ESG for EMD: Toward Best Practice
Our Emerging Markets Debt team has been analyzing ESG indicators for sovereigns for eight
years, and started working with Sustainalytics five years ago to build ESG coverage of its
corporates universe. During that time, it has continually enhanced its ESG processes, and over
recent months the team has drawn some conclusions about best practice in integrating ESG
into emerging markets debt, and particularly corporate bond strategies. In this paper, we explain
the importance of integrating sovereign governance indicators into corporate-issuer analysis;
selecting ESG indicators for financial materiality; gathering proprietary corporate governance
data; and engaging with bond issuers that present particularly high ESG risks.
|
White Paper
|
JANUARY 2019
ROB DRIJKONINGEN
Co-Head of Emerging Markets Debt
NISH POPAT
Senior Portfolio Manager, Emerging Markets Debt
ESG for EMD: Toward Best Practice
Our Emerging Markets Debt team has been analyzing ESG indicators for sovereigns for eight
years, and started working with Sustainalytics five years ago to build ESG coverage of its
corporates universe. During that time, it has continually enhanced its ESG processes, and over
recent months the team has drawn some conclusions about best practice in integrating ESG
into emerging markets debt, and particularly corporate bond strategies. In this paper, we explain
the importance of integrating sovereign governance indicators into corporate-issuer analysis;
selecting ESG indicators for financial materiality; gathering proprietary corporate governance
data; and engaging with bond issuers that present particularly high ESG risks.
Downloaded from www.hvst.com by Jack Petersen (id:111894) on 03/15/2019
ESG FOR EMD: TOWARD BEST PRACTICE
2
Investors have long been aware of the impact that certain material environmental, social and governance (ESG) indicators can have on
asset risk and performance. They increasingly expect to see these risks addressed across their entire portfolios—including in emerging
markets sovereign and corporate debt—and they are beginning to differentiate between their external asset managers based on the
sophistication and effectiveness of their approach to ESG investing.
We have been integrating ESG indicators into our sovereigns investment process for eight years. We started working with the ESG
research and rating agency Sustainalytics five years ago to build coverage of our corporates universe, and have since expanded our
partnerships with MSCI and TruCost. Over recent months, as we have taken further steps to enhance our own processes, we have
drawn some conclusions about best practice in integrating ESG into emerging markets debt strategies, and particularly corporate bond
strategies, based on our experience.
First, because corporate risk is especially sensitive to sovereign risk in emerging markets, relevant sovereign ESG indicators should feed
into corporate ESG analysis just as sovereign credit indicators feed into corporate analysis, and that sovereign ESG analysis should be
as rigorous and comprehensive as possible.
Second, more rigorous empirical research continues to show that ESG indicators are linked to credit spreads and credit ratings, but it
also helps us to isolate those indicators that are directly relevant and material to companies.
Third, investors can get very useful data from ESG research providers, but it is important to rely on proprietary data and analysis
whenever possible, particularly on governance indicators, because there is evidently no standardization in what the research providers
are measuring and therefore no guarantee that their data is relevant to investors’ concerns.
And fourth, in addition to integrating ESG indicators into the investment process, engagement on ESG issues is both critical and
effective for bondholders.
We discuss these four conclusions in more detail in this article.
Executive Summary
•
Our Emerging Markets Debt team has been analyzing ESG indicators for sovereigns for eight years, and started working with specialist
ESG research providers to build coverage of its corporates universe five years ago.
•
As we have enhanced our ESG processes over the years, we have begun to draw some generally applicable conclusions about best
practice with respect to integrating ESG into emerging markets debt, and particularly corporate bond strategies.
•
In this article we present four of those conclusions as a roadmap to best practice:
–
Sovereign ESG indicators should feed into corporate ESG analysis just as sovereign credit indicators feed into corporate
credit analysis.
–
Empirical research is helpful not only to show the materiality of ESG indicators, but also to isolate those indicators that are the most
directly relevant and material to companies.
–
Proprietary ESG data may better reflect an investor’s views on financial materiality, particularly when it comes to governance
indicators; and proprietary scoring will ensure that indicators are weighted to reflect views on materiality across different sectors.
–
Engagement on ESG issues should not be confined to shareholders with voting rights: it is both critical and effective for bondholders,
as it improves an investor’s ability to identify significant ESG risks and monitor their trends within companies, thereby supporting the
right investment decision at the right time.
