The unintended consequences of sovereign ESG benchmarks

The equity domain has seen a proliferation of ESG benchmarks for many years and more recently this has extended to corporate bond benchmarks. Sovereign or government bond index providers like Bloomberg Barclays, ICE BofA, JP Morgan and FTSE Citi have been waking up as well to create ESG, ESG-tilted, climate and green bond indices for both developed and emerging markets. This has been largely driven by a demand from asset managers and owners to do something with ESG integration for government bond portfolios but may have unintended consequences for investors and sovereign issuers.

First of all it is important to understand that government bonds are typically used to finance the public sector and critical elements of society, i.e. education, health care, social services and national infrastructure. In the Netherlands this amounted to 70% of all expected 2020 national government expenditures¹ and in many emerging markets this is lower but growing, particularly in the light of COVID-19 and the need to upgrade national health care capacity.

Second, it is essential to understand how an ESG benchmark is constructed, as data sources and methodologies vary widely. The benchmark provider commonly takes data from either one or multiple 3rd party ESG data providers such as MSCI, Sustainalytics or RepRisk (who may themselves use external data from for example the World Bank), directly from an underlying data provider like the World Bank or UN, or utilize news scraping methodologies. Subsequently, scores are amalgamated through custom re-weighting, scaling and/or exclusions leading to an altered benchmark universe.

Consequences for emerging markets
The implications for developed high income countries where governments can borrow at near 0% interest rates are relatively nimble but the implications for emerging and especially lower income countries to finance the public sector and development are more precarious.

In one of the major emerging market sovereign bond indices, the methodology employed results in the exclusion of seven countries from the benchmark universe and a sizably lower weight for many other lower income countries. Out of these, six countries are from Africa where countries have made sizable efforts in the last few years to establish access to international capital markets only to lose it in ESG benchmarks.

In effect, ESG scores correspond very strongly with wealth as measured by GDP per capita; the lower the GDP per capita the lower the ESG score. Ironically, for certain countries this is a double hit: their weight in indices is low because they have a low debt/GDP ratio and exhibited fiscal prudence in the past. On top of that, they face a scaled down weight or exclusion altogether.

The United Nations Sustainable Development Goals (UN SDGs), which virtually all countries have signed up for with the aim to meet certain government targets by 2030, are increasingly being integrated into the investment process by investors. Both the growing adoption of ESG benchmarks for emerging markets and the growth of passive index tracking raises the barrier for low income countries to finance progress towards 2030 UN SDGs.

Investors be aware!
Investors should be aware of the risks of outsourcing ESG integration decisions to 3rd parties and anchoring it in benchmarks:

• Responsibility/appropriateness. Certain government bond indices use corporate reputation risk data which does not adequately capture incidents which are attributable to the government or affect the risk for a sovereign bond investor. For example, we observed instances of corporate corruption by Western firms in Africa (e.g. airline crash, mining related environmental damage or supply chain issues) which are now effectively impacting benchmark membership and sovereign borrowing access and costs.
• Counterintuitive. Often anti-corruption campaigns where legitimate court processes are followed create news articles and actually lower the ESG score.
• Wash-out effect. In the case of news scraping algorithms, large countries often have a range of irrelevant news articles in the data set, thereby increasing the average ESG score.
• Freedom of press. On the contrary, countries or corporates which supress news flow will generally have overstated ESG scores.
• Timeliness. ESG data is often backward looking and sometimes provided with a lag of up to two years making it slow to capture changes in circumstances.

Many asset owners and asset managers are likely unaware of the unintentional consequences of benchmark choices for themselves, let alone the sovereign issuer. ESG scores do not tend to look at whether one is compensated for risk and the ESG trend as opposed to the ESG level is a more valuable metric to look at.

Solutions and considerations

• Understand the underlying methodology and intentionality before choosing a benchmark. Similarly, sovereign issuers should be aware of how ESG benchmarks operate and how they are scored.
• Stimulate thematic bond issuance. Targeted use of proceeds investments, for example through green bonds or supranational issuance, transparency and reporting can mitigate this concern and risk. Voices are emerging to think about sovereign bond covenant tied to UN SDG progress.
• Give consideration to having a dialogue with sovereign bond issuers around ESG concerns. The UN Principles for Responsible Investment (UN PRI) Sovereign Debt Advisory Committee, which PGGM is part of, has a paper forthcoming following their inaugural guide on ESG integration for sovereign debt². Exclusion should be applied in the portfolio when all forms of engagement and dialogue have been exhausted, whilst leaving the door open should there be a policy shift.
• Consider active investing around a benchmark. If an investor truly wants to make a positive impact on the world then they should reward and deploy capital to those countries that are striving to improve in terms of ESG criteria rather than just those that already score highly (and have low yields).

Links:
1. https://www.rijksoverheid.nl/onderwerpen/prinsjesdag/inkomsten-en-uitgaven-van-het-rijk-2020
2. https://www.unpri.org/fixed-income/a-practical-guide-to-esg-integration-in-sovereign-debt/4781.article