Manufactured confusion over the role of ESG in investment has led to oversimplification and capital misallocation
Supported by increasingly promising regulatory frameworks, asset managers and allocators must consciously resist this trend rather than fall prey to its deceptive allure
By embracing complexity, and swapping easy falsehoods for harder truths, we can reset how we think about ESG and put our capital to work in a more effective and beneficial manner
“Whenever a theory appears to you as the only possible one, take this as a sign that you have neither understood the theory nor the problem which it was intended to solve.”
“There’s always an easy solution to every human problem – neat, plausible, and wrong.”
H. L. Mencken
At the beginning of 2022 we wrote that a turning of the tide was underway in the ESG investment movement. We now find ourselves knee-deep, with the current drawing strong. Amid a flurry of regulatory activity, asset managers are scrambling to explain quite how strategies marketed (and priced) as ‘ESG’ are having the real-world impact they claim. Compounding the pain, many such products – which were rushed to market during an explosion in demand as asset valuations soared during Covid – have begun to underperform as their underlying sector biases are left strewn on the rocks of inflation, higher rates, and geopolitical upheaval. Amundi, the largest asset manager in Europe, has announced that it will downgrade almost all of its €45 billion worth of “deep green” Article 9 funds to the less stringent Article 8, citing an “evolving regulatory environment”. Vanguard has withdrawn from the Net Zero Asset Managers Initiative (NZAMI) citing a need to “provide clarity to investors […] about the role of index funds”. Meanwhile, the UK Financial Conduct Authority’s (FCA) upcoming climate-related disclosure requirements have prompted confused articles about ESG and greenwashing to pop up in tabloid newspapers. It is a chaotic scene, and one which is understandably difficult to navigate as an investor. How did we end up here? How does ESG really relate to investment? How does Hosking Partners integrate it into our process, and why do we think we are – broadly – headed in the right direction? From the murk we would like to provide some clarity.
First of all, we will state an assumption up front: We assume that unless otherwise specified, and in line with the fiduciary principle to act in their clients’ best interests, an asset manager’s primary investment objective is to achieve a return on clients’ capital. For active managers such as Hosking Partners, this return is generally measured relative to an index or benchmark, although other approaches do exist. This assumption is important. Much of the current confusion around what ESG investing should or shouldn’t be is caused by conflating arguments that only function once you have accepted this assumption with arguments against the assumption itself. The debate over single versus double materiality, which asks whether or not companies should report non-financial data, revolves around this issue. Unfortunately, measuring externalities – the unpriced social byproducts of economic activity that ESG analysis attempts to describe – is an extremely complex problem. They are long-term, they are often largely intangible, and their materialisation represents a market failure. There are essentially two paths that an investor interested in the relevance of externalities can take. Either they can accept the above assumption, and work harder to incorporate the analysis of potential externalities into their valuations. Or, they can challenge the assumption and sacrifice the undiluted pursuit of returns to deliver a different measurable output – exclusively or additionally – which attempts to affect the size of the externality itself. Both options are legitimate but, as we will see, each lends itself to different types of investment, and should garner different expectations from asset owners. As such, asset managers must be honest about which path they are on and how they intend to walk it. Cherry-picking the most attractive parts of each has become commonplace, and is not only deceiving investors but contributing to a worrying misallocation of capital.
In today’s always-online media milieu, confusion around ESG comes thick and fast. This was exemplified in a recent comment piece in a UK tabloid, in which the author expressed outrage that so-called “sin stocks” have higher ESG ratings than the FTSE 100 average. This represents a basic error in the way many people think about ESG, which is to broadly equate it with ethical investing. We touch on this point again later, so for clarity we will recap the basic difference: Ethical investing explicitly considers your chosen values – be they religious or otherwise – before it considers returns. So-called ESG investing is not supposed to do this; instead, it is supposed to consider social externalities as an input into returns. The categorisation of “sin stocks” – which is ethical in nature – challenges our assumption by supposing that asset managers should bias their selection according to something other than expected returns. ESG ratings, on the other hand, are treated as a quantified assessment of material risk to a specific business, and therefore should exist ex post our assumption to hold value as a comparative tool. Playing one argument off the other inevitably results in bewilderment. This mixing of ethics and economics, of the qualitative and the quantitative, and of social impact and financial performance, are at the heart of the ESG muddle. They are also the reason that ESG has become an increasingly politicised issue, as political actors that question our assumption conflict with those that share it.