•
Overall, we are increasingly persuaded that a serious approach to ESG in emerging markets is primarily a proprietary approach,
achieved by partnering with third-party research providers, not relying on them.
Downloaded from www.hvst.com by Jack Petersen (id:111894) on 03/15/2019
ESG FOR EMD: TOWARD BEST PRACTICE
3
Country Risk Counts
We believe that it is important to integrate sovereign ESG risks into the analysis of corporate ESG risks, but in order for that to be
effective it is also important to make sovereign ESG analysis as comprehensive and rigorous as possible.
We empirically observe a relationship between ESG indicators and subsequent changes in the credit spreads of hard currency sovereign
bonds, and for eight years ESG scoring has been a part of our country credit model. However, while we have always believed that
environmental factors were material to sovereign credit analysis, in the past a shortage of appropriate environmental data has led us
to weight our analysis toward social and governance indicators.
In fact, until recently, our country credit model score gave only a 5% weight to the environmental score, out of its 40% weight to
ESG scores overall. Moreover, that 5% came entirely from one indicator: Energy Intensity, as measured by the nominal U.S.-dollar
GDP achieved for each kilowatt hour of energy usage. More recently we have seen an improvement in the quality and availability of
environmental data, and therefore also in our ability to demonstrate correlation with sovereign financial performance over time.
As Figure 1 shows, our “E” scores for issuers appear to be less clearly correlated with credit spreads than our “S” and “G”
scores. Nonetheless, they do have predictive power and we think it is important to recognize that heightened investor scrutiny of
environmental issues is likely to increase their impact. For example, institutional investors in France are now required to report the
carbon footprints of their portfolios; and in emerging markets debt specifically, the launch of JPMorgan’s “JESG” suite of indices in
April 2018, which apply an ESG overlay to its hard currency, local currency and corporate bond benchmarks, using Sustainalytics and
RepRisk scoring, is likely to channel more investor capital to issuers that perform better on these indicators.
To enrich the environmental sensitivity of our model, we took eight widely recognized environmental indicators—such as CO
2
Emissions per Capita, CO
2
Emissions per GDP, Proportion of Electricity Production from Coal Sources, positioning in the Notre Dame
Global Adaptation Index (ND-GAIN), and alignment with the UN Sustainable Development Goals—and looked at their relationships
with changes in credit spreads over three, six, nine, 12, 18 and 24 months.
The results led us to increase the weighting of “E” indicators from 5% to 15% by including five additional indicators that our tests
showed had a material impact on spreads. The new indicator weights were added at the expense of lower weights for some of the “S”
and “G” indicators.
FIGURE 1. PREDICTIVE POWER OF NEUBERGER BERMAN’S ESG SCORES ON HARD CURRENCY SOVEREIGN CREDIT SPREADS, 2000
– 2018
Source: Neuberger Berman, Bloomberg. A negative lagged correlation (a higher bar) indicates that a lower ESG or macro score has predicted wider credit spreads. Data
covers the period Q1 2000 to Q1 2018.
Correlation of ESG and macro scores with
change in spreads over six time horizons
0
-0.04
-0.08
-0.12
-0.16
-0.20
24 months
18 months
12 months
9 months
6 months
3 months
Our improved ESG
score has become more
predictive than our macro
factor score over the
medium to long term.
Macro
Combined ESG
Correlation of E, S and G scores with
change in spreads over six time horizons
0
-0.04
-0.08
-0.12
-0.16
-0.20
Governance
Social
Environment
24 months
18 months
12 months
9 months
6 months
3 months
For hard currency sovereigns, as Figure 1 shows, the newly enhanced ESG scores that we apply appear to be just as predictive of future
credit spreads as our macro scores, which cover indicators such as GDP growth, inflation, sovereign fiscal data and foreign exchange
reserves. In fact, as the time horizon lengthens beyond 18 months, the ESG indicators appear to exhibit even stronger predictive
Downloaded from www.hvst.com by Jack Petersen (id:111894) on 03/15/2019
ESG FOR EMD: TOWARD BEST PRACTICE
4
power—perhaps reflecting that they are more structural than cyclical. We have also looked into the materiality of ESG factors in the
more complex dynamics of local currency bonds, and our results indicate that there are similar relationships to those we see in hard
currency markets.
