Ellipse: regulatory reporting and data analytics platform

An integrated regulatory data and analytics platform

Project Ellipse explores how supervision could become insights based and data driven using an integrated regulatory data and analytics platform. If implemented, regulatory authorities, as the ultimate end users of the platform, would be able to digitally extract, query and analyse a large quantity of data from diverse sources. These data could then be relevant to current events in real time and visible via dashboards, informing them of early supervisory actions that may need to be taken.

An overview of Project Ellipse

5 Oct 2021 An integrated regulatory data and analytics platform

Project Ellipse is a proof of concept (POC) launched by the Bank for International Settlements Innovation Hub to explore how supervision could become insights-based and data-driven using an integrated regulatory data and analytics platform.

Project Ellipse is being developed in two phases.

In Phase 1 of the project, the BIS Innovation Hub partnered with the Monetary Authority of Singapore (MAS), the Bank of England (BoE) and the International Swaps and Derivatives Association (ISDA) to explore the concept of cross-border digital regulatory reporting, using a machine executable data model.

Reporting platforms built on common data models offer the possibility that global financial reporting entities can fulfil cross-border reporting obligations using a common input layer. This would reduce compliance burdens placed on those financial institutions to respond to template-based regulatory reporting requests from different supervisory regimes for similar exposures. It would also enable home and host supervisors of these global reporting entities to compare exposures in a more consistent and transparent way.

To find a common understanding of data collected for regulatory purposes across reporting regimes, the project reviewed reporting requirements for retail mortgages in the UK and Singapore. It then derived a subset of granular data attributes from those requirements and modelled them using ISDA's Common Domain Model (CDM), which is an open-source, standardised, machine-readable and machine-executable model used for over-the-counter (OTC) derivatives, cash securities, securities financing and commodities.

By using the CDM, the PoC demonstrated the feasibility of extending an existing globally applicable derivatives data model to retail mortgages. Executable code generated from the model's definitions enabled the automation of regulatory mortgage data for Singapore and the UK, referencing the same common model.

Phase 1 of our project illustrates the possibilities of this process and the efficiencies gained when adopting machine executable reporting using common data models. It also increases the volume of granular data available to supervisors, which is needed to enable the use of advanced analytics.

You can view our Phase 1 data model, a demonstration and technical documentation.

Building on this use case, Phase 2 will explore the integration of granular data sets with unstructured data, using artificial intelligence (AI) and machine learning (ML) to extract insights from these data sources to highlight correlations between current events and supervisory metrics. Insights extracted from the mined data would be displayed as early warnings for supervisory attention via dashboards.