Ellipse: regulatory reporting and data analytics platform

What is Project Ellipse?

Project Ellipse is a proof of concept (PoC) that aims to demonstrate the functionalities and feasibility of an integrated regulatory data and analytics platform. An integrated platform would enable supervisors to digitally extract, query and analyse large and diverse sources of structured and unstructured data that are relevant in real time and to current events.

What problem does Project Ellipse address?

Regulatory reporting for financial institutions and for supervisors is still manual, highly aggregated, template-based and hence static and inflexible in its use. Data are also sourced from largely disparate legacy systems, which often results in a heterogeneity of data points for any given product or transaction, both within a bank and across banks as these systems describe the data differently. These issues are further exacerbated at the global level, where internationally active banks need to duplicate similar data in different ways to fulfil reporting obligations across jurisdictions. In addition, at times of heightened risk, such as the current Covid-19 pandemic, the need for up-to-date data increases, but given the challenges of compiling and receiving the data, supervisors may not have consistent or the timeliest data to make informed judgments.  

Project Ellipse intends to address these common pain points by demonstrating how an integrated platform can:

  • reduce compliance burdens placed on financial institutions by moving away from template-based regulatory reporting requests
  • be nearer to "real-time" and relevant to current events to support supervisory judgments and actions, both locally and globally
  • support a move towards newer digitally enabled architectures to replace traditional concepts and processes of data collection
  • enable predictive insights and early warning by integrating big data analytics

The roadmap: what will we be doing?

We are working with our host central bank, the Monetary Authority of Singapore, to deliver a PoC that demonstrates the functionalities of the platform with today's data collection infrastructure. The PoC uses mortgage exposures, which comprise a wide range of products that describe these data differently. Such exposures are also prudentially and systemically relevant to supervisors globally.

The PoC involves two phases:

Phase 1: Interoperable reporting platform using a common data model

Phase 1 will build a reporting platform using a common data model. Integrated reporting platforms built on common data models offer the possibility that global financial reporting entities can fulfil different cross-border reporting obligations using a common input layer. This would reduce compliance burdens placed on financial institutions to respond to template-based regulatory reporting requests from different supervisory regimes. It would also enable supervisors to compare exposures in a more consistent and transparent way.

Phase 2: Data enrichment and advanced analytics to enable predictive insights and early warning

Building on the mortgage use case, data on mortgages at origination can be augmented by risk assessments over the life of the mortgage via borrower income shocks, real-time assessments of collateral value changes and (more recently) loan moratoriums. Phase 2 will build on the platform by enriching the internal supervisory metrics using big data, such as news, macroeconomic indicators and property sales by ways of example. Advanced analytics using tools such as machine learning will also be applied to harvest insights and correlations of the enriched data sets to enable predictive and early warning indicators of risk. Machine learning techniques for supervisory use would also increase the capabilities of supervisors to oversee the use of artificial intelligence and machine learning by the financial sector.

What will we deliver?

The common data model and automated reporting rules will be open-sourced for our stakeholders. We will also share the methods and tools used to normalise and segment the big data sources enabled by machine learning techniques. Above all, the project is intended to provide the central bank and regulatory communities with an implementation reference model of a regulatory data and analytics platform to support a move towards newer digitally enabled architectures that replace traditional concepts and processes of data collection and analytics.