For nearly 40 years, the Bloomberg Terminal has helped the global financial and business ecosystem conduct research, communicate, and transact business together. The Terminal provides data, news, analytics, and other functions our clients need to evaluate, test, and ultimately act upon investment decisions. Whether those decisions are about making and saving money or mitigating risks such as climate risk or impacts on their supply chains, we help them do that efficiently and effectively.
Geospatial at Bloomberg basically covers three different areas of ownership. First, the data —everything starts with data. I maintain a team to keep hundreds of geospatial datasets. The second pillar is around insights, and so we deliver automated news and alerts based on the intersections of geospatial data. That’s a way to really communicate the value of ‘geo’ without having to go to the map. It’s a premium sweet spot for us. And then the third is building map applications. I have a flagship tool and a line of maps and other areas that have been integrated throughout the Bloomberg Terminal. At the end of the day, my group has a full suite — everything from data insights to applications.
Extreme weather is impacting financial markets in a number of ways, and in real time. We have traders, hedge funds and quantitative scientists trying to come up with strategies to buy stocks or change their positions in markets based on the forecasts. In the financial world, that becomes a money-making or money-saving decision in real-time workflow. The portfolio managers — or those managing a bunch of stocks or studying a bunch of companies — are looking at it from a risk perspective, where they want to understand how an extreme weather event could impact their holdings that have assets or supply-chain dependencies in the affected location. So, they are looking at it from a medium-term perspective — quarterly earnings or how their stocks are going to perform next year.
However, I also see a trend emerging in the financial industry around climate reporting. There are two main initiatives — both with acronyms. One is ESG — environmental, social, and corporate governance – where companies are categorized on the good they do for the world. Are they polluters? How diverse is their board? So, people are making investment decisions based on how companies are following regulations or making a positive social impact in the world. This means a lot of clients are looking at long-term data on companies from a sustainability standpoint.
The second is an initiative called Taskforce for Climate Financial Disclosures (TCFD), which is chaired by Michael Bloomberg. In 2017, the TCFD released recommendations on climate-related financial disclosurse that are designed to share relevant, forward-looking information that can be used in financial decision- making. Today, more than 1,500 organizations globally have formally expressed support for the TCFD recommendations. Just recently, the UK became one of the first countries to make TCFD-aligned disclosures mandatory. Given the widespread support for the TCFD, many companies are now interested in a lot of other climate and weather analytics in alerts and news delivery. We see a lot of demand in this area and take pride in this work. It’s like influencing the entire industry to use geospatial analysis for decision-making.
I was invited to collaborate with the United Nations Environment Programme Finance Initiative’s TCFD Banking Pilot Project to help banks develop methodologies for reporting climate-related physical risk scenarios. Specifically, our group established a methodology around physical risk in climate reporting using geospatial tools. That was done just a couple of years ago, and now it’s starting to become mainstream. More and more people are using mapping data and visualizations to both process their climate exposures and provide beautiful maps in climate reporting.
There are three types of clients — the analysts who do research, the traders who transact, and the portfolio managers who manage all of that research and transactions. At the end of the day, we are trying to embed ourselves as deep as possible within the workflows of these core audiences.
Geospatial data represents a new dimension for these audiences to analyze. I think we have found the most success with analysts who conduct research. They will go to a map and see a path of a cyclone or a wildfire, and then might say ‘I didn’t know there were three power plants near this wildfire. Who owns them, and how much risk may they face from the hazard?’. Similarly, for portfolio managers who may cover 20 different stocks or companies, they might have their whole portfolio shown on the map, tracking events like extreme weather or COVID, and trying to understand if their holdings are being impacted by these real-world geographic events. That’s how people have been using geospatial data in the Bloomberg Terminal.
Geospatial data has recently gained popularity in the financial world for three reasons. First, the use cases weren’t as relevant earlier; but in the last five years, the severity of these extremes has increased. Now, use cases are being sought to understand how financial markets are going to be impacted from the events. This year, we’ve seen a record number of hurricanes and tropical storms in the Atlantic. We also faced massive wildfires and smoke in the Western United States. Then there is the pandemic. These are having a heavy impact on the economy, and therefore, there is a pressing need to understand all this in real time.
