepest problem right now is not to know there is a factor or there's no factor. It's very likely that we will never find an easy factor. Because to find a factor, we should be sure of the direction they bring with the factor. So as on average, we can document this on paper. Why? By constriction, the ESG factor, we don't know if the premium will be positive or negative because they have this opposite effect.
On the one hand, the negative effect due to the willingness to compromise financial performance. And on the other hand, the positive alpha due to the changing tests for ESG. So in fact, you don't know. For value of growth, for years, you knew that value would dominate. So when you look at the value factor, you should find a positive premium for the value factor. If you find a negative premium, you have a problem. For ESG factor, you don't know. Because if you are in a period where the sentiment is very high, then it may be the ESG factor is delivering like crazy. But if you have a normal five, 10-year, the ESG factor can have a negative... As we say [ 00:40:59] negative.
I want to press you a little more on this specifically, because I mean, one of the interesting comments within the paper is you really dissect the E, S and G. And even within E, you concede that the carbon premium, not E premium, the environmental premium, is a bit more pronounced. I think one of the recent and controversial side effects of all this regulation and legislation, and even frameworks like the TCFD is that certain asset like non-carbon intensive sectors, namely technology, big tech, emerge as huge beneficiaries of asset flows, just from a screening perspective. How do you think that this is skewing the returns? We're still in the fairly early phase of all this regulation, so why wouldn't it continue to skew and perpetuate returns towards firms that are less carbon intensive?
First of all, I don't know. That's an open-end question, and we will see, I don't know exactly what can be the future. But let me please tell you the following, think about a tech company, so what you're saying is the tech company [ 00:42:15], I'm not sure. Think about to do huge mainframes used to collect the data by companies, I don't want to cite any one of the companies, but all these companies, they are managed in the cloud. A cloud is not on the air, right? For the cloud, which consumes a lot of electricity not like mining the Bitcoin. But you're using a lot of electricity, to buy, you have to use the computer, you have to produce the computer. So I'm not sure that if any of these companies which look like they are less carbon intensive or so less carbon intensive.
Let me give you a different perspective on this, there is another paper that I have been working on, and what we have seen is that in fact, what flows through to the tech companies may have completely different storytelling, which has nothing to do with ESG. It can have to do with the safe asset haven. In fact, and this is another paper, in another paper that I've been working on, we asked the question during extreme weather events, what companies, what industry emerge as a safe haven? And guess what? It's the tech company. So the tech stocks, they have the same property as treasury bills from the US or fiat money. So you see it has nothing to do with ESG. And this is a little that we have been finding on 20 years, not only the last five years, that ESG. So just so that I understand the buzz to our tech company, I'm not sure that the buzz is because people believe they are less common policy.
Personally, I don't believe at all that they are less company policy. Because if you add up all the electricity use, and buildings, and the transportation of the product that they sell, et cetera, I'm not sure they will be far from, although, a well-known entering the industry. And this is where, for example, scope one, scope two, scope three, that's a project that has to be continued and met much more efficient in measuring the initials from the companies on the whole value chain of the company, not only the direct cost. And as you know, scot-free, which is a broad measure of emission, it's the one measure with the less precision.
And because it is measurement with less precision, companies belonging to the tech industry will benefit from this, let's say, confusion around the right level of emission attached to the whole check [ 00:45:16]. So this is about the tech story that you talked about. Both speaking, what about continuing front flows toward ESG. We were there for the one or two last years, we have been there, really, and considering we see the steep increase into this [ 00:45:32], whether it's permanent or part-time, I'm not sure. At some point, those who wanted relocate, they will have relocated, and that's it.
I wanted to finish up with this question about how you see, it's called the chase for the ESG factor evolving from here? It seems to me, at least as a practitioner, that there are four... There might be more dominant, prevalent approaches to tackling this problem. I mean, the first is, you take ratings from third party data providers, you regress outcome and risk factors, and you look at the residual, and more often than not, there's just not enough residual in there to do anything with it. And the second one is, many people are scraping using natural language processing or other tools, unstructured data, even in the ESG space, and trying to divine something out of all that noise.
I think the problem is, it's less about producing an ESG factor and more about a sentiment indicator, which by the way, decays really quickly. It's just hard to use that unless you're a very short-term investor, at least from my view, maybe I might be wrong. The third is, you attack this from a very theoretic, academic Markovian perspective, you talk about the efficient frontier. But that's very conceptual. And then the fourth is proprietary approaches of which we're seeing more, which is to say, "Look, let's take the data from these third party data providers, let's throw out the ratings and let's reconstitute them. Let's retrain the data, let's recondition them, kick out all these biases and statistical anomalies. What's your view? I mean, how do you see this evolving?
So in fact, I like very much your fourth approach. My hope is that my paper will be outdated and completely useless in future for years. In fact, what happened is that ESG came up in the last three, four years ago. I've been working on corporate social responsibility for more than 10 years. But then the birth of ESG showed up the recent years. That's the human being, right? People just use the tools that they have to analyze this new issue, which is ESG. And I think that it is the wrong direction. I think it's a wrong direction. Why? Because that's not the point, the point of ESG is not to generate more or less financial performance, that's not the point. So which means that why my factor I control? Because everyone would ask about size value, blah, blah, blah. I would love to see a paper on the ESG factor, where you don't have at all data size, no value, nothing.
You have all the metrics which are really related to ESG, the level of the emissions, the discrimination by the company, the gender gap in the company, the human rights within the country, where the company has its headquarters, et cetera. So I'd love to see people providing data on these characteristics of the companies... Not characteristics, in the way academics and practitioners, our understanding of today that is profitable next year. By the way, when you look at characteristics, most of them, the very large majority are based on accounting documents. Unfortunately, we don't have yet this information related, really, to the ESG published on the main databases by default. And my hope is that it becomes the case. And they prefer that it is on a voluntary basis by the company to publish all the variables, which we really have that sense of where the company is standing in terms of ESG. We'd probably like the investors direct the money toward these companies, or at least the projects undertaken by these companies.
Well, that's a really good way to end this. So it's been fascinating to talk about your latest paper, Chasing The ESG Factor, and how investors may be mispricing ESG. And what the implications are for identifying ESG risk premia, and why more rigorous approaches are absolutely essential to understanding ESG in the context of quantitative finance. So I'd really like to thank you for your time and insights. I'm Jason Mitchell, co-head of responsible investment at Man Group here today with Abraham Lioui, professor of finance at EDHEC Business School. Many thanks for joining us on A Sustainable Future. And I hope you'll join us on our next podcast episode. Thanks so much, Abraham.
Thank you very much, Jason, for the opportunity to discuss these important issues.