In 2023 the UK Financial Conduct Authority’s Sustainability Disclosure Requirements (SDR) come into force. This policy obliges asset managers with over £5 billion in assets under management to report against a number of qualitative and quantitative standards to help investors understand how the manager, and the products they offer, incorporate sustainability into their investment process. The FCA’s requirements draw heavily on those laid out by the Taskforce for Climate-Related Disclosures (TCFD), but has an expanded scope that targets not only how managers deal with E but also S and G under the umbrella term of “sustainability”. The lever is a series of formal labels (Impact, Focus, Improver) which will be granted to investment entities and products that meet the FCA’s qualifying standards. In the words of the FCA’s own Director of ESG, Sacha Sadan, the idea is to “raise the bar” by making it harder for a manager to claim to follow an ESG-related strategy without being clear about how that strategy functions and impact it may or may not have on performance. The FCA’s policy is a particularly interesting example of a wider regulatory theme designed to help investors navigate the ESG muddle described above. Unlike the EU, the FCA understands that its role is not to try to directly influence the behaviour of investors to further a political agenda. Instead, the FCA is attempting to mediate the manner in which investment products are marketed to try to protect the integrity of an investor’s decision making. Broadly speaking, this is the right approach. However, the proposed methodology remains exposed to several of the same flaws it is trying to ameliorate. These issues pertain to several of the major inconsistencies that still exist around ESG, and which continue to fuel the confused picture described above.
The great asset class discrepancy
The first issue is that the FCA states that it does not want its new rules to favour certain asset classes over others. This is an understandable position to take on the surface given the organisation’s hands-off approach. However, it fails to acknowledge that investment in certain asset classes is demonstrably easier to link to a decisive impact on the behaviour of an underlying entity than in others. For example, investment in primary markets directly and immediately increases the capital available to a company in a way that trading shares between secondary market participants does not. Furthermore, participation in the former is often explicitly tied to specific terms or covenants which define the manner in which raised capital should be deployed, while the latter is not. Although there is a theoretical connection between secondary market activity, the value of a firm’s equity, and as such its cost of capital, this connection is fairly weak in the absence of widely coordinated action by a majority of shareholders and targeted regulatory support. For example, even as increasing numbers of investors exclude fossil fuel stocks from their equity portfolios – particularly in the past year – earnings and accordingly share prices have soared. As the ex-Head of Sustainability at Blackrock Tariq Fancy has put it, “10% of the market not buying your stock is not the same as 10% of your customers not buying your product”. As such, it is not enough for a public equities product to simply claim that an exclusionary, best-in-class, or other metrics-driven mandate delivers a sustainable outcome. Although it may seem counter-intuitive, the reality is that owning or not owning an oil company’s shares has very little real-world impact on the emissions that firm does or does not generate. As investors David Blitz and Laurens Swinkels write in a 2019 study on the effectiveness of exclusionary mandates, “if one investor lowers the carbon footprint of their portfolio, another will have a higher footprint by definition”.
The standard of evidence required for a secondary market product to claim sustainable credentials should therefore be held significantly higher than for primary market alternatives. This is particularly important because the overwhelming majority of investment activity happens in secondary markets (88% in 2021). It is also important because as both retail and institutional investors are increasingly exposed to soundbites describing the amount of investment that is needed to help the world reach net zero, we need to be clear about what the word ‘investment’ actually means in that context. In brief, what it does mean is primary market allocations, government spending, corporate capital expenditure, and research and development. What it does not mean – for the same reasons as discussed above – is manipulating the public equity allocations in your pension pot to provide more exposure to wind and solar companies. While the latter may help one benefit from the energy transition, the former is what will actually make it happen in the first place. This is an important point, because the generally good intentions of ESG-focused investors will never be realised until we are more honest about the varying degrees of agency that capital flows into different asset classes actually have to make an impact.