Second, real-time data around these events is easily available, unlike in the past. Today, one can subscribe to data feeds from weather and government agencies and build a global solution around that. Solutions like that didn’t exist 10 years ago. I don’t think my group would have had the resources do some of the initiatives we are doing today a few years ago.
Finally, there is the globalized supply-chain factor. If there is a typhoon heading to an island in Southeast Asia that could affect manufacturing facilities, there is a ripple effect for that supply chain — whether it’s the stock price of that particular company, or those of secondary or tertiary companies that rely on the products made in those facilities — all are going to be affected somewhat.
Weather and climate events — as well as their supply chain impacts — are driving a lot of day-to-day usage, and that’s new for us. The commodity space at Bloomberg is perhaps the most traditional user of maps. We have a dedicated user base that looks at oil flow or tracks vessels around the world to predict the supply and demand of oil for different countries and companies. We also cover natural gas pricing, agricultural pricing, mining production, and more. We have many commodities with pricing datasets, and clients are, in some cases, working with weather metrics and trying to do different analytics.
I am most excited about the stock and equity user base because that’s the biggest audience in finance, and their use of geospatial data has really exploded in the last few years.
There is quite a story to this. In January, I was chatting with one of our colleagues in Singapore, and he directed us to sources of data that pointed to some interesting COVID developments. To stay ahead of the markets, my group launched a COVID mapping tracker as early as January 29. At that point, there were only 7,000 cases globally and maybe a handful in the United States. At that moment, we didn’t have a single request from a client for that kind of data tracking, but we were able to demonstrate to the financial world that we recognized the moment. The Bloomberg COVID tracker drove record usage and received a lot of recognition. In a way, it really influenced the rest of the Bloomberg Terminal team to take action and build in more COVID-related analysis.
As the months went by, we started to get inputs from clients demanding more data and analytics. For instance, there was demand for data around supply chain impacts, recoveries, and specific companies or factories. So, as soon as the data became available, we continued to add it in the Terminal. This helped our clients to research and understand the evolving situation.
It’s a combination of various things that enables us to process the data into compelling investment trading signals.
We make our data digestible and action-oriented by combining open source and in-house technologies. We use PostgreSQL extensions, such as PostGIS and TimescaleDB, to load data into a queryable form that allows our clients to derive time-series based insights. These insights can then be consolidated with data from public sources like OpenStreetMap. We also leverage Elasticsearch to optimize query performance. The result is a unified experience that combines information from multiple sources, regardless of their internal, external, public, and/or proprietary origins. Our infrastructure is deployed in Docker containers and orchestrated with Kubernetes, both state-of-the-art, open source, cloud native technologies with vast engineering communities to which we’re proud to belong.
Plus, we’re part of an organization that has an Engineering department that has grown to over 6,000 engineers globally. My team can leverage these resources to build optimizations, services, machine learning algorithms, and relationships across other datasets that Bloomberg has.
I am deeply involved with the Earth Observation community and the ancillary industry. I think the biggest challenge for satellite imagery is still the cost. When you talk to the financial community, there is a high appetite for geospatial data. While on one hand I see this demand continuing to grow, on the other hand I see challenges remaining with it — for instance, the skillsets required to process that imagery.
I wish I had the full answer to that. What I can recommend is more research and analysis that can be done with some of the weather and climate data that my team has collected. As more people quantitatively backtest and see if there is any impact in the prior events, we can gain a better understanding and communicate more insights to users. We are at a moment where the use cases, the data with history, and the technology to process and backtesting are just starting to converge. My mission is to communicate that this type of analysis is now possible and available here at Bloomberg. We plan to work with academia, scientists, government agencies, and clients to see where they are finding impact and changes and to build more solutions.
We have published two white papers – which are publicly available – that show the relationships of tropical cyclones’ impact on manufacturing companies and the impact of snowfall on retailers. In both cases, we have actually found real-time, real-world financial impact with sharp ratios and technical studies that can inform a trader or an investor when to buy and sell stocks, as it relates to those type of events.
