The belief that more information leads to better insights is eroding.
Today’s insights come from finding signals amid the “noise” caused by excessive amounts of information. Spatial finance whereby geospatial data informs financial practice is one approach poised to uncover such signals. Three structural trends are driving the rise of spatial finance: climate change, technology and machine learning, and the demand for environmental, social, and governance (ESG) performance indicators.
Climate change is altering the geographical dispersion of certain environmental risks, and geographically diversified assets will need to adapt to this new reality. Generally speaking, geographical diversification can mitigate the capital risks associated with climate change. However, if the geographical locations of assets correlate with the locations of climate risks, diversification alone is insufficient to mitigate the risks. For example, if a structured financial product has exposure to mortgages across numerous states in the U.S., it can be said to be geographically diverse. Yet if all the mortgaged properties are in low-lying coastal regions with high exposure to storm surges and sea-level rises, then the climate risks remain.
In the next decade, alternative environmental data from satellites and geotagged social media data could be useful for financial analysts to better understand risk. S&P Global Ratings has conducted a study demonstrating this. We used satellite data and machine learning to assess whether there was any link between the locations of U.S. public water utilities and their financial performance. We found that public water utilities located in U.S. regions with evergreen forests and perennial ice and snow had better all-in-coverage ratios than utilities located elsewhere. In other words, ecosystems that maintain good water quality and facilitate natural water storage also appear to support the debt metrics of the U.S. public water utilities we analyzed. We believe that the techniques used in this study represent the future of our ESG analysis.
“We found that public water utilities located in U.S. regions with evergreen forests and perennial ice and snow had better all-in-coverage ratios than utilities located elsewhere”
Climate change and spatial finance
Climate change is a global priority challenging our financial systems. This is partly due to the 2015 Paris Agreement, which, in the words of the UN, aims “to combat climate change and to accelerate and intensify the actions and investments needed for a sustainable low carbon future”. In addition, some financial regulators are becoming increasingly vocal about the need to integrate climate change risks into investor decision-making. For example, Christine Lagarde, president of the European Central Bank (ECB), wants climate change to be “mission critical” for the ECB. However, there is little consensus on how the ECB will achieve this.
Mark Carney, a long-term proponent of action against climate change, will become the UN’s special envoy for climate action and finance after stepping down from the Bank of England. His responsibilities include channeling private finance toward opportunities that support action against climate change. While the policy response will play out differently depending on the region, it is not the only driver of financial risks that companies might face related to climate change.
In May 2019, the Bank of England published a framework for assessing the risk of physical climate change (for example, droughts and floods) for financial professionals. The framework suggests using hazard maps and catastrophe models as tools. The insurance industry already uses geospatial natural catastrophe models to calculate the probability of loss associated with environmental events like floods. Outside insurance, however, there is limited inclusion of geospatial models in other areas of finance.
Environmental and social issues often need to be examined within the local context. For example, if a company uses lots of water in a region where water is abundant, there is no scarcity risk.
However, if the region is becoming increasingly water stressed, water scarcity becomes an emerging risk. Companies and investors are calling for better information to support commitments to reporting according to the Taskforce for Climate-Related Financial Disclosures. There is often a lack of reliable data available for investors with exposure to emerging markets, and satellite data combined with machine learning could help to bridge that gap.
Technology and machine learning are making spatial finance “Mission Possible”
The ability to conduct spatial analysis is increasingly possible for financial analysts thanks to the proliferation of spatial information and data, cloud computing, and machine-learning algorithms.
The cost of earth observation–sending satellites into space to monitor terrestrial conditions–has been steadily declining. Euroconsult forecasts that the market for earth observation data and services will reach $12.1 billion by 2028, and 52 countries will have at least one earth observation satellite in orbit.
Governments with advanced space programs are now generating additional revenue streams by exporting satellites to other countries, such as Japan’s recent deal with Vietnam to supply weather satellites to support Vietnam’s natural disaster prevention efforts. Commercial enterprises are also investing in this space, including those launching satellite constellations that monitor the earth’s entire surface. All this spatial economic activity is generating minable information that could have value for financial analysts.
Cloud computing enables analysts to process large calculations that previously would have taken too much time and computing power. Newly available engines such as Google Earth Engine and ArcGIS Pro allow for spatial analysis using large amounts of computing power to be done online via the cloud. We expect computing power to continue to improve in the next decade.
To extract value from this growing source of information, analysts need the right tools. Machine learning and artificial intelligence are now more accessible to analysts with limited programming experience. The financial sector is still experimenting with applications of these new tools, and spatial finance is one of the new frontiers that may enhance financial, as well as our ESG analysis.
A new way of doing ESG analysis: linking nature to debt metrics
To understand how the physical impact of climate change translates into a credit impact, it is important to consider how environmental factors affect credit metrics. To this end, we conducted a study using satellite data and machine learning to analyze public water utilities in the U.S. We found that utilities located in regions with evergreen forests and perennial ice and snow had greater all-in-coverage ratios than those located elsewhere (see charts 1 and 2).
The all-in-coverage ratio is a ratio of freely available cash to debt service, recognizing fixed costs as debt-like in nature (see chart 3). This ratio is an important driver of a utility’s financial profile, with a higher all-in-coverage ratio indicating stronger credit quality (see “U.S. Public Finance Waterworks, Sanitary Sewer, And Drainage Utility Systems: Rating Methodology And Assumptions,” published Jan. 19, 2016).