Confusing these different types of investment has inflated demand for sustainable investment strategies in the secondary markets with the weakest claim to real-world effect. This new demand has in turn created substantial financial incentive for asset managers – and other market participants – to overplay their credentials, which reinforces the misunderstanding. Studies have shown that funds which change their name to match in-trend investment strategies (for example, adding “Growth” in an upswing) enjoy abnormal inflows of 28% over the next year even if their holdings remain constant. This phenomenon has been rife in ESG over the past several years, and has resulted in a swathe of supposedly ‘sustainable’ funds that appear indistinguishable – in strategy, impact or holdings – from less deceptively marketed alternatives. Regulation such as the FCA’s, which is designed to protect consumers from such claims, should be bolder in explaining that certain forms of investment are fundamentally easier to link to real-world impact than others, and require asset managers’ justifications of their strategies to account for this inequality.
The false promise of metrics
This brings us to the next problem with the FCA’s approach, which is inconsistency over the role and utility of quantitative analysis in managing ESG strategies. At first sight it appears that the FCA stand out from their regulatory peers in their overall scepticism towards metrics-based approaches, highlighting – for example – that “exclusion on its own is not enough” to justify a sustainable label. Encouragingly, they are preparing complementary regulation designed to reign in the ESG ratings agencies, who have profited immensely from the dark art of presenting opinion as fact. The FCA also accepts that the very concept of a sustainable investment product is largely subjective, stating that in the absence of an objective standard, “if someone thinks their product is sustainable they need to define what they think sustainability is themselves". So far, so good. The problem is that after all that, the regulation will still require firms to publish a collection of mandatory numeric metrics and targets, which are identical regardless of the nature or mandate of the fund. The inevitable implication is that this data tells investors something useful about how sustainable a product is or isn’t (it probably doesn’t), and worse, that it can be used comparatively on an apples-for-apples basis (it almost certainly can’t). Furthermore, because of the primacy the investment industry affords to data-driven interpretations of reality, the oversimplified world these metrics describe ends up trumping the more balanced qualitative analysis that is generally required to properly evaluate the complexity of ESG-related materiality. This dynamic only serves to perpetuate the idea that long-term, intangible value drivers can be reduced to a set of figures, and benefits funds with deceptively simplistic quantitative approaches at the expense of those that champion nuance and pragmatism.
The most common metrics-driven strategies for ESG-labelled funds are positive and negative screening. Negative screening or exclusion, which involves avoiding investments in firms that breach some pre-defined red line, is by far the most common. As of March 2022, exclusionary screens of some form were incorporated into 75% of index funds and 69% of active funds bearing an ESG label. The opposite approach – only investing in the ‘best’ ESG performers – is positive screening, although it accounts for about ten times less in ESG-labelled AUM than exclusion. The boundaries of inclusion or exclusion are usually defined by one or more quantitative metrics, which are selected to align underlying securities to the theory or idea to which the fund – and its clients – subscribe. Unsurprisingly, similar methods are used by funds with ethical mandates to exclude companies operating in certain sectors (e.g. alcohol). The key difference is that for ESG-labelled funds that supposedly accept our assumption, using these metrics to define portfolio construction should also be linked to a clear rationale for enhancing alpha. Unfortunately, conclusively proving or demonstrating this link is elusive. In their 2021 paper ‘Honey, I Shrunk the Alpha’, Giovanni Brun (et al.) show that “while many ESG strategies have positive returns, adjusting these returns for risk shrinks alpha to zero”. Similarly, with specific reference to exclusion, the evidence is clear that more constrained portfolios underperform their unconstrained peers. To disguise this disappointing reality, asset managers and ratings agencies have instead performed a methodological bait-and-switch whereby ESG value drivers are quantified to look like the traditional financial factors we are all familiar with such as growth, quality, momentum and so on. This has made these strategies far easier to sell, often at higher fees, despite the non-existent evidence for either sustained outperformance or – for secondary market products in particular – real-world impact.
We have mentioned several times so far that ESG factors are essentially little more than long-term, often intangible value drivers. In some cases, these are relatively easily quantifiable. For example, at Hosking Partners we might consider the methane leakage rate of an oil and gas E&P, or the replacement value of assets located in a potential future flood plain. We may look at the relative turnover of minority ethnic employees at a local retailer, or the remuneration structure of an investment bank. In all these examples, the analysis is designed to determine if these factors represent material risks that should be discounted against the valuation of a potential investment. Different analysts will look at the same data and exercise judgement in ascribing varying degrees of importance to what it tells them, just as they do with free cash flow forecasts. When an asset owner allocates to an active fund, they are implicitly expressing confidence in the particular way that firm or portfolio manager exercises that judgement. As such, the judgement of value drivers related to ESG should be considered holistically alongside that related to other long-term factors and indeed traditional financial metrics. All are as potentially relevant as each another, and the weighting of that relevance will vary on a company-by-company basis. As LSE Professor Alex Edmans has put it, “considering long-term factors when valuing a company isn’t ESG investing, it’s investing”.
