FactSet, Snowflake Take Aim at Financial Data Mapping Challenges

FactSet will this week unveil Concordance, a new service for performing outsourced data mapping of traditional, alternative, and clients’ proprietary data. It leverages its existing operations and technology, as well as its relationship with cloud provider Snowflake to provide mapping of data stored in mutual clients’ cloud environments.

The Concordance service aims to link standard reference data to alternative data and a firm’s own internal datasets—in the same way that FactSet already links identifiers and other metadata to “traditional” market data—so that all relevant information can be mapped to the companies and securities it relates to.

For any data—such as new “alternative” or proprietary client data—that does not immediately match the reference data in FactSet’s security master file, clients can submit the unmatched data via the vendor’s Concordance API, and the vendor will accurately link it to reference data using AI algorithms, then send it back to clients with the metadata required to link all of a client’s data with minimal effort. The data matches are checked and validated by FactSet’s data teams, who also manually link any data that does not match automatically.

Once data is mapped and linked, it can be used to power clients’ security masters, data lakes in the cloud, as well as creating new usage opportunities, such as an input to customer-relationship management (CRM) systems, Jonathan Reeve, global head of content and technology solutions at FactSet, tells WatersTechnology.

The big challenge to achieving this is that many firms have assembled proprietary datasets which have value but do not conform to FactSet’s reference data. Previously, to use its own data in conjunction with FactSet’s market data, a firm would have needed to link the content manually, which—at a conservative estimate—could take one minute per company, whereas the new service aims to match up to 80% of data instantaneously, with the remainder matched manually by FactSet’s operations staff.

“Think of the service as concentric circles of capabilities: At the center of the circle is a service we call Data Management Solutions—a reference data file that allows you to connect securities with all standard identifiers and reference data, full descriptions, our FactSet ID, other vendor identifiers, and entity/issuer identifiers, so that you have the relationship between a company and the securities it has issued,” Reeve says.

Having accurate reference data allows users to connect and search for companies across different industries and sectors, he says. “The next ring of the circle is our Open:FactSet platform, where we bring in ‘alternative’ content… and link it all. That allows people to navigate not just ‘traditional’ datasets, but also new data types,” he adds.

For example, one company already using the service is an alternative data vendor that participates in the Open:FactSet Marketplace of alternative data. This vendor scrapes content from the web, and in the past needed to source and link reference data to its content by hand to make the data usable. Now, the vendor can achieve that almost immediately using the Concordance service, Reeve says.

FactSet will offer Concordance as a standalone service, but Reeve believes it offers even greater value when applied to data being stored and shared in Snowflake’s cloud. “When you put that in Snowflake, you have the ultimate data-linking capability in the ultimate data-sharing platform,” he says.

“This mapping problem—whether internal or between different providers—plagues every industry, but in particular financial services,” says Matt Glickman, VP of customer product strategy at Snowflake, who spent almost 25 years at Goldman Sachs prior to joining Snowflake in 2015. Addressing the data mapping challenge is as “strategic” to Snowflake as expanding its ecosystem of data and services, because it will help clients not just integrate but also search for and identify potential services on its cloud.

“I could take a dataset that uses natural-language extraction to generate sentiment, and tie it to classification data in FactSet, and then I have a classified dataset that I can join with other data I have from FactSet within minutes or hours, without having to undertake a big project,” Glickman says. This enables firms to use the data not just to support trading and investment decisions, but also to support new business development.

For example, another potential—and non-traditional—use-case for Concordance is that of an unnamed European financial firm that uses it to augment client data in its CRM system with reference data and other information, so that the firm can link FactSet’s market and fundamental data to its database of customer data.

“Everyone has client data in a CRM system,” Glickman says. “What if you could enrich your CRM system to identify which clients are ESG (environmental, social, and governance)-friendly, or which companies in a client’s supply chain have experienced Covid-related earnings disruptions?”

“It transforms the client’s CRM platform from being a call-tracking system to a lead-generation system,” Reeve adds.