Research Insight w Cover

msci.com

Research Insight

Optimizing Environmental, Social and Governance Factors in

Portfolio Construction: Analysis of three ESG-tilted strategies

Zoltán Nagy ([email protected])

Doug Cogan ([email protected])

Dan Sinnreich ([email protected])

February 2013

Abstract Institutional investors wanting to integrate Environmental, Social and Governance (ESG) factors

in their investment strategies need the right tools to measure portfolio risk characteristics and

performance. MSCI‟s BarraOne and Barra Portfolio Manager can provide this utility with

Intangible Value Assessment (IVA) ratings from MSCI ESG Research.1 In this study, we

examine the use of IVA ratings with the Barra Global Equity Model (GEM3) to build optimized

portfolios with improved ESG ratings, while keeping risk, performance, country, industry and

style characteristics similar to conventional benchmarks, such as the MSCI World Index. The

currently available dataset of IVA scores allowed us to compare three strategies during the

period between February 2008 and December 2012, using current IVA ratings methodology. Of

the three strategies, we found the best active returns during this period were achieved by

overweighting firms whose IVA ratings improved over the recent time period, showing ESG

momentum. Underweighting assets with low ESG ratings also raised portfolio performance

during this period. The highest ESG rated assets had more uneven performance; they did better

in periods of limited risk appetite during this volatile market cycle.

1 MSCI ESG Research is produced by Institutional Shareholder Services Inc. or its subsidiaries (“ISS”). ISS is a wholly owned subsidiary of MSCI Inc.

MSCI ESG Research msci.com © 2013 MSCI Inc. All rights reserved. Please refer to the disclaimer at the end of this document Research Insight

Optimizing ESG Factors in Portfolio Construction

2 of 23

Introduction

For the last two decades, institutional investors have debated whether considering

Environmental, Social and Governance (ESG) factors can lead to better financial returns.

Skeptics maintain that so-called “Socially Responsible Investing” (SRI) does not support

fundamental analysis, restricts the investable universe, and runs counter to tenets of Modern

Portfolio Theory. Proponents argue that markets do not efficiently price ESG factors because

they address long-term risks that have not been absorbed by the economy, and that alpha

generation is possible as markets begin to recognize these undervalued influences.

Academic studies and industry analyses have supported both sides of this argument, although on

balance they find that investors employing ESG factors do not impose a significant performance

penalty, that investors can achieve comparable risk-adjusted returns to non-ESG tilted strategies,

and that investors may be able to enhance their returns through the use of certain ESG strategies.

For example:

Two reports by Mercer found that 20 of 36 peer-reviewed studies showed evidence of a

positive relationship between ESG factors and financial performance, while five showed

mixed results and eight showed a neutral relationship. Three others found negative

results. 2

A recent Deutsche Bank study of 56 peer-reviewed papers found that 89 percent linked

consideration of ESG factors with market-based outperformance, and that application of

exclusionary SRI screens was responsible for any negative results.3

These findings support an evolving view of ESG investing strategies. What started as a “values-

versus-profits” proposition, focused on screening out objectionable stocks from an SRI

perspective, has moved toward “best-in-class” investing that attempts to seize on mispriced

market risks and opportunities.4 Another commonly cited view is that ESG factors form a

fundamental measure of risk that correlates with the cost of capital – equity or debt – and that

companies with high ratings tend to outperform on operational and accounting-based metrics.5

Still, there are lingering questions about the alpha-generating power of ESG-tilted strategies.

One line of argument is that the market has evolved to the point where ESG factors – led by

governance and followed by environmental factors – are now efficiently priced so it is harder to

gain a stock-picking advantage as opportunities have become narrower and are moving away

from values-driven investors.6 At the same time, many large asset owners may have

2 See, for example, Mercer Investment Consulting, “Shedding Light on Responsible Investment: Approaches, Returns, Impacts,” November 2009; and Mark Fulton,

Bruce Kahn and Camilla Sharples, “Sustainable Investing: Establishing Long-term Value and Performance,” Deutsche Bank Climate Change Advisors, June 2012.

3 Mark Fulton, Bruce Kahn and Camilla Sharples, “Sustainable Investing: Establishing Long-term Value and Performance,” Deutsche Bank Climate Change Advisors,

June 2012.

4 Jeroen Durwall, Kees Koedijk and Jenke Ter Horst, “A tale of values-driven versus profit-seeking social investors,” Journal of Banking and Finance, January 2011.

5 See, for example, Sadok El Ghoul, Omrane Guedhami, Chuck C. Y. Kwok, and Dev R. Mishra “Does corporate social responsibility affect the cost of capital?”

Journal of Banking & Finance, September 2011; Subdeer Chava, “Environmental Externalities and the Cost of Capital,” Georgia Institute of Technology, 2011; and

Rob Bauer and Daniel Hann, “Corporate Environmental Management and Credit Risk,” Maastricht University, European Centre for Corporate Engagement, June 30,

2010.

6 Lloyd Kurtz and Dan DiBartolomeo, “The Long-term Performance of a Social Investment Universe,” The Journal of Investing, Fall 2011.



3 of 23

fundamental considerations related to asset allocation and risk tolerance that may limit their

ability to pursue alpha-oriented ESG strategies.

However, if the size and diversification of institutional investor portfolios is such that they

effectively own a cross-section of the global economy, possible externalities associated with

portfolio holdings effectively become internal.7 And if not owning those companies exposes

them to a possible performance penalty, then they have the option of building a strategy that tilts

the portfolio to capture positive ESG exposures while minimizing negative ones.8

Applying MSCI ESG Research Ratings in Portfolios

Two recent studies have investigated the results of using such a “best-in-class” strategy from the

application of ESG ratings available through MSCI ESG Research. Intangible Value

Assessment (IVA) scores (originally developed by Innovest Strategic Value Advisors, a firm

MSCI acquired in 2010) evaluate sector-specific ESG material risks and opportunities. IVA

scores are distributed on a seven-point letter scale („AAA‟-„CCC‟) to reflect the relative strength

of companies' management of these risks:

RCM, an investment management company owned by Allianz Global Investors, found

that a strategy of investing in the top two of five tiers of IVA-ranked companies added

1.6 percent a year in excess returns for the period December 2005 through September

2010. The study found that returns from portfolios of European companies represented

the largest and most consistent spread between best-in-class and worst-in-class

companies, reflecting greater integration of ESG factors in Europe. The volatility of this

ESG-tilted portfolio did not increase relative to the MSCI World Equal Weighted Index

(encompassing 24 developed markets) during this period.9

Credit Suisse used IVA ratings over a slightly more recent time series and produced

similar results. From May 2006 through May 2012, this study found that companies with

an IVA rating of „A‟ or above outperformed the MSCI World Equal Weighted Index.

