What Big Data Says About ESG

The Environment Agency Pension Fund uses data science to prove the merits of environmental, social, and governance investing.

Do environmental, social, and governance (ESG) considerations help or harm investment performance? Big data might have the answer.

In paper commissioned by the Environment Agency Pension Fund (EAPF), Henley Business School professor and data scientist Andreas Hoepner explored the effects of ESG investing—and divestment from “sin stocks” in particular—on the risk and returns of a portfolio.

“[Data science] is not only crucial for potential divestment of sin stocks, but also for investing purely in environmentally responsible firms.”“A core component of our responsible investment policy is to ‘apply evidence-based decision making in the implementation of responsible investment,’” wrote Faith Ward, responsible investment and risk chief at EAPF. “Financial data science can help investors consider the perennial and thorny question as to whether ESG integration in investment processes can be beneficial to returns and risk of their portfolio.”

Hoepner argued that data science offered the best insight into this question, because it involves “large scale, deep data, and advanced statistical analysis”—allowing researchers to identify what is happening in the real world as opposed to what should happen according to economic theory.

“While both divestment campaigners and financial economists tend to argue on a normative (‘should happen’) level, modern technologies such as algorithmic learning and big data analytics allow financial data scientists to address crucial societal questions from a more descriptive (‘is happening’) level,” he explained.

His findings? ESG factors can be used to identify signals for higher return and reduced downside risk, so long as investors possess the “necessary sophistication.”

Even the exclusion of “sin stocks” like tobacco firms—supposedly high-performing equities—is actually beneficial to investment portfolios, Hoepner found. This is in direct contrast with recent research by Wilshire about the California Public Employees Retirement System’s tobacco divestment: The consultant reported that the fund missed out on roughly $3 billion in returns since exiting tobacco stocks in 2000.

“While one might, simplistically, consider tobacco firms as ‘safe’ value stocks, deeper analysis shows that they actually exhibit riskier growth stock characteristics within their industry, which itself is closely associated with value investing,” Hoepner wrote. “In other words, while consumer goods are likely a safer investment during economic downturns than the average industries, tobacco stocks are actually less likely safe than other consumer goods.”

Furthermore, when analyzing “realistic” value-weighted portfolios instead of the standard equal-weighted portfolio benchmark, Hoepner found no evidence of outperformance by sin stocks, either globally or in the US.

“[Data science] is not only crucial for potential divestment of sin stocks, but also for investing purely in environmentally responsible firms,” he concluded. “While financial economists would expect that this should increase portfolio risk, it is actually significantly reducing the worse case risk of investable pension fund portfolios.”

In a separate report, the BlackRock Investment Institute has argued that climate risks had been “underappreciated and underpriced because they are perceived to be distant.” The researchers added investors that engage with climate change research and technology will also be able to reap financial rewards.

Read the full paper, “Financial Data Science for Responsible Investors.”

Related: Infographic: The Market for Vice & Big Data’s Big Rise