Building on previous research that examined the relationship between environmental, social, and governance (ESG) factors and stock performance, the authors study the impact of ESG factors on portfolio returns in the US market for the period 2007–2012. Previous researchers have found that excluding segments of a universe is detrimental to the risk-adjusted returns, but the authors, when using ESG factors to narrow an investment universe, find a positive effect.
What’s Inside?
The authors hypothesize that environmental, social, and governance (ESG) ratings have predictive power on the return and risk profile of stocks and that incorporating these factors into the investment process should improve portfolio performance. They examine the relationships among ESG ratings, returns, risk, and risk-adjusted returns. They also eliminate the lower tail of ESG scores from a stock universe and examine the return and risk-adjusted return distributions of randomly generated portfolios. They conclude that ESG investing does not impose an opportunity cost and that asset managers can enhance their stock selection and portfolio construction by incorporating ESG factors.
How Is This Research Useful to Practitioners?
As ESG investing becomes more mainstream, investment professionals are increasingly confronted by the types of questions the authors address.
Previous researchers have found that randomly excluding a set of stocks from an investable universe imposes a cost. The authors demonstrate that excluding stocks with the worst ESG ratings does not impose a cost. Instead, eliminating lower-tail ESG companies tends to improve portfolio risk-adjusted returns.
They also find a strong negative correlation between ESG ratings and stock volatility. High-ESG stocks tend to be in the low-volatility group. In addition, the best-performing stocks have better ESG profiles, suggesting that an asset manager, by excluding the worst ESG stocks, can reduce portfolio risk and increase the probability of identifying these best-performing stocks. The authors do not find a strong correlation between ESG ratings and median stock returns except during the peak of the financial crisis.
To determine the independent effects of the ESG factor, the authors untangle volatility. Based on their tests, they conclude that the ESG factor is a positive contributor in its own right.
This research will be relevant to those who are currently investing using ESG factors or who are considering doing so. It will also be of interest to investors and fiduciaries who are considering ESG investing but have concerns about the impact on portfolio returns or risk. The authors also include a good overview of other studies in the area of ESG investing, which may be helpful to those seeking additional information on the topic.
How Did the Authors Conduct This Research?
To conduct their research, the authors use Thomson Reuters Corporate Responsibility annual data files from 2007 through 2012. The ESG coverage is compared with the Russell indexes. The data availability is better for larger companies; 85% of the Russell large- and mid-cap indexes have ESG rating coverage, compared with 18% for the Russell 2500 and only 3% for the Russell 2000 index.
The authors divide the universe based on ESG ratings and then analyze the returns of the two groups. The universe tails are established for both the 5% and 10% cutoff levels. They then divide the dataset based on returns and analyze prior ESG ratings. To avoid look-ahead bias, they use a six-month lag and capture annual return and volatility data for the subsequent two-year periods following the lag. They use a similar method to analyze risk-adjusted returns. Risk-adjusted returns are calculated as annual return/annualized monthly standard deviation. Risk is calculated relative to broad market volatility as measured by the Chicago Board Options Exchange Volatility Index (VIX) and the standard deviation of the S&P 500 Index.
When running simulated portfolios from the modified universes, the authors create 100 sample portfolios of 40 stocks—each chosen through a random selection process. They also calculate the correlation between ESG ratings and subsequent stock performance, as well as the correlation between ESG ratings and subsequent stock volatility.
The study has some limits. For example, the authors focus on US data from one ESG source, the study has limited small-cap ESG coverage, and the study period is a relatively short window that includes the 2008–09 credit crisis. The authors state that the focus on US data provides better control over market effects and ESG standards.
Abstractor’s Viewpoint
This research represents a strong contribution to the body of literature focused on socially responsible/ESG investing. ESG investing has seen robust growth and continued interest from the investor community, but concerns persist that reducing an investment universe negatively affects portfolio construction and returns. The authors tackle this issue and find there is a positive effect when excluding poorly ranked ESG companies from a universe. This study should help alleviate concerns that ESG investing necessitates a performance cost. I would be interested to see this study applied to a non-US dataset and expanded to include additional years.