However, it also found that companies with an IVA rating of „CCC‟ outperformed the

index. The paper went on to explain possible factors for this countervailing result. One

of the possible factors is that the IVA ratings had a greater proportion of companies in

Europe with high ratings that were disproportionately exposed to the European debt crisis

in 2011 (which emerged after the RCM study was concluded). Second, companies with

smaller market capitalizations and lower price-earnings ratios also had lower average

ratings in the IVA data set. This study concluded that higher-rated companies performed

better during this period when markets were in a defensive, risk-off condition, and

companies with improving ESG ratings showed the most momentum during this period in

raising their market values.10

7 This is sometimes referred to in the ESG literature as a “universal owner” concept. For a seminal example, see James P. Hawley and Andrew T. Williams, “The

Emergence of Fiduciary Capitalism,” Corporate Governance, Vol. 5, No. 4 (1997). Available at SSRN: http://ssrn.com/abstract=11430

8 For more background, see Remy Briand, Chin-Ping Chia, Roger Urwin, MSCI Inc., “Integrating ESG in the Investment Process: From Aspiration to Effective

Implementation,” MSCI Research Insight, August 2011.

9 RCM Sustainability White Paper, “Sustainability: Opportunity or Opportunity Cost?” Allianz Global Investors, July 2011.

10 Credit Suisse Strategy Paper, “ESG Investing Themes: Are „good‟ stocks good investments?” Credit Suisse, Oct. 11, 2012.



4 of 23

These findings set the stage for this MSCI Research Insight paper where we analyze the effect of

ESG ratings on portfolio performance for the period of February 2007 through December 2012,

and the effect of ratings changes starting in February 2008.11

For the ratings, we apply

Intangible Value Assessment (IVA) scores that evaluate sector-specific ESG material risks and

opportunities on a 10-point scale, which are converted into final letter grades of „AAA‟ to

„CCC.‟ We adopt the MSCI World Index as our performance benchmark and as the asset

universe for the optimized portfolios. The Barra Global Equity Model (GEM3) is used to build

and analyze three families of optimized ESG-tilting strategies.

Unlike the majority of studies that seek to test the alpha generation of companies with high or

low ESG ratings, we use the model to create portfolios that have region, sector and investment

style risk profiles similar to the benchmark‟s exposures. The risk model also allows us to

separate systematic sources of active return – that is, common factor contributions – from asset

specific return sources associated with IVA scores. While our study was designed primarily as

an enhanced indexing exercise, focused on achieving benchmark returns comparable to the

MSCI World Index, we also found three possible strategies during the observed period that can

raise ESG ratings and improve active returns with minimal effects on benchmark tracking error.

We acknowledge that this research is based on a short sample of data, spanning nearly six years.

As additional data becomes available to extend the time period, investors may wish to conduct

further research on these strategies to compare the results presented here.

Three Strategies that Lead to an ESG Tilt

We explore three optimized strategies that implement an ESG tilt of the MSCI World Index,

based on the IVA scores of underlying portfolio holdings.

The first strategy is called an “ESG worst-in-class exclusion” approach (hereafter

referred to as “ESG exclusion”). It is based on excluding the companies with the lowest

current ratings („CCC‟), which results in a narrower investment universe. We first

analyze the performance of this restricted market cap weighted portfolio. As a second

step, we further enhance this pure exclusion strategy by overweighting stocks with high

current ESG ratings and underweighting those with low current ratings inside the smaller

universe, while maintaining other exposures of the portfolio very close to the

benchmark‟s exposures.

The second strategy is called a “simple ESG tilt” approach. Here we do not exclude any

stocks based on their ESG ratings. Rather, we overweight stocks with high current ESG

ratings and underweight those with low current ratings, while maintaining other

exposures of the portfolio very close to the benchmark‟s exposures.

The third strategy is called an “ESG momentum” approach. Here again, we do not

exclude any stocks based on their ESG ratings. Instead, we overweight stocks that have

improved their ESG ratings during the preceding 12 months over the time series, and

underweight stocks that have decreased their ESG ratings over the same period. We

11 This period was selected to allow back-testing of a time series using current IVA ratings methodology.



5 of 23

expect the resulting portfolio to reflect companies whose ESG trajectory is positive, even

though it will be less tilted towards companies with high current ESG ratings than the

simple ESG tilt portfolio referenced above.

The implementations of these strategies share some common characteristics, especially regarding

the investment constraints that were applied in the construction of the portfolios. The goal of

these constraints is to assure that the optimized portfolios have region, sector and style

characteristics close to the benchmark‟s characteristics and close to each other, and thus

minimize the effect of these common factors on portfolio performance.

In this section, we provide a high-level comparison of these strategies. In subsequent sections,

we give more details on each strategy, and on the sources of their performance. General

optimization methodology and settings are described in greater technical detail in the Appendix.

Table 1: Comparison of ESG Strategies, February 2008 – December 2012.12

In Table 1, we compare risk and performance characteristics of our three approaches relative to

the MSCI World Index. Due to the one-year lag in applying ratings changes in the ESG

momentum strategy, we show statistics for the common time period ranging from February 2008

through December 2012. All three strategies achieved positive active returns, with important

asset specific contributions (except for the ESG tilt strategy), meaning that ESG scores provided

these active returns once other systematic contributions were factored out.

At the same time, the resulting portfolios were tilted towards companies with higher ESG scores.

ESG exclusion and ESG tilt strategies led to a higher average ESG portfolio rating than ESG

momentum strategies, but their risk-adjusted performance trailed by a distinct margin. We

conclude that during this period it would have been possible to raise the ESG tilt of the MSCI

World Index by one ratings notch, from „BBB‟ to „A,‟ without harming portfolio performance in

terms of active returns, while maintaining a small tracking error.

