ESG investing is dead! Long live ESG investing!

By Henry Bee and Nigel Nguyen


  • “Tree hugging” investing is taking the investment world by storm, and the ESG score is fast becoming a part of every firm’s investment process.
  • ESG investing may have a positive impact on the world, but may not stuff investors’ pockets.
  • With the right tools, ESG investing can produce the superior risk-adjusted return, without the guilt.


ESG investing: trending higher than Taylor Swift

Sustainable investing has been a hot button topic lately. Big institutional investors want to show the world that they can profit while making a positive impact all around. The ESG score – short for environmental, social and governance – is a ranking method to tell investors how much “good” a company does.

Now that investors are paying attention, companies also need to showcase their commitments to socially responsible operations or risk getting overlooked. Many have even developed separate sections in their annual reports to discuss ESG issues.

Practicing ESG is good ethics, but does it make a good investment strategy? We use the technology behind Cassia’s new COPILOT platform to answer two burning questions:

  1. Can a passive index of highly ranked ESG companies match the performance of the S&P 500?
  2. Can we add value to the ESG selection process to improve its risk-adjusted return?


Sustainable investing. From scratch.

With data provided by Sustainalytics, we take the top 50 stocks as ranked by ESG scores from the S&P 500, equal-weight them, and pitch this portfolio against the S&P 500 to see how it fares. Why 50 stocks? The largest socially responsible funds select around 50 stocks – funds like the Parnassus Core Equity, Neuberger Berman Socially Responsible, and Calvert Equity, which are all listed on the Forum for Sustainable and Responsible Investment.

Look. We know. Equal-weighting isn’t ideal. Certain sectors may have higher ESG scores than others, leading the investor to become overexposed to one sector, resulting in an unbalanced portfolio.

So we test an alternative weighting method. The goal here is to spread out the underlying exposures, as much as possible. In other words, we want to overweight the ESG stocks that behave most differently from the others. We use Cassia’s very own multivariate GARCH model and ICA (independent component analysis) to forecast the underlying drivers of ESG stock returns. Knowing these factors, we can then use the minimum correlation algorithm to optimize the weights to create a more balanced exposure. We expect this method to have a lower correlation to the S&P 500. And possibly even generate some alpha.

Look at how unique the holdings are for Cassia’s MinCor (minimum-correlation) ESG portfolio. Unlike equal-weight, Cassia’s MinCor ESG portfolio only holds 17 stocks, on average. Some examples include Seagate (computer hardware), Kohl’s (department stores), Edwards Lifesciences (medical equipment), Staples (office supplies), and Gaps (clothing retail). consumer discretionary).


We crunched the numbers, and…

Based on available data, we performed walk forward analysis, rebalanced once a quarter with a 2-day trade delay from March 2014 to February 2016. Due to the rather short time horizon, we broke up the data into three chunks — a widely-used technique called cross validation in statistics and machine learning. We want to see if there are any patterns in the three sub-periods.

Cassia’s minimum correlation ESG portfolio outperformed the simple equal-weight one, and outperformed the S&P 500 in two out of the three sub-periods; in both absolute return and Sharpe ratio.

Perhaps more importantly, the balanced exposures resulted in a much lower correlation with the S&P 500, offering investors unique return streams (in all three sub-periods, correlation with the market hovers around the 68% to 75% range). Important for you factor-investing geeks out there. And for anyone looking to charge a fee – Hey ESG fund companies, I’m looking at you!

Because otherwise, things are pretty grim in ESG land.

The equal-weighted ESG portfolio underperformed the S&P 500 in all three sub-periods. You know, lower returns, higher volatility, longer and more severe drawdowns, and a 95% correlation to the market.

And this is before fees! Imagine the underperformance after the 1-2% fees. Our numbers suggest that due to a more constrained universe, ESG may systematically underperform the market. And that using an alternative weighing method can make ESG investing more attractive for investors.


We will explore sector-neutral weighting in part 2 of this series. We expect that a sample of 50 stocks with similar sector exposure as the S&P 500 can enhance the robustness of the ESG portfolio. We will also compare sector-weighted ESG against Cassia’s minimum correlation method. Stay tuned!

If you have any questions or would like to schedule a walk-though of the type of institutional quant strategies Cassia has developed, just get in touch. And be sure to follow us on twitter.

Build your practice on the backs of 3 Nobel prizes — Prospect Theory, MPT, and GARCH. 

See what GARCH can do for you.
See what GARCH can do for you.
Build your practice on the backs of 3 Nobel prizes — Prospect Theory, MPT, and GARCH.