Frequently Asked Questions

It is based on well-established models of investment risk measurement. It basically measures the investment exposure to a set of ESG factors and uses that to determine what portion of the total risk was used to reach the ESG exposure.

An ESG index might have a tracking error (which means risk) of 80 bps (100bps is 1% in financial jargon). The model tells us that 60 bps of that is used to achieve its sustainable impact. This 60bps represents 75% of the total risk. This seems ok, indices are usually designed to have low tracking error. And ESG branded indices can be expected to spend most of it to achieve impact.

An active ESG fund could have much higher risk by design, let's say 500bps of tracking error. The model tells us that the fund has an active ESG weight of 25bps. This not only means that the fund is not particularly impactful, but also that of all the risk it takes on only about 5% is spent on impact.

It quickly tells you how much ESG impact there is in a portfolio, which is very helpful when comparing one portfolio to another across regions and investment styles.

In a bigger picture, more investment combined with more active ESG weight translates to more impact. The more there is impact, the more the cost of capital shifts in favour of sustainable firms at the expense of unsustainable firms. See our white paper for a full explanation http://www.aurielequities.com/auriel-equities-in-the-news/

The idea was to select a small set of ESG indicators to capture the cross-sectional variance of the company's impact against sustainable development.

It is a well based concept in mathematics and in finance. A multidimensional space can be represented with a finite set of variables. An abstract and multidimensional concept like sustainable development can be represented via a limited set of indicators or factors.

There are plenty of examples. Is any one person happy for example? It’s a difficult and multifaceted question, but it can be reliably approximated by looking at if the person is warm, safe and satisfied. Does the person have adequate safety nets for the future? Is the person loved? Does the person feel competent and confident? Does the person have a chance to fulfil individual wishes and dreams? None of these individual questions are necessarily decisive nor is the list exhaustive. But to determine if any one person is 'happy' this list already goes a long way.

It is the very same when measuring sustainable development. We try to capture the key ways in which large corporations influence sustainable development. We don't capture all the nuances, but we believe our list goes a long way in establishing overall sustainable impact.

The model requires enough indicators to 'span the sustainable development space'. There is no one indicator that can tell how sustainable a company is. At the same time, there is no need to measure the companies' sustainability hundreds of different ways. We argue that the number of indicators needed is quite low for two reasons.

• ESG indicators themselves relate to one another

Companies’ climate action can be proxied by looking at the company's investment in clean energy and the company's carbon emissions. Makes perfect sense, except that a company's investment in clean energy will ultimately reduce their carbon emissions, resulting in double counting. There is no need to look at both, one is enough.

• Companies' actions to address sustainable development correlate with one another

If company wants to manage their climate impact, they could invest in more efficient production mechanisms and better environmental tracking and accounting systems. The very same systems are very likely to address company's other emissions and water use as well.

As a result, there is a 'Goldilocks zone' of sustainable development indicators. There must be enough indicators to cover the key angles, but there can't be too many for a model to still be manageable.

The overall aim of selecting the indicators was to 'span the sustainable development space' as described earlier. But there is some other important requirements too.

• Indicators are sector and region neutral

The ESG indicators tend to be sector and region specific. Charitable contributions for example are more relevant in India and access to medicines is only relevant for healthcare and pharmaceutical companies. The selected indicators in our model are relevant for every company regardless of their size, sector or the country they operate in.

• All indicators are commonly accepted and recognized

Sustainability measurement has developed a lot during the past few years. There are various commercial models and rating systems in existence. But more importantly there are plenty of national (like stock exchange listing requirements) and international initiatives (GRI, SASB, GIIN) that define sets of ESG indicators. All the indicators used in our model are well established through these national and international initiatives.

• All are outcome indicators

ESG indicators come in various shapes and sizes, and while all of them have their place and their significant use cases, we wanted to focus only on so-called outcome indicators. For example, we wanted to measure a company's carbon emissions (typical output indicator), not their plans around clean technologies or their investments in energy efficiency (which are typical input indicators). This is because we want to measure only companies’ actual impact, not their plans and intentions.

• All indicators are widely used and can be estimated if not properly reported

Relying on very basic financial information guarantees that information is widely available globally. When the model uses more specific information regarding climate for example, that is not part of the basic financial accounting framework, we use the very simple indicators like carbon emissions. All the indicators in the model that are not universally available can be reliably estimated.

