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.

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).

Figure 1

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).

Chart 1

Chart 2

Chart 3

It could be that operating in regions with evergreen forests and perennial ice and snow leads to lower operating and capital costs associated with water treatment and reservoirs, or the results could be driven by other factors, such as local economic growth (see the "Related Research" section below for relevant articles on water utilities in California and the Rocky Mountains).

Nevertheless, our analysis demonstrated that some ecosystems appeared to be positively linked to water utilities' strong financial performance. For example, the proportion of evergreen forest in a county's land cover was consistently among the top most important variables in the analytical models. This finding is consistent with the existing scientific literature, which has argued that forests in the northern hemisphere naturally absorb nitrogen pollution and improve water quality. We observed a similar positive link for the utilities in countries with perennial ice and snow cover.

Perennial ice and snow are a feature of the western mountainous regions of the U.S. During winter, precipitation freezes and expands the ice and snow, which act as natural seasonal store of water. The drought in California from 2011 to 2018 was in part due to low precipitation during the winter months that then stressed the water supply needed in spring and summer. In addition, the geographic regions facing water scarcity are among the most economically robust, leading to growth in water demand that in turn contributes to scarcity risk. California responded to the drought with emergency regulation of compulsory water conservation, whereby residents were required to restrict water use, especially for landscaping.

Climate change will increase the severity of wildfires and shrink the area of perennial ice and snow, which could in turn undermine the credit quality of utilities reliant on these ecosystems. Mountains are considered the water towers of the world, and are particularly vulnerable to climate change. Currently, 1.9 billion people depend on water from perennial ice and snow ecosystems.

The findings of our study are unsurprising given the science linking the abovementioned ecosystems to improved water quality and natural water storage. However, our methodology--using satellite data and machine learning to supplement traditional methods of credit analysis--is unique, and may presage the future of our ESG analysis. This methodology might be useful for uncovering emerging risks to credit quality, while the issuer's management of emerging risks and local economic conditions remains relevant.

Spatial Finance Points To A More Sustainable Way Forward

The pace of change and innovation in financial services is accelerating in both the supply of finance and the demand for improved transparency. One new trend that has emerged is that of lenders tying interest rates to sustainability targets and overall ESG performance scores (see "Why Linking Loans To Sustainability Performance Is Taking Off," published Sept. 3, 2019). Tying interest rates to environmental and social performance creates the demand for independent verification that targets and performance goals are being met.

Governments are using satellites to monitor illegal forest clearance and illegal fishing. However, deforestation and overfishing are not always regulated or enforced, and therefore responsible investors are already starting to make use of tools derived from satellite data to meet environmental goals. In 2019, asset managers such as Robecco and ACTIAM deployed deforestation monitoring tools to support their own zero deforestation targets. Companies with highly exposed supply chains are also demonstrating their credentials by tracking trucks to and from processing mills and farms. While this strategy may not be the panacea to environmental woes and is not 100% foolproof, it is a clear trend we are seeing in the responsible investing market.

There are broader applications of spatial finance than sustainability. Other innovative applications include monitoring ports to get an indication of trade volumes, or using night-time light indicators to estimate local gross domestic product where data are not available. These applications will likely only become more sophisticated and integrated into the financial system.

Spatial finance is poised to provide investors and financial professionals with information on economic activities before government statistical accounts are made available and on emerging markets where less information is available, and to monitor assets and infrastructure projects almost in real time. After a slow start, spatial finance has finally sprung into action and is only set to grow in the next decade.

Related Research

• Clearing The Smoke: Which California Municipal Utilities Are Most Exposed To Potential Severe Or Catastrophic Wildfire Damage?, Nov. 20, 2018
• Why Linking Loans To Sustainability Performance Is Taking Off, Sept. 3, 2019
• Many Rocky Mountain Water And Wastewater Utilities With Strong Credit Profiles Face Growth-Related Infrastructure Needs, Dec. 18, 2017
• U.S. Public Finance Waterworks, Sanitary Sewer, And Drainage Utility Systems: Rating Methodology And Assumptions, Jan. 19, 2016
• California's Water System Illustrates The Near-Term Impacts Of Long-Term Climate Change, May 20, 2014

Appendix: Study Methodology

Our study used satellite data and machine learning that we believe represent the future of ESG analysis. We examined environmental factors mentioned in the scientific academic literature that have long been associated with protecting sources of drinking water. We leveraged data from the National Space Agency's Landsat satellite missions, county data on agricultural practices and water use, and traditional socioeconomic indicators that we factor into our credit ratings. Then, we assessed if these factors supported the strong financial ratios that feed into our credit ratings on public water utilities in the U.S.

We calculated the all-in-coverage ratios for 2,079 public water utilities in the U.S. for 2016, 2017, and 2018. These utilities supply water to 770 counties across 48 states. We calculated the median value of the all-in-coverage ratio for counties with multiple utilities. There was a positive skew in the distribution of the all-in-coverage ratios (see chart 4), which implies that most of the utilities we rate have strong ratios.

Chart 4

We processed geographical information on land cover from NASA's Landsat satellites using cloud computing capabilities provided by Google Earth Engine.

The statistical analysis in this study used a supervised machine-learning algorithm called gradient-boosting regression that uncovered the environmental factors supporting credit quality.