›  Strategy 

ESG Data-Overload: Artificial Intelligence to the rescue

Today any ESG manager has to access, understand and analyse vast amounts of data, from worker safety standards to greenhouse gas emissions, to meet their commitment (and investors’ expectations) to embed ESG analysis into their portfolio management.

Article also available in : English EN | français FR

A big challenge to successful environmental, social and governance (‘ESG’) investing is the ability of portfolio managers to use data efficiently to drive value creation. Today any ESG manager has to access, understand and analyse vast amounts of data, from worker safety standards to greenhouse gas emissions, to meet their commitment (and investors’ expectations) to embed ESG analysis into their portfolio management.

This need to meet investors’ expectations has placed pressure on research providers who are expected to mine quantities of information to deliver timely, reliable and comprehensible ESG data. As more companies disclose data using different metrics and frameworks, investors are forced to compare and measure vast datasets. This can also make it difficult to discriminate relevant data from ‘greenwashing’ tactics deployed by companies. ESG analysts are forced to overcome logistical obstacles so that they can screen public data and company information to find accurate and timely ESG data.

How can this data overload be tamed? Step forward innovative solutions, like artificial intelligence (AI), machine learning and satellite imaging. Asset owners are now forced to look beyond traditional datasets and practices.

Objective, scalable and transparent, these techniques can help make ESG data collection into a coherent, unified process. They allow research providers to improve the quality and rigour of their data, while offering asset managers a platform that has the functionality to improve the granularity and quality of information available to their investors.

Despite the integration of AI within ESG analysis being in its early stages, we are starting to see AI implemented within portfolio construction. Pioneers such as TruValue labs mine big data and apply AI to their research output to assess, monitor and rate ESG behaviours and criteria, in real time. There are also examples of big data being used as part of a portfolio’s risk mitigation strategy. For example, satellite imagery is being used to monitor for natural disasters, such as flooding.

AI and machine learning enhances the credibility and authority of ESG integration. The software can help avoid claims of ‘greenwashing’ and ensure ESG analysis remains credible and reliable. The industry has worked to promote the adoption of ESG criteria and educate investors about the long-term opportunities ESG data presents. But the enthusiasm for ESG means that the industry is susceptible to promoting products that only superficially incorporate ESG analysis. Under the spotlight of AI and big data, there will be nowhere to hide.

That said, integrating AI into ESG analysis will not be without its own obstacles. Portfolio managers must assess critically the ethical conundrum of AI. For instance, what measures need to be in place to prevent data errors, leaks or manipulation? How can we protect the ESG data analysts we do need against the speed and processing capabilities of robotics? These are important questions that the industry must address. We do not want to undermine our achievements for the sake of short-term gains.

Last year, we at CPR Asset Management worked with independent advisory firm, INDEFI, to study the changes in ESG integration practices among institutional investors and distributors in France and Europe. We found that investors use multiple sources of information in their ESG analysis. While this helped them uncover new insights, perspectives and datasets, it also overwhelmed investors with the sheer scale of extra-financial data. The impact of this ‘data fatigue’ was twofold; it decreased the robustness of their analysis and increased the likelihood of missing important, and costly, controversies.

Understanding, evaluating and monitoring data is key to the longevity and credibility of the ESG movement. As we face a tide of new ESG data and frameworks, we can’t compromise on providing investors with quality data. It is our responsibility to find solutions that are cost-efficient, and which can evolve in unison with the ESG landscape.

Technology offers a promising tool to help us better determine the financial performance of portfolios. The algorithms driving AI and big data exceed human capabilities in many respects. We should embrace these approaches in an ethical and responsible manner.

Tegwen Le Berthe , June 2019

Article also available in : English EN | français FR

Send by email Email
Viadeo Viadeo


Strategy CPR AM has recently launched CPR Invest – Global Disruptive Opportunities | A look back at an accelerating phenomenon: disruption

The recently theorised phenomenon of "disruption" is defined as a process whereby a product, a service or a solution disrupts the rules on an already established market. Technological progress, along with the globalisation of trade and demographic changes are now helping to (...)

© Next Finance 2006 - 2020 - All rights reserved