ESG integration gets into particular trouble when it tries to quantify factors that are resistant to simplification, especially when that quantification leads directly to inclusion or exclusion in a strategy. An obvious example is the aggregated score-carding that ESG ratings agencies have made common, whereby tens or even hundreds of value drivers that should be considered individually are rolled into a single data point. Here, the sum is most certainly not greater than the parts. These scores are nothing more than simplistic summaries of a subjective opinion. Nevertheless, entire funds are constructed by screening securities with good scores, and subsequently marketed as sustainable. This is the equivalent of building a portfolio solely on the buy recommendations of a single analyst. No diligent investor would would buy such a product, but funds screened by ESG ratings sell like hot cakes. In a vicious cause-consequence cycle, companies have in turn begun manipulating their own behaviour to earn a better score. While in some cases this may encourage companies to adopt genuinely more responsible behaviour, more often than not it results in questionable decisions such as altering remuneration structures to favour ESG over other – perhaps more relevant – long-term value drivers (for a more detailed discussion of this particular issue, see ‘A focus on..ESG-linked remuneration'). While the arrival of long-needed regulation may bring some semblance of utility to ESG ratings, we would nevertheless do well to recall Goodhart’s Law that “when a measurement becomes a target, it ceases to be a good measure”.
The problem is not limited to aggregated ESG scores. There are also problems with the individual metrics that contribute to them. Unhelpfully, the confusion is most pronounced where the data has the most superficial claim to objectivity. Emissions data are a major culprit, primarily because the narrative that ‘higher emissions are always bad’ drowns out more nuanced analysis, which is often – disappointingly – dismissed out of hand as regressive. We can illustrate the importance of nuance when considering emissions data by examining one very common metric, carbon intensity (tons of CO2 per million USD in revenue). In its portfolio weighted form (WACI) this metric also constitutes one of the FCA’s mandated disclosures. The idea is to determine which companies are more reliant on emissions to generate revenue, which seems on the surface like a reasonable approximation for how sustainable that company is. In turn, the weighted combination of that same data is used to describe the sustainability of an investment portfolio.
This problem is the emergent result of two underlying issues. Firstly, we need to think about the numerator (emissions) and the denominator (revenue) that form the basic intensity ratio, and how they interact. The latter, gross revenue, is units sold multiplied by price per unit. For different industries and sectors, the balance of this equation changes substantially. Some sectors sell a few expensive products at high margins, others sell millions of cheap products at low margins. If you just take revenue, two companies from opposite ends of this spectrum look the same. Calculating gross emissions is a more complicated process, but for the purposes of this discussion we will focus on embedded CO2 intensity. This measures how much CO2 goes into a product on a kilogram-for-kilogram basis. What’s particularly interesting is that a company having a high carbon intensity by revenue does not necessarily mean its products have a high embedded CO2 intensity. For example, cement and steel have CO2 intensities of 1kg/kg and 1.6kg/kg respectively, so to make 1kg of cement you emit an average of 1kg of CO2. IT hardware, on the other hand, emits an average of over 100kg per kilogram of product. Producing a MacBook emits 170kg of CO2 per 1.3kg device. This is not to suggest that the cement industry is less of a problem for the energy transition than the IT industry. But what it does show us is that sectors that produce on the high-volume, low-margin end of the spectrum tend to display substantially higher carbon intensities simply as a by-product of the way the metric is constructed.