12 To be consistent in comparing results, the figures presented in this table are for February 2008 – December 2012 only, corresponding to the time period available for

the ESG momentum strategy. Portfolio ESG score is calculated over the indicated time period as the market capitalization weighted average score of portfolio

constituents, where missing values are replaced by zero. The benchmark portfolio is the MSCI World Index and had an average IVA score of 5.4 during the period

February 2007 – December 2012, as well as during the period February 2008 – December 2012.

ESG strategyESG

ExclusionESG Tilt

ESG

Momentum

Active return (annual, %) 0.10 0.05 0.35

Common factor contribution (annual, %) 0.06 0.03 0.08

Asset specific contribution (annual, %) 0.05 0.01 0.27

Tracking error (ex-post, annual %) 0.45 0.46 0.36

Information ratio 0.23 0.10 0.97

Average improvement in ESG score 1.27 1.21 0.46

Average relative improvement in ESG score (%) 23 22 8



6 of 23

Details of ESG Worst-in-Class Exclusion Approach

The ESG worst-in-class exclusion approach comes closest to traditional forms of Socially

Responsible Investing. However, instead of excluding entire sectors based on objectionable

business practices, for this exercise we only excluded „CCC‟-rated companies from the

investment universe, representing approximately 7 percent of MSCI World stocks on a market

capitalization basis. This approach takes advantage of IVA ratings that reflect ESG risks of

companies relative to their industry peers; therefore, the excluded companies are less likely to

concentrate in specific industries. No industries are excluded entirely.

As shown in Table 2, the average active weight of any sector was less than 1 percent above or

below the benchmark sector weight. The largest deviation was for Health Care, with an active

sector weight 2.05 percent below the benchmark value.

Table 2: Active Sector Weights of the MSCI World ex „CCC‟ portfolio, February 2007 – December 2012.

We consider this first strategy as a reference point for the ESG exclusion approach (referred to as

Strategy 1). Strategies 2-4 take the process one step further, and apply an optimized tilt towards

higher rated stocks and away from lower rated stocks inside the reduced investment universe.

We also aim to maintain country, industry and style exposures of the portfolio close to the

benchmark‟s exposures, and vary the investor‟s aversion to tracking error only (see the Appendix

for execution details). Summary statistics for the full period are gathered in Table 3.

Sector

Average

Active Weight

(%)

Minimum

Active

Weight (%)

Maximum

Active Weight

(%)

Energy -0.64 -1.10 -0.08

Materials 0.03 -0.30 0.32

Industrials 0.25 0.04 0.49

Consumer

Discretionary0.66 -0.05 1.29

Consumer Staples 0.17 -0.96 0.61

Health Care -0.86 -2.05 -0.37

Financials -0.99 -1.91 -0.35

Information

Technology-0.15 -0.62 0.35

Telecommunication

Services0.94 0.73 1.16

Utilities 0.58 0.05 1.05



7 of 23

Table 3: Summary Statistics of ESG Exclusion Strategies, February 2007 – December 2012.

Excluding „CCC‟-rated companies (Strategy 1) without any further adjustment led to a low

tracking error of 0.66 percent and a small negative active return of -0.16 percent. This was due

mainly to country biases that contributed -0.39 percent to the annual active return (see Figure 1

below). However, the exclusion itself contributed positively, as evidenced by the 0.13 percent

asset specific contribution, meaning that eliminating „CCC‟-rated companies raised portfolio

performance once the other residual factors were factored out. The average improvement in the

portfolio ESG score was modest, however, rising only 0.6 points, from 5.4 to 6.0 (or 12 percent),

on a scale of 0 to 10.

Strategies 2-4 yielded better improvements in average ESG portfolio scores, ranging from 6.6 to

7.6. This put the realized ESG tilt well into the „A‟-rated category. At the same time,

benchmark tracking errors could be reduced. However, Strategy 4, which provided the highest

ESG portfolio tilt, increased the tracking error and significantly reduced active returns, asset

specific contributions and information ratios.

These results suggest that the lowest-rated ESG stocks tended to underperform during the study

period relative to the benchmark, since eliminating them from the portfolio raised active returns.

This supports our hypothesis that worst-in-class ESG-rated stocks could potentially be

eliminated without significantly changing risk and performance characteristics relative to the

MSCI World Index. However, this does not resolve whether employing other ESG tilt strategies

can provide additional performance benefits. We address this question next.

ESG

Exclusion

1

ESG

Exclusion

2

ESG

Exclusion

3

ESG

Exclusion

4

Active return (annual, %) -0.16 0.02 -0.01 -0.20

Common factor contribution (annual, %) -0.30 0.04 0.04 0.01

Asset specific contribution (annual, %) 0.13 -0.02 -0.05 -0.21

Tracking error (ex-post, annual %) 0.66 0.42 0.52 1.02

Information ratio -0.24 0.05 -0.02 -0.19

Average improvement in ESG score 0.64 1.19 1.42 2.20

Average relative improvement in ESG score (%) 12 22 26 41

Turnover (annual, %) 4 20 20 20



8 of 23

Figure 1: Return Decomposition of ESG Exclusion Strategy 1, February 2007 – December 2012.

-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5C

um

ula

tive

Act

ive

Re

turn

Co

ntr

ibu

tio

n

Styles Industries Countries Asset Specific Total Active



9 of 23

Details of ESG Tilt Approach

We now analyze the effect of an optimized tilt towards higher-rated stocks in the entire MSCI

World universe without excluding low-rated stocks (see the Appendix for optimization details).

Summary statistics of these strategies, with varying tracking error levels, are presented in Table 4

and compared to the ESG Exclusion Strategy 2, showing the best results.

Table 4: Comparison of Statistics of the ESG Tilt and ESG Exclusion Strategies, February 2007 – December 2012.

We found that allowing for larger tracking error in an effort to maximize ESG portfolio scores

did not lead to superior returns over the time period (see especially Strategy 3). Also, strategies

with much more limited active risk produced a minimal negative active return, while improving

the 5.4 ESG score of the MSCI World Index by at least 1.1 points. Even so, these low-risk ESG

tilt strategies did not match the performance of the lowest risk exclusion strategy, as defined in

the previous section as ESG Exclusion 2 in Table 4 above.

