We recently sat down with Mike Lubansky, Dun & Bradstreet’s Leader of Alliances Strategy, focusing on Capital Markets, Compliance, and Tax Solutions, to talk about FIBO, the increasingly important data standard being developed in the financial industry.
In our first blog, Mike shared his views on Financial Industry Business Ontology (FIBO), its value and importance in Capital Markets and the challenges firms face without a standard. He then discussed how FIBO enhanced with third party entity data helps these companies do risk aggregation for meeting regulatory requirements.
Read on for a deeper dive into making analytics comes to life for financial institutions with an ongoing proof of concept for a major US financial institution, and what data visibility possibilities exist in the future for this industry.
You recently participated on a panel at the Enterprise Data World event titled “Harmonizing Diverse Derivative and Entity Data for Powerful Analytics using FIBO.” Can you share an overview of the discussion?
The discussion was very consistent with what we covered so far in this Q&A (blog 1):
- The emerging role and importance of ontology and linked data, not just for FIBO but also for data management best practices overall
- The utility of the FIBO model, and how integrating FIBO into client environments becomes one step easier when data is already mapped to the standard as it emerges
- The importance of FIBO and tying business entity data to transactions for greater insights, risk analysis and aggregation
- The kind of entity data that’s important including core business data, business hierarchy connections and the increasingly adopted LEI, all essential for assessing counterparty risk and exposure
- Lastly, we discussed a Proof of Concept (POC) currently underway for one of the oldest financial institutions in the United States, and being conducted with Dun & Bradstreet and Cambridge Semantics. The focus of the POC is to determine the value of using both FIBO and external third party entity data to tie into their interest rate swap transactions and perform risk analysis.
Very interesting. So you are actively involved in testing the FIBO standard with D&B data for this major bank’s risk analysis needs. Can you expand more on the POC, the approach and findings so far?
For background and what got us where we are today, the EDM Council has been sharing the development of FIBO with the Capital Markets industry. It caught the attention of the bank’s Chief Data Scientist and head of data governance, who wanted to learn what value FIBO could bring to their internal data management practices. The Office of Financial Research also expressed interest in seeing what value FIBO could provide in practice. As a result, the bank partnered with the EDM Council to do a FIBO POC focused on interest rate swap transactions.
The first phase of the POC included a small number of interest rate swaps. It focused solely on ensuring the bank could map to FIBO, attach meaning to fields and easily bring the transactions into Cambridge Semantics, their semantics platform.
The next phase involved bringing in larger number of records to do analytics. The scope also expanded to connecting interest swap transactions to business entity data, including externally sourced data. This is where Dun & Bradstreet (D&B) got involved.
- D&B’s Partner Innovation Center used the ontology mapped to the FIBO business entity model to establish master records, connecting the D&B D-U-N-S® Number to a counterparty involved in transactions. Then the master records were enriched with a set of data about each counterparty, including legal ownership data.
- Cambridge Semantics loaded the data and mappings into the Cambridge Semantics Anzo platform. All the data types were harmonized using FIBO mappings.
- Cambridge Semantics and the bank developed graph database visualizations including swap transactions matched with entities and related entities. The graph database connected well with the concept of FIBO for viewing inferences and connections revealed by applying ontology to linked data. Not just rows and columns of data, the graph database made it easy to query vast webs of interconnected data sets.
- The bank was able to get a clear picture of aggregated exposures and see the interconnectedness of counterparties with D&B’s corporate linkage. Dashboards and analytics included aggregated country risk exposure from different counterparties, heat maps showing transitive exposure and total exposure by the top most parent of a corporate family tree.
This has been the initial work.
So what’s next?
The specific scope of the next phase is still TBD. The team is interested in exploring more proof points:
- If this mapping allows for the harmonization of other data sources and types of transactions like credit default swaps
- Other types of entity linkage, for example related companies with less than 50% ownership. (The first phase was just legally owned corporate family relationships.)
- The connection between LEI and D-U-N-S Number, as another form of entity resolution and to help in regulatory reporting where LEI is becoming a requirement
- Additional platform dashboard views for analytics and insights that can be drawn from loading transactions and other data about related entities onto the platform
- Testing scalability by loading a greater number of transactions onto the platform
- Benefits and speed of enhanced inference capabilities inherent in graph database technology as compared to traditional relational databases
There’s a lot of opportunity ahead for FIBO. Banks and regulators are paying attention to the POCs being done. The EDM Council is encouraging interim steps such as FIBO-Vocabulary to encourage early adoption as the FIBO standard is built out across the different working groups for the various types of instruments, entities and other content that would benefit from application of ontology.
Additionally, FIBO is converging with other trends in the financial industry such as blockchain, a distributed ledger technology that holds promise in many applications. As this technology develops, it will be important to apply standards. FIBO fits the need for application of standards in this emerging technology.
As the FIBO standard gains greater adoption, communication within and across banks and with regulators will provide enhanced transparency, analytic insights and reporting efficiencies. Over the long term, FIBO will help the entire financial system in preventing a financial collapse similar to that of 2008.
Learn more about Dun & Bradstreet's perspectives and the types of data that organizations in Capital Markets need to help them make better decisions.