ESG in sovereigns: a dynamic model for enhancing sovereign debt investing

• We believe that all three factors—E, S, and G—can affect the long-term risk/return profile of sovereign debt issued in both developed and emerging markets.

• To assess sovereign ESG risks and opportunities, we take a multifaceted approach involving historical data input from dozens of nongovernmental and supranational sources. 

• In addition to looking backward in time, we also look forward, using a dynamic approach for capturing the trajectory of sovereign sustainability in our model’s signals, including governments’ efforts—or failures—to institute positive reforms.

• Our model highlights sovereign ESG exposures for our portfolio managers and analysts, helping to build risk-aware cases for tactical positions and for identifying strategic opportunities, particularly where the model identifies sovereign outliers relative to their country peer group.

Robust sovereign environmental, social, and governance (ESG) analysis and systematic ESG integration have lagged in the rapid development of contemporary practices of sustainable investing. Beyond the usual challenges of ESG data availability and lack of consistent assessment standards and frameworks, development in this space has been slow given the widespread focus of sustainable investors on corporate ESG risks, and a general tendency to conflate sovereign ESG specifics with macroeconomics and geopolitics. Given that global sovereign bonds represent an estimated $66 trillion¹ of the $100 trillion² bond market worldwide and are a staple of long-term-oriented investors, the asset class deserves more attention.

Analysis of sovereign creditworthiness has historically focused on elements such as economic growth, fiscal and monetary policy, external accounts, debt dynamics, and institutional strength. Whereas the materiality of sovereign governance (often termed political risk) has long been appreciated by traditional analysis, environmental and social factors have been secondary or conflated with other macroeconomic concerns.

We believe that all three factors—E, S, and G—can affect the long-term risk/return profile of sovereign bonds, as well as the volatility of spreads. We chose to explore a model-driven approach to benchmarking sovereigns on ESG factors for two reasons: first, because we believe that current economic data doesn’t fully reflect the value of ESG factors to national economies or the exposure of governments to risk because of them; and second, because we think that distilling the handful of potentially material ESG factors into a signal would supplement our investment teams’ views.

"Robust sovereign environmental, social, and governance (ESG) analysis and systematic ESG integration have lagged in the rapid development of contemporary practices of sustainable investing."

Our model of sovereign ESG analysis was constructed as a collaboration between our dedicated ESG team and several investment teams that regularly allocate capital to sovereign debt, including global fixed income, emerging market debt, Japanese fixed income, Asia (ex-Japan) fixed income, and Canadian fixed income. This lends conviction to our approach and keeps it relevant to our investment practices, providing a method of measuring the materiality of ESG factors that’s scalable and responsive to the real-time experience of our portfolio managers and investment analysts.    

We’re not the first asset manager to develop a thoughtful model of sovereign ESG analysis. Indeed, a handful of leading firms have sought to bring ESG insight into their sovereign debt frameworks, and our efforts in this regard have some resemblance to those of our peers. In particular, we share a number of beliefs with other firms, including that:

• Our model primarily aims to highlight sovereign ESG exposures for our portfolio managers and analysts, helping to build risk-aware cases for tactical positions and for identifying strategic opportunities, particularly where the model identifies sovereign outliers relative to their country peer group
• Improving trajectories on sustainability factors, particularly with respect to governance, signal potential risk mitigation or potential for stable and improving economic performance
  
• Environmental factors can help investors identify whether sovereigns are using their natural resources sustainably
• Strong ESG scores can signal potential for greater resilience in the face of domestic or external challenges

But we also have some important differences:

• We think that social and environmental factors rise in significance over time, and that our view of the factors most important to sovereign ESG performance should adjust relative to time horizon
• We think that given the slow-moving nature of sovereign ESG indicators, it’s important to build a view on a country’s current momentum into the assessment
  
• Dialogue with sovereigns is critical, even if it can be complicated and requires diplomatic sensitivity. Our model incorporates information from these engagement efforts, which can lead to dramatic adjustments in the signals it generates

Our dynamic sovereign ESG model provides group-relative rankings for 200 countries based on a multidimensional scoring method focused on distinct E, S, and G factors. The model incorporates historical data, real-time assessments of sovereign idiosyncrasies, and forward-looking momentum adjustment factors. The combination of these perspectives reflects the careful calibration of active investment insight with fundamental ESG data.    

Defining ESG materiality

Based on our deep experience in analyzing ESG risks, we chose a subset of available indicators that we believe are relevant aspects in assessing sovereign risks. For each factor, we aggregate several key performance indicators (KPIs) into group-relative country scores, and we expect, over time, to adjust KPIs as conditions warrant.

We feel the environmental indicators selected enable assessment of how a country is managing its resources, and its energy and emissions profile. The social indicators distill a view of a country’s investment in the potential of its citizenry through education, promotion of technology, a more balanced distribution of wealth, and legal protections. Last, we’re confident that a country that’s transparent, controls corruption, and fosters stable government should be stronger over the long term.