ESG Tilt

1

ESG Tilt

2

ESG Tilt

3

ESG

Exclusion

2

Active return (annual, %) -0.01 -0.03 -0.19 0.02

Common factor contribution (annual, %) 0.02 0.03 0.01 0.04

Asset specific contribution (annual, %) -0.03 -0.06 -0.20 -0.02


Information ratio -0.03 -0.06 -0.19 0.05






10 of 23

Figure 2: Return Decomposition of ESG Tilt Strategy 1, February 2007 – December 2012.13

Taking ESG tilt Strategy 1 as an example and looking at its performance over the entire time

period, we observe a cyclical behavior in our sample (see Figure 2). Most notably, there was a

two-year period of outperformance of the strategy between May 2008 and April 2010, relative to

the MSCI World Index. This corresponded with a “flight to quality” in stock selection at the

beginning of the period that favored companies with strong long-term reputations and high ESG

ratings.14

The outperformance of high ESG-rated assets continued as global markets recovered

and assumed more of a “risk-on” mentality. The outperformance ended as market signals turned

neutral, completing a cycle of risk-off, risk-on and neutral phases.15

The European debt crisis and U.S. credit downgrade in the summer of 2011 once again reset

market expectations. A new cycle began with high ESG-rated assets in Europe particularly

subject to downward pricing pressure in the first risk-off phase. As the market bottomed and

entered a new risk-on phase, high ESG-rated assets outperformed in the first half of 2012,

13 Vertical lines denote May 2008 and April 2010.

14 For historical comparisons, see Robert Eccles, Ioannis Ioannu and George Serafeim, “The Impact of Corporate Culture of Sustainability on Corporate Behavior and

Performance,” Harvard Business School Working Paper 12-035, Nov. 4, 2011; and Sandra A. Waddock and Samuel B. Graves. "Finding the Link Between

Stakeholder Relations and Quality of Management," Journal of Investing,Winter 1997.

15 Risk on, risk off has recently become a common way to describe market behaviour, but the delimitation of the different periods is not well-defined. See Costabile,

Marossy “Risk On, Risk Off in a Multifactor World” for a more quantitative approach to the problem. They describe and estimate a global regime switching model

with three market states: risk-on, risk-off, and neutral. In this paper, we use the risk-on/risk-off/neutral periods defined therein.

-1.2

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6C

um

ula

tive

Re

turn

Co

ntr

ibu

tio

n (

%)




11 of 23

followed by choppier performance in the second half of the year. In other periods of this time

series, the asset specific contribution of high ESG-rated assets was mostly negative; hence, we

realized a slight negative contribution from an ESG tilt over the entire time series.

Analysis of the asset specific component

The asset specific component – reflecting the effect of ESG tilt net of any other systematic risk

source – can be further decomposed into contributions from both overweighted and

underweighted stocks. For the purpose of this paper, we selected the decomposition of Strategy

1 shown in Figure 3 below, but the results are qualitatively similar for the other strategies.

Figure 3: Decomposition of the Asset Specific Component of ESG Tilt Strategy 1, February 2007 – December 2012.

Figure 3 provides further insight to this cyclical behavior of the asset specific contribution into

contributions, and allows us to disentangle the role of higher- and lower-ESG rated stocks. The

intended result of employing this ESG tilt strategy is a “symmetric” positive contribution to

asset specific returns, whereby lower-ESG rated assets tend to underperform (and thus have a

positive contribution due their underweighting), while high-ESG rated assets tend to outperform

and are overweighted to boost returns. However, our results over the observed time period show

an “asymmetry,” where underweighted assets tended to underperform as intended with this

strategy, while overweighted assets did not outperform consistently and showed a mostly

negative and more uneven, cyclical contribution. In line with our previous observation, the

positive contribution was concentrated mainly in the “risk-off” period, but not exclusively.

-1.5

-1

-0.5

0

0.5

1

1.5

Feb-07 Aug-07 Feb-08 Aug-08 Feb-09 Aug-09 Feb-10 Aug-10 Feb-11 Aug-11 Feb-12 Aug-12

Cu

mu

lati

ve A

ctiv

e R

etu

rn C

on

trib

uti

on

(%

)

Overweighted Assets Underweighted Assets Total Asset Specific



12 of 23

In general, this finding suggests that the market‟s overall reaction to companies with above

average and below average ESG practices and ratings was not treated equally during this period.

The market placed a more consistent pricing penalty on companies with lower ESG scores than it

rewarded companies with higher ESG scores.

We offer some possible explanations for this asymmetric result. Due to investors‟ aversion to

loss, downside ESG risks may be incorporated more quickly into stock prices than long-term

upside ESG opportunities. Lower ESG ratings take into account event-driven “bad” news like

accidents, product recalls, supply chain interruptions, governance controversies and regulatory

decisions that pose immediate threats to financial performance.

Conversely, higher ESG ratings reflect and anticipate “good” news that is more secular and long-

term in nature and, hence, more subject to market discounting. Examples include:

implementation of sustainable resource practices, progress toward low-carbon energy use,

increases in employee training and motivation, consumer-led product innovations and

shareholder-led governance reforms. It follows that company efforts to enhance their ESG

performance may take longer to appreciate in market value than their ESG shortcomings, which

are more instantly recognized and priced into the market.

To further investigate this phenomenon we look at the sector and region breakdown of the asset

specific component. The results are presented in Table 5 and Table 6.

Table 5: Decomposition of the Cumulative Asset Specific Contribution by Sectors, ESG Tilt Strategy 1, February 2007 –

December 2012.

In the Telecommunication Services and Utilities sectors, the stock specific contributions of

overweighted and underweighted stocks were symmetric and all positive (highlighted in brown

in the table above) during the study period. This was the intended result of implementing an

ESG strategy in portfolio construction.

SectorOver-

weighted

Under-

weighted

Total Asset

Specific

Energy -0.02 0.05 0.03

Materials -0.42 0.00 -0.42

Industrials -0.22 -0.18 -0.40

Consumer Discretionary 0.48 -0.40 0.08

Consumer Staples -0.06 0.13 0.07

Health Care -0.23 0.47 0.23

Financials -0.62 0.78 0.16

Information Technology -0.30 -0.11 -0.41

Telecommunication Services 0.05 0.04 0.09

Utilities 0.14 0.21 0.35



13 of 23

In the Industrials and Information Technology sectors, the asset specific contributions of over-

and underweighted stocks were symmetric but negative – opposite of the intended result

(highlighted in orange in Table 5).

