Request Free Trial
July 13, 2022

Dealing with income bias in sovereign debt ESG scores

New research by FTSE Russell finds that low income countries score lower on ESG for sovereign debt, but this improves when ‘income bias’ is removed. How should managers deal with this bias, and ensure capital is going to the right countries?

Although the use of sustainability metrics in sovereign fixed income markets is becoming more mainstream, there is a lack of consensus on how to appropriately assess countries’ environmental, social and governance performance. This is an area that has been researched by the World Bank, and one of its recent studies found a high correlation between sovereign ESG scores and economies’ national income. 

Taking this a step further, FTSE Russell has released a new report, which adjusts the sovereign E, S and G scores for income bias using an ex post approach. This enables differentiation between countries’ ESG performance, independent of their income level.

The paper finds that low income economies score higher after removing income bias in each of the E, S and G pillars. The income bias is most pronounced for G scores, but is weaker for S and particularly E scores.

Julien Moussavi, senior research lead, sovereign ESG at FTSE Russell, says: “Income bias is really, really ingrained in the governance performance.” One key reason why the bias is so visible for G scores is the issue of corruption, which occurs more in low income countries, he says. However, he highlights that within low income countries, some are less corrupt than others. 

“That’s why we tried to tackle this bias, in order to see which countries are the best among the poorest ones,” he says. This research can help asset managers to discriminate within an emerging market portfolio which countries are performing better according to their income level. 

Moussavi says the E pillar is the most difficult to assess – because topics such as biodiversity and climate change are still relatively new, there is a lack of transparency, standardisation and forward-looking data. 

Filtering the bias

According to Carl Vermassen, a portfolio manager at Vontobel Asset Management who specialises in emerging market fixed income, while it is clear there is a strong link between economic wealth and ESG scores, how the paper proposes to deal with it “is overkill”.  

As ESG levels and income are related, “by only focusing on the residuals, you end up filtering out something that is the result of – and at least very correlated to – all the good/bad policies of the past”, he says. He adds the report “goes too far” by disregarding the fact that having a high ESG score is the result of “doing a proper job and developing your country into a high income economy”.

Especially for high scoring countries, “you are erasing something that is most probably the result of good policies from the past”, Vermassen says, arguing that, at least for the majority of countries, “you do get to the status of being a ‘rich’ country by having good policies”.  

Senior portfolio manager at BlueBay Asset Management Jana Velebova-Harvey also says to gain a full picture, FTSE Russell’s analysis needs to be put into the context with the likely future ESG trajectory of sovereign issuers, which ex post data can shed only limited light on.

This is where engagement with sovereign issuers and forward-looking granular analysis of underlying data needs to step in, she says. This was also concluded in a World Bank paper that highlights simplistic income adjustment of underlying data may lead to overcorrection, and suggests putting greater emphasis on individual countries’ progress over time rather than between-country comparisons and peer group analysis.

In response to these critiques, Moussavi says: “We welcome a debate about ways in which investors can address these issues, and agree that further research is required to better understand how income-adjusted sovereign ESG scores could be used in a forward-looking assessment.”

This study examines one possible approach to deal with the income bias, and FTSE Russell provides income-adjusted scores as an alternative to its standard scores. The adjusted scores underline that it is also necessary to compare countries within their income group. 

Different methods for calculating ESG

Natasha Smirnova, senior sovereign analyst, emerging markets fixed income at PineBridge Investments, says ESG in sovereign is still a relatively new field that is developing fast, with more research needed into various factors that can improve methodologies. She says: “It remains a significant challenge to select ESG indicators, with investors tending to pick metrics that appear the most relevant for their processes and values, resulting in disparate outcomes.” 

To overcome this challenge, Smirnova says PineBridge identifies and focuses on factors that lead economic development, rather than lag per capita income, to get a clearer picture of ESG in emerging markets.

She says: “Our analysis focuses on ESG factors that can be mapped to the macro fundamentals and feed into vulnerability indicators. This approach results in a better, lower correlation between the ESG scores and per capita income, a flatter distribution for the overall ESG scores, with much less ‘punishment’ for lower-rated sovereigns.”

Vontobel’s Vermassen says best policy should find a middle ground between being “primarily backward-looking and almost fully erasing the past”. He says Vontobel puts a lot of emphasis on the performance progress by using a ‘momentum indicator’, which calculates what the recent score change is compared to the score from one, two and three years ago. 

This method rewards countries that take good measures and make a difference within the term of a typical electoral cycle. This way, the scores are more dynamic on ESG dimensions that are typically slow moving, and it enables poor countries with typically low ESG scores to improve their ranking in a material way. 

Other ESG issues for fixed income

Smirnova says the largest issue for emerging markets is data: “There are a vast number of countries and possible ESG indicators to consider, and the available data is often patchy and hard to compare due to varying methodologies.” 

PineBridge has seen some improvement in data collection and ESG metrics across emerging markets, she says. Some countries, notably Uruguay, are making particularly good progress in speeding up data verification, which is typically a very slow process. 

However, government engagement with investors remains a challenge in emerging markets, she says, and “we are advocating for the establishment of ‘ESG representatives’ to help guide investors to the relevant government ministry or department”.

BlueBay’s Velebova-Harvey says there are many further dilemmas for sovereign issuers, including: how do you offset strong performance within one category with a poor track record in another? What’s the best way of measuring a country’s performance in the environmental dimension? Which sovereign ESG provider ratings accurately represent the reality, given notoriously low correlation among data providers? Should ESG risk and opportunities be factored in the same way when investing into a short-dated sovereign T-bill vs a 100-year bond?

She says: “Ongoing challenges in ESG sovereign analysis mean it still remains more of an art than a science. Analysis needs to be multi-dimensional and quantitative data needs to be overlayed with forward-looking qualitative analysis, supplemented by frequent engagement evaluating the future path of a sovereign. Relying simply on backward-looking, quantitative data is to the detriment of both issuers and investors alike.”

Looking ahead

Moussavi says looking ahead, he thinks asset managers should try to take into account these income-adjusted datasets in order to find investment opportunities. He says initially, ESG assessments were risk management exercises and managers wanted to minimise ESG risk by investing in the best performers.

However, if managers want to make “impact investments or responsible investments”, they need to rethink the way they are allocating capital. Moussavi says that capital should go to countries for “good reasons”, such as for the transition into a low carbon economy. Looking at datasets such as FTSE Russell’s can help managers to find opportunities in low income countries that are actually performing well with regard to ESG. 

Velebova-Harvey says arguably it is emerging markets that need the majority of global investment to meet Sustainable Development Goals by 2030. According to a report by the United Nations Conference on Trade and Development, $5-7tn per year is needed globally to meet these goals, with estimates of $3.3-4.5tn per year for developing countries, mainly to fund basic infrastructure, food security, climate change mitigation and adaptation, health, and education.


Article updated on July 15 to specify Carl Vermassen’s area of expertise

A service from the Financial Times