The details of how an ESG policy is implemented in a portfolio may have a direct impact on its performance.
There is a key distinction between an ESG approach based on negative screening by industry and one based on relative comparisons of the firms in each industry.
For example, an investor using a negative screen may choose to exclude coal mining companies from their investment universe. Another investor may use ESG ratings to rank coal mining companies, and choose to invest in the ones that have the best overall ranking within the sector. In the first case, in a year in which coal mining companies outperform the market, the investment portfolio may lag a broad market index.
In the second approach, the portfolio is neutral with regard to the systematic sector exposure, but favours those companies with better ESG policies – i.e., those that do less harm to the environment, treat their workers better, and are better managed.
In our study, we constructed diversified portfolios designed to track the US Investment Grade Corporate Bond Index and imposed either a positive or negative tilt to different ESG factors. We found that:
- In comparing the excess returns of a high-ESG portfolio with a low-ESG one, the high-ESG investment outperformed steadily using data from both Sustainalytics and MSCI. The return advantage over the past seven years averaged 0.29% per year and 0.42% per year, respectively.
- In addition to the overall ESG scores, we also tested the effects on performance of the separate E, S and G scores from these two providers. Despite their different approaches to evaluating issuers, a very similar pattern is observed for both: Governance had the strongest link, followed by Environment. Social scores had the weakest link with performance; for one provider, the high-S portfolio slightly underperformed the low-S portfolio.
- One explanation for the steady outperformance of high-ESG bonds over the past seven years could have been that increasing interest in sustainable investing has driven up the prices of these bonds, potentially making them less attractive going forward. We carried out tests to measure this “ESG spread premium” and found that this has not been the case – high-ESG bonds (as measured by either provider’s scores) have not become more expensive.