In the remaining sectors, the asset specific contributions from over- and underweighted stocks

were asymmetrical during the observed time period, from February 2007 through December

2012, producing mixed results for asset specific returns. In the Energy, Materials, Consumer

Staples, Health Care and Financials sectors, the underweighting of low-rated companies led to a

positive contribution to asset specific returns, but overweighting higher rated ones also led to a

negative contribution. The opposite phenomenon was observed only in the Consumer

Discretionary sector, also producing mixed results. On balance, these sector results provide

further evidence that low ESG-rated companies have seen a greater market penalty with respect

to risk-adjusted performance over this time period than high ESG-rated companies have seen

rewards for their positive behavior and practices.

Further analysis may be conducted to explain such symmetric or asymmetric results, but it is

beyond the scope of this paper to do so. As a general observation, ESG ratings may be seen as a

proxy for forces external to the market that are subject to near-term regulatory control. Over the

February 2007 – December 2012 time period, we observe that high-rated ESG assets tended to

outperform at times of heightened anticipation of regulatory risks and then underperformed as

these regulatory risks subsided.

As a further exploration of this hypothesis, two industry sectors are briefly examined here. For

Utilities, the symmetric positive result came at a time of persistent regulatory expectations to

constrain carbon emissions from power production. During the study period, there was an initial

rise in market expectations with the introduction of U.S. climate change legislation and progress

toward an international control agreement, and then a similar fall in these expectations as these

policy initiatives stalled.

Conversely, the Financial sector, which had the greatest asymmetric mixed result over the study

period, faced a considerable degree of government intervention during the recent financial crisis.

During this period, low ESG-rated financial services firms benefited portfolio performance

because they were underweighted, while high ESG-rated firms harmed portfolio performance

because they were overweighted. Government efforts taken during this time rescued the most

highly distressed firms, and legislative reforms were enacted to rein in high-risk banking

practices. However, these reforms did not fundamentally alter the structure and business lines of

financial services firms. Accordingly, those firms with more traditional and conservative lending

practices – which generally had higher ESG ratings over the period – did not realize a distinct

competitive advantage as a result of these subsequent government interventions.



14 of 23

Table 6: Decomposition of the Cumulative Asset Specific Contribution by Regions, ESG Tilt Strategy 1, February 2007 –

December 2012.

In terms of regions, we again found diverse results, as shown in Table 6. In the EMEA region,

contributions were symmetrical and positive over the study period, as intended. But they were

asymmetrical for the Americas and Pacific regions, producing mixed results. A more granular

analysis reveals that asset specific contributions were symmetrical and positive – among others –

in Germany, France, Italy, Denmark, Austria, Portugal and New Zealand. Asymmetrical results

were found in the United States, United Kingdom and Switzerland.

We observe that the positive symmetric results came mainly in the EMEA region, which

arguably has the most extensive regulation of ESG factors and, hence, the most internal market

pricing signals. Conversely, the Pacific region arguably has the least extensive regulation of

ESG factors. In this region, high ESG-rated companies that were overweighted contributed to

portfolio outperformance, while low ESG-rated companies that were underweighted contributed

to underperformance. The Americas region was in-between these two poles of comparison.

This region had the worst results in terms of total asset selection, with the overweighting of high

ESG-rated companies contributing significantly to portfolio underperformance.

Details of ESG Momentum Approach

The high latency that may characterize the incorporation of positive ESG ratings into stock

prices, as explored earlier, is a motivation for a third and final ESG investing strategy addressed

in this paper. In the ESG momentum strategy, we replace the ESG ratings with the change in

ratings over the past 12 months as our input variable for portfolio optimization. By doing this,

we aim to overweight, relative to the MSCI World Index, companies that increased their ESG

ratings during the recent past and underweight those with decreased ESG ratings (see the

Appendix for technical details). Due to the 12-month cycle over which ratings changes are

calculated, our analysis for this strategy starts in February 2008.

RegionOver-

weighted

Under-

weighted

Total Asset

Selection

Americas -1.81 0.95 -0.86

EMEA 0.09 0.54 0.63

Pacific 0.48 -0.50 -0.02



15 of 23

Table 7: Comparison of ESG Momentum Strategies with ESG Tilt Strategy 1, February 2008 – December 2012.16

Results corresponding to a range of tracking error levels are summarized in Table 7. We find that

during the observed period low active-risk ESG momentum strategies performed better on a risk

adjusted basis than the lowest-risk ESG tilt strategies. Another observation is that the

contribution of the asset specific component was also higher than in the simple ESG tilt

strategies (see Figure 4 for the time series return attribution). This may reflect that markets are

more likely to react to news of companies showing improvement in their ESG scores than to

those who had already attained top ratings in their sectors.

16 To be consistent in comparing results, the figures presented in this table are for February 2008 – December 2012 only, corresponding to the time period available for

the ESG momentum strategy. They correspond with figures shown in Table 1.

ESG

Momentum

1

ESG

Momentum

2

ESG

Momentum

3

ESG Tilt

1

Active return (annual, %) 0.35 0.40 0.29 0.05

Common factor contribution (annual, %) 0.08 0.10 0.09 0.03

Asset specific contribution (annual, %) 0.27 0.30 0.21 0.01


Information ratio 0.97 0.92 0.35 0.10






16 of 23

Figure 4: Return Decomposition of ESG Momentum Strategy 1, February 2008 – December 2012.

On the other hand, this relatively better performance came at the price of a smaller improvement

in the portfolio‟s overall ESG rating. For Strategy 1, the gain with respect to the MSCI World

Index was only 0.46 points for the ESG momentum strategy, compared to 1.18 points for the

simple ESG tilt strategy.

Comparing the asset specific contributions of the ESG tilt and ESG momentum strategies on

Figure 3 and Error! Reference source not found., we find some similarities and differences. On the

one hand, we again observe an asymmetry between contributions from underweighted assets that

received ratings downgrades over the time period and contributions from overweighted assets

that were upgraded. As Error! Reference source not found. shows, downgraded assets that were

underweighted tended to produce a consistent positive contribution to portfolio performance,

whereas upgraded stocks that were overweighted had a small, more cyclical and overall negative

contribution over the time period.

