Is sustainability data a constructive input in the quant process?
Bart van der Grient: ”We believe investors can benefit from the use of sustainability data in various ways. In terms of risk, climate change is a growing concern for investors, for instance. For one, we can take measures to restrict the carbon footprints of our portfolios to guard against carbon risk. Regarding stock selection, we do find that corporate sustainability performance indicators can tell us something about returns, if you look at the right segments of the market and focus on financially material issues.”
Our decarbonized value signal provides a good illustration of risk mitigation.
Kristina Ūsaitė: “Our decarbonized value signal provides a good illustration of risk mitigation. If you look at conventional value strategies, based on book-to-price and earnings-to-price measures, you typically find portfolios that consist of many ‘brown’ companies. This is intuitive as these metrics are typically tilted towards asset-heavy sectors like energy, materials and utilities.”
“To address this, we designed an innovative methodology that adjusts the valuations of high-polluting firms by making them less attractive, based on their environmental footprints. We found this results in a ‘greener’ value signal that removes a large tilt to ‘brown’ companies at the stock selection stage. As a result, further environmental constraints at the portfolio construction phase are potentially more easily satisfied. We also observed that this approach hardly impacts the value premium.”
Sebastian Schneider: ”In our models we use ESG signals that we have established to have some stock selection power. That said, we continuously look to improve the robustness of these signals. The one issue we face is that most data providers have short datasets and their sample periods coincide with the rise in sustainable investing.”
“As a result, we are performing robustness checks to see if the recent history of increased sustainable investing has driven multiple expansion in stocks that are ESG leaders. If we can pick up a premium from ESG signals that is not driven by multiple expansion, then we can be more confident in the robustness of those signals.”
What are some of the interesting areas in your research agenda?
K.Ū.: “We are looking into ESG sentiment data that is compiled by various providers. Our team is conducting an investigation into whether these datasets can help us predict ESG score updates, or even provide alpha or distress signals. To do this, we are assessing if there are any differences between stocks with positive and negative sentiment, either in outperformance or risk-adjusted returns.”
“But the jury is still out. Firstly, these datasets are relatively short. Secondly, there could be a lot of noise in the data. Therefore, we have to scrutinize if financially material ESG issues are reflected in the sentiment scores. Although we are still working towards a final conclusion, the initial results suggest there is no strong evidence of alpha signals. Some of the results are also short-lived and may not be easy to capture.”
Bart van der Grient
We are also doing research on how to enhance the integration of SDGs in our process.
B.v.d.G.: “There is increasing interest in impact investing and how investors can support the achievement of the United Nations Sustainable Development Goals (SDGs). A recent research project focused on a quantitative application using both numerical and textual datasets from various data providers. The researchers used sophisticated methods to analyze the text in order to assess the activities of companies, and to see if they could subsequently draw any links to the SDGs. This has provided new insight on alternative ways of assigning SDG scores to companies.”
“We are also doing research on how to enhance the integration of SDGs in our process. For instance, if you build a portfolio that targets specific SDGs based on an impact theme, how does this affect its factor exposures, the extent to which its carbon footprints can be reduced, as well as its general behavior and characteristics? These are some of the questions we are addressing. This will allow us to apply research to translate client sustainability preferences into live portfolios.”
Are there any intriguing developments in terms of forward-looking data in this space?
S.S.: ”We are currently exploring forward-looking climate metrics. Current data like carbon emissions tend to be more backward-looking. Several data providers are attempting to measure climate-related risks and opportunities associated with companies, based on scenarios that have yet to materialize. The relationship between these metrics and valuations is just one angle we are viewing it from. What we see in some of our early research is that higher climate-related risks are reflected in lower valuations on an industry level.”
K.Ū.: “Information from patents could be a useful signal. As an example, high-polluting companies could be investing in research to generate innovative ideas to decarbonize their operations. These details could be picked up in green patents. In fact, there are some data providers that link patent database information to the different SDGs. So you could have a forward-looking perspective on how companies could contribute to the SDGs.”
B.v.d.G.: “In general, there is a lot of forward-looking sustainability data, but it is not always tangible for our process. For instance, there is a lot of information on policies and sustainability targets, but in the end, we need to be able to quantify and measure a company’s performance. In the future, we believe there will be a higher focus on incorporating a combination of forward-looking and quantifiable data, which will give investors more information about the opportunities or risks inherent in a sustainable solution.”