Building responsiveness into the model

We adjust the assessment of each ESG factor based on our real-time investment experience. These adjustments aren’t limited to evaluations by our portfolio managers. They also include the ESG team’s analysis of controversies, our assessment of significant changes in sovereign-specific idiosyncrasies, such as overreliance on individual commodities or low-cost labor, or regression in governance norms. In addition, they can incorporate findings from bilateral engagements conducted by our ESG and portfolio management teams with representatives of sovereign issuers. This makes the results more reflective of our credit and investment teams’ expertise and less dependent on historical data sets that struggle to capture real-time developments on a quarterly and annual basis.

Momentum is a factor of sovereign sustainability

The model would be incomplete without a quantitative component designed to capture countries’ attempts to institute long-term structural changes to enhance their ESG resiliency. This momentum adjustment draws on two primary data sources: the survey-based Worldwide Governance Indicators, which map sovereign trajectories according to ESG criteria; and a separate academic data source that summarizes and tracks countries’ vulnerability to climate change and other global challenges in combination with their readiness to improve their resilience. We believe that capturing directional change, whether positive or negative, makes the signals from our ESG assessment model more country-specific and captures the direction of sometimes slow-moving sovereign reforms.

Factor weightings are also time sensitive

The model’s weighting of momentum-adjusted ESG factors is time sensitive. This reflects our conviction that the materiality of different ESG factors shifts over time, and that this matters for bond investing where the maturity date of issuance can be both relatively short and multidecadal. Whereas the five-year horizon for E, S, and G is weighted 20%, 35%, and 45%, respectively, a timeframe greater than five years puts an incrementally heavier weight on the E (25%) and S (40%) factors relative to the G (35%) factor. The primary reason for this weighting change is our conviction that the E and S factors become more prominent drivers of economic hardship and wealth creation than the G factor over longer timeframes.

"Australia's strongest ESG pillar is the social factor, but the country has shown weakness in certain environmental KPIs focused on resource conservation, underscoring its vulnerability to climate change in light of weak climate policy." 

When we compare baseline scores with short- and long-term momentum-adjusted scores, we can locate sovereigns of interest to bond investors with an investability assessment that may improve or decline in line with ESG signals over time. While causes may vary, the larger weight applied to the E and S factors for long-term scores is a key driver of positive and negative shifts, as is the magnitude of the momentum factor.

South Africa, for example, earns “moderate” baseline and near-term momentum-adjusted scores, but exhibits weak environmental KPIs—particularly around energy and water resource management and consumption. As the E factor weighs more on the long-term score, and because the country’s readiness to adapt to worsening water shortages driven by climate change appears likely to deteriorate, its long-term momentum-adjusted ESG score falls to “weak.”

Australia is another case—and one that bears watching from the perspective of our model's dynamic blending of historical and forward-looking data. Australia's strongest ESG pillar is the social factor, but the country has shown weakness in certain environmental KPIs focused on resource conservation, underscoring its vulnerability to climate change in light of weak climate policy. Before the wildfires of 2019 revealed deficiencies in the government’s response to fire-related risks, Australia received a positive momentum adjustment in our model. Now, however, we expect this adjustment will move to neutral or negative unless Australia undertakes rapid structural changes to improve its climate resilience.

Group-relative scoring

Finally, we think it’s important that ESG analysis takes into consideration the fact that countries have radically different economic profiles, dictating the extent to which governments are able to manage sustainability risks and capitalize on available resources over the medium- and long-term. Based on historical data covering GDP, income, government corruption, and market liquidity, we group countries into four categories of economic development, which helps ensure we’re making apples-to-apples sovereign comparisons. This is important for our boutique in-house fixed-income investment teams, which need to leverage sovereign ESG information in the context of regional and emerging-market credit investing, or make strategic asset allocations to specific regions. We then assign scores to each sovereign for each ESG factor based on its deviation from the median of its peer group’s results.    

Armed with assessments for the E, S, and G factors, and momentum-adjusted near- and long-term overall ESG scores, our sovereign debt investors can make more informed investment decisions that contemplate the materiality and overall balance of ESG risk factors. The model standardizes our methods for incorporating our views on disruptive sovereign events, including regulatory regime changes, local political developments, geopolitical fluidity, and other real-time adjustments.    

Comparing ESG results with sovereign credit ratings

One application of the model is spotting ESG-driven credit risks, which we do by simply pairing the model’s output with credit ratings data. For example, if we compare Moody’s local currency credit rating with our model’s output for the 47 countries in the two emerging market country groups, we get surprising results. Some sovereigns with high credit ratings according to Moody’s rank among the weakest countries in terms of ESG factors. Conversely, some high performing countries from an ESG perspective are well below investment grade in traditional rankings.