On the other hand, the asymmetry between over- and underweighted assets‟ contributions was

much smaller in the ESG momentum strategy than in the simple ESG tilt strategy, and more in

line with the intended results to improve portfolio performance. Underweighted assets in the

ESG momentum strategy contributed 1.6 percent in positive cumulative active returns over the

study period, compared to 0.6 percent for the ESG tilt strategy shown in Figure 3. While

overweighted assets in the ESG momentum strategy still contributed to portfolio

underperformance, the effect was smaller than in the ESG tilt strategy. As shown in Figure 5,

the negative cumulative active return was 0.4 percent over the study period, compared to a

negative return of 1.2 percent for the ESG tilt strategy.

-1

-0.5

0

0.5

1

1.5

2

Feb-08 Aug-08 Feb-09 Aug-09 Feb-10 Aug-10 Feb-11 Aug-11 Feb-12 Aug-12

Cu

mu

lati

ve A

ctiv

e R

etu

rn C

on

trib

uti

on

(%

)




17 of 23

This decomposition of ESG momentum results reinforces our earlier finding that companies

associated with negative ESG factors – as reflected in low ratings or ratings downgrades –

consistently underperformed the market benchmark over the study period, while those with high

ratings or ratings upgrades had more uneven market performance. This lends further support to

the hypothesis that downside ESG risks, as reflected in low ratings or ratings downgrades,

receive a more instantaneous market reaction than long-term upside ESG opportunities, as

reflected in high ratings or ratings upgrades. Moreover, over the study period the market

appeared to be more sensitive to ratings changes than to the effect of absolute ESG scores.

Ratings downgrades appeared to compound market perception of potential adverse material

impacts among the affected firms, while ratings improvements drew added investor recognition

to positive ESG factors, with less market discounting among companies showing such ESG

momentum.

Figure 5: Decomposition of the Asset Specific Component of ESG Momentum Strategy 1, February 2008 – December 2012.

Finally, while the ESG momentum strategy limited the overall ESG tilt of the portfolio, the

trade-off came in exchange for a higher asset-specific contribution and information ratio over the

study period. This strategy rewarded companies demonstrating progress in managing ESG risks

and realizing ESG opportunities, even if they had not yet achieved top-level scores.

Accordingly, for this study period, institutional investors tilting toward companies making

progress in addressing ESG material risks would have seen a lower increase in ESG rating for

their portfolio overall, relative to the simple tilt strategies. On the other hand, those investors

would have achieved higher active returns and lower tracking error.

-1

-0.5

0

0.5

1

1.5

2

Feb-08 Aug-08 Feb-09 Aug-09 Feb-10 Aug-10 Feb-11 Aug-11 Feb-12 Aug-12

Cu

mu

lati

ve A

ctiv

e R

etu

rn C

on

trib

uti

on

(%

)

Overweighted Assets Underweighted Assets Total Asset Specific



18 of 23

Conclusion

We examined possible implementations of ESG-tilt strategies based on MSCI‟s ESG Research

Intangible Value Assessment (IVA) scores. Our observations are limited to the historical period

that we studied (February 2007 through December 2012), and do not necessarily reflect the

results or performance of other periods or future periods. As additional data becomes available

to extend the time series, investors may wish to conduct further research on these strategies to

compare the results presented here.

Using the MSCI World Index as a benchmark, and the Barra Global Equity Model (GEM3) as a

risk model, we ran a series of backtests for the period February 2007 – December 2012 with the

same constraints on systematic portfolio exposures, but with increasing tracking error levels. We

found that during the observed period that it was possible to build portfolios with equivalent risk,

performance, industry, country or style characteristics as conventional benchmarks, while also

raising ESG portfolio scores. We introduced three families of ESG strategies, all based on the

IVA dataset.

The first ESG strategy – based on the exclusion of low-rated („CCC‟) companies before the

application of the ESG tilt – led to a small improvement in ESG scores, and positive

performance results once common factor attributions were stripped out. Applying a further ESG

ratings tilt in addition to the „CCC‟ exclusion reduced the residual contributions and produced a

more positive active return for this lowest risk strategy, raising the ESG portfolio rating from

„BBB‟ to „A.‟

When we applied optimized ESG tilt strategies to the entire MSCI World universe, we found that

the active return was generally small, and slightly negative. At the low-risk end of the ESG-

tilted spectrum, a 44 basis point tracking error strategy attained a one-notch increase in overall

ESG portfolio rating to „A‟ with returns that matched the market. Decomposing active

performance into systematic and stock-specific components, we observed that the cumulative

performance in the ESG tilt strategy was mainly due to the cyclical nature of the stock-specific

component.

Between May 2008 and April 2010, and more recently in the first half of 2012, the active

performance of higher-rated ESG companies turned significantly positive. We observed that

during this period high-rated ESG assets started to outperform when the market was in a

defensive, “flight to quality” mode and extended their gains when more of a “risk-on” market

mentality took hold. The outperformance ended when market signals turned neutral, completing

a regime switching cycle. We hypothesize that downside ESG risks may be incorporated into

stock prices more quickly than long-term ESG opportunities. This is because downside risks

tend to be more event-driven and instantly recognized than long-term ESG opportunities that are

evolving, discounted, and correlated with changing market and geo-political conditions.

Based on this observation, we implemented a third series of strategies based on the change in

company ESG scores. Our analysis revealed that during the observed period the ESG

momentum strategy tended to deliver the best risk-adjusted performance of the strategies we

analyzed. We found that when we decomposed the ESG momentum strategy, ESG ratings

downgrades had a consistent negative effect on asset specific performance over the time series,

while ESG ratings upgrades had a more positive effect during “risk off – risk on” periods in the



19 of 23

market regime switching cycle. While the ESG momentum strategy came with a lower ESG tilt,

it still supported positive movement on ESG scores at the portfolio level.

Our study‟s main conclusion is that asset managers would have been able to employ ESG factors

to attain higher ESG portfolio scores with low active risk, and still achieve moderate benchmark

outperformance over the time period from February 2008 through December 2012 in which the

three strategies are compared.