In our view, ESG indicators are material to the performance of emerging markets sovereign and local currency bonds, then, and it is
important to keep testing one’s scoring approach to ensure that it remains sensitive to the full range of material ESG indicators as
markets evolve. In addition, it is important to recognize that sovereign ESG-related risks are material to the performance of emerging
markets corporate bonds, too.
While our peers in developed corporate bond markets may have started to differentiate between regions and countries when it comes
to credit analysis since the financial crisis, they are still likely to apply the same ESG analysis, regardless of country, for each corporate
issuer. Similarly, not all third-party ESG research providers embed sovereign risks into their corporate scores, and those that do seem
to emphasize them in some sectors over others. That is ill-suited to emerging markets, however, where credit risk in general is much
more correlated with sovereign risk, and ESG risks in certain sectors exhibit substantial regional variability. For example, a sovereign
widening in AA- rated Qatar or in B+ rated Turkey has tended to be immediately followed by a widening in their corporate bonds’
spreads, albeit that corporate bonds in some countries or sectors might widen less or more, depending on macroeconomic, liquidity
and other factors.
The depth of our experience with integrating ESG indicators into our country credit model has allowed us to plug them into our scoring
for corporates. We have done so by allocating a substantial weight in our corporate governance score to the four sovereign indicators
that are most material for the business environment: Rule of Law (based on World Bank indicators); Corruption (based on Transparency
International’s Corruption Index); Ease of Doing Business (based on World Bank indicators); and Banking Sector Risk (based on
Standard & Poor’s analysis).
Salient and Material
Alongside our work on the relationship between ESG indicators and sovereign spreads, the research we have been doing since 2014
continues to show that ESG indicators are directly linked to corporate credit spreads and credit ratings. We show evidence of that below.
However, we think an investor’s objective should really be to identify the specific ESG indicators that genuinely make a difference to asset
prices (a finding that should determine which indicators are selected for issuer analysis); and how those indicators differ across countries
and sectors (a finding that should determine the weighting of those indicators when they are integrated into issuers’ ESG scores).
When we tested the new environmental indicators in our country credit model for materiality, for example, we found that whereas the
country score on Proportion of Electricity Production from Coal Sources showed a significant correlation with future credit spreads,
Proportion of Electricity Production from Renewable sources did not. This may reflect the relative importance that investors assign
to one indicator over the other, or it may reflect the fact that it is too early for risk premiums to respond to the renewables variable.
Whatever the reason, the result points to the importance of empirical testing of these indicators for materiality.
Our indicator-selection process took 75 environmental and social indicators from MSCI and Sustainalytics and regressed them
against nine financial metrics, and credit spreads, from the hundreds of emerging markets corporate bond issuers rated by those two
research providers. We found 24 indicators to have a material relationship on those metrics and spread (from Water Stress and Carbon
Emissions to Human Capital Development and Privacy & Data Security). Our regression results also enabled us to see how different
indicators are more or less relevant to different sectors, which informs subsequent weightings when it comes to ESG scoring. For
example, we found Biodiversity and Land Use to be material for the Utilities sector, but not for Financials; whereas Privacy and Data
Security was material for Financials, but not for Utilities.
Some of these indicators we also regard as “salient”—that is, they had no, or only a weak, relationship with financial metrics or
spreads in the regression analysis, but they have a clear and substantial impact on some or all of a company’s stakeholders, and could
over time become financially material via changing reputational and regulatory expectations. Examples would include whether or not
a company has a Community Involvement Program; this can be impactful and reputation-enhancing even if the regression analysis
suggests they have not historically significantly affected businesses’ cash flows or creditworthiness.
Downloaded from www.hvst.com by Jack Petersen (id:111894) on 03/15/2019
ESG FOR EMD: TOWARD BEST PRACTICE
5
Once we identified the indicators that met our thresholds for materiality and salience on a sector-specific basis, we were able to
dismiss the rest—the majority of indicators—from our scoring.
