Risk Data Aggregation and Risk Reporting (RDARR) is critical to compliance in financial services and is manifested in several initiatives such as BCBS 239, DFAST and CCAR. I will be co-hosting a webinar with Zaloni on this topic called Governance of the Big Data Lake with a focus on RDARR for Financial Services on Tuesday, December 1st at 1 pm eastern standard time.
I have had the opportunity to work with Ben Sharma, Zaloni’s CEO, in the past and I am excited about the combined story we can tell, and best practices we can share with you. Here are a few topics, critical to RDARR compliance, that we will cover during the webinar:
- Agree on Critical Data Elements (CDEs)
RDARR programs need to have a clear understanding of CDEs. For example, Gross Domestic Product (GDP) is a key macroeconomic variable that is used to drive stress tests. RDARR programs need to have a clear understanding of the authoritative source for GDP as well as related terms such as nominal GDP growth rate and real GDP growth rate.
- Identify data owners
RDARR programs need to identify data owners who will be accountable for business definitions, data lineage and data quality for CDEs.
- Establish processes to ingest new data into the Big Data Lake
Because of the volume and complexity of RDARR, financial institutions are increasingly turning to Big Data Lakes based on technologies such as Hadoop. RDARR programs need to have clearly defined processes to govern what data gets ingested into the Big Data Lake and who approves.
- Manage data lineage
RDARR programs also need to rely on metadata technologies to support regulatory requirements for data lineage all the way back to source systems.
- Improve data quality
Financial institutions need to have strong processes to improve the quality of RDARR data. This includes identifying data stewards, profiling CDEs in source systems and resolving data issues.
- Data Sharing Agreements
Financial institutions need to implement Data Sharing Agreements to support informal arrangements like data quality service level agreements (SLAs). For example, if the Risk team is attesting the BCBS 239 reports then they need Data Sharing Agreements from source system data owners that the quality of CDEs meets acceptable thresholds.
- Govern risk models
RDARR programs need to tie model governance into the broader data governance platform. This means that risk models will have assigned owners. In addition, data lineage needs to show traceability from risk models to input variables and all the way back to source systems.
- End User Computing (EUC)
RDARR programs also need to define owners and manage data lineage for EUC assets that are used in risk models. These EUC assets include SAS datasets, Microsoft Access databases and Microsoft Excel workbooks that are used as inputs into risk models.
During the webinar we will touch on each of these topics. Hope to see you there. In addition, in a future blog, I will go into further detail on how Zaloni’s flagship product, Bedrock, addresses these critical requirements for RDARR governance in financial services.