Interview with Antonio Celeste, Sustainalytics
Sustainalytics, a leading global provider of ESG and corporate governance research and ratings, has launched ESG Signals, an innovative quantitative risk management and portfolio tool that provides securities-level financial risk and opportunity signals based on environmental, social and governance (ESG), trading and financial data.
Developed in collaboration with Advestis, ESG Signals combines ESG data and quantitative modeling to help portfolio managers and heads of research identify and monitor portfolio risk and opportunities. ESG Signals can also be used as a portfolio construct, enabling index providers to build and modify current or new indices. Combining seven years of research on 1,600 companies and trading/financial data from Advestis, ESG Signals facilitates the investment selection process by identifying an opportunity, assessing a neutral-weighting, or highlighting a risk signal for each individual security being analyzed.
3BL Media recently spoke with Antonio Celeste, Director of Institutional Relations at Sustainalytics, on the launch of their innovative analytical tool, ESG Signals.
3BL: What makes ESG Signals an innovative product/process?
AC: Our approach in incorporating ESG factors into machine learning is a unique and new concept.
ESG Signals focuses more on short-term performance versus longer-term fundamental analysis and the rules are meant to have an impact on a short-term time horizon. Depending on the output signals, rebalancing can occur based on observed variables as opposed to investing for longer term outlooks or shifts in fundamental beliefs.
To test the findings, Sustainalytics and Advestis applied ESG Signals to a large cap, market-weighted index. Our results showed the re-weighted indices outperforming their benchmark between 110 and 430 basis points annually, depending on the frequency of rebalancing adopted.
3BL: What impact do you think ESG Signals will have on Sustainalytics’ client base? Does this innovative tool constitute a shift in the way Sustainalytics will model their business going forward or should this be considered a complement to the current suite of offerings?
AC: We view ESG Signals as a complement to our business, not a replacement of current products. Sustainalytics currently has a global footprint with 320 people on four continents and coverage for over 7,000 companies and sovereigns. Traditionally, Sustainalytics clients have been focused on long-term investing based on fundamentals and research. One trend we have seen over the last five years has been a movement towards quantitative trading and more requests from investors regarding variables impacting performance and the details behind them. This led to our partnership with Advestis to develop ESG Signals.
ESG Signals will likely generate interest from both quantitative and fundamental investors. While ESG Signals may initially have a larger appeal to quantitative trading given the shorter-term time horizon, fundamental investors will look at the variables as well to complement their strategic approach.
3BL: How do you define “Big Data?” Is the focus on Big Data a long-term trend in the market for investors?
AC: Big data refers to the large amount of data used as inputs to investment decisions. ESG Signals looks at over 500 variables- momentum variables/financial variables/etc.- per company. We analyze over 1,600 companies, looking at all those variables per company, at different data points per year. This translates to a huge amount of data. Machine learning is needed to consolidate this information into output. ESG Signals analyzes thousands of correlations between variables over time and applies the machine learning to extract meaningfully predictive risk/opportunity signals.
The focus on big data has substantially increased over the last few years and should continue to do so going forward.
3BL: What are the variables/indicators that you most focus on when helping portfolio managers identify opportunities? How do you define signals?
AC: ESG Signals doesn’t assign weight to variables. Some are more recurring than others (corporate governance, controversy, energy transition, etc.). We use machine learning to check the interaction with variables and plan to publish a study on the individual variables in the future.
Signals are the results generated by ESG Signals. They are derived by observing the data factors that made stocks increase/decrease in price in the past and the movement’s correlation with past data. ESG Signals generates a three-level output—opportunity, neutral, and risk—for each company every day.
3BL: You currently have over 1,600 companies that you are following with ESG Signals. How did you choose those companies and do you have a target for increased institutional coverage?
We chose the top 1,600 companies according to market cap from developed countries. Within ESG Signals we tested the application with these companies and calculated sector rules. We do plan on expanding our universe to include small caps, emerging markets, bonds, etc.
3BL: In your analysis, frequent rebalancing by an investor (one to seven days) correlates to increased outperformance against the index. Do you foresee investors in the ESG sector rebalancing/trading to this extent, and is there a concern that through frequent trading the manager could have a substantial deviation from the index they are following?
AC: Yes, in principle an ESG investor can apply a weekly or daily rebalancing. We tested the impact of daily and weekly balancing on return, and plan to test this on deviation from the index.