Governance Needs a Spotlight
The investment industry owes a lot to specialist ESG research and rating providers. They developed the early methodologies for
analysis of ESG investment risk factors, and their work is a big reason why data disclosure by corporates has improved and broadened
so much over the years, helping to change the challenge from data availability to data consistency. Having said that, we understand
that they are serving a wide range of clients and that they are trying to provide data that will be of interest to investors as varied
as quantitative equity strategists, short-term oriented hedge fund portfolio managers, corporate credit analysts and stand-alone
engagement service providers. It is therefore unsurprising that we use only a subset of the environmental and social data that they
provide, and it is essential that we are confident that we understand what that data is measuring—in the same way we are confident
that we know what Transparency International is measuring with its Corruption Index in the sovereigns space, for example.
On governance data, we are more purist—here we think a best-practice approach in emerging markets corporate bonds is to use
an analysis framework that is as proprietary as possible, for three reasons. Firstly, these indicators are the most directly material to
financial risk and, for that very reason, they ought to be covered as part of normal credit analysis; secondly, ESG research providers
appear to be measuring very different things to get to their governance scores; and thirdly, we believe the weighting that these
research providers allocate to governance indicators in their ESG scores understates its significance in our investment universe.
Many commentators have observed the inconsistent rating methodologies used by the ESG research providers at the overall ESG rating
level. It is well understood that the primary reason for this is the heritage, geographic location and initial client set of the various
providers. A European-headquartered provider grounded in a stakeholder framework focuses on different social factors than a
U.S.-headquartered competitor, for example.
What has not been given sufficient comment is the extent to which these research providers disagree on corporate governance.
When we compared the governance scores from MSCI and Sustainalytics on our own universe of emerging markets corporates there
appeared to be no relationship between the two sets of data in this sample (see Figure 2). We have also regressed these two data sets
against corporate bond spreads in our universe, and again found a correlation close to zero.
FIGURE 2. ON GOVERNANCE, ESG RESEARCH PROVIDERS SEEM TO MEASURE DIFFERENT THINGS
MSCI and Sustainalytics governance scores for 234 issuers in the Neuberger Berman emerging markets debt universe
Source: MSCI, Sustainalytics.
MSCI
0
10
20
30
40
50
60
70
80
90
30
40
50
60
70
80
90
100
Sustainalytics
Downloaded from www.hvst.com by Jack Petersen (id:111894) on 03/15/2019
ESG FOR EMD: TOWARD BEST PRACTICE
6
Given our concerns about the appropriateness for emerging market corporate debt investors of the governance methodologies used
by both Sustainalytics and MSCI, we decided to put the third-party governance data entirely to one side and build our own proprietary
governance data set. This decision was further informed by the preponderance of equity-focused indicators in the governance
ratings of both providers. By building our own proprietary governance data set, we believed we would be more likely to identify the
governance characteristics that are financially material for emerging market corporate credit.
Our fundamental credit research analysts developed five indicators for our proprietary governance score and then assessed issuers
one-by-one against these indicators to build out a proprietary governance data set. The five indicators include Senior Management
Quality (covering indicators such as experience, reputation, key-person risk, and turnover); Ownership and Board Quality (controlling
shareholder quality, board independence, audit effectiveness); Corporate Strategy (anti-competitive behavior, track record); Financial
and Accounting Strategy and Disclosure (complexity and aggressiveness of tax and financing structures); and Regulatory and Legal
Track Record (class actions, fines, regulatory actions, bribery, corruption, environmental or social claims). We have assessed more than
90% of our universe based on this proprietary data.
Figure 3 shows the correlation of our scores on these five indicators with the credit spreads of the 400-plus issuers that we have
covered. The closer the correlation is, the more the points on the scatter plots should cluster along a line from the top left-hand corner
to the bottom right-hand corner.
Source: Neuberger Berman. Data as of June 27, 2018. Neuberger Berman governance scores are generated by a proprietary Governance Assessment Tool, based on the
four sovereign governance indicators and five corporate governance indicators shown.
FIGURE 3. IDENTIFY AND THEN FOCUS ONLY ON MATERIAL OR SALIENT “G” INDICATORS
20
40
60
80
100
0
200
400
600
800
1000
Governance Score
G-Spread
Neuberger Berman governance scores for 460 issuers
plotted against those issuers credit spreads
Neuberger Berman Governance Assessment Tool (GAT) Indicators
Sovereign
Corporate
Rule of Law
Senior Management Quality
Corruption
Ownership and Board Quality
Ease of Doing Business
Corporate Strategy
Banking Sector Risk
Financial and Accounting Strategy
and Disclosure
Regulatory and Legal Track Record
We believe taking a proprietary approach to selecting ESG indicators is best practice, then. But we also believe that a proprietary
approach to generating ESG scores from those indicators is best practice. Only a proprietary approach can ensure that “E”, “S” and
“G” indicators are weighted in such a way as to reflect an investor’s own view of their materiality and saliency for different sectors.
