As the 2020 ESG (Environment, Social, and Governance) season begins, we appear to be entering the third era of ESG investment integration.
The first generation of ESG investors used data on topics such as product involv
ement (alcohol, tobacco, gambling) or business practices (anti-union, involvement in Burma) to screen out “bad” companies. These investors often relied on a single data provider and simple guidelines (e.g., <5% of revenue is OK, more than 5% of revenue is bad).
The second generation of ESG investors decided that a company’s sustainability performance should be related to its riskiness and/or its financial performance. They used multiple ESG data sets to scan large universes of companies such as the Russell 3000 or the MSCI ACWI. It was hard to reconcile the disparate signals from these data sets—each was based on its own methodology towards measuring sustainability. It was also hard to get coverage across an entire investment universe. As a result, this approach required some finesse and finagling. An analyst or portfolio manager might have to rely on his or her own instincts or insights about whether or not a particular company would fit into a given investment approach.
The third generation of ESG integration has now begun. The existing data sets have been broadened to improve their coverage and their providers have clarified their methodologies. New data sets are available that offer insights that were not previously available. A wider range of asset owners are requesting investment products that have sustainability-related claims. This has prompted the creation of passive ETFs, single theme funds (e.g., gender lens, decarbonized), and various types of hedge funds (including long-short and short-only offerings).
As is often the case with new theories, proof of their validity is not yet available. While many claims of investment outperformance, risk avoidance, and social impact are being made, few participants in the ESG space seem able to share evidence that supports these claims. Academic studies are lagging far behind. Most seem to still be mulling era 1 or 2 issues. This is not an unusual situation for the money management market. Many past “hot” investment ideas have turned out to be money-losing duds.
It is probably impossible to list all of the themes that are currently being pursued. There seem to be hundreds of competing theories for how best to generate and use ESG data. Here are few of those that have received the most attention:
|
Theory |
Why It Might Work |
Issues and Concerns |
|
Machine Learning |
Natural Language Processing can look for signals of ESG-related opportunities or risks. By going outside the scope of most traditional ESG data sets, these systems offer a chance to trade ahead of the market. |
Only a small number of companies have frequent signals. Both false positives and false negatives are hard to identify in advance. Only a limited number of investors can use a system before its benefits would be arbitraged by the market. |
|
Materiality |
Certain ESG factors may be tied to a company’s success. An investor can combine data on these issues with traditional financial and market information to get a better long-range view of company’s future performance. |
Various groups have attempted to identify which factors are material. However, their assessments disagree and there is little empirical support for any of these systems. In many cases, only a few companies report each factor. This makes it hard to do systematic research or to make consistent decisions across an entire industry. |
|
Engagement |
Invest in companies that have weak ESG performance. Engage with them to improve their policies and reporting. Benefit from the increased attractiveness of the company to other ESG-interested investors. |
Companies may not respond well to pressure from investors on business-related matters. It may take several years for the benefits of ESG-related changes to take effect and be noticed. Most investors don’t have such a long-term investment horizon. |
|
Factor Analysis |
Dump ESG data into a quantitative model and uncover significant factors. Structure a portfolio to take advantage of the results. |
Given the lack of data (most ESG indicators are not available for most companies) and the inconsistent way that ESG data is generated and reported, the quality of ESG data may be too poor to use in quant models. ESG factors may not be stable over time, as they are driven by social issues and current topics. |
|
Passive |
Include ESG factors in the list of things that can be used to “tilt” a portfolio. Position the resulting portfolio as attractive to groups of investors who care about a particular ESG-related theme. |
The restrictions associated with tilting portfolios generate tracking errors and can increase the costs of managing the portfolio. This may cause passive ESG funds to systematically underperform their benchmarks. |
|
Aggregation |
Combine together a lot of different ESG data sources. Create a new rating that incorporates the information from the underlying sources, but has broad coverage and an improved ability to predict future market performance. |
Averaging disparate sources of ESG data does not give good results. (This was tested in era 2.) A new method for aggregation is required—one such as CSRHub that uses Big Data methodologies to properly weight and combine a range of sources. The resulting ratings may not contain alpha (but could provide an estimate of consensus that could be used to generate alpha via other means). |
It takes at least three to five years to determine if an investment approach has promise. It takes another ten years to be sure that the approach will survive the test of market cycles and changes in market structure. The first and second eras of ESG integration did not produce any huge winners or star funds. The third era has many new ideas and approaches. Even without solid academic foundations, we can hope that one or more of them turn into a mainstream path for ESG integration.
To learn more about CSRHub, our ESG/CSR metrics or how you can improve your ESG scores, contact us here.
Bahar Gidwani is CTO and Co-founder of CSRHub. Bahar has built and run large technology-based businesses for many years. Bahar holds a CFA (Chartered Financial Analyst) and was one of the first people to receive the FSA (Fundamentals of Sustainability Accounting) designation from SASB. Bahar worked on Wall Street with Kidder, Peabody, and with McKinsey & Co. He has founded several technology-based companies and is a co-founder of CSRHub, the world’s broadest source of corporate social responsibility information. He has an MBA from Harvard Business School and an undergraduate degree in physics and astronomy. He plays bridge, races sailboats, and is based in New York City.
CSRHub is the largest ESG and sustainability rating and information platform globally. We aggregate 230 million data points from 650+ data sources including leading ESG analyst databases. Our patented algorithm aggregates, normalizes, and weights data to rate 18,000+ companies in 141 countries across 134 industries. We track 97% of world market capitalization. We cover 12 subcategories of ratings and rankings across the categories of environment, employees, community and governance. We show underlying data sources that contribute to each subcategory’s ratings. CSRHub metrics are a consensus view (any 2 sources may have about a 30% correlation so we make sense of the disparate data). We tag companies for their involvement in 17 Special Issues. We provide Macro-enabled Excel dashboard templates, customizable dashboards, and an API. Our big data technology enables 85% full coverage of data across our rated companies and robust analyses. We provide historical ratings back to 2008.
