Growing concerns over the future of the planet, corporate greed and social inequality have led to a boom in socially responsible investing (SRI).
According to a recent BNP Paribas Securities Services survey, some 75% of asset owners and 62% of asset managers hold more than a quarter of their investments in funds incorporating environmental, social and governance (ESG) factors into their strategy. That’s up from 48% and 53% respectively just two years ago.
To meet this soaring demand, funds and wealth management sectors are devising innovative strategies. French firm Ecofi Investissements, for example, recently launched a fund managed with artificial intelligence, which extracts performance from ESG data and making active stock selections from this.
Focused on Eurozone equities, its mathematical algorithms look at a stock’s ESG performance as well as market data to determine whether it is likely to generate strong investor returns.
“Our algorithm looks at all the variations in ESG scores since 2012. It understands variables such as sector biases and how a mining company will always have a higher E-score than a bank,” says Christophe Geissler, chairman of French asset manager Advestis. “We look at a company’s carbon footprint and its tax responsibilities compared to the index and find correlations between ESG scores and price returns.”
The SRI filter is used at the beginning of the process, as is the screening out of companies involved in areas such as controversial arms, tobacco, gambling, coal and tax havens. At the end, human eyes have the last look to ensure the portfolio is fully ESG compliant.
“We hope AI will bring consistent performance through managing ESG data,” says Geissler. In the past, he points out, there has been a perception that there is a trade-off between ESG and performance; if you wanted good ESG metrics then you had to give up getting strong returns. Geissler insists that is not the case: “ESG variables contain more valuable information about the future performance of a company than accounting or balance sheet variables do. It means less portfolio risk and more performance.”
Using Algorithms
Paris-based asset manager Ossiam is another firm applying machine learning to ESG data as part of its Ossiam World ESG Machine Learning UCITS ETF (OWLP). It tracks a selection of equities in developed markets using an algorithm that ranks companies and shows a distinct link between their ESG characteristics and potential future financial performance.
“The majority of investors are convinced they need to improve the ESG side of their portfolios but are scared about potentially sacrificing returns,” says Carmine De Franco, head of fundamental research at Ossiam. He wants to find the ESG profiles and patterns in each company that present an financial opportunity – or identify risk.
“A human being would, because of the amount of data and indicators per company, find this difficult to do. Machine learning can do this more robustly,” he adds.
James Purcell, head of sustainable and impact investing at UBS Global Wealth Management, agrees that using machine learning is important in getting bigger and better ESG data. “Getting as many alternative sources of data in areas such as corporate culture helps you make more informed decisions.” At present there are a number of ESG data sets such as company annual and sustainability reports, but these are self-reported and there can be issues around their consistency and biases. However, Purcell adds: “I don’t think we are at the stage where we can fully take an automated approach. There are many nuances in sustainability which need a human and personal element.”
Wealthy Investors Matched With ESG Products
Indeed, UBS has just completed a pilot programme to help align ESG investments with personal investors' choices.
It created a database using information from sustainable data providers, which helps it score the ESG strengths of 10,000 instruments across equities and bonds. They are rated against seven areas such as climate change and corporate governance.
A pool of UBS’s wealthiest investors then rated their preferences for these seven ESG sectors to create a personalised sustainability score. This is then compared with the bonds and stocks to create the best match.
“Previously some investors struggled to fully engage with ESG because some companies may do well in certain areas of sustainability such as pollution but not so well on others, such as human rights,” says Purcell. He says the pilot had helped investors feel better able to express their personal ESG preferences. Most were familiar with exclusions but not making positive selections in causes that matter most to them.
He adds: “They now think more deeply about ESG and the fact that every bank or retail firm has, for example, a water footprint and needs a strategy to deal with it. They are looking behind the headline ratings and finding out how companies are directly approaching ESG.” The scheme will be rolled out to other investors in the future.
Finland's Nordea Asset Management is also seeking closer direct engagement with businesses, Its STARS fund range invests only in firms meeting its ESG standards and are selected on both their financial and sustainability performance.
“Active ownership is vital. We engage in dialogue with our investee companies on a range of issues, including how they may review their own performance on sustainability indicators,” says Marjo Koivisto, co-head of responsible investments. “We also use our voting right on our holdings and are represented on several nomination committees. We are also working in investor coalitions on climate data and building a proprietary ESG data platform where we deploy the latest scientific data methods.”