20 of 23

Appendix: Backtest Parameters

The benchmark portfolio was the MSCI World Index. The company universe was also the MSCI

World Index, except for the ESG exclusion strategies, where the universe was the MSCI World

portfolio without the „CCC‟-rated stocks.

The long-horizon Barra Global Equity Model (GEM3L) was used as a risk model.

Stock-level alpha was calculated as follows. The IVA scores (for ESG tilt and exclusion

strategies), or the changes in IVA scores (for ESG momentum strategies) were transformed into

alpha following the heuristic method described by Richard C. Grinold, in “Alpha is Volatility

times IC times Score.”17

Namely, when defining the alpha, the stock‟s standardized score was

adjusted by the specific volatility forecast of the stock.

We applied the following constraints on country, industry and style characteristics:

Maximum +/-1% region weight deviation relative to the benchmark. The three regions are North

America, EMEA and Asia Pacific

Maximum +/-1% GICS® sector weight deviation relative to the benchmark

Maximum +/-0.25 GEM3 style factor active exposure relative to the benchmark

We set a lower and upper bound on the weight of benchmark stocks in the optimized portfolio at

0.001 percent and 5 percent (except for the ESG exclusion strategies).

The portfolios were rebalanced quarterly starting in February 2007, with a 5 percent turnover

limit.

The risk aversion parameter ran through the values of 1, 4 and 6. The ratio of common factor to

asset specific risk aversion was set to 1.

17 Richard C. Grinold, “Alpha is Volatility times IC times Score,” The Journal of Portfolio Management, Summer 1994



21 of 23

Bibliography

Bauer, Rob; and Hann, Daniel. “Corporate Environmental Management and Credit Risk,”

Maastricht University, European Centre for Corporate Engagement, June 30, 2010.

Briand, Remy; Chia, Chin-Ping; and Urwin, Roger. “Integrating ESG in the Investment Process:

From Aspiration to Effective Implementation,” MSCI Research Insight, August 2011.

Chava, Subdeer. “Environmental Externalities and the Cost of Capital,” Georgia Institute of

Technology, 2011

Costabile, Audrey; and Marossy, Zita. “Risk On, Risk Off in a Multifactor World”, MSCI Stress

Testing Market Report, August 2012

Credit Suisse Strategy Paper. “ESG Investing Themes: Are „good‟ stocks good investments?”

Credit Suisse, Oct. 11, 2012.

Durwall, Jeroen; Koedijk, Kees; and Ter Horst, Jenke. “A tale of values-driven vs. profit-seeking

social investors,” Journal of Banking and Finance, January 2011.

Eccles, Robert; Ioannu, Ioannis; and Serafeim, George. “The Impact of Corporate Culture of

Sustainability on Corporate Behavior and Performance,” Harvard Business School Working

Paper 12-035, Nov. 4, 2011.

El Ghoul, Sadok; Guedhami, Omrane; Kwok, Chuck C. Y.; and Mishra, Dev R. “Does corporate

social responsibility affect the cost of capital?” Journal of Banking & Finance, September 2011.

Fulton, Mark; Kahn, Bruce; and Sharples, Camilla. “Sustainable Investing: Establishing Long-

term Value and Performance,” Deutsche Bank Climate Change Advisors, June 2012.

Grinold, Richard C. “Alpha is Volatility times IC times Score,” The Journal of Portfolio

Management, Summer 1994.

Hawley, James P.; and Williams, Andrew T. “The Emergence of Fiduciary Capitalism,”

Corporate Governance, Vol. 5, No. 4 (1997). Available at SSRN:

http://ssrn.com/abstract=11430

Kurtz, Lloyd; and DiBartolomeo, Dan. “The Long-term Performance of a Social Investment

Universe,” The Journal of Investing, Fall 2011.

Mercer Investment Consulting. “Shedding Light on Responsible Investment: Approaches,

Returns, Impacts,” November 2009.

RCM Sustainability White Paper. “Sustainability: Opportunity or Opportunity Cost?” Allianz

Global Investors, July 2011.

Waddock, Sandra A.; and Graves, Samuel B. "Finding the Link Between Stakeholder Relations

and Quality of Management," Journal of Investing, Winter 1997.



22 of 23

Biographies

Zoltán Nagy is a member of MSCI‟s Applied Research team based in the Budapest office; he

focuses on portfolio management and risk-related research for asset owners and investment

managers. Dr Nagy joined MSCI in 2008, at first working on the development of new index

methodologies and on other index-related research.

Prior to entering finance, Dr. Nagy was a post-doctoral researcher at the University of Algarve,

Faro, Portugal. His area of research was Quantum Integrable Sytems. He has several

publications in this field.

Dr. Nagy holds a PhD in Theoretical Physics from the University of Cergy-Pontoise, France, and

an engineering degree from the Ecole Polytechnique, France. He is also a CFA® charterholder.

Douglas Cogan is a Vice President of MSCI ESG Research and a member of MSCI‟s Asset

Owner Consultant team, serving clients throughout North America. He started his career at the

Investor Responsibility Research Center in 1982 and served as Deputy Director of IRRC‟s Social

Issues Service from 1996 through 2005. He then assumed ESG leadership positions at

Institutional Shareholder Services and RiskMetrics Group, which MSCI acquired in 2009.

Mr. Cogan is the author of several books and many reports on energy and environmental topics.

His 1992 book, The Greenhouse Gambit, was one of the first to address the business and

investment implications of climate change. He has also written reports on divestment actions

and fiduciary responsibility, and connections between corporate governance and climate change

for the Investor Network on Climate Risk.

Mr. Cogan holds a Bachelor of Arts degree in Political Economics from Williams College. He is

associated with MSCI‟s Boston office and is based in Plainfield, New Hampshire.

Dan Sinnreich is an Executive Director at MSCI in the Risk Management Analytics product

management team. He joined Barra in 2003, initially managing Barra‟s fixed income risk

analytics products. He later served as the lead technical product manager for several multi-asset

class product releases. Mr. Sinnreich is currently focused on expanding the Risk Management

Analytics business in the Americas.

Prior to his current position at MSCI, Mr. Sinnreich was a product manager for an electronic

mortgage software startup in Silicon Valley, and designed enterprise-wide management reporting

applications at Oracle Corporation. Mr. Sinnreich started his career with KPMG‟s management

consulting practice and received his Bachelor of Arts degree in Management Information

Systems from Auburn University. He is based in Berkeley, California.



