MarkLogic and ModelDR are challenging traditional approaches to BCBS 239 compliance with a solution that combines MarkLogic’s non-relational, multiple model database platform and ModelDR’s data point modelling software. The companies detailed their approach in a recent webinar, Sustained Compliance with BCBS 239, and said banks implementing the solution could lower costs, increase agility and ensure data quality while achieving compliance.
The webinar was moderated by Diane Burley, chief content strategist at MarkLogic, and joined by experts Chris Atkinson, solutions architect at MarkLogic, and Greg Soulsby and Simon Roberts, directors at ModelDR.
Atkinson started the conversation explaining the limitations of traditional data management and regulatory compliance processes. He said: “The most common approaches are to coalesce as many data sources as possible into a data warehouse, or to stitch together siloed data. A data warehouse needs a canonical data model that is very rigid. Silos offer more agility, but their cost is significant and if siloed data is put into a data warehouse it becomes brittle. Alternatively, multiple feeds can be sourced many times and calculations can be performed many times. In this scenario, consumers become servers and sources of data to others, but this is expensive, rigid and it is difficult to get a congruent view of data.”
Noting the need for banks to achieve BCBS 239 compliance, but also reduce costs, increase business agility and improve data quality – all contributors to a lower total cost of ownership – Atkinson highlighted a data point modelling solution that is used by the European Banking Authority to turn regulations into specifications. Banks, he said, could also gain from using the solution, which is relatively quick to implement and provides ongoing flexibility and agility for regulatory responses.
Looking at the practicalities of this type of solution, Soulsby described a data point model as a business rather than technology model that provides a base for concerns such as data quality, validation, storage, integration, reporting requirements definition and the generation of XBRL taxonomy. The model is combined with the MarkLogic multiple model database that can store multiple types of data in one database and support queries across the data, making it a good tool for data aggregation.
Atkinson explained: “The clever part is stitching regulatory requirements to data in the database. The data needs to be mapped and metadata attributes need to be created in the database and tagged to an ontology such as FIBO.” Combining the source metadata, ontology, an enterprise data model, and ModelDR data point models for diverse elements such as BCBS 239, MiFID II, data quality, FIBO and Murex with NoSQL and semantic queries it is possible to run queries across the models to generate a view of data that could be used for regulatory or other reporting purposes.
Soulsby says: “The combination of the MarkLogic database platform and ModelDR data point model allows data to be ingested dynamically without ETL tools and queries to be run on the fly in real time against any database in the bank. BCBS 239 suggests banks need these kinds of capabilities.”