Breakout B: New Tools in ESG Analysis

BREAKOUT B: New Tools in ESG Analysis

Rapporteur: Max Sorto


Panelists:

  • Jim Hawley, TruValue Labs, Moderator

  • Mark Kriss, Macroclimate: Geofinancial Engineering Initiative, 

  • Erik Allen, Rho AI: The Application of Machine Learning to Sustainable Finance


In the New Tools in ESG Analysis breakout, two speakers and the moderator led a lively discussion about how technology could spur innovation in ESG research and potentially change financial markets. Although the speakers focused on methane leaks and ESG data quality, all participants agreed that these promising technologies could reduce time-pricing lags on externalities for all investors and input missing data on financial reports.

One speaker displayed satellite images depicting methane gas sensors, and demonstrated how one could use the sensors’ locations to attribute methane leaks to public companies. The speaker then used two case studies, BP Deepwater Horizon and the Aliso Canyon gas leak, to illustrate how public knowledge of an issue can directly affect stock price. In the case of BP, the public was immediately made aware of the explosion and BP’s share price tumbled accordingly. For Aliso Canyon, however, it took six weeks for the leak get disclosed to the public, and it was only revealed because the LA County filed a lawsuit against SoCal gas, the site’s overseeing utility and a subsidiary of Sempra Energy. After the lawsuit was filed Sempra’s stock price begin to fall, and it fell further after the LA times estimated the financial damages to the company. The speaker predicted that financial consequences are not appropriately quantified when the environmental impacts are invisible. New technologies will help shareowners pinpoint location and timings of leakages; the financial markets will adjust appropriately and poor-acting companies would be held accountable.

The other speaker exhibited how machine learning drastically improves ESG data quality. Using Sustainable Accounting Standards Board (SASB) data as a backdrop, his analysis uncovered numerous missing company sustainability parameters and predicted values with high degrees of accuracy. Furthermore, the speaker used the same models to test existing SASB data to identify metrics that seem incorrect – likely where the company inaccurately reported their information. If machine learning tools were to successfully predict sustainability data, it would reduce investor dependency on company self reports and improve data quality.

Participants in the breakout brainstormed the types of clients that would use these tools, and seemed to agree that the improved access and timing of data retrieval could have large implications on the industry. Currently, many ESG investors use data in their long term oriented portfolios, but these technologies could also make sustainability data more useful for investors with shorter time horizons.

In the New Tools in ESG Analysis breakout, two speakers and the moderator led a lively discussion about how technology could spur innovation in ESG research and potentially change financial markets. Although the speakers focused on methane leaks and ESG data quality, all participants agreed that these promising technologies could reduce time-pricing lags on externalities for all investors and input missing data on financial reports.

One speaker displayed satellite images depicting methane gas sensors, and demonstrated how one could use the sensors’ locations to attribute methane leaks to public companies. The speaker then used two case studies, BP Deepwater Horizon and the Aliso Canyon gas leak, to illustrate how public knowledge of an issue can directly affect stock price. In the case of BP, the public was immediately made aware of the explosion and BP’s share price tumbled accordingly. For Aliso Canyon, however, it took six weeks for the leak get disclosed to the public, and it was only revealed because the LA County filed a lawsuit against SoCal gas, the site’s overseeing utility and a subsidiary of Sempra Energy. After the lawsuit was filed Sempra’s stock price begin to fall, and it fell further after the LA times estimated the financial damages to the company. The speaker predicted that financial consequences are not appropriately quantified when the environmental impacts are invisible. New technologies will help shareowners pinpoint location and timings of leakages; the financial markets will adjust appropriately and poor-acting companies would be held accountable.

The other speaker exhibited how machine learning drastically improves ESG data quality. Using Sustainable Accounting Standards Board (SASB) data as a backdrop, his analysis uncovered numerous missing company sustainability parameters and predicted values with high degrees of accuracy. Furthermore, the speaker used the same models to test existing SASB data to identify metrics that seem incorrect – likely where the company inaccurately reported their information. If machine learning tools were to successfully predict sustainability data, it would reduce investor dependency on company self reports and improve data quality.

Participants in the breakout brainstormed the types of clients that would use these tools, and seemed to agree that the improved access and timing of data retrieval could have large implications on the industry. Currently, many ESG investors use data in their long term oriented portfolios, but these technologies could also make sustainability data more useful for investors with shorter time horizons.