• Carbon efficiency
  • Tonnes for carbon equivalent emitted per unit of revenue
  • Based on publicly available company reporting on GHG emissions (scope 1&2)
  • Based on company reporting or regression where reported numbers are not available
  • This is usually called 'carbon footprint'

• Waste efficiency
  • Tonnes of solid waste generated to create one unit of revenue
  • Based on publicly available company reporting on created waste
  • Based on company reporting or regression where reported numbers are not available
  • This is similar to standard 'carbon footprint' sometimes called 'waste footprint'

• Water efficiency
  • Litres of fresh water used to create one unit of revenue
  • Based on publicly available company reporting on fresh water use and withdrawal
  • Based on company reporting or regression where reported numbers are not available
  • This is similar to 'carbon footprint' sometimes called 'water footprint'

• Gender equality
  • Percentage of women in top management
  • Board level data used by default, backed up by executive/top management data where available

• Executive pay
  • Ratio of companies' top management compensation compared to average employee compensation

• Board independence
  • Percentage of independent board members
  • Based on board level data which is typically widely reported as a stock exchange listing requirement

• Environmental good
  • Revenue from environmentally positive products i.e. clean technologies, for example:
    • Wind turbines, solar panels
    • Water purification, waste management
    • Electric vehicles
  • Pre-defined by existing industry classification systems
  • Drawn largely from the sustainable development goals

• Social good
  • Revenue from socially positive products, for example:
    • Education
    • health care
    • access to financial services
  • Pre-defined by existing industry classification systems
  • Drawn largely from the sustainable development goals
  • Geographical component sometimes used i.e. access to financial services for example is counted only if it is provided in places where there is a lack of access to financial services

• Avoiding environmental harm
  • positive signal for avoiding companies with revenue from environmentally negative products i.e. dirty technologies
    • Coal mining and coal based power generation
    • Factory farming
  • Pre-defined by existing industry classification systems
  • Drawn largely from the sustainable development goals

• Avoiding social harm
  • positive signal for avoiding companies with revenue from socially negative products
    • Tobacco, alcohol
    • junk food
    • Weapons
  • Pre-defined by existing industry classification systems

• Economic development
  • Revenue in least developed countries
  • Used as a proxy for an economic impact
  • Based on companies geographical spread of operations and countries GDP per capita

• Avoiding water scarcity
  • Litres of water used in water scarce parts of the planet
  • Based on the companies use of freshwater and their operations in water scarce parts of the planet

• Employment
  • How many employees company has in areas with high unemployment rates
  • Based on companies reporting on employees and publicly available unemployment rates

• Tax gap
  • Tax gap i.e. how much companies pay taxes compared how much they would be expected to pay based on the geographical spread of their operations
  • Measured as companies' theoretical tax rate based on the geographical spread of their operations and their reported tax rate

The data largely comes from companies themselves. We use intermediates (there is plenty of global companies that provide this service) that collect the data points from company reporting and filings for convenience. But we do not (and will not) use third party interpretations or opinions of what is sustainable and what is not.

Some data points are not reported by all companies. Carbon emissions or water use for example while encouraged, is not mandatory to report in most jurisdictions in the world. In these cases, we use existing company reporting and our in-house modelling (again, no 3^rd parties here) to estimate the data points that companies do not provide. As a side note, we wholeheartedly support any initiative that increases the quality and the quantity of the companies' ESG reporting. The better the company provided information is, the better results we get.

By design most of our data points are 100% available as part of standard financial reporting. What companies produce and where they operate for example. Our governance data points of board composition are usually a part of stock exchange listing requirements. For the few indicators where our coverage is not 100%, i.e. companies do not report a specific data point, we have ways to estimate missing values. If company produces certain amount of electricity by using natural gas for example, we can approximate how much water it used based on industry averages for the same electricity generation method.

At the moment, we have coverage of around 10,000 companies across 46 countries covering around 90% of the investable market in each country.

No problem. Most of the indicators we need are publicly available, and we can estimate the ones that are not. However, the more data that needs to be estimated, the greater the uncertainty. For instance, if a small cap Indian semiconductor manufacturer doesn't report water, we can estimate reasonably accurately its water use by assuming that their facilities have the average water efficiency of the industry and looking at their output to estimate water use. If they don't report gender balance, we would have to estimate it with the average semiconductor manufacturer in India, which is a bit more of a guess than the water, but still reasonable to use.