Establishing data controls in financial services for effective model data management

Co-authored by: 

Vishal Gossain - EY Risk Analytics and Balance Sheet Management Leader | EY Canada

Anil Sood – EY AI Governance and Model Data Management Leader | EY Canada

Ishant Arora, Manager, Risk Consulting, Financial Services Risk Management, EY Canada

Gauri Rajgopal, Senior, Risk Consulting, Financial Services Risk Management, EY Canada 

Ensuring good data quality and robust data controls in financial services is necessary for effective management of model data. Key financial regulations or guidelines, such as BCBS 239 Principles and OSFI’s E-23 Guidelines on Model Risk Management and Data Maintenance Notes for IRB institutions emphasize the need for effective data controls.

Model data management is a growing discipline aimed to ensure data is fit for use in models and meets regulatory and management expectations. Data controls help organizations prevent data issues and subsequent regulatory fines and loss of customer trust.

Ensuring compliance and effectively managing business operations requires organizations to make investments in designing and implementing a robust data controls strategy and procedures. An effective controls strategy ensures reasonableness of risk model results by focusing on the people, process and technology aspects:

• Data controls ownership along with clear definition of roles and responsibilities to operationalize and monitor the controls
• A structured process to design controls across the data lifecycle, such as upstream data capture, data consolidation and pre-processing
• ·Consistent implementation of controls across the underlying data and technology platforms that handle business data

To effectively operationalize the data controls strategy, it requires partnership across business units: the chief data office, technology and risk functions.