Again, we think this is most pertinent to governance indicators. When we analyzed the weights given to “E”, “S” and “G” indicators
in the scores that third-party providers assigned to our universe of securities, we found that both MSCI and Sustainalytics weight
governance at around 30%. That may reflect the materiality of those indicators to equity performance, but we do not believe it reflects
their materiality to creditworthiness. At Neuberger Berman, we weight governance approximately 50% higher than that, on average.
We believe the payoff for this proprietary effort is set out in Figure 4, which shows how Neuberger Berman’s proprietarily selected and
proprietarily weighted ESG scores correlate with credit spreads and third-party credit ratings for our universe.
Downloaded from www.hvst.com by Jack Petersen (id:111894) on 03/15/2019
ESG FOR EMD: TOWARD BEST PRACTICE
7
FIGURE 4. THE MATERIALITY OF PROPRIETARY ESG SCORES AND PROPRIETARY WEIGHTINGS
Neuberger Berman ESG scores for 303 issuers plotted against those issuers’ credit spreads and credit ratings
Source: MSCI, Sustainalytics, Neuberger Berman. The first chart shows the proprietary NB ESG scores, and the second chart shows the average proprietary NB ESG score
for each credit-rating bucket, for the 313 JPMorgan Corporate Emerging Markets Diversified (CEMBI-D) issuers that provide full ESG data. Data as of October 1, 2018.
Mapping of ESG Scores vs Spreads
Average ESG Score per Credit Rating
0
100
200
300
400
500
600
700
800
900
1000
25
35
45
55
65
75
85
ESG Score
ESG Score
Spread
Credit Rating
45
50
55
60
65
70
AA
AA-
A+
A
A-
BBB+
BBB
BBB-
BB+
BB
BB-
B+
B
B-
CCC+
Once financially material ESG indicators have been selected and weighted, best practice requires that they be integrated into a scoring
system that has a meaningful and sector-specific impact on security selection.
Investors will approach integration in different ways. For our part, we feed ESG scores into our internal credit rating process, and have
recently moved toward a sector-specific, standard deviations-based integration approach: we upgrade or downgrade by one notch
when an ESG score is more than one standard deviation above the mean score of its sector; and we downgrade by two notches when
an ESG score is more than two standard deviations below the sector mean, thereby re-rating asymmetrically.
Under this methodology, ESG indicators affect the internal credit ratings of more than a third of the issuers in our benchmark index.
The resulting comprehensive risk rating is a key part of how we calculate the fair value of the bonds in our universe.
Better Engagement and Reporting, Even on Difficult ESG Issues
The final piece in the best-practice jigsaw is company engagement. This should not be confined to shareholders who have voting
rights. Our
previous article
on ESG in emerging markets corporate bonds mentioned the importance of engagement. In some ways it is
just a natural extension of the importance we assign to on-the-ground research in emerging markets, which was the topic of
another
article
we published earlier this year. We continually try to enhance our engagement program; we also believe that insights gained
from good ESG scoring can inform that program, and vice versa.
Whenever possible, our portfolio managers and economist will discuss with Sovereign Debt Offices, finance ministers and other
officials why we demand a certain yield premium to account for ESG risks and concerns. In addition, we engage mandatorily on ESG
matters when an issuer is involved in a severe controversy (as categorized by Sustainalytics and/or MSCI), or when an issuer’s ESG
score is low enough to trigger a downgrade in our proprietary risk rating methodology. We have long believed that it is only through
engagement that the underlying dynamics that lead to a low ESG score can be identified, and potential solutions proposed by company
management can be properly assessed. Furthermore, the quality of the response we get from an issuer to our engagement program
feeds into our ESG scoring system. Analysts have the discretion to suggest an additional one-notch downgrade in the event of a poor
response, but a response can also help them to identify, early on, a trend for improvement within a company.
