How Artificial Intelligence is transforming ESG data and indices
Thomas Kuh, PhD, Head of Index, Truvalue Labs (San Francisco) writes on the fourth industrial revolution and how it is affecting ESG investments.
We are in the early stages of the Fourth Industrial Revolution (4IR), “a fusion of advances in artificial intelligence (AI), robotics, the Internet of Things (IoT), 3D printing, genetic engineering, quantum computing, and other technologies.” According to Klaus Schwab, Founder and Executive Chairman of the World Economic Forum, who coined the term, 4IR “is characterised by a fusion of technologies that is blurring the lines between the physical, digital and biological spheres.” Preceded by periods defined by the emergence of transformative technologies – steam and water power, electricity and assembly lines, and computerization – 41R is expected to disrupt virtually every industry and fundamentally alter our day-to-day lives as the pace of change accelerates.
It is not a coincidence that ESG investment has taken root in mainstream finance during the Digital Age. In a white paper, ESG Research in the Information Age, I argue that the Internet provides two critical conditions for understanding the dynamic growth of ESG investing. First, it provides daily information essential for analyzing ESG issues at corporations. Second, it provides stakeholders a voice in shaping the narrative on a corporation around sustainability.
ESG research analyzes how effectively companies manage risks and opportunities associated with sustainability. These issues vary across industries – mining companies face vastly different sustainability issues than banks – so ESG research is a complex undertaking. In recent years, ESG research benefited from the work of organisations like CDP, which collects company-reported carbon emissions data, and the Sustainable Accounting Standards Board (SASB), which identifies which material ESG issues for each industry.
ESG research has been indispensable to the growth of sustainable investing, yet critics point out limitations of the research related to transparency, timeliness and subjectivity. AI technologies enable firms like Truvalue Labs to distill ESG signals from unstructured data. By applying natural language processing (NLP) and machine learning (ML) to news media, industry publications, NGO reports, and other digital sources we can generate analysis of the perspectives/sentiment of corporate stakeholders on ESG issues. It takes machines to sift through the proliferation of data to find the salient signals.
This “New World” data directly addresses the criticisms of traditional ESG research. The analysis is:
- Transparent: Based on the industry developed SASB framework, rather than a proprietary approach
- Timely: Daily signals incorporated into sustainability scores reflect the current situation, instead of annual ESG ratings
- Consistent: AI produces consistent results, not subject to analyst bias [nor influenced by unaudited, company-reported information]
Globally, assets using sustainable investment strategies are estimated to represent more than USD30 trillion. [GSIA] According to Morningstar, European sustainable equity fund flows gathered EUR37 billion in the first half of 2019, nearly equaling the record of EUR38 billion for all of 2018. Though a small proportion of the overall market, sustainable funds had net inflows while conventional equity funds realised net outflows during the period. Through Q2 2019, sustainable ETFs attracted a record EUR5 billion in net flows – more than in all of 2018 – while all sustainable index funds brought in EUR6.8 billion.
In response to investor demand, the number of ESG funds in the market has grown dramatically since 2016, when 124 funds were launched (104 active, 20 passive), to 2018 when 305 were launched (262 active, 43 passive). As of August 2019, there were 102 ESG ETFs available to European investors, 88 equity funds and 14 bond funds.
As demand for ESG investments accelerates, AI-driven sustainability data offers the opportunity to create innovative benchmarks that reflect insights hidden in the mass of unstructured data. And it offers ETF sponsors a chance to create differentiated funds to capitalize on the market’s appetite for authentic sustainable investments.