Now, let’s take these two observations and look at the bigger picture. Many of the products that rely on this sort of high volume, low margin model are commodities like cement, steel, aluminium, and concrete. These low margin commodity-centric businesses are doubly misrepresented by carbon intensity, because in addition to the structural disadvantage described above, their scores also exhibit volatility unconnected to underlying emissions as commodity prices fluctuate. While demand for some of these commodities may decelerate between now and 2050 as new technologies develop, gross demand will grow substantially and, in some cases, intensify (e.g. aluminium and steel). These are industries that are called “hard to decarbonise”. The manufacturing bases tend to be in emerging economies and are primarily powered by fossil fuels like coal. In many cases, although technology is progressing, at present viable alternatives are either non-existent or remain prohibitively expensive, especially for sale into emerging markets where demand is rising fastest. But the difficulty of decarbonising these industries is matched by the reward, in emissions terms. Cement alone is responsible for almost 8% of global CO2 emissions, a significant amount for a single material. Hard to decarbonise industries are not limited to commodities; aviation is another. As you might expect, aviation companies have much higher than average carbon intensities.
A pattern is emerging, and it is a worrying one. Metrics such as carbon intensity are being used to infer sustainability, but actually describe a cross-sectional snapshot of a world in which the largest gross emitters are the worst, regardless of the fact that the largest emitters also have the most potential impact to deliver by decarbonising. These metrics ignore change over time, efficiency of production, and the wider societal importance of the underlying product. When applied to influence investment flows, the result is either counter-productive or illusory. Primary market participants that screen investments using metrics like carbon intensity in the name of pursuing a sustainable strategy are denying capital to the parts of the market that most need it to decarbonise. This puts these sectors in a strange Catch-22 position: they require investment to achieve decarbonisation breakthroughs, but are denied it for the very reason that the breakthrough has not yet been achieved. This is counter-productive. In secondary markets, constructing portfolios by screening according to these metrics appears low-carbon on paper, but as we have already discussed, trading secondary assets has a weak claim to affecting the actual behaviour of companies. In these markets the most common result of these strategies is deception. Investors are sold the idea that they are having an impact when they aren’t. This is illusory.
Getting real on engagement
There is another lever that secondary market participants can pull to have a direct effect on the behaviour of a company: engagement. Engagement can take the form of voting or communication, both of which can be approached individually or collaboratively with other investors. The larger your shareholding in a company, the more agency you have to affect change. This means that although highly diversified active funds (such as our Hosking Partners portfolio) and passive index funds can still make useful contributions to help influence company behaviour, particularly when collaborating with others, the overall ability to drive direct change is frequently limited by their minority positions. This is why so-called impact investors generally rely on the combination of engagement with large, concentrated positions to deliver effect when operating in secondary markets. That said, according to the Global Impact Investing Network (GIIN), only 3% of impact funds’ AUM is invested in public equities, versus 31% in real assets and 60% split approximately evenly between private equity and private debt. This reflects the fact that impact funds themselves recognise the asset class discrepancy we discussed earlier. If you have an impact mandate, why concentrate your portfolio in the part of the market where it is hardest to realise?
Importantly, impact investors generally define their mandate as something other than solely delivering alpha. This differs from activist investors, who use similar tactics to try to enhance returns, often by encouraging changes that benefit shareholders. When managers intend to achieve an outcome other than return on clients’ capital, this generally comes at the cost of alpha. For example, studies show that venture funds with both societal and financial goals earn around 5% lower returns. More broadly, across the entire spectrum of private equity impact funds, average IRRs are 100bps below peers with an undiluted mandate. As a result, impact investors may temporarily operate in direct opposition to an activist – one has a short-term focus on shareholders, the other a long-term focus on wider society. This should not come as a surprise, because the sort of long-term, intangible value effects associated with social externalities are rarely priced into a security to the same degree as more clearly measurable short-term financial factors. When an impact investor encourages a board to increase capital spending aligned to energy transition, or replace low-cost suppliers with higher-cost ethically sourced alternatives, these actions may depress earnings in the near to medium term due to their associated costs. Longer-term, these may well be recouped where the targeted externality proves genuinely material to performance and its elimination provides a competitive advantage. But nevertheless, the materiality of these long-term and intangible factors remains extremely difficult to measure and therefore the decision to invest in an impact fund remains a highly subjective and qualitative affair. Where a fund has not explicitly declared a non-financial mandate, it should therefore retain our assumption as the primary objective of its engagements: that the primary objective of an asset manager is to deliver a return on clients’ capital.