According to Moody’s, about 30% of past sovereign defaults are directly related to institutional and political weakness, ranging from political instability and weak budget management and governance problems to political unwillingness to repay lenders. We conducted a comparison between Moody’s sovereign credit ratings and our ESG sovereign scores that demonstrates a low correlation between the two signals. From this, we conclude that while governance indicators may carry weight, ESG profiles overall don’t appear to be a substantial consideration in Moody’s approach to measuring credit risk, meaning that ESG assessments from our model may well be additive to traditional credit ratings.    

Given the sensitivity of our model, sovereign ESG performance may become key drivers in a forward-looking credit upgrade or downgrade thesis for an emerging market or, in the case of core strategies, the model can provide assurance regarding the relative steadiness of ESG factors affecting long-term positioning.

Costa Rica, for example, is a frontier country that stands out across E, S, and G factors—and has positive momentum—yet its debt rates a “B–” from multiple credit rating agencies. Lithuania is an A-rated sovereign by most credit rating agencies, but may deserve higher, judging from the resiliency suggested in its ESG scores.

Conversely, Bermuda, which is in the developed country category, receives “A” credit rating, but our models indicates only a moderate near-term ESG score and trends toward weak in the long term. This is due in large part because of Bermuda’s high emissions, which are driven by the fact that over 90% of electric power generation is derived from fossil fuels. Bermuda is currently is in the midst of its first Integrated Resource Plan, which provides a series of options to aggressively develop utility-scale renewable energy projects focused on wind, solar, and biomass. Bermuda’s generation mix must be monitored to assess whether it can accomplish its emission reductions goals, which would likely push its momentum-adjusted long-term trajectory in a positive direction.

Comparing ESG results with sovereign valuations

Another ready use of our model involves looking at the relationship between ESG performance and yield to connect to bond valuation metrics. The poor ESG performance of certain countries should theoretically be properly reflected in spread data, but surprising disconnects can appear when we see lower- or higher-rated countries trading at unexpectedly tight or wider spreads. In general, it’s reasonable to seek wider yield spread for debt issued by weaker sovereigns or sovereigns with higher risk exposures, including ESG risks.    

Case study: Brazil and the promise of structural reform

Our sovereign ESG model has also been helpful in building conviction for certain non-consensus investment views in our fixed-income portfolios. A recent example can be found in our fixed-income team’s assessment of Brazil’s government debt.

Historically, Brazil has been relatively challenged from an ESG perspective, from the tailings dam scandals of Vale to the corruption expose of state-owned Petrobras. However, from a traditional political risk evaluation of sovereign debt issuers, Brazil has been perceived to be a relatively safer credit than some emerging market peers. While we had generally maintained exposure to Brazil’s debt, we lacked a compelling reason to increase our position until the governance factor received a strong boost in 2018 with the election of Jair Bolsonaro, who swept into power on the promise of anti-corruption and structural reform.

The key point to note here is that nothing on the surface of our model’s historical data inputs at the time—whether from World Governance Indicators published by the World Bank, Transparency International’s assessment of sovereign corruption levels, or any of a variety of indicators around Brazil’s level of fiscal transparency and financial governance—would have changed the country’s sovereign ESG scoring. But we saw the potential for significant changes in governance, with the focus on anti-corruption, to translate into significant long-term benefits for Brazil’s competitiveness, government effectiveness, control of corruption, and the rule of law.

Since this decision, further developments with the international controversy over the 2019 wildfires in the Amazon have led to another shift in our model’s signal. Applying an adjustment to the environmental factor in light of these issues—before seeing them show up in the historical environmental KPIs—results in a decline in Brazil’s long-term momentum-adjusted score from moderate to weak.    

We believe that ESG risks and opportunities will be with us for the long term, and that they will become increasingly important for investors over time. Accordingly, we’ve sought to define our approach to a sovereign ESG scoring model that’s meaningful, easy to understand, and quantifiable, and at the same time not duplicative of the fundamental economic information that our investment teams employ in their research and portfolio construction processes.

As we recognize the unique political pressures on sovereigns and how they’re forced to use policy tools to handle profoundly disruptive conditions that are thrust on them, we think it’s important to capture the trajectory of potential future change in our assessments. With sometimes significant time lags in official data publications, momentum-adjusted scores and real-time adjustments from our proprietary interpretations of qualitative as well as quantitative data are critical elements of meaningful analytical judgment. Active assessment of ESG data and anticipated changes are therefore indispensable to our project. With this framework for identifying latent risks and opportunities, we believe the integration of ESG factors in sovereign analysis adds a long-term perspective that captures the best of what active fixed-income approaches have to offer.

1 Global Government Debt Chart Book,” Fitch Ratings, January 2019. 2 SIFMA Fact Book 2018, Securities Industry and Financial Markets Association (SIFMA), September 2018.