I have spoken to hundreds of clients around the world ranging from ones in the quantitative data science field to the much more technical GIS/geo users and learned these big hedge funds and quantitative shops are using more and more programmatic resources or financial analysts who write code or are data scientists.
When I really dug deeper, I discovered that many in the financial industry found geospatial data to be too messy or unstructured to be handled. The data points weren’t connected to a company identifier or didn’t have a history. So, all the things that a traditional financial person wanted was difficult to find. Since they ended up spending too much time looking for it, they moved onto the next type of alternative — or non-financial — data. This has evolved over the last few years, and we are getting more demand for geospatial data. However, the same challenges continue to exist.
We, the geospatial industry as a whole, need to think about history and the point-in-time aspects of geospatial data. The number of use cases from point-in-time data really expands the spatial analysis capabilities from what we are looking at now to a much deeper insight. For this type of technology or data to continue to emerge in the financial sector, we need cleaner data that is more accessible and has a temporal history correlated with location data.
Three thoughts come to my mind — big data, real-time, and moving beyond the map. The map is nice, but connecting the data is powerful. And that’s where I look to use geospatial technology and the infrastructure that I have invested in to build new relationships in finance — to derive and build new insights that aren’t being reported by anyone or anywhere else. And we will deliver those insights without a map; instead it could be an alert, a text message, an email, a data feed, a chart, or a textual news story. There are a lot of ways to deliver the value of geo without just a map, which is what a lot of people can’t think about. That’s the core of my mission.
The other two aspects – analyzing all this big data in real time, and then delivering those insights in real time — is the real challenge. Add to that the ubiquitousness of the small sensors that you are talking about, and it’s going to take the geo engine and geo analytics to a place that we can’t even imagine today. I think the use cases that come out of real-time data, big time data, real-time alerting capabilities are the kind of geo analytics that will become the core of most functionalities.
I am really excited because I think businesses are starting to recognize in totality the value and the foundational framework that geospatial data and analysis bring to their organization and decision-making.
That’s an interesting question. I learned that when there is volatility, especially a crash, our clients need data and insights more than ever and faster than ever. We find that companies and industries in the field of delivering insights and information to users who are trying to save money, can do well in these times. And our demand skyrockets.
It’s hard to prepare for a crazy, volatile moment, even though we responded quickly to COVID. It’s important to be agile, to react and to understand how the markets are changing from a business or a product standpoint, to be able to reposition ourselves, and listen to the market.
My background in geography started about 20 years ago when I joined the United States Air Force as a ground radar technician. My mandate was to find the optimal places to set up a new radar sites and maintain their operational readiness. During my military career, I had several real-world deployments, including the Iraq War in 2003. Those are some real life experiences for me that were very geographically related.
After my time in the Air Force, I wanted to continue learning about geography. Frankly, it was also good timing — Google Maps had just come out in 2005, and GPS and Garmin were starting to gain popularity. I went to James Madison University and received a degree in geographic science with a focus on GIS. While attending college, I took those tools and real-life experiences to work for a local city where we enabled GPS in E911 database.
Then I took that skill to the financial industry. The degree and my background made me in high demand from government contractors and agencies, but I purposely went towards the financial community because I saw that there was no solution in the market that leveraged geospatial insights. I was hired by SNL Financial in 2008. I pitched my way into the company by explaining what mapping and GIS can do for them, and they created a position for me. I built a web-based map application, which we launched to Wall Street in 2009. It was the first big curated map app for the finance world. Of course, there was Esri and MapInfo and other platforms like that, but there wasn’t a bespoke solution for banking, insurance, media, metals, and a whole variety of industries. It’s been a long road from that point to now, where people are starting to take on this type of information.
I was hired five years ago by Bloomberg to build a real-time, data-feeding analytical tool for the financial industry. My dream today is to introduce a second form of orientation in decision-making for the financial community. This community has been doing time-scale analysis with charts for a long time but missing the spatial analysis component. I am looking to make spatial analysis a key part of the financial community’s analysis and then bring all that time and space together to generate insights.