A good engagement program will add another layer of proprietary tailoring to an ESG scoring system, then. But it is also necessary to
take ESG analysis beyond the corporate sector that tends to be well-covered by the third-party research agencies, to cover the state-
owned enterprises that represent a big share of bond issuance in emerging markets.
Downloaded from www.hvst.com by Jack Petersen (id:111894) on 03/15/2019
ESG FOR EMD: TOWARD BEST PRACTICE
8
Our engagement effort is increasingly getting traction. For example, a year ago we received a full written response from a very large
state-owned enterprise to a series of very difficult and potentially uncomfortable ESG-related questions. The company in question had
experienced a series of safety incidents, culminating in a serious accident; it operated in a number of countries with histories of human
rights abuses; and an ex-member of its senior leadership team is under investigation for corruption that led to a loss of assets.
We wrote to ask how they had adapted their safety, human rights and anti-corruption policies in the light of these facts. The response
detailed a safety program that we would credit with having prevented serious accidents at the firm; confirmed the terms of the
company’s human rights policy; and highlighted its internal anti-corruption regulations and training.
Finally, a best-practice approach to ESG, regardless of the asset class, requires full and transparent reporting of ESG risk exposure to
clients. Of course, this will draw on the independent scores provided by third-party ESG research providers, as a benchmark. Firms
such as MSCI and Sustainalytics are facilitating this by working to integrate their scores into widely used fixed income portfolio
management and reporting tools such as BlackRock Aladdin. We think that reporting should also include full explanations as to why
portfolio managers invest with issuers that exhibit poor ESG scores. From our perspective, for example, that may be because we feel
the risks are all reflected in market prices, or because we see or expect a substantial improvement in ESG performance, or because we
see potential for change through our engagement program.
Toward Best Practice
The commitments we received, in writing, from the management team of the state-owned enterprise we mentioned above show that
even the largest emerging markets issuers recognize the importance of ESG issues, and of engaging with their keystone investors
on those issues. We believe this is a significant aid to prudent risk management in markets where formal ESG data is frequently
inconsistent and third-party ESG data providers find it difficult to achieve comprehensive coverage.
But even where the availability of third-party data is abundant and growing, we believe that any asset manager or investor that
wants to move toward best practice for ESG in emerging markets needs to take a proprietary approach to generating ESG scores, a
proprietary approach to integrating them into the investment process, and an engaged approach to scrutinizing issuers that present
high ESG risks and, where possible, helping them to address their challenges.
In short, a serious approach to ESG in emerging markets is increasingly a proprietary approach, achieved by partnering with third-party
research providers, not relying on them.
Downloaded from www.hvst.com by Jack Petersen (id:111894) on 03/15/2019
Neuberger Berman
1290 Avenue of the Americas
New York, NY 1
0104-000
1
www
.nb.com
T0376 01/19 279412 ©2019 Neuberger Berman Group LLC. All rights reserved.
This material is provided for informational purposes only and nothing herein constitutes investment, legal, accounting or tax advice, or a recommendation to buy, sell
or hold a security. Information is obtained from sources deemed reliable, but there is no representation or warranty as to its accuracy, completeness or reliability. All
information is current as of the date of this material and is subject to change without notice. Any views or opinions expressed may not reflect those of the firm as a
whole. Neuberger Berman products and services may not be available in all jurisdictions or to all client types.
This material may include estimates, outlooks, projections and other “forward-looking statements.” Due to a variety of factors, actual events or market behavior may
differ significantly from any views expressed. Investing entails risks, including possible loss of principal. Investments in hedge funds and private equity are speculative
and involve a higher degree of risk than more traditional investments. Investments in hedge funds and private equity are intended for sophisticated investors only.
Indexes are unmanaged and are not available for direct investment.
Past performance is no guarantee of future results.
This material is being issued on a limited basis through various global subsidiaries and affiliates of Neuberger Berman Group LLC. Please visit
www.nb.com/disclosure-
global-communications
for the specific entities and jurisdictional limitations and restrictions.
The “Neuberger Berman” name and logo are registered service marks of Neuberger Berman Group LLC.
Downloaded from www.hvst.com by Jack Petersen (id:111894) on 03/15/2019
Next
More from Money Management Institute Sustainable Investing Community