This is especially the case for diversified secondary market participants, for whom the primary utility of engagement should be to encourage long-term performance. As mentioned previously, this is not to say that minority shareholders cannot participate in active engagement with a co-incidental societal impact. At Hosking Partners we routinely engage with companies to encourage them to act in a certain manner. These engagements cover a broad range of topics, from capital allocation policies, to energy transition strategy, to supply chain integrity and human rights. As an investor with an average holding period approaching ten years, we believe that the sort of corporate behaviour that produces long-term outperformance generally overlaps with being a productive and progressive member of society. That said, all of our engagements are linked by the golden thread that we believe the change we are advocating for will be of material benefit to the long-term value of the company, and therefore to our clients’ returns. This golden thread prevents us from confusing or diluting our mandate, or being unrealistic about our ability or responsibility to effect change.
Furthermore, let us not forget that generating strong returns for pensioners and other investors provides deep value for wider society in and of itself. As we have discussed elsewhere, a transformative societal change like the energy transition cannot occur in a world of shortages and financial hardship. Investing in secondary markets may not be the most efficient way to make your money ‘do good’ directly, but it does contribute to the sort of economic growth that makes the doing of good elsewhere more probable. Unfortunately, the manufactured confusion over what ESG investing means, the misrepresentation of subjective opinion with oversimplified data, and unrealistic claims to impact are instead contributing to the misallocation of capital, much of which is hidden from asset owners by nothing short of dishonesty and illusion on the part of unscrupulous managers. While broadly we support the FCA’s approach to correct this problem, rightly aimed as it is on restoring the primacy of qualitative analysis, we feel it could go further in its approach to the asset class discrepancy, the distortionary effect of metrics, and the variable utility of engagement as an effective tool for change.
The Hosking Partners way
At Hosking Partners, we integrate ESG analysis in the most pragmatic, honest, and value-accretive manner we can. We embrace the importance of long-term, intangible drivers of companies’ valuations. We assess and re-assess the trends, externalities, risks, and opportunities which affect us all. We use the simple elegance of the capital cycle lens to distil the world into a collection of ideas that we believe will generate our clients the best relative long-term return on their capital, and thus encourage economic growth. We are also always honest about what we are. As a highly diversified manager that mostly operates in the secondary market, we do not claim to have an impact unless we can demonstrably explain how that impact was achieved. We are not afraid to question received wisdom. We are deeply supportive of efforts to minimise the negative externalities associated with issues such as climate change, but we acknowledge the complexity of such efforts rather than trying to sell a false simplicity. We do not rely on quantitative screens as a crutch, but rather use data as a tool to challenge or reinforce deep, qualitative analysis.
In place of the oversimplifications, illusions, and counter-productive strategies that have come to blight ESG investing, we offer the following suggestions to forge a constructive way ahead:
A diversified public equities product can occupy a key position in a sustainable allocator’s portfolio, but its primary objective should be to generate outperformance rather than impact, as the latter is more effectively accomplished through primary markets.
Where a manager or ratings agency uses a metrics-based approach to justify sustainability credentials or judgements, it should be rigorously informed by a qualitative discussion that frames the terms of reference of those metrics and is honest about the real-world impact of their usage. Furthermore, the impact those metrics have on stock selection should be clearly linked to a thesis for outperformance.
As such, when assessing a public equity manager’s sustainable credentials, allocators should focus on the qualitative way in which the manager thinks about long-term, intangible value, rather than rely on misrepresentative ESG metrics which generally offer an incomplete or even deceptive description of a portfolio’s approach.
Consideration and analysis of ESG factors should form a critical part of all high-quality investment processes, because ESG factors are a subset of a wider pool of long-term, intangible value drivers. However, ESG should not be placed on a pedestal over other such factors, and the importance attributed to the valuation impact of each should be weighed by its assessed materiality to the underlying investment.
Engagement is a fundamental responsibility for all asset managers, because it offers a lever to effect change. However, managers should be realistic and honest about both their mandate and their agency, and align the goals of their engagement accordingly.
The turning of the tide is underway. As the regulators stir, our industry is beginning to question the oversimplified version of the world in which it has found itself. Increasingly, we see our peers adopting the sort of nuanced, pragmatic approach to ESG integration that we have long championed. As we move forward we hope this approach will not only help to steer sustainable capital flows in a more effective manner, but also continue to weave fibres into the mandate that remains our golden thread: delivering long-term outperformance for our clients.
References for any data or quotations included in this article and elsewhere in this report are available on request.
30 December 2022
The state of ESG heading into 2023