AI and Big Data Enabling ESG Investors to Clear the 'Poor Disclosure' Hurdle

AI and Big Data Enabling ESG Investors to Clear the “Poor Disclosure” Hurdle
New environmental, social, and governance investment themes are appearing

July 2018, London. Big data and artificial intelligence (AI) are helping environmental, social, and governance (ESG) investing to gain a firm foothold in quantitative investment―an option previously severely restricted by limited disclosure, according to the latest The Cerulli Edge―Global Edition.

The combination of big data and AI is enabling investors to access vast amounts of information from objective sources, bringing greater frequency, granularity, and real-time analysis to ESG investment, says Cerulli Associates, a global research and consulting firm.

“Enhanced data is enabling asset managers to integrate ESG, leading to new investment themes, such as ESG momentum strategies, and investment in line with the sustainable development goals,” says Justina Deveikyte, associate director, European institutional research, at Cerulli.

Deveikyte notes that, although the days of responsible investment merely entailing negatively screening of sin stocks are long gone, ESG has struggled to fully capitalize on quantitative investment, primarily because weaknesses in the data have made it difficult to build strategies. Unlike financial reporting, there are no universal guidelines to steer or compel corporate ESG reporting. Piecemeal information and incomplete data are the norm.

She adds that data limitations have affected quant investment in various ways. For example, large companies are better at disclosure than small companies. Because most ESG models penalize nondisclosure, this can result in a large-cap bias in portfolios.

“The use of big data and AI is radically transforming data gathering. These changes have opened up ESG investment to quant funds, which are busy developing new algorithms to systematically evaluate companies,” says Deveikyte.

Cerulli cites the example of one asset manager that draws in raw and unstructured data to build an initial “input layer” that combines companies’ own reporting with external information mined from news outlets, non-governmental organizations, social media, and Google trends. It uses more than 50,000 sources in 15 different languages to create a daily view on more than 7,000 companies in the world. The technology sifts out fake news. Another example is a hedge fund that gets around the fact that only 2,500 companies in the world report their water usage by using regression models to infer what water companies would have used based on data from similar companies, in similar sectors and geographies.

“Vastly improve access to data is changing the way asset managers integrate ESG and enabling new investment themes. For example, some asset managers are using the data flow to introduce more short-term, reactive strategies, traditionally difficult in ESG because of the slow pace of change, accentuated by the lag in reporting data,” says Deveikyte.

Looking for more information? Contact Us.