Welcome to the 12 Days of BCBS 239! On this first day, our banker friend G. Sibley starts with a fundamental: the data dictionary.
BCBS 239 addresses how banks should manage the data used in risk analysis. Our experience is that the most logical place to start is setting and agreeing definitions of terms. This logic is reflected in the BCBS 239 Principles:
- BCBS 239 Principle 3 calls for data to be defined consistently across an organization.
- BCBS 239 Principle 9 calls for banks to inventory and classify their risk data items.
The first step on G. Sibley’s road to BCBS 239 compliance is to ask a series of questions, including: Which data elements matter most to risk analysis and reporting? What do we call them? Are these meanings consistent across the enterprise? Are there instances where different businesses refer to the same thing by different names, or different things by the same name? It’s important to document these instances.
Let’s dive into what “defined consistently” means. It’s very common for different areas of the business to refer to the same attribute by different names. This is ok in the context of risk data aggregation, so long as this is identified and recorded. What matters most from a risk data aggregation perspective is that each of these attributes named in “local dialect” is mapped to a single underlying attribute definition which outlines the shared meaning of those terms. The same concept applies to terms used to express different meanings in different contexts – the local business areas don’t need to change how they refer to specific data attributes, but the differences need to be known. Only once you have a common data dictionary can you get to the business of managing and aggregating data effectively.
The good news is that you don’t have to start from scratch. There are templates and platforms to accelerate you and industry standards, such as FIBO, the Financial Industry Business Ontology, created by the membership of the EDM Council. FIBO provides a common standard reference point for data alignment.
Look out for day two tomorrow.
By Steve Engdahl, SVP, Product Strategy, Goldensource