Massimiliano Tondi, Senior Portfolio Manager at KBI Global Investors, highlights ESG’s unique diversification characteristics.
With ESG issues becoming increasingly important to asset owners, KBI Global Investors decided to use its proprietary factor framework to answer some key questions.
Is ESG a new alpha factor or is it just a new name for an already existing risk premium? Does an ESG exposure provide diversification benefit when compared to the more traditional factors?
To do this, we compared ESG performance and its correlation to existing styles. We also explored the possibility to replicate the ESG exposure synthetically.
The increasing importance of ESG
Investors have significantly increased their ESG exposure in recent years and this has changed the shape of stock indices, with companies willing to dedicate resources to improve their overall ESG score. Likewise, we can see an increased ESG coverage by ESG info provider[1].
From 2020, the constituents of the MSCI World have shown a marked tendency to improve their ESG ratings. The number of companies with an above average ESG score (AAA, AA, A bucket) has increased by 21, while the below average score (BB, B, CCC bucket) has decreased for 77 companies.
For the MSCI Emerging Market Index, the trend is the same. In total, 25 companies have been added to the above average ESG bucket, while the below average category has 39 fewer stocks.
The European sustainable funds universe comprised 3,196 funds in 2020, including an increase of 505 new sustainable funds and the repurposing and rebranding of at least 253 conventional funds.
Given the constant upswing in AUM in sustainable investing and the increasing commitment of investors to ESG, it’s worth asking if the factor has become more expensive.
To answer the question, we calculated the ratio between the median forward P/E of the top ESG quintile (industry neutral) divided by the median forward P/E of the bottom ESG quintile (industry neutral) for both the two major developed regions (North America and Europe) and the most important emerging region (Asia). The ESG baskets are created using the stocks that are part of the specific MSCI regional index.
Since 2019, the relative expensiveness of the factor has fallen for North America and Europe, and now sits at a more normal level. In Asia, the factor has remained consistently cheap relative to its long-term average.
While the increased focus on ESG is more a developed markets phenomenon so far, the factor is not historically expensive even there. While investors seem increasingly willing to pay a premium for holding ESG committed companies (leaders), and/or they require a discount for holding less committed ESG stocks (laggards), we’ve by no means reached the extremes of this trend.
How does ESG compare to other style factors?
To evaluate ESG in the factor world, we assess it against a subset of style factors. The style factors are a blend of multiple signals[2] to reduce the noise of a specific metrics.
- Value – average performance of dividend yield, B/P and EBITDA/EV.
- Quality – average performance ROE, ROA, debt/equity.
- Momentum – 1Y performance momentum factor.
- Growth – average of the 1Y EPS, 4Y EPS and sales growth.
- Size – companies sorted by logarithm of the market cap (smaller is better).
- Low risk – average performance of 6 months volatility and earning volatility.
In North America, we examined performance over different time periods for both the long only and the long-short factors at both an index level and in a dividend yield universe.
The trend in each case is almost identical with ESG being one of the strongest factors during the period 2007-2021. The efficiency increases when we examine a shorter, more recent time (Jan 2019-Feb 2021), clearly due to the increasing awareness of how ESG can drive future returns. The comparison of long-only versus long-short highlights that the outperformance is generated not only from shorting the lower end of the ESG spectrum, but also from holding the companies with higher ESG scores.
We repeated the above analysis for Europe. It is important to highlight that North America and Europe have almost an opposite ESG distribution. The former is more skewed to stocks with lower ESG ratings, while the latter tends to have stocks with higher ratings. This skew might explain why in Europe the long-only part (holding only the higher ESG rated names) is performing better from a risk/return point of view compared with North America.
In Asia, overall ESG effectiveness is muted compared with developed markets. The two most effective styles from 2013 to 2021 have been quality and growth, which is also the case in a dividend yield universe. A possible explanation is the initial ESG focus by investors in developed markets and only more recently to emerging markets.
A correlation analysis between ESG and style factors
Given the strong performance of ESG, we investigate if we can create synthetically the ESG premium, by combining the different style elements.
We first ran a correlation between ESG and the other risk premia, to assess its consistent over time, finding that in North America, the correlation for both the index and dividend universe is changing over time. Through this analysis, we also found that quality was negatively correlated with ESG between 2013 and 2015 for dividend-paying stocks, while the correlation at an index level was slightly positive. Interestingly, we can also see the spike after the 2020 market correction, which saw ESG negatively exposed to value and small size, and positively to momentum, quality and growth.
The same analysis in Europe shows that over the past six years ESG has a negative exposure to value and size and a positive one to quality and growth. More recently, the correlation between ESG and the different styles is between -0.2 and +0.2 for the dividend universe and slightly wider for the index universe. Between 2011 and 2014, we see the ESG return pattern is correlated with Value at the index level while it’s close to zero for a dividend universe.
