How data science helps sustainable investors
How data science helps sustainable investors
Data science is increasingly driving investment analysis through the lens of environmental, social and governance (ESG) factors.
It’s been an established part of Schroders investors’ toolkit since the Data Insights Unit (DIU) was established in 2014. The team uses new alternative and unconventional data sets to build investment tools and provide a research service to investors across the firm.
From how it helps us track thawing permafrost to our analysis of the make-up of company boards, these are just few examples of data-driven sustainability insights in practice.
What can data science tell bond investors?
The data insights unit has worked with Schroders’ US municipal bonds team and sustainability experts in North America on county-level data since 2019.
One of the outcomes of this collaboration is our Schroders Municipal US Sustainability Explorer (MUSE).
There are thousands of issuers around each state in the US, each with a unique profile from a fundamental perspective, but also from a sustainability standpoint.
This proprietary investment tool allows analysts to access dozens of data points across ESG factors and assign an overall sustainability score to over 3,000 counties in the US.
What makes this tool unique is that it affords us a forward-looking lens into the population that supports the municipalities that comprise each county.
For example, the below chart below shows the differences in MUSE sustainability rankings for each county across the US. Green shading reflects high sustainability rankings, red reflects low rankings.
David Knutson, Head of Integrated Research for the Americas, says: “The pandemic has made credit analysis even more challenging. The longer-term negative impact of the shutdowns means at some point certain state and local municipal issuers will face difficult challenges. This research model has allowed us to compare issuers based on factors that could impact on an issuers’ ability to maintain itself, such as access to healthcare.”
What can data about the make-up of company boards tell us?
Looking at company governance is another important part of active, long-term investing.
Investors can easily determine who is on a board, how long they have been on it and their background, but we have a tool which allows for the creation of peer groups of similar companies across industries.
It is important to compare how companies measure up against each other in a number of different metrics, such as what proportion of the board is female and what other commitments board members have.
It also allows investors to see who has worked together in the past and for how long. This can be a positive attribute for a board and lead to greater cooperation. However, it can also be a signal to us to watch out for the possibility of so-called “group-think”, which can be less positive for company governance.
Moreover, analysis of this kind of data can lead to otherwise obscured insights. For example, a new board member coming from a completely different sector or background to the company could signal that it is considering moving into a new sector or providing a new service.
We applied this analysis to one of the UK’s big banks. The resulting chart below shows how the individual company compares on each of the board data metrics based to the aggregate of a peer group (in this case UK banks in the FTSE 350). The visualisation makes it easy to compare this particular bank to its competitors.
Figure 2: Data insights for a UK bank’s board
For example, in comparing the bank with its peers:
- Average sector experience is approximately 1.4 times greater than the peer group
- This board is larger than the average, as there are 1.2 times as many directors serving on the board compared to the peer group
- The data suggests the board has been relatively stable, given its greater score on the “years since last board appointment” metric.
- The bank has fewer female representatives on the board than its peers do
- Board members dedicate slightly less time to their board responsibilities compared to peers
- The tenure of the current chair is much shorter than for peers.
How does data prompt increased engagement
As an active manager, contact with the companies in which we invest is a daily occurrence and an important part of sustainable investment. Here are two instances where the use of data science has led to deeper questioning or offers of information.
1. The supermarket chain with unrealistic expansion claims
As part of an initial public offering (IPO), a supermarket chain promised that it would double the size of its network in five years by opening 1,000 new stores a year. This seemed a very ambitious strategy. Investors wanted to know if the targets were realistic, especially as it is important to pay fair value for a security at the IPO stage.
Using geospatial data science, we mapped the existing store locations of the chain and of its competitors around population centres. This allowed us to assess whether there would be demand for expansion or if the competition was already too fierce.
As an example, Figure 3 below illustrates our findings for Turkey. The bubbles indicate number of stores in that area – so the bigger the bubble the greater the number of stores in that spot. The colour of the bubbles indicate whether these stores belong to the supermarket (dark green) or competitors (dark purple). This map makes it easier to identify which areas you’d be cannibalising your own store front if you expanded into them, areas where there would be too many competitors and areas where there could be opportunities for expansion.
Figure 3: Geospatial mapping of supermarkets in Turkey
Based on our analysis, we calculated that the company could realistically only open a maximum of 3,000 new stores over the five year period. This is 2,000 fewer than claimed. The company’s management was then asked how its targets and calculations measured up against ours.
This type of analysis can lead to a deeper conversation with company management as well as being a prompt for greater disclosure over other elements of operations.
2. The Russian oil, gas and gold mines at risk from thawing permafrost
We created a tool to analyse weather and geographic trends in oil and gas production in Siberia – particularly thawing permafrost.
Risks from this include the release of bacteria and viruses from the ground. The release of gases such as carbon dioxide and methane into the atmosphere contributes to climate change.
Thawing permafrost represents a significant physical climate risk to companies operating in permafrost areas. For example, as soils shift, infrastructure is threatened and floods become more likely.
Our tool covers the main companies that operate in Russia and the locations of all their oil, gas and gold mines versus surface temperatures in those locations. It means our investors are better informed in their engagements with those firms at risk.
The figure below shows that for the first company’s mine located in the region of Tambeyskoyo, only 44.4% of days in 2020 experienced temperatures below zero degrees. This is substantially lower than the 75.3% of days in 2019.
Investors can then use this knowledge to inform their view of the risk that thawing permafrost may pose to this company.
James Ferguson, Research Analyst Emerging Market Equities, says: “The DIU’s dashboard provides us with insights that are hard to replicate. It helps us track the rate of permafrost thawing in different locations and identify which companies could be at risk from the environmental shift. As a result of this analysis, we have engaged with six companies that operate in at-risk areas to understand how they view this physical climate risk”.
In summary: the increasing adoption of data science techniques in tandem with the rise of ESG
Quantifying the potential impacts of sustainability risks is becoming ever more important to delivering sustainable investment returns for our clients. There are many ESG metrics and an ever-growing pool of ESG-related data which can provide rich insights for our investors.
At Schroders, we believe that unconventional data should be used alongside so-called “traditional” data – the quantitative data from sources such as financial statements, management presentations and press releases – as part of their fundamental investment process. These more advanced techniques can be used to analyse sectors and offer evidence or research points to help investors better actively challenge and encourage companies.
Unstructured Learning Time
- 16 years of returns: History's lesson for investors
- Omicron pushing US Fed into action
- What’s driven upgrades to our long-term return forecasts?
- Sustainable Investment Report Q4 2021
- Could bust follow UK’s post-Covid boom?
- Podcast: The crisis even Elon Musk can't afford to solve
The views and opinions contained herein are those of Schroders’ investment teams and/or Economics Group, and do not necessarily represent Schroder Investment Management North America Inc.’s house views. These views are subject to change. This information is intended to be for information purposes only and it is not intended as promotional material in any respect.