23 of 23

Client Service Information is Available 24 Hours a Day

[email protected]

Notice and Disclaimer This document and all of the information contained in it, including without limitation all text, data, graphs, charts (collectively, the “Information”) is the property of MSCl Inc. or its

subsidiaries (collectively, “MSCI”), or MSCI‟s licensors, direct or indirect suppliers or any third party involved in making or compiling any Information (collectively, with MSCI, the

“Information Providers”) and is provided for informational purposes only. The Information may not be reproduced or redisseminated in whole or in part without prior written permission

from MSCI.

The Information may not be used to create derivative works or to verify or correct other data or information. For example (but without limitation), the Information many not be used to

create indices, databases, risk models, analytics, software, or in connection with the issuing, offering, sponsoring, managing or marketing of any securities, portfolios, financial products or

other investment vehicles utilizing or based on, linked to, tracking or otherwise derived from the Information or any other MSCI data, information, products or services.

The user of the Information assumes the entire risk of any use it may make or permit to be made of the Information. NONE OF THE INFORMATION PROVIDERS MAKES ANY

EXPRESS OR IMPLIED WARRANTIES OR REPRESENTATIONS WITH RESPECT TO THE INFORMATION (OR THE RESULTS TO BE OBTAINED BY THE USE THEREOF),

AND TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE LAW, EACH INFORMATION PROVIDER EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES

(INCLUDING, WITHOUT LIMITATION, ANY IMPLIED WARRANTIES OF ORIGINALITY, ACCURACY, TIMELINESS, NON-INFRINGEMENT, COMPLETENESS,

MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE) WITH RESPECT TO ANY OF THE INFORMATION.

Without limiting any of the foregoing and to the maximum extent permitted by applicable law, in no event shall any Information Provider have any liability regarding any of the Information

for any direct, indirect, special, punitive, consequential (including lost profits) or any other damages even if notified of the possibility of such damages. The foregoing shall not exclude or

limit any liability that may not by applicable law be excluded or limited, including without limitation (as applicable), any liability for death or personal injury to the extent that such injury

results from the negligence or wilful default of itself, its servants, agents or sub-contractors.

Information containing any historical information, data or analysis should not be taken as an indication or guarantee of any future performance, analysis, forecast or prediction. Past

performance does not guarantee future results.

None of the Information constitutes an offer to sell (or a solicitation of an offer to buy), any security, financial product or other investment vehicle or any trading strategy.

MSCI‟s indirect wholly-owned subsidiary Institutional Shareholder Services, Inc. (“ISS”) is a Registered Investment Adviser under the Investment Advisers Act of 1940. Except with

respect to any applicable products or services from ISS (including applicable products or services from MSCI ESG Research Information, which are provided by ISS), none of MSCI‟s

products or services recommends, endorses, approves or otherwise expresses any opinion regarding any issuer, securities, financial products or instruments or trading strategies and none of

MSCI‟s products or services is intended to constitute investment advice or a recommendation to make (or refrain from making) any kind of investment decision and may not be relied on as

such.

The MSCI ESG Indices use ratings and other data, analysis and information from MSCI ESG Research. MSCI ESG Research is produced ISS or its subsidiaries. Issuers mentioned or

included in any MSCI ESG Research materials may be a client of MSCI, ISS, or another MSCI subsidiary, or the parent of, or affiliated with, a client of MSCI, ISS, or another MSCI

subsidiary, including ISS Corporate Services, Inc., which provides tools and services to issuers. MSCI ESG Research materials, including materials utilized in any MSCI ESG Indices or

other products, have not been submitted to, nor received approval from, the United States Securities and Exchange Commission or any other regulatory body.

Any use of or access to products, services or information of MSCI requires a license from MSCI. MSCI, Barra, RiskMetrics, ISS, CFRA, FEA, and other MSCI brands and product names

are the trademarks, service marks, or registered trademarks of MSCI or its subsidiaries in the United States and other jurisdictions. The Global Industry Classification Standard (GICS) was

developed by and is the exclusive property of MSCI and Standard & Poor‟s. “Global Industry Classification Standard (GICS)” is a service mark of MSCI and Standard & Poor‟s.

About MSCI MSCI Inc. is a leading provider of investment decision support tools to investors globally, including asset managers, banks, hedge funds and pension funds. MSCI

products and services include indices, portfolio risk and performance analytics, and governance tools.

The company‟s flagship product offerings are: the MSCI indices with approximately USD 7 trillion estimated to be benchmarked to them on a worldwide basis1; Barra multi-asset class factor models, portfolio risk and performance analytics; RiskMetrics multi-asset class market and credit risk analytics; MSCI ESG

(environmental, social and governance) Research screening, analysis and ratings; ISS governance research and outsourced proxy voting and reporting services; FEA

valuation models and risk management software for the energy and commodities markets; and CFRA forensic accounting risk research, legal/regulatory risk

assessment, and due‐diligence. MSCI is headquartered in New York, with research and commercial offices around the world.

1As of June 30, 2011, based on eVestment, Lipper and Bloomberg data.

Americas Europe, Middle East & Africa Asia Pacific

Americas Atlanta

Boston

Chicago Montreal

Monterrey

New York San Francisco

Sao Paulo

Stamford Toronto

1.888.588.4567 (toll free) + 1.404.551.3212

+ 1.617.532.0920

+ 1.312.675.0545 + 1.514.847.7506

+ 52.81.1253.4020

+ 1.212.804.3901 + 1.415.836.8800

+ 55.11.3706.1360

+1.203.325.5630 + 1.416.628.1007

Cape Town Frankfurt

Geneva

London Milan

Paris

+ 27.21.673.0100 + 49.69.133.859.00

+ 41.22.817.9777

+ 44.20.7618.2222 + 39.02.5849.0415

0800.91.59.17 (toll free)

China North China South

Hong Kong

Seoul Singapore

Sydney

Tokyo

10800.852.1032 (toll free) 10800.152.1032 (toll free)

+ 852.2844.9333

798.8521.3392 (toll free)

800.852.3749 (toll free)

+ 61.2.9033.9333

+ 81.3.5226.8222

mailto:[email protected]

Research Insight w Cover

Documents

Transcript of Research Insight w Cover