In Asia, the correlation between ESG and the other styles is more spread out in the dividend space, while it’s more stable in the index universe.
To evaluate the statistical relationship between ESG and risk premium, we ran a regression for the full period using the long-short returns. In the different regions the R2 and the adjusted R2 are extremely low (ranging from 4% to 35%) meaning our styles don’t explain our ESG return pattern in a meaningful way. This implies that it would be difficult to create synthetic ESG using a weighted average of the different risk premia.
Our analysis concludes that the ESG factor carries an information set not already available in other factors, meaning it offers potential for diversification. These results are crucially important in a real-world portfolio management context. This is particularly true during periods of market distress. During this challenging time, ESG exposure has often provided diversification benefit when compared with other factors.
A unique source of alpha
Both institutional and retail investors are paying more attention to ESG investing, as we can clearly see from the increase in related AUM over recent years. That could mean, however, that the factor could become more expensive, which could negatively affect future performance.
By looking at the three main regions in developed and emerging markets (North America, Europe, and Asia) both at index level and in a dividend universe, we can understand the relative cost of ESG. Since 2019, the relative expensiveness of the factor has fallen for North America and Europe and is now at a more normal level. In Asia, meanwhile, the factor has remained consistently cheap relative to its long-term average.
While the increased focus on ESG is more of a phenomenon in developed markets so far, the factor has not proven historically expensive even there.
The factor is quite efficient (information ratio) in North America and Europe, with its effectiveness improving over time, while in Asia it is falling behind quality and growth. This could be due to the early focus of investors to force ESG on developed companies (thanks to stronger pressure from policymakers).
We tried to create a synthetic ESG, but this wasn’t a success. First, the correlation structure between ESG and the traditional styles is not stable but varies over time. Second, the explanatory power of the regression is quite poor, meaning we can only capture a small amount of the ESG variability using traditional risk premium.
Our conclusion is that ESG is a unique source of alpha and exploits an information set that’s not captured by conventional factors and could improve diversification in a way that can’t done with traditional metrics.
[1] For this analysis ESG scores are provided by MSCI ESG Research LLC. MSCI ESG Research products are designed to provide in-depth research, ratings, and analysis of environmental, social and governance-related business practices to companies worldwide.
[2] Each single factor is built monthly on regional, industry neutral basis for both the index and a dividend universe. We use the simple average to calculate the final factor return vs the universe and we take into count the impact of turnover.
Massimiliano Tondi, Senior Portfolio Manager at KBI Global Investors, highlights ESG’s unique diversification characteristics.
With ESG issues becoming increasingly important to asset owners, KBI Global Investors decided to use its proprietary factor framework to answer some key questions.
Is ESG a new alpha factor or is it just a new name for an already existing risk premium? Does an ESG exposure provide diversification benefit when compared to the more traditional factors?
To do this, we compared ESG performance and its correlation to existing styles. We also explored the possibility to replicate the ESG exposure synthetically.
The increasing importance of ESG
Investors have significantly increased their ESG exposure in recent years and this has changed the shape of stock indices, with companies willing to dedicate resources to improve their overall ESG score. Likewise, we can see an increased ESG coverage by ESG info provider[1].
From 2020, the constituents of the MSCI World have shown a marked tendency to improve their ESG ratings. The number of companies with an above average ESG score (AAA, AA, A bucket) has increased by 21, while the below average score (BB, B, CCC bucket) has decreased for 77 companies.
For the MSCI Emerging Market Index, the trend is the same. In total, 25 companies have been added to the above average ESG bucket, while the below average category has 39 fewer stocks.
The European sustainable funds universe comprised 3,196 funds in 2020, including an increase of 505 new sustainable funds and the repurposing and rebranding of at least 253 conventional funds.
Given the constant upswing in AUM in sustainable investing and the increasing commitment of investors to ESG, it’s worth asking if the factor has become more expensive.
To answer the question, we calculated the ratio between the median forward P/E of the top ESG quintile (industry neutral) divided by the median forward P/E of the bottom ESG quintile (industry neutral) for both the two major developed regions (North America and Europe) and the most important emerging region (Asia). The ESG baskets are created using the stocks that are part of the specific MSCI regional index.
Since 2019, the relative expensiveness of the factor has fallen for North America and Europe, and now sits at a more normal level. In Asia, the factor has remained consistently cheap relative to its long-term average.
While the increased focus on ESG is more a developed markets phenomenon so far, the factor is not historically expensive even there. While investors seem increasingly willing to pay a premium for holding ESG committed companies (leaders), and/or they require a discount for holding less committed ESG stocks (laggards), we’ve by no means reached the extremes of this trend.
How does ESG compare to other style factors?
To evaluate ESG in the factor world, we assess it against a subset of style factors. The style factors are a blend of multiple signals[2] to reduce the noise of a specific metrics.
In North America, we examined performance over different time periods for both the long only and the long-short factors at both an index level and in a dividend yield universe.
The trend in each case is almost identical with ESG being one of the strongest factors during the period 2007-2021. The efficiency increases when we examine a shorter, more recent time (Jan 2019-Feb 2021), clearly due to the increasing awareness of how ESG can drive future returns. The comparison of long-only versus long-short highlights that the outperformance is generated not only from shorting the lower end of the ESG spectrum, but also from holding the companies with higher ESG scores.
We repeated the above analysis for Europe. It is important to highlight that North America and Europe have almost an opposite ESG distribution. The former is more skewed to stocks with lower ESG ratings, while the latter tends to have stocks with higher ratings. This skew might explain why in Europe the long-only part (holding only the higher ESG rated names) is performing better from a risk/return point of view compared with North America.
In Asia, overall ESG effectiveness is muted compared with developed markets. The two most effective styles from 2013 to 2021 have been quality and growth, which is also the case in a dividend yield universe. A possible explanation is the initial ESG focus by investors in developed markets and only more recently to emerging markets.
A correlation analysis between ESG and style factors
Given the strong performance of ESG, we investigate if we can create synthetically the ESG premium, by combining the different style elements.
We first ran a correlation between ESG and the other risk premia, to assess its consistent over time, finding that in North America, the correlation for both the index and dividend universe is changing over time. Through this analysis, we also found that quality was negatively correlated with ESG between 2013 and 2015 for dividend-paying stocks, while the correlation at an index level was slightly positive. Interestingly, we can also see the spike after the 2020 market correction, which saw ESG negatively exposed to value and small size, and positively to momentum, quality and growth.
The same analysis in Europe shows that over the past six years ESG has a negative exposure to value and size and a positive one to quality and growth. More recently, the correlation between ESG and the different styles is between -0.2 and +0.2 for the dividend universe and slightly wider for the index universe. Between 2011 and 2014, we see the ESG return pattern is correlated with Value at the index level while it’s close to zero for a dividend universe.
In Asia, the correlation between ESG and the other styles is more spread out in the dividend space, while it’s more stable in the index universe.
To evaluate the statistical relationship between ESG and risk premium, we ran a regression for the full period using the long-short returns. In the different regions the R2 and the adjusted R2 are extremely low (ranging from 4% to 35%) meaning our styles don’t explain our ESG return pattern in a meaningful way. This implies that it would be difficult to create synthetic ESG using a weighted average of the different risk premia.
Our analysis concludes that the ESG factor carries an information set not already available in other factors, meaning it offers potential for diversification. These results are crucially important in a real-world portfolio management context. This is particularly true during periods of market distress. During this challenging time, ESG exposure has often provided diversification benefit when compared with other factors.
A unique source of alpha
Both institutional and retail investors are paying more attention to ESG investing, as we can clearly see from the increase in related AUM over recent years. That could mean, however, that the factor could become more expensive, which could negatively affect future performance.
By looking at the three main regions in developed and emerging markets (North America, Europe, and Asia) both at index level and in a dividend universe, we can understand the relative cost of ESG. Since 2019, the relative expensiveness of the factor has fallen for North America and Europe and is now at a more normal level. In Asia, meanwhile, the factor has remained consistently cheap relative to its long-term average.
While the increased focus on ESG is more of a phenomenon in developed markets so far, the factor has not proven historically expensive even there.
The factor is quite efficient (information ratio) in North America and Europe, with its effectiveness improving over time, while in Asia it is falling behind quality and growth. This could be due to the early focus of investors to force ESG on developed companies (thanks to stronger pressure from policymakers).
We tried to create a synthetic ESG, but this wasn’t a success. First, the correlation structure between ESG and the traditional styles is not stable but varies over time. Second, the explanatory power of the regression is quite poor, meaning we can only capture a small amount of the ESG variability using traditional risk premium.
Our conclusion is that ESG is a unique source of alpha and exploits an information set that’s not captured by conventional factors and could improve diversification in a way that can’t done with traditional metrics.
[1] For this analysis ESG scores are provided by MSCI ESG Research LLC. MSCI ESG Research products are designed to provide in-depth research, ratings, and analysis of environmental, social and governance-related business practices to companies worldwide.
[2] Each single factor is built monthly on regional, industry neutral basis for both the index and a dividend universe. We use the simple average to calculate the final factor return vs the universe and we take into count the impact of turnover.
Share via:
Recommended for you