ESG integration has historically only been associated with fundamental strategies, but this perception is slowly changing as several quant managers are now integrating ESG factors into their valuation models and investment decisions.

Overview of quantitative strategies

Quantitative (quant) strategies harness data, using mathematical models and statistical techniques to outperform their benchmarks.

Quant managers define models and rules that make investment and/or portfolio weighting recommendations. Quant managers will for example make predictions on future asset price movements and/or company fundamentals, based on technical and/or fundamental data, both historical and forecast.

The investment process can be typically divided into the following three stages:

1) Analysing data and statistical testing

  • Some quant managers use statistical techniques to identify relationships between datasets over different investment horizons, and look for patterns, correlations and/or factors that drive asset price movements. Other quant managers will use valuation techniques to identify mispriced securities.

2) Building models and back-testing

  • Quant managers write algorithms, which form the basis of their models. Back-testing shows how they perform using historical data, to indicate whether they are likely to generate superior returns.

3) Implementing strategy

  • If the back-testing is considered successful, quant managers will implement the model. Changes in market conditions have the potential to make purely statistical approaches defunct and may require managers to restart the process, identifying new relationships and developing new algorithms.

Computers can run the models and produce suggested investment decisions. Systematic rules and portfolio construction techniques, along with integrated risk management tools, lead to portfolio weighting recommendations. Some models are integrated into managers’ trade order management systems to facilitate execution. Many quantitative managers have risk management procedures in place to ensure that model output reflects the investment teams’ strategy and intentions.

As ESG data becomes more prevalent, statistically accurate and comparable, more managers are likely to perform statistical techniques to identify correlations between ESG factors and price movements that can generate alpha and/or reduce risk.

The quant managers that perform ESG integration have constructed models that integrate ESG factors alongside other factors, such as value, size, momentum, growth, and volatility. ESG data and/or ratings are included in their investment process and could result in the weights of securities being adjusted upwards or downwards, including to zero. There are two main approaches to integrating ESG factors into quantitative models. They involve adjusting the weights of:

  • securities ranked poorly on ESG to zero, based on research that links ESG factors to investment risk and/ or risk-adjusted returns;
  • each security in the investment universe, according to the statistical relationship between an ESG dataset and other factors.

New Amsterdam Partners’ portfolio construction process uses the first approach. They found a positive correlation between ESG ratings and risk adjusted returns since the 2008 financial crisis, and their quant model adjusts the weights of stocks rated poorly on ESG to zero for all portfolios. Asset Management also see a correlation between ESG and company/investment performance, and reduces the weights of stocks that do not pass their sustainability process to zero. Auriel Capital and Analytic Investors provide examples of the second approach. Auriel Capital uses their research on the statistical relationship between ESG factors and investment returns to create an ESG score that adjusts the weights of securities in their portfolio. Due to their research into the links between ESG ratings and future risk, Analytic Investors’ investment process uses a risk-scaling process to ratchet down the stock-specific maximum position limit as a company’s ESG rating falls.