Missing link between carbon intensity and ESG ratings drives ‘dilution’

The carbon intensity of companies hardly affects the ESG ratings they receive from rating agencies MSCI, Moody’s and Refinitiv, according to an analysis by index provider Scientific Beta.

The provider of factor investing strategies and indices has been investigating the effect of adding additional ESG objectives to a low-carbon benchmark. The outcome of that experiment was surprising, research director Felix Goltz told a Scientific Beta conference in Amsterdam recently.

“It turned out that adding other environmental themes to the index leads to a huge dilution of the original goal of reducing carbon intensity,” said Goltz.

Scientific Beta found a dilution of on average 92% across analysed portfolios, meaning that 92% of the carbon intensity reduction investors could have reached by solely weighting stocks to minimise carbon intensity is lost when adding ESG scores as a partial weight determinant.

The dilution is because there appears to be no correlation between companies’ scores on the ‘E-component’ (environment) of ESG and their CO₂ intensity. Scientific Beta defines the latter as CO₂ emissions per million euros of sales.

Indeed, ESG-raters do not base their E-rating solely on companies’ CO₂ emissions and their plans to reduce them. They also look at things like environmental pollution, water use, plastic waste and the impact of production processes on biodiversity.

This automatically leads to E-scores that do not correlate perfectly with CO₂ intensity. But Goltz did not expect the link to be sometimes completely absent.

Moreover, at Refinitiv and Moody’s, there is even a slight (albeit statistically irrelevant) positive relationship between carbon intensity and E-rating. This means that companies with higher CO₂ intensity score slightly better on ‘environment’ than those with lower emissions. Only with MSCI is there a slight negative correlation, of -0.24.

Many pension funds apply both ESG screening and a CO₂ filter, but they generally do not ‘mix’ the two. Instead, two separate filters are used.

This kind of approach largely avoids the kind of dilution found by Scientific Beta in its ‘mixing’ experiment, said Goltz. If an investor first excludes the 10%, 20%, 30%, 40% or 50% of companies with the worst E-scores, and subsequently applies a CO₂ filter to the remaining companies, investors hardly suffer from the non-correlation between CO₂ intensity and e-scores. The reverse is also possible.

According to Scientific Beta, only 8% of the carbon reduction objective survived the inclusion of ESG scores in portfolio weighting schemes.

Financial relevance

Asked to comment on Scientific Beta’s conclusions, ESG rating firm MSCI referred to a blog on its website. In it, the company writes that it considers CO₂ intensity in its E-rating only when a company operates in a sector where climate risk is “financially relevant”. MSCI cites CO₂-intensive sectors such as oil and gas producers, cement makers and utilities as examples in its blog.

Sustainalytics, which is not mentioned in Scientific Beta’s study, also stresses that CO₂ emissions only matter for a rating if it they also are financially relevant. Sustainalytics’ so-called climate rating is based not only on a company’s actual CO₂ emissions at a given time, but also on that company’s plans to further reduce those emissions.

‘Our climate ratings are forward-looking. We look at what they are doing to become climate-neutral and bring their policies in line with the Paris Agreement,’ said director Arnold Gast. Each company also gets a separate score on carbon intensity, emissions trend (how fast a company has reduced its emissions in recent years) and greenhouse gas risk management.

Refinitiv did not respond to questions from IPE.

The world of ESG ratings, scores and other data products is very heterogeneous, including in what they seek to measure and the methodological approaches taken. ESG ratings are due to be regulated in the EU, with MEPs having recently come to the decision that they will push for the regulation to encourage raters to address the impact of a rated entity on the environment and society in general more than is currently the case.

Read the digital